Re: [SHR] Scrolling (was Re: [SHR] Miscellanious minor issues)
On Thu, 29 Jan 2009 17:20:40 -0500, Joel Newkirk wrote: > single > icon image and toggling (visible when touched) just a single-image frame > behind it, and removing associated fade-in/fade-out animations makes a > significant difference. Pop open the Illume Config (via spanner/wrench) > and try finger-scrolling the top icon bar right to left - it's a > tremendous > difference when everything's structurally and graphically simplified. > > j Sorry, when I referred to 'single icon image' that was as opposed to multiframe animated icons. (I realized after sending that this might sound stupid, since the default icons are all static images) I've not seen them in any distro or package yet for the FR, but Enlightenment supports them. I've used an animated blinking shell prompt before as my terminal icon. http://www0.get-e.org/Resources/Animated_Icons/ - just stick them (they're .edj files) in /usr/share/enlightenment/data/icons and put the name in for icon in the .desktop file... (with the understanding that loading down your Freerunner with a bunch of animated icons is guaranteed to slow down your scrolling...;) j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [android] inital thoughts
On Fri, 23 Jan 2009 00:21:26 -0500, Charles Pax wrote: > On Thu, Jan 22, 2009 at 9:06 AM, Russell Hay wrote: > >> Just installed the latest android image, and it's looking nice, >> responsiveness is where 2008.8 used to be, and my main gripes from a 1hr >> play are; >> >> - very low call volume, despite setting it to full >> - to accept calls, you have to use use the key sequence as follows: > power >> button- choose keyboard-hit the phone icon >> >> Just an inital impression - and I appreciate it's a port so these are >> workarounds for a lack of a keyboard! >> > > If the rumors [1] of the G2 are true, we won't have to worry about ugly > workarounds; a touch screen only interface will be in the upstream > > -Charles Pax > > [1] http://i.gizmodo.com/5135926/android-g2-photos-thinner-and-no-keyboard However, the G2 does have the five hardware buttons (plus a trackpoint, looks like) that 'cupcake' is said to require... j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
[SHR] Scrolling (was Re: [SHR] Miscellanious minor issues)
On Thu, 29 Jan 2009 20:48:28 +0100, kimaidou wrote: > I don't totally disagree, but for now scrolling in SHR is painfull. I tried > Qextended, and the scroll (eg in contacts) is smoothy and efficient ! I > also > proposed shortcuts in another trac for contacts : > http://trac.shr-project.org/trac/ticket/292 > > 2009/1/29 Johny Tenfinger > >> On Thu, Jan 29, 2009 at 19:37, kimaidou wrote: >> > About the scrollbar, it is much to thin to be reachable with the > finger. >> I >> > proposed an enhancement in the trac : make the right scrollbar twice > (or >> 3 >> > times) bigger, so that all soft needing it become more finger friendly >> >> In elementary (shr-settings), scrollbar is not reachable at all, >> because it is not supposed to. For me, scrollbar in contacts should >> look and behave like one from elementary, so I don't agree. It's only >> wasting of space, and there are better ways to navigate than scrollbar >> and finger scrolling (like alphabetic filtering in contacts etc.) I'm split between camps on this. I don't want to see scrollbars usually, as they are either so small as to be ornamental, or so large as to waste huge swaths of valuable screen real-estate. But simple finger-scrolling doesn't cut it sometimes. For example, I tried to use neon to examine some image files from the default.edj theme. Unfortunately that theme contains 400+ images, and the one I wanted was around 360 in the list - I gave up after about 60 seconds of finger scrolling, somewhere around 'J'. What we need, IMHO, is not just alphanumeric jumps, but standardized presentation of, activation of, and interaction with such controls. Reusable that way for file requesters, addressbook, whatever. Meanwhile, the 'appears on demand' scrollbar in some Elementary apps looks promising, if only it were a bit wider so as to be friendly for fat fingers like mine. BTW - scrolling is made more painful by the theme... Raster pointed this out to me - the illume theme in SHR uses multiple layered png images with transparency for icons, for example. (the button, the icon, the hilight, the shine, with anternates for disabled or selected states) Using a single icon image and toggling (visible when touched) just a single-image frame behind it, and removing associated fade-in/fade-out animations makes a significant difference. Pop open the Illume Config (via spanner/wrench) and try finger-scrolling the top icon bar right to left - it's a tremendous difference when everything's structurally and graphically simplified. j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [All] To build a better music player
On Thu, 29 Jan 2009 18:09:39 -0500, Dylan Reilly wrote: > I am still in the process of making changes, but I have some packages > if anyone wants to try it. This version has the option of using a > gstreamer back-end (the default with the packaged config file) instead > of mplayer. By using gstreamer, CPU and memory load have been reduced, > and I was able to rig it such that track transitions occur without the > delays that were present using mplayer. Furthermore, I added tag > reading support for Ogg and Flac through mutagen. > > The program itself is [1]. The package does not require all the other > dependencies since technically one does net need to use gstreamer. The > dependencies for gstreamer are: [2], gst-plugin-playbin, > gst-plugin-mad, python-pyalsaaudio, and possibly others that I missed. > Any idea where I can find python-pyalsaaudio? not found on SHR, can't find a package anywhere. (except armv7a at angstrom) j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [All] To build a better music player
On Fri, 30 Jan 2009 09:31:21 +, Al Johnson wrote: > On Friday 30 January 2009, Joel Newkirk wrote: >> Any idea where I can find python-pyalsaaudio? not found on SHR, can't > find >> a package anywhere. (except armv7a at angstrom) > > http://openmoko.truebox.co.uk/repos/mazikeen/fso-testing/ipk/armv4t/python- > pyalsaaudio_0.3-ml0_armv4t.ipk Thanks. Unfortunately it's still unhappy, due to the change to python2.6. :( j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Dialup On Demand (was: [SHR] Miscellanious minor issues)
On Sat, 31 Jan 2009 11:34:42 +1100 (EST), "NeilBrown" wrote: > On Sat, January 31, 2009 2:55 am, Michael 'Mickey' Lauer wrote: >> Ok cool, but how does this work internally? We want the framework to be >> notified, not the ppp process (which we don't want to run all the time) > > On demand network connection is awkward to do in a completely general way. > > When an application makes a TCP connection, the kernel needs to choose a > local address for "this" end of the connection. It chooses that > based on the IP address of whichever interface will be the first > hop of the outgoing connection. > > If you don't have that interface established, then there is no address > or route, so the kernel will give up quite quickly - there is no where > for user-space to hook in and make things work. > > One option is to always use a VPN through some base station. > I think you can get OpenVPN to establish a connection on demand. > So the kernel sees the OpenVPN interface and sends a packet to that. > OpenVPN then brings up whatever network can be found and forwards the > packet over the VPN and on the the Internet. > > You could probably set up something vaguely similar that works > locally without needing a remove VPN server or even a VPN. What about (which I expect will not be a candidate for an 'official' solution;) get frameworkd talking to ip_queue or netlink. For example, we could: Create a lowest-priority default route that hits lo, like: route add default gw 127.127.127.127 dev lo metric 100 use ip_queue.ko. Write a userspace handler for the iptables QUEUE target. (It's been a few years but I've written a couple, in c, use libipq - pretty simple) The handler will receive every packet that is sent to it with a firewall rule like: iptables -A INPUT -d ! 127.0.0.0/8 -i lo -m limit --limit 1/sec -j QUEUE iptables -A INPUT -d ! 127.0.0.0/8 -i lo -j DROP Now whenever another route doesn't exist, the kernel presents outbound packets to the queue handler. (up to one per second, and unspecified --limit-burst is 5 - the second rule drops what doesn't get queued) The handler gets to tell the kernel whether to accept or drop the packets, in this case it'd be simplest probably to drop them. In the meantime, however, it can tell the network manager (or whatever mechanism) to try to bring up an interface. If that's successful then its newly-created route takes precedence. (providing it has a lower metric than 100...) It can also take a moment to examine the packets it's handed and decide - based on whatever criteria desired - if it really needs to bring up an interface or not. Like "if it's DNS fire it up, but if it's broadcasts just ignore it". (those criteria could be handled in iptables actually, but others would be harder, the things that frameworkd can offer like "is a voice call in progress?" [if so, forget gprs] or "am I preparing to suspend?", etc) j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Testing] 12-hour clock
On Sun, 1 Feb 2009 15:13:47 -0600, The Digital Pioneer
wrote:
> Hi, guys. Does anyone know how to set the phone to use a 12-hour clock?
I'm
> one of those weird Americans who don't like 24-hour clocks. :P
>
{applies to all non-qtopia distros AFAIK}
The clock is laid out in the illume theme, such as
/usr/share/enlightenement/data/themes/illume.edj - I've been working on
customizing the theme to suit my tastes, one of the things I did was add
the date below the time. I experimented with AM/PM but felt it used too
much valuable space, given that the same space is being competed for by GSM
widget, battery widget, left/right arrows, taskname, etc. (or, I had to
make the font for the time too small)
If you extract the theme with "edge_decc illume.edj"[1] then look through
the .edc file to find "MOD: CLOCK". There's a script within the clock
module section that includes lines like these:
timer(60.0 - second, "clock_cb", 1);
if ((hour < 10) && (minute < 10))
{snprintf(buf, 10, "0%i:0%i", hour, minute);}
else if ((hour < 10) && (minute >= 10))
{snprintf(buf, 10, "0%i:%i", hour, minute);}
else if ((hour >= 10) && (minute < 10))
{snprintf(buf, 10, "%i:0%i", hour, minute);}
else if ((hour >= 10) && (minute >= 10))
{snprintf(buf, 10, "%i:%i", hour, minute);}
set_text(PART:"e.text.label", buf);
(the appropriate changes are left as an exercise for the reader ;)
recompile the theme with './build.sh' inside the theme dir, move the
created .edj file (presuming no errors) up one folder to replace the
original (note that enlightenment is quite prone to segfaulting when you do
this, so do it via SSH!!) then restart enlightenment with "killall -HUP
enlightenment" and it will reload with the altered theme. (note that
unless you use a different name and change enlightenment config to use the
'new' theme, it will be overwritten by package upgrades)
I'm planning to revisit the clock soon, I want to try to incorporate
external flags that are honored to determine whether to display day, and
whether to use 12- or 24-hour format. If successful I'll post the full
clock module group for anybody who wants to incorporate it in their edje
theme.
j
[1] or whichever theme you're using - note that if 'default.edj' is present
it is normally overridden by the actual selected theme, so changing the
clock there [which is the analog one with 120+ png files] likely won't make
any change to your clock. You may need to "opkg install edje-utils" to
have edje_decc & edje_cc. (the latter is used by build.sh to rebuild the
.edj)
--
Joel Newkirk
http://jthinks.com (blog)
http://newkirk.us/om (FR stuff)
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR] no route to host
On Mon, 02 Feb 2009 18:49:00 +0100, Vinzenz Hersche wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > hello there, > > i've the shr unstable from today and got a problem; i was connected on > ssh to my phone and want to test the wifi-connection (so i've got 2 > connections, usb and wifi). after i want to connect to wifi, my shell > on the computer freeze. > > so i tried to connect a second time, but there's just a failure-message: > > skams...@skamster-laptop:~$ ssh r...@192.168.0.202 > ssh: connect to host 192.168.0.202 port 22: No route to host > skams...@skamster-laptop:~$ > > this problem is also there after a reboot. > > i think, wifi changed something in /etc/interfaces, but i don't know > what. does someone know, what i must change to connect with usb to my > phone? > > thanks a lot for ideas, hope, this failure isn't there in the future.. :) > > greets Wifi /shouldn't/ touch /etc/network/interfaces, usually the only programs that will ever try to alter it are network managers. What procedure did you use when trying to bring up wifi? I've previously used "ifup eth0" from console or terminal, but with newest SHR (Jan31 or so opkg upgrade) there IS NO ETH0 most of the time, resulting in "No such device" errors. Also note that if you've been connecting and disconnecting the FR from the host computer many times since the host was rebooted it may have problems. Initially my usb0 was the third network device (the '3' in the first column above) on my host, but currently it's 12. I've seen it get "too high" before and only a reboot of the host would allow it to connect again. (sorry, I don't remember where it started failing - IIRC it was at 32 but don't shoot me if that's way off) I recently encountered the same problem - unplug from USB, power-cycle freerunner & wait for boot to finish, plug in USB, wait 5 secs, unplug USB, wait 5 secs, plug back in, and I had a connection. (I don't know if it's the FR or Ubuntu but if I boot or reboot the FR while attached to USB it usually doesn't connect until unplugged and re-attached, so I'm already used to unplug/wait/replug) If that fails, then reboot the FR, open the terminal on the FR and check the output of "ip a s usb0" - allowing for the randomly-selected MAC address, it should look pretty much like: 3: usb0: mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 22:7d:18:7d:5a:a8 brd ff:ff:ff:ff:ff:ff inet 192.168.0.202/24 brd 192.168.0.255 scope global usb0 inet6 fe80::207d:18ff:fe7d:5aa8/64 scope link valid_lft forever preferred_lft forever and "ip r" should show you: 192.168.0.0/24 dev usb0 src 192.168.0.202 default via 192.168.0.200 dev usb0 Presuming that's correct, do the same on the host computer, which should at least show interfaces usb0, though it may be down or have no IP address at this point. If it's down, try "ifup usb0" on the host (root or sudo) or "ip l s usb0 up" and take a look at the data again. If the interface is up on the host but has no IP, try "ip a a 192.168.0.200/24 dev usb0" (again, root or sudo) and see if you can connect. j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: YAIKT - Yet Another Illume Keyboard Thread
On Mon, 2 Feb 2009 11:55:12 +0100, Michael Zanetti wrote: > Hi all! > > I'm using the illume keyboard on my OM2008.12 powered Freerunner. I have > created my own keyboard layouts that perfectly fit to my needs: > The problem now is, that when I'm writing a message using my thumb and > switch > between the different layouts I need to slide the keyboard up once to > switch > from the qwerty layout to the numbers but I have to slide up twice to get > back > to the qwerty one. This is because I have to hop over the full-featured > layout > that I cannot use with my fingers. Just checking - are you aware that you should be able to slide down as well? if you have three keyvboard layouts defined, the 'other two' are always a single stroke away, up or down. j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [paroli] update week 6/09
SHR-unstable, 01/29, paroli downloaded 01/31 or 02/01: root INFO read config file ./paroli.cfg root INFO read config file /etc/paroli/paroli.cfg root INFO read config file /home/root/.paroli/paroli.cfg root INFO init gui root WARNING can't use backend paroli : No module named etk root WARNING can't use backend csdl : No module named gui root WARNING can't use backend sdl : No module named guip Traceback (most recent call last): File "./paroli", line 146, in tichy.init_gui(None) File "../paroli-core/tichy/__init__.py", line 71, in init_gui import guip ImportError: No module named guip What is missing? j On Mon, 02 Feb 2009 15:43:06 +0100, Mirko Lindner wrote: > Hi all, > > in the good tradition of weekly updates we will try to to get back on > track of keeping you posted on current developments in the paroli and > GUI section. > > For the code base of paroli(d) the past weeks development has been > focused on the core on the one hand and finishing the first set of > design files on the other. > > biggest changes: > - a launcher was introduced allowing the launching of > paroli-applications from within paroli itself > - the dbus interface is optional and both modes should work > - closing application windows from within the window-manager does not > kill the process anymore > - introduction of a preliminary audio service allowing to mute calls (a > test service for non-fso environments was also added) > - a reworked version of the dialer was added > - works on an error-message service have begun > - works on a service checking for changes in the fso have begun > and many small fixes in the code > > for more info go to http://www.paroli-project.org > (also check our api section at http://www.paroli-project.org/api/ ) > > Hoping that the core offers everything we need for now, we hope to be > able to focus our attention on the applications again. > > Main points here are the integration of opimd, a more resource efficient > list generation as well as more a more generic and reusable code-base in > the apps. > > /mirko > > website: http://www.paroli-project.org > bug tracker: http://www.paroli-project.org/trac > source: http://git.paroli-project.org > > _______ > devel mailing list > de...@lists.openmoko.org > https://lists.openmoko.org/mailman/listinfo/devel -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [All] FR survey and locations
On Tue, 3 Feb 2009 21:46:09 -0200, Pablo Miño wrote: > How do I find out what version my neo is? Do I have to open it? http://wiki.openmoko.org/wiki/Finding_hardware_revision j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: when buying a freerunner, ...
Note also that the 'buzz' only affects some people, with no certain pattern yet discerned, and is fixable with a minor hardware alteration that OM is working out the details of providing to any and all A5/A6 owners. (change one resistor, add one capacitor) It's worth checking (but no guarantee) if other people in your area and/or on your carrier have reported problems, at http://spreadsheets.google.com/pub?key=p2ZQdcRSVg9XfYLwBVsZicg j On Sun, 8 Feb 2009 09:52:40 +0100, Yorick Moko wrote: > the fix for the sd card is introduced in A6 (that is what they are > selling and producing at the moment), there also was some minor > led-power-drain correction. > the fix for the buzz will be in A7, which is not yet in production > > On Sun, Feb 8, 2009 at 2:39 AM, wrote: >> there were issues with noise/buzz on the speaker that would be solved > with >> some soldering. >> >> If I would buy a new freerunner now, would it have the fix ? If none, > when >> would that fix become standard ? If all current, after which 'serial > number' >> would the fix be included ? >> >> Thx >> W >> >> ___ >> Openmoko community mailing list >> community@lists.openmoko.org >> http://lists.openmoko.org/mailman/listinfo/community >> > > _______ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR - latest unstable] endless ringing
On Tue, 10 Feb 2009 18:16:43 +0100, Tony Berth wrote: > Dear List, > > there are cases where someone calls me, I try to pick-up the call but the > button 'answer' doesn't react! As a result the phone keeps ringing for > ever! > > Even if I re-start the xserver the phone keeps ringing! Any idea why that > happen? > > Thanks > > Tony The changelog for FSO milestone5 includes: Fixed a major bug in the RingToneAction. This (and removing a race in the RingToneAction) fixes problems with neverending vibration and audio ringtone on short calls. Ergo, the problem appears to be fixed upstream and should arrive in SHR when frameworkd there is updated to match M5 FSO. j -- Joel Newkirk http://jthinks.com (blog) http://newkirk.us/om (FR stuff) ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: SHR missing rootfs (was: Re: bearstech SHR missing rootfs)
On Tue, 17 Feb 2009 17:57:40 -0800 Michael Shiloh wrote: > SHR project currently resides at: http://shr-project.org/trac > this link has instructions on getting and installing your SHR > image. > > Just before I fixed the wiki page it occurred to me to double-check: > > http://shr-project.org/trac points to "how to install" which says to > get images here: > > http://build.shr-project.org/shr-testing/images/om-gta02/ > > which contains only the kernel and not the rootfs. > > The wiki page is very helpful in that it points to a separate page > which discusses the different versions of the rootfs, but the links > point to the same place above which has only the kernel. > > What now? I'd been editing it after reading your previous post and pointing those four links at the build.shr-project.org folders, and sure enough the testing images are not complete there. However, the unstable images ARE at that site, and unstable is so far beyond testing that at this time installing shr-testing is probably not a good idea. I've got the wiki pointing at the same four targets (testing and unstable for gta01 and gta02) with what are currently valid links. I suspect that the links for Testing will need to change eventually. * [http://shr.bearstech.com/shr-testing/images/om-gta02 Neo Freerunner testing] * [http://shr.bearstech.com/shr-testing/images/neo1973 Neo 1973 testing] * [http://build.shr-project.org/shr-unstable/images/om-gta02 Neo Freerunner unstable] * [http://build.shr-project.org/shr-unstable/images/om-gta01 Neo 1973 unstable] j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: (SHR) again
On Sun, 22 Feb 2009 17:19:29 +0100 Francesco de Virgilio wrote: > Pardon, I answer myself[1]: the GPS is powered on autoagically on > boot, so... is there a way to default power it off on boot? > > Thanks ;) It IS powered off by default. 'Auto' means that frameworkd will power up the relevant subsystem only if/when something requests the resource. In the case of the GPS, fso-gpsd is not starting up the gps, but is waiting for something to access its 'fake' libgpsd, and when that happens it requests the gps resource from frameworkd, so that it powers up the GPS. Similar with WiFi - until something requests the resource it is completely powered down. (except - like other non-gps resources on the FreeRunner - it doesn't have a special external handler like fso-gpsd, it depends on the resource being properly requested via dbus) The GSM chipset is also handled this way - with SHR, the GSM chipset isn't powered up until something requests it. (like ophonekitd, Paroli, Zhone, etc) But ophonekitd is launched by default when X starts up, so the visible effect is that GSM is powered on automatically when booted. The result is that if you were to disable ophonekitd, you could boot the FR and it would initially conserve maximum power by leaving GPS, GSM, Wifi, and Bluetooth powered down. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GUI for SHR fast charge mode: possible?
On Sun, 22 Feb 2009 17:37:36 +0100 Francesco de Virgilio wrote: > as in the topic name, I've flashed SHR unstable and noticed that the > battery.py GUI I used to open on OM 2008.12, with SHR doesn't work > anymore, showing the following error: SHR unstable (and FSO MS5) use 2.6.28 series kernels, and the sysfs structure has been rearranged: > IOError: [Errno 2] No such file or directory: > '/sys/devices/platform/bq27000-battery.0/power_supply/bat/capacity' /sys/class/power_supply/battery/capacity > IOError: [Errno 2] No such file or directory: > '/sys/class/i2c-adapter/i2c-0/0-0073/charger_type' /sys/class/power_supply/adapter/type (I think...) > IOError: [Errno 2] No such file or directory: > '/sys/devices/platform/bq27000-battery.0/power_supply/bat/voltage_now' /sys/class/power_supply/battery/voltage_now > For me the fast-charging mode is really important to use the Neo in my > car, which adaptor for default provides only 100 mhA. Is there a way > to set quickly fast-charge mode on SHR unstable? /sys/class/power_supply/ac/device/force_usb_limit_dangerous Note that these are not the only paths that will work, as there are a buttload of symlinks within the sysfs structure such that some of these will appear at several different points under /sys/. But the original paths from the 2.6.24 kernel won't work. What I've listed here are (AFAIK) the shortest paths to each of these four. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
[ubuntu host] NetworkManager improvements
I know several people here have had problems with the NetworkManager bugs in Intrepid. I installed Xubuntu 8.10 on my new netbook the other day, upgraded everything, and found that if I used the networkmanager applet to define the interface 'auto usb0' with manual IP of 192.168.0.201/255.255.255.252 (the subnet I use for my FreeRunner, two IPs plus network and broadcast) it 'just works' again. On my desktop Ubuntu (gnome) however, I've still been unable to get it to handle it correctly. (I'm seeing some odd misbehavior in the gnome NM applet, it's possible that either xfce's nm-applet or NM itself is why my netbook works, but gnome nm-applet may still be broken) Has anyone else been successful with Ubuntu NetworkManager recently? And did you revert from a manual config via /etc/network/interfaces to allowing NM to handle usb0? BTW, if anyone is interested I have a working udhcpd set up where my FR hands out 192.168.0.201/30 to whatever asks for an IP on usb0, so I can just jack in and go, more often than not, on a new host. The daemon listens only on usb0, and hands out only an IP and subnet with no routes, which works fine with NetworkManager default - it finds usb0 without a problem on a 'virgin' host and asks for DHCP, and the FR provides. :) Oh, udhcpd is one of the functions rolled into busybox, so it's already on the FR by default and just needs config and launch. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: SHR first impression : it's slow ?
For those looking for the 'Advanced' settings: Grab and drag the row of icons at the top from right to left and you should find, altogether: Display Look Screen Input Language Settings Advanced Extensions AFAIK the arrows don't even work, when you tap the right-hand arrow you are effectively clicking on the edge of the 'Input' icon. BTW - the speed with which those icons scroll is how fast the main Illume icons /should/ be able to be dragged, but with the Illume theme as it currently stands (with FSO M5 and SHR at least) each 'desktop' icon is adorned with multiple transparent PNG files layered to create pretty buttons, and scrolling becomes dramatically slower than what you are probably familiar with from the base ASU theme with 2008.x. I (and surely others) am working on a leaner, faster theme - any eye-candy that distinctly impacts the user experience should NOT be default, and in this case we desperately need a simple and fast theme as the default or the immediate impression users will get is: "Damn, this is slow!" j On Mon, 23 Feb 2009 15:49:55 +0100 Yorick Moko wrote: > i have shr unstable from 09/02 > and i can confirm that there is no "advance" entry under "screen" > the top bar only shows: > "display" "look" "screen" and "input" > > I know there used to be an option to change the software rendering, > but can't find it on this image... > > y > > > On Mon, Feb 23, 2009 at 3:44 PM, Xavier Cremaschi > wrote: > >> I can hit the wrench and see: > >> > >> "Display" "Look" "Screen" "Input" > >> No "advanced" at all. > > > > You can scroll horizontally. I have lots of item in this bar in my > > SHR (unstable installed yesterday) > > > > Xavier. > > > > > > ___ > > Openmoko community mailing list > > community@lists.openmoko.org > > http://lists.openmoko.org/mailman/listinfo/community > > > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: NetworkManager works just fine (was: [ubuntu host] NetworkManager improvements)
On Mon, 23 Feb 2009 15:35:41 +0200 "Patrick C. F. Ernzer" wrote: > Hello, > > On Sun, 22 Feb 2009 21:13:17 -0500 > Joel Newkirk wrote: > > [...] > > Has anyone else been successful with Ubuntu NetworkManager recently? > [...] > > Not Ubuntu but NetworkManager in general. As said on the wiki page, > one needs to nail down the MAC address for NM to dish out the same IP. > > I use the following on many boxes, so the version numbers can be off a > bit. > > NetworkManager-0.7.0-1.git20090102.fc10.i386 > > create /usr/local/sbin/freerunner-usb-add.sh > create /etc/udev/rules.d/80-freerunner.rules > cd /etc/udev/rules.d/ > chcon --reference=40-multipath.rules 80-freerunner.rules > chmod 744 freerunner-usb-add.sh > udevcontrol --reload_rules > > Depending on how your firewall is set up, you may or may not have to > issue the following > > iptables -A POSTROUTING -t nat -j MASQUERADE -s 192.168.0.0/24 > iptables -I FORWARD 1 -s 192.168.0.202 -j ACCEPT > iptables -I INPUT 1 -s 192.168.0.202 -j ACCEPT > echo 1 > /proc/sys/net/ipv4/ip_forward > > RU > > PCFE Thanks. With the current Ubuntu there have been problems where NM tries to take control of usb0 despite being told not to, or refuses to follow manual setting within its own GUI, but also it has had problems where it tries to point default route out usb0, or brings eth0 down and back up, or other odd behavior. Lately most of that seems fixed. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
GUI responsiveness (was Re: SHR first impression : it's slow ?)
On Tue, 24 Feb 2009 12:32:09 +0100 Helge Hafting wrote: > Yorick Moko wrote: > >> I (and surely others) am working on a leaner, faster theme - any > >> eye-candy that distinctly impacts the user experience should NOT be > >> default, and in this case we desperately need a simple and fast > >> theme as the default or the immediate impression users will get > >> is: "Damn, this is slow!" > >> > > > > nice to hear that > > ugly and fast beats pretty and damn slow any day for me > > (pretty and fast would also be acceptable ;-)) > > Pretty and fast should be possible then. > > There is no need for multiple layers of transparent icons. They can > be collapsed into one layer with a single transparent icon, looking > _exactly_ the same. > > Effects when a icon is actually selected is another story, but that > sort of thing should not need to impact scrolling. > > Helge Hafting Therein lies the problem, in a sense. (or a large part of it) With the default.edj theme (Illume doesn't override it for Fileman, which includes Illume icons) every icon on the 'desktop' initially displays just the icon image, be it png, jpg, animated edj, whatever. When you touch the screen to scroll it, it will highlight the touched icon even if you don't actually select it. When it highlights it, it makes visible a 'background' png behind the icon, and two or three layers of transparent pngs on top of the icon, to give the 'glass button with an icon embedded in it' effect. Even when not visible (IE, on at least all but one icon at a time) those extra pngs are there, their positions are calculated AFAIK and their bitmaps are loaded. (again, AFAIK - those two are internals of Enlightenment and I'm guessing) But to make the user experience worse, whenever those extra pngs are made visible or invisible, it uses an animated fade-in/fade-out. So every time you drag to scroll, it's busy animating a fade-out on the previously highlighted icon, animating a fade-in on the one under your finger, and scrolling all the transparent and invisible PNGs. The effect is quite attractive, if only the FR had the horsepower to manage it while running a phone, GPS, and frameworkd. :( With the present state of my altered Illume theme (serenity.edj) I've trimmed the icons down to just the 'icon' image itself and a single png that appears behind it when highlighted. Outside the theme itself I've disabled dropshadows and changed rendering, and disabled the battery applet display (pending debugging - it sucks CPU apparently) and it reduced Enlightenment cpu usage dramatically. But I found significant further savings by tweaking icons. I've been using some Oxygen icons, and because I always have the launcher at 'extra large' (3 icons across in portrait) I started out with the 256x256 oxygen icons. Which are beautiful, but suck resources like crazy. I figured changing to 128x128 would help but the problem was almost exactly the same, while it looked worse on screen. So I did some investigating and testing and found that if I prescale my icons to the actual displayed size on the FR it responds great. In this case, that's 116x116 pixels. (that's the icon itself at 'extra large' - Illume displays the name below that, and the two of them are in IIRC a 162x142 tile) So by prescaling to 116x116 pixel icons in Gimp I get the same memory and CPU (hence UI responsiveness) as with much smaller (blotchier after scaling) images. 'Large' launcher icons appear to be 76x76, and 'medium' 36x36. This is theme-dependent though, as well as preferences dependent - within the edje theme they can be scaled or resized or whatever. So a prepackaged icon theme would probably need to be scaled to match a particular Illume theme's resulting sizes... Even then, if the user has utilized the useful 'scaling' settings it will probably differ. Still pondering the best answer here. So right now, with the rescaled icons, removing excess icon adornments, and disabling dropshadows and battery the SHR Illume is /almost/ as smooth scrolling as 2008.x. I also found that disabling animation of the slipshelf drop-down etc helped significantly in making it feel more responsive. Overall, it feels like a completely different phone when I switch back to the Illume theme. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GUI responsiveness (was Re: SHR first impression : it's slow ?)
On Wed, 25 Feb 2009 09:19:37 +0800 HouYu Li wrote: > Sounds amazing. But where can we get the serenity.edj theme??? > Nowhere right now, I'm working on it. I've gotten a few thing just how I want them, others as yet untouched, and occasional bugs. Started with illume.edj, imported some pieces of default.edj to make overriding them easier, and started in changing the look. Probably 2/3 done now, but I've kept restarting as I changed from 2008.x to FSO to SHR and so on and kept resyncing my changes to the newest Illume changes. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GUI responsiveness (was Re: SHR first impression : it's slow ?)
On Wed, 25 Feb 2009 12:43:59 +0100
Helge Hafting wrote:
> Joel Newkirk wrote:
> > With the default.edj theme (Illume doesn't override it for Fileman,
> > which includes Illume icons) every icon on the 'desktop' initially
> > displays just the icon image, be it png, jpg, animated edj,
> > whatever. When you touch the screen to scroll it, it will highlight
> > the touched icon even if you don't actually select it.
> Hm - it shoudln't highlight unless it actually is selected. :-/
> > When it highlights it, it
> > makes visible a 'background' png behind the icon, and two or three
> > layers of transparent pngs on top of the icon, to give the 'glass
> > button with an icon embedded in it' effect. Even when not visible
> > (IE, on at least all but one icon at a time) those extra pngs are
> > there, their positions are calculated AFAIK and their bitmaps are
> > loaded.
>
> I'm not sure if I understand that. Only one icon looks like a glass
> button - sure. Now, I understand that illume may have precomputed the
> glass button look for every icon there is, spending some memory. But
> why should that need any cpu when scrolling? The glass button effect
> isn't applied to the other icons, so those glass images should just
> sit in memory somewhere untouched?
>
> > (again, AFAIK - those two are internals of Enlightenment and I'm
> > guessing)
As I said, I'm guessing, but when I removed the extra PNG images and
leav just one, enlightenment average CPU drops and the display is more
responsive. The glass button effect /is/ applied to every icon,
it's just that the parts ('parts' in edc syntax) relevant to the effect
are flagged as non-visible by default. I'm assuming that even when a
element in the GUI is flagged visible=0 that it still calculates its
position onscreen, so that if you have 20 icons in Illume then you're
trying to scroll at least 100 transparent png images. ('icon' image,
background, shine, highlight, shadow - there are actually 9 png images
specified, plus the icon image itself) Pasted at the end is the entire
group "e/fileman/desktop/icon/fixed" from default.edc. You can see that
each icon actually contains eight png images, and twelve programs,
including six that specify animated transitions.
it also doesn't help that Enlightenment is advanced enough to perform
some very nice scaling tricks - like specifying that 6 pixels inward
from each edge should not be stretched away from that edge, so for
example you can have a 13x13 PNG that functions as a 6-pixel border,
with the central pixel being scaled as large as needed, while the edges
are only stretched in one dimension, along the edge. Wonderful feature
to have, but I suspect that the calculations involved in this scaling
and other nice effects E offers are at least a slight detriment to the
(integer) FR user experience...
> > But to make the user experience worse, whenever those extra pngs are
> > made visible or invisible, it uses an animated fade-in/fade-out. So
> > every time you drag to scroll, it's busy animating a fade-out on the
> > previously highlighted icon, animating a fade-in on the one under
> > your finger, and scrolling all the transparent and invisible PNGs.
> > The effect is quite attractive, if only the FR had the horsepower
> > to manage it while running a phone, GPS, and frameworkd. :(
>
> I see no fade effect. When I click an icon, it gets the "glass"
> effect. It appear with a slight delay, but there is no visible "fade
> in". One minute there is just the icon, the next moment it is
> "glassed". So if much work goes into this - then it is all wasted.
If you watch an icon closely when you press your stylus against it, you
can usually perceive the fade-in taking place, particularly if your FR
is straining, in which case you can sometimes see a few distinct delayed
steps. The linear transition is set to occur in 0.2 seconds fading in,
and 0.1 seconds fading out - so it is quite brief. I believe it
abides by the "Framerate" setting in Illume config (the spanner), such
that a 30fps setting and a 0.2 second fade equal a 6-frame animation.
You can see it in the first two program sections below, "go_active" and
"go_passive". The thumb_gen series with the decelerated and sinusoidal
transitions aren't used on the Illume desktop it seems, but if you for
example open the wallpaper settings in Illume config you can observer
thumbnails that 'zoom' out from zero size when displayed, to become
clickable 'icons'.
> >
> > With the present state of my altered Illume theme (serenity.edj)
> > I've trimmed the icons down to just the
Re: GUI responsiveness (was Re: SHR first impression : it's slow ?)
On Thu, 26 Feb 2009 11:31:43 +0100
Helge Hafting wrote:
> Joel Newkirk wrote:
> > On Wed, 25 Feb 2009 12:43:59 +0100
> > Helge Hafting wrote:
> >
> >> Joel Newkirk wrote:
> >
> [...]
> > As I said, I'm guessing, but when I removed the extra PNG images and
> > leav just one, enlightenment average CPU drops and the display is
> > more responsive. The glass button effect /is/ applied to every icon,
> > it's just that the parts ('parts' in edc syntax) relevant to the
> > effect are flagged as non-visible by default. I'm assuming that
> > even when a
>
> Urrgh - such inefficient coding. Making a 'movie' per icon - every
> time
> - and just not showing it for most of them. :-( The sane way is to
> only do the _needed_ calculations. Either animate a single icon when
> the effect is actually used, or generate all the frames once and just
> store them till needed. I wonder how they come up with such stuff.
> This is not only a problem on weak processors - it wastes energy on
> the good ones as well. :-(
Well as I said, I don't know to what degree and in what fashions this
may be optimized within E - I'm assuming that it prescales and doesn't
scale on the fly, but that the scrolling needs to take all visible and
non-visible parts into account - I don't know this however, I'm
guessing.
> Maybe we ought to use a modified duke nukem as an app launcher
> interface instead of enlightenment. Duke has a _better_ framerate for
> scrolling and zooming - in 3D!
>
> Shoot at icons to start apps. Fire at the process list to kill. kill
> -9 using a bigger gun. ;-)
:) Thanks, and to BillK for the enlightening and entertaining link. :)
With my car broken down this morning, any amusement is welcome.
> [...]
> > Wonderful feature
> > to have, but I suspect that the calculations involved in this
> > scaling and other nice effects E offers are at least a slight
> > detriment to the (integer) FR user experience...
> >
> I wish people though more about efficiency. One can have all sorts of
> wonderful effects by precomouting some stuff _once_, and then use
> plain bit block transfers. 1990 game machines was weaker than the FR,
> but that did not seem to be a problem then.
My philosophy has always been colored by my time with the Amiga. I
'grew up' believing that clever hardware and clever software could
overcome limited horsepower, and that the immediacy of feedback
provided by a responsive GUI is often worth more than a doubled clock.
(at 7mhz it 'felt' as fast as some 3ghz beasts I've used, simply
because everything was tweaked for smooth consistent responsiveness)
> [...]
> > If you watch an icon closely when you press your stylus against it,
> > you can usually perceive the fade-in taking place, particularly if
> > your FR is straining, in which case you can sometimes see a few
> > distinct delayed steps. The linear transition is set to occur in
> > 0.2 seconds fading in, and 0.1 seconds fading out - so it is quite
> > brief. I believe it abides by the "Framerate" setting in Illume
> > config (the spanner), such that a 30fps setting and a 0.2 second
> > fade equal a 6-frame animation.
>
> I have now set the framerate to its minimum of 5, and turned off
> animations where I could. At least the keyboard appears more quickly
> now.
>
> Icon animation and only two icons - selected and
> unselected.
What has me concerned on this front, looking more long-term, is that OM
have exhibited a tremendous reluctance to make the Illume/Enlightenment
config accessible, so that changing framerate and disabling animations
of windows opening, Top Shelf opening/closing, etc are currently not
possible for a new owner 'out of the box'. Hopefully that'll be
addressed before the platform reaches more mainstream (read:
non-hacker) customers... And not by simply disabling such features,
though that might be preferable to enabled and unconfigurable.
> > Thanks for the encouragement. :) It's already improved my user
> > experience, but in my poking about I've broken things as well,
> > which is why I'm not offering the .edj to anyone yet. I plan to
> > start from a clean extraction from illume.edj and default.edj once
> > more, applying only the changes confirmed to be beneficial and not
> > cause E to segfault.
>
> Great!
> I hope this will go into the distributions, at least as a selectable
> alternative. Eye candy is nice - but only as long as it doesn't
> create performance problems.
>
> Scrolling is slow enough even if I grab the iconless corner - so that
> no icon actually change state. (None was
Re: GTA03 Touchscreen Capacitive (was Re: OT: iPhone howto)
On Thu, 26 Feb 2009 11:49:01 +0100 Matthias Apitz wrote: > El día Thursday, February 26, 2009 a las 11:38:45AM +0100, Steve > 'dillo Okay escribió: > > > > > On Feb 26, 2009, at 11:20 , Matthias Apitz wrote: > > > > [...] > > > > >> > > > > > > Default: unuseable for terminal/UNIX input > > > Numbers: unuseable for terminal/UNIX, not even for SMS > > >Terminal: useable for terminal/UNIX, but only with stylus > > > > Hmmmit seems as if the problem with the default is that it > > doesn't know UNIX keywords. > > Is it possible to populate the dictionary with UNIX keywords. Who > > decided that the predictive keyboard would be the default ? > > for UNIX you need chars and not predictive words, I think; > > > Or alternatively, how do you set the Terminal keyboard to be the > > default ? > > # cd /usr/lib/enlightenment/modules/illume/keyboards > # mv Default.kbd Default.kbd.orig > # cp Terminal.kbd Default.kbd > # killall -HUP enlightenment > > matthias I usually rename Default.kbd to Alpha.kbd so it's still accessible, but Terminal is default. The problem with either is that opkg upgrade is prone to replace default.kbd. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Opkg.org segfaulting packages
On Fri, 13 Feb 2009 11:59:04 +0200 "Risto H. Kurppa" wrote: > Hi! > > I'm using the opkg.org repository: > echo "src/gz opkgorg http://www.opkg.org/packages"; > > /etc/opkg/opkg-feed.conf > > Trying to install the following apps using the repository cause the > installation to segfault: > liquidwar wireshark-common wireshark mokomaze pythm > > Installing from direct URL:s (opkg install http://...) works on some > packages. There might also be some others, I'll report if I see any. > > Anyone know what's wrong and could the authors fix the packages? Sorry, this thread got overlooked until I just ran a message search for subject 'opkg'. :( The problem seems to be happening in the script that creates the Packages.gz file - several packages have nothing but filename and md5, which doesn't seem right, but the problem-causers (like the wireshark/tshark packages I contributed) seem to be problems parsing the control file within the ipk. Specifically, it doesn't seem capable of ignoring empty lines at the end of the file, but rather appends them to the last field. In the case of the wireshark packages that field is Source. (in addition, somehow there's a stray 'T' on a line by itself in the control file for wireshark - that was just my stupid error, though I wouldn't have anticipated it would break the feed) I'm fixing the wireshark packages to have no empty lines at the end of the control file. Odd. After the hourly script run, wireshark package is missing from the feed, though it still appears on the opkg.org site. Will see what happens. Looking further, it seems that after I upload the new ipk, it no longer lists the package link (its new opkg.org URL) under 'direct-download link' in the opk upload section, likely the reason it's left out of the feed update. I pasted the original data for tshark back in and updated the prepended "0_", will see what happens at midnight. (It'd be nice if opkg.org could look up a named dependency locally and fill in the link - I had to edit the dependancies manually to reflect filename changes when it prepended 0_ and 1_) Also odd: I'd swear there was a package rating on Wireshark, but after the correction it's back to 0. Regardless, in all of this I decided to rebuild with the latest stable Wireshark. (the present opks at opkg.org are 1.0.99+svnr26030, 1.1.2 stable is available now, I'll stay out of svn this time) Hopefully I'll be able to build, test, and package it by the end of the weekend. > I noticed some extra empty lines in > http://www.opkg.org/packages/Packages - does it matter? > > r Yeah, in some instances the empty lines are clearly causing a problem. ;) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Opkg.org segfaulting packages
On Sat, 28 Feb 2009 00:12:27 -0500 Joel Newkirk wrote: > On Fri, 13 Feb 2009 11:59:04 +0200 > "Risto H. Kurppa" wrote: > > > Hi! > > > > I'm using the opkg.org repository: > > echo "src/gz opkgorg http://www.opkg.org/packages"; > > > /etc/opkg/opkg-feed.conf > > > > Trying to install the following apps using the repository cause the > > installation to segfault: > > liquidwar wireshark-common wireshark mokomaze pythm > > > > Installing from direct URL:s (opkg install http://...) works on some > > packages. There might also be some others, I'll report if I see any. > > > > Anyone know what's wrong and could the authors fix the packages? > > Sorry, this thread got overlooked until I just ran a message search > for subject 'opkg'. :( Apologies, please disregard that message. I accidentally sent a half-composed message, after I'd folded it into a reply to a later thread on the same subject. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [2008.12] Opkg.org repository packages that won't install
On Fri, 27 Feb 2009 11:58:26 +0200
"Risto H. Kurppa" wrote:
> Some packages in opkg.org are somewhat broken and won't install using
> the opkg.org repository in 2008.12. Could you please try this and let
> me know what happens. After this we could try to let the authors know
> what there's something wrong in their packages. If you are author of
> some of these packages please check that your package has all the
> required information.
>
> #adding repository & updating
> echo "src/gz opkgorg http://www.opkg.org/packages";
> > /etc/opkg/opkg-feed.conf opkg update
>
> #trying to install packages that are in opkg.org repository but return
> 'An error ocurred, return value: 2'
> opkg install centerim
> opkg install libboost-signals
> opkg install mokoconv
> opkg install neon
> opkg install shortom
> opkg install gpssight
> opkg install openmiaocat
> opkg install voicenote
>
> #these will segfault:
> opkg install liquidwar
> opkg install wireshark-common
> opkg install wireshark
> opkg install pythm
> #pythm requires many dependencies so this might also give you an
> dependency error instead
>
> #Also fbreader often fails :(
> opkg install \
> http://mikeasoft.com/~mike/openmoko/enca_1.9-r3_armv4t.ipk \
> http://www.mikeasoft.com/~mike/openmoko/fbreader_0.8.2a-r7+elleopatches_om-gta02.ipk
>
> fbreader also available on opkg.org and enca in
> http://downloads.freesmartphone.org/fso-testing/feeds/armv4t/enca_1.9-r3_armv4t.ipk
> but you get md5sum mismatches :/
>
> My results: http://pastebin.com/m27e0d226
>
> I want to use the repository instead of 'direct' opkg install http://
> adresses to be able to make Kustomizer work even if some packages are
> updated to a newer version and the filename changes. See
> http://wiki.openmoko.org/wiki/Kustomizer
>
> Thanks!
>
>
> r
>
Sorry, I overlooked your earlier related thread until I ran a message
search for subject 'opkg'. :(
Reading through Packages.gz, several packages have nothing but filename
and md5, which doesn't seem right. It seems on closer examination that
those are all ones listed with the 'return value: 2' error. Likely the
error occurs because it in fact knows no package named, for example
"voicenote", only a package with a filename "voicenote_0.3_arm.ipk".
Those packages, manually downloaded, also fail to extract via "ar x"
with the error "invalid ar magic" - I smell a connection, where the
feed-building script cannot extract the archive to parse the control
file. I'm unsure what the cause of the 'ar x' failure is.
The segfaulters (like the wireshark/tshark packages I contributed) seem
to be emerging when the hourly repo update script parses the control
file within the ipk. Specifically, it doesn't seem capable of ignoring
empty lines at the end of the file, but rather appends them to the last
field parsed. In the case of the wireshark packages that field is
Source. (in addition, somehow there was a stray 'T' on a line by
itself in the control file for wireshark - that was just my stupid)
Liquidwar has "${SRC}\n" as the source, and pythm also has a blank line
embedded in the "source" field. So it seems the segfaults are packages
with spare newlines in the control file... Does the official syntax
for the control file permit or prohibit newline-at-end?
I'm fixing the wireshark packages to have no empty lines at the end of
the control file. ;)
Odd. After the hourly script run, wireshark package is missing from
the feed, though it still appears on the opkg.org site. Will see what
happens. Looking further, it seems that after I upload the new ipk, it
no longer lists the package link (its new opkg.org URL) under
'direct-download link' in the opk upload section, likely the reason
it's left out of the feed update. I pasted the original data for
tshark back in and updated the prepended "0_", will see what happens at
midnight. (It'd be nice if opkg.org could look up a named dependency
locally and fill in the link - I had to edit the dependancies manually
to reflect filename changes when it prepended 0_ and 1_)
UPDATE: Nope, it turns out it placed it near the top of the Packages.gz
file instead of alphabetically, and I missed it at first.
Also odd: I'd swear there was a package rating on Wireshark, but after
the correction it's showing 0.
Regardless, in all of this I decided to rebuild with the latest stable
Wireshark. (the present opks at opkg.org are 1.0.99+svnr26030, 1.1.2
stable is available now, I'll stay out of svn this time) Hopefully
I'll be able to build, test, and package it by the end of the weekend.
> I noticed some extra empty lines in
> http://www.opkg.org/packages/Packages - does it matter?
>
> r
Yeah, in some instances the empty lines are clearly causing a
problem. ;)
j
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: howto: making ipk/opk
On Fri, 27 Feb 2009 23:32:49 +0200 Aapo Rantalainen wrote: > A) There are many broken packages on opkg.org. They can be downloaded > and installed and they work. But adding opkg.org repository and > installing packages from repository cause Segmentation Faults and Not > found errors (opkg error -2). I'm getting these with the FSO-M5 repo right now. > B) People has made cool little scripts and python-things, but they > can't release them because packaging is so mysterious. > > > I want share to you one simple way to make package. > Read full text: > http://cc.oulu.fi/~rantalai/freerunner/packaging/ > > briefly: > Make directory and file-structure under it. > Put things in CONTROL/control > Use ipkg-buildpackage-script (v 1.1 2001/07/26) by Oliver Kurth > > > -Aapo Rantalainen > > P.S. If you know some other easy way to make ipk/opk please share it > with us. What's wrong with good old 'ipkg-build'? Script included in your path if using the toolchain, under /usr/local/openmoko/arm/bin. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
[FSO ms5] opkg segfaults on frameworkd
trying to opkg upgrade I get a segfault on frameworkd. Retrying with "-V3" the output ends with: opkg state set to Installing Package: frameworkd;0.8.4.9+gitrc9e5d80540d505fe64d2acae7bbd0645625446f2;armv4t; Running script /tmp/opkg-jnPXxs/frameworkd-mYo6Qv/preinst opkg: fork failed Segmentation fault r...@iota:~# There IS no preinst script in the frameworkd package... Why is it trying to run one? (opkg is 0.1.6+svnr197-r1) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: unable to get the boot menu
On Mon, 2 Mar 2009 20:56:59 +0100 Bram Mertens wrote: > So now I have what I believe to be everything I need on the SD but I'm > unable to boot it! > > I've tried several times to follow the instructions on > http://wiki.openmoko.org/wiki/Booting_the_Neo_FreeRunner as well as > those on http://wiki.openmoko.org/wiki/Bootloader but I cannot get the > device to display the boot menu. > > From the above I understand that the following should suffice: > 1) shut down the device and disconnect it from the computer > 2) remove the battery for at least 30 secs > 3) press and hold the AUX button > 4) press and hold the power button until the boot menu appears > > The first thing I see is "OpenMoko", then the initialization messages > scroll by and FDOM is started. (The AUX button flashes once shortly > before "Openmoko" is displayed on the screen.) Just to clarify and be sure: 3) Press and hold down AUX button 4) while AUX is still held down, press and hold power button until the boot menu appears. Then release both buttons. That should present you with the NOR boot menu. The NAND boot menu (the one that can be readily overwritten) is accessed by: 3) Press and hold down Power button 4) while holding down power button, and /before/ booting begins (IE, within 5-6 seconds should be fine) press and hold AUX, then release both buttons after boot menu appears. I've never found the powerdown process to be so sensitive... The issue with external power (step #1 refers to AC power as well as USB-port from a host computer) is that when you plug in, it often starts booting without needing a power press. In this situation, however, if you're holding down the AUX button when you plug it in you'll end up with the boot menu. Other than removing the battery to effect a hard-reset, I've never performed step #2 above. Generally you'd want the NAND boot menu to actually boot, since it's presumably at least as new as the NOR, but potentially much newer, possibly drastically different (Qi vs uBoot). Use the NOR boot menu when flashing, however. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Om2009 release plan
On Mon, 02 Mar 2009 09:14:56 -0700 Angus Ainslie wrote: > The following features already have an owner and will be taken care of > before the end of March: > phone calls incoming and outgoing > sms incoming and outgoing > simple phone book (no images) > call log > charging > suspend and resume > alarm clock > resume speed < 2 seconds > boot time < 2 minutes > screen lock > battery indicator (gta01 and gta02 battery) > gsm indicator > > Then there are still features looking for someone to help bring them > in before the end of March: > Settings application > gprs & edge > user changeable ring tones > bluetooth > wifi > led indication for missed calls or sms > sliding in UI Two questions. What about package management GUI - is one planned to be included? Can you clarify "sliding in UI"? j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Fwd: unable to get the boot menu
On Tue, 3 Mar 2009 00:31:54 +0100 Bram Mertens wrote: > I've tried several times again. > > After removing the battery for several minutes the device > automatically boots when the battery is replaced. By the time I've > turned the device around after replacing the back cover the "openmoko" > logo is already displayed. (The device was not connected to any > computer.) > > So I've shut the device down again by pressing the power button for > several seconds. > > Without removing the battery but waiting some 30 seconds: > 1) press and hold AUX > 2) press and hold power > after about 4 seconds the AUX button lights up, about 2 seconds later > the "openmoko" logo is displayed and after that the device boots > I released both buttons as soon as the "openmoko" logo was displayed. > > Again power down the device > 1) press and hold power > 2) press (and hold) AUX approx 2 seconds after pressing power before > the openmoko logo is displayed > > again the device boots without displaying any kind of boot menu > > Is it possible to reconfigure the wait time of the boot menu via the > device? I can't find anything in the settings menu or any terminal > application that would allow me to execute commands. > > Regards > > Bram Maybe someone else will have a brighter idea, but all I can suggest at this point is unplug, remove battery, then hold down AUX while inserting battery. If that doesn't work, but it just boots right up instead, then try to confirm that the AUX button is functional, as I'm starting to suspect a bum button. If that's the case I've no idea what solution is feasible. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Om2009 release plan
On Mon, 02 Mar 2009 11:40:21 -0700 Angus Ainslie wrote: > On Mon, 2009-03-02 at 18:31 +0100, Florian Hackenberger wrote: > > Thanks for keeping us informed of the release schedule! One > > question related to paroli comes to mind. I tested paroli (from > > SVN) with FSO 5.1 and noticed that paroli is meant to be running > > full screen. Is that a temporary requirement? Because it would > > basically rule out the use of a window manager like illume and > > would result in having to implement WM, launcher (.desktop), > > systray, ... functionality in paroli, which would obviously be > > quite a duplication of effort. Could someone please explain the > > paroli plans with respect to the fullscreen requirement? > > > > Cheers, > > Florian > > > > Full screen is how paroli is intended to be used. It will not be a WM, > launcher etc. It will still co-exist quite nicely with illume and that > is how I've been running it for the last few days. Providing a stable > phone platform is what paroli is about. > > Once the UI spec is published it should become a little clearer. > > Angus I've always envisioned a sideways swipe switching from 'desktop' to fullscreen 'phone'/'home' GUI... :) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Om2009 release plan
On Tue, 3 Mar 2009 10:08:31 +0800 HouYu Li wrote: > I think "package management GUI" can be included in "Settings > application". And for the wifi feature, I recommend the knjMokoWifi > which using PHP-GTK. It works very well and with comprehensive > interface. Package management is something I think of as a fairly independent thing, rather than part of "Settings". Granted its UI could be called up from within a settings panel or something, but its purpose, usage, and design seem distinctly apart from what I think most people consider "settings". Regardless, given the variety and nature of the things itemized I wondered if it had accidentally been omitted, and if so which list it was on. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: SHR testing and TangoGPS problem
On Thu, 5 Mar 2009 14:48:49 -0500 Adam Jimerson wrote: > I tried to upgrad TangoGPS one my SHR testing install because it > didn't seem like it was working, and fso-gpsd said that it was > running, but when I tried to do it even with -force-depens it removed > the version the SHR testing came with and gave me this error: > > r...@om-gta02 ~ $ opkg -force-depends install libsqlite3-0 > libpixman-1-0 libgl > ib-2.0-0 > http://pcl210-00.fit.vutbr.cz/openmoko/libexif_0.6.17_armv4t.ipklibdb > us-glib-1-2 libdbus-1-3 libcurl4 libcairo2 > http://pcl210-00.fit.vutbr.cz/ope nmoko/tangogps_0.9.6-r0_armv4t.ipk > Collected errors: > * Warning: Cannot satisfy the following dependencies for tangogps: > * gtk+-fastscaling (>= 2.10.14) * > > Do I need to install a older version of tangogps inorder to get it to > work? You should be good to go. What you got was a warning instead of an error regarding gtk+-fastscaling - exactly what you asked for with the "-force-depends", which makes dependency errors into warnings instead. That said, you may have believed it not installed if the icon didn't appear - if that's the case then edit /usr/share/applications/tangogps.desktop and try changing the "application" category to "applications". BTW - a quick search of wiki or list archives should turn up this answer, keywords like "tangogps" and "fastscaling"... ;) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Detecting ethernet gadget connections
On Fri, 06 Mar 2009 11:04:40 +0100 Helge Hafting wrote: > Daniel Benoy wrote: > > I'm working on a script that will detect which network interfaces > > are connected and mess with the routing accordingly, but I'm having > > trouble detecting whether my USB ethernet gadget connection is up > > or down. > > > > > > ethtool when it's up: > > lisa:~# ethtool usb0 > > Settings for usb0: > > Link detected: yes > > > > ethtool when it's down: > > lisa:~# ethtool usb0 > > Settings for usb0: > > Link detected: yes > > > > > > Unlike on the host side, the usb0 interface doesn't appear and > > disappear, allowing udev scripts to bring up/down the interface. > > > > Anyone know if there's a way to detect that a network connection > > has actually been established? > > > > Is this necessary? > Just set a higher metric on usb0, then routes thorough eth0 will be > preferred when available. The default route through usb0 will only be > used as a last resort when there is no other way. > > Such a setup works very well on my laptop. It doesn't on the > freerunner yet, because the "ip" utility in busybox currently can't > set metrics. But port that (or use debian with its binary ip utility) > and you'll be fine. > > Helge Hafting You should be able to find iproute2 for freerunner... I know I've got an ipk of 2.2.26 at http://newkirk.us/om - my "OM stuff" attic, but I think it's in the feeds as well. Apart from that, I dug into this exact situation before, including dealing with usb0, wrote an article about it at my blog. (http://jthinks.com/better-freerunner-networking) You can use the standard 'route' command to set metrics, so iproute2 isn't required. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Troubles with opkg.org repo
On Sun, 8 Mar 2009 15:30:36 +0200
"Risto H. Kurppa" wrote:
> On Sat, Mar 7, 2009 at 10:09 AM, Tobias Kündig
> wrote:
> > This is a little summary of what I do: http://www.opkg.org/repo.html
>
> Thanks Tobias!
>
> At #openmoko tried to figure out what's wrong. Someone pointed out
> that ftp uses ascii mode by default - I suppose you're uploading at
> least the Packages -ascii file. How about packages.gz. Man ftp says
> that ascii upload of binary files might corrupt it -> this might be
> the problem - using binary format shouldn't corrupt the .gz
>
> Another find was that 22 packages out of 93 were other than Debian
> Binary when checked with file *ipk. The list file types listed below:
>
> DEBIAN BINARY – CORRECT!!!
{snipped package names}
> DATA
> EMPTY
> GZIP
> So things seem to fail because:
> - Empty file is an empty file, not much to discuss about that. Only
> need to find out the reason for that: did the authors really upload an
> empty package or did the upload fail but an empty file was created?
Nothing to add here.
> - ar isn't able to extract gzip files so all files recognized as gzip
> will not get listed properly in the Packages file. Again the reason
> for the wrong file type should be found: did the packages package it
> wrong or did the type/mime/something get corrupted along the way,
> during the upload or on the server?
It is perfectly acceptable for an ipk to be a tarball (.tar.gz) instead
of an ar archive. 'opkg install' happily works with them. So the
'fix' needs to be in create_repo.
j
{!!untested!!}
#!/bin/sh
cd /root/opkg.org/packages/
echo -n >Packages
for file in `ls -1 *.ipk *.opk`; do
echo $file
sum=`md5sum $file | awk '{print $1}'`
pkgtype=`file $file | cut -d' ' -f2`
case "$pkgtype" in
Debian)
ar x $file control.tar.gz
tar xzf ./control.tar.gz ./control
cat control|grep -e "[:alnum:]" >> Packages
echo Filename: $file >> Packages
echo MD5Sum: $sum >> Packages
echo >> Packages
echo >> Packages
rm
-f /root/opkg.org/packages/control.tar.gz /root/opkg.org/packages/control
;;
gzip)
tar xzf $file ./control.tar.gz
tar xzf ./control.tar.gz ./control
cat control|grep -e "[:alnum:]" >> Packages
echo Filename: $file >> Packages
echo MD5Sum: $sum >> Packages
echo >> Packages
echo >> Packages
rm
-f /root/opkg.org/packages/control.tar.gz /root/opkg.org/packages/control
;;
*)
echo "Problem extracting $file"
;;
esac
done
gzip -c /root/opkg.org/packages/Packages
>/root/opkg.org/packages/Packages.gz
#Upload
ftp -v -n ftp.opkg.org < /root/opkg.org/ftp2.input
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: Troubles with opkg.org repo
On Sun, 8 Mar 2009 15:30:36 +0200
"Risto H. Kurppa" wrote:
> On Sat, Mar 7, 2009 at 10:09 AM, Tobias Kündig
> wrote:
> > This is a little summary of what I do: http://www.opkg.org/repo.html
>
> Thanks Tobias!
>
> At #openmoko tried to figure out what's wrong. Someone pointed out
> that ftp uses ascii mode by default - I suppose you're uploading at
> least the Packages -ascii file. How about packages.gz. Man ftp says
> that ascii upload of binary files might corrupt it -> this might be
> the problem - using binary format shouldn't corrupt the .gz
>
> Another find was that 22 packages out of 93 were other than Debian
> Binary when checked with file *ipk. The list file types listed below:
>
> DEBIAN BINARY – CORRECT!!!
{snipped package names}
> DATA
> EMPTY
> GZIP
> So things seem to fail because:
> - Empty file is an empty file, not much to discuss about that. Only
> need to find out the reason for that: did the authors really upload an
> empty package or did the upload fail but an empty file was created?
Nothing to add here.
> - ar isn't able to extract gzip files so all files recognized as gzip
> will not get listed properly in the Packages file. Again the reason
> for the wrong file type should be found: did the packages package it
> wrong or did the type/mime/something get corrupted along the way,
> during the upload or on the server?
It is perfectly acceptable for an ipk to be a tarball (.tar.gz) instead
of an ar archive. 'opkg install' happily works with them. So the
'fix' needs to be in create_repo.
j
{!!untested!!}
#!/bin/sh
cd /root/opkg.org/packages/
echo -n >Packages
for file in `ls -1 *.ipk *.opk`; do
echo $file
sum=`md5sum $file | awk '{print $1}'`
pkgtype=`file $file | cut -d' ' -f2`
case "$pkgtype" in
Debian)
ar x $file control.tar.gz
tar xzf ./control.tar.gz ./control
cat control|grep -e "[:alnum:]" >> Packages
echo Filename: $file >> Packages
echo MD5Sum: $sum >> Packages
echo >> Packages
echo >> Packages
rm
-f /root/opkg.org/packages/control.tar.gz /root/opkg.org/packages/control
;;
gzip)
tar xzf $file ./control.tar.gz
tar xzf ./control.tar.gz ./control
cat control|grep -e "[:alnum:]" >> Packages
echo Filename: $file >> Packages
echo MD5Sum: $sum >> Packages
echo >> Packages
echo >> Packages
rm
-f /root/opkg.org/packages/control.tar.gz /root/opkg.org/packages/control
;;
*)
echo "Problem extracting $file"
;;
esac
done
gzip -c /root/opkg.org/packages/Packages
>/root/opkg.org/packages/Packages.gz
#Upload
ftp -v -n ftp.opkg.org < /root/opkg.org/ftp2.input
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: New splash screen with lots of logos
On Tue, 10 Mar 2009 23:55:01 +0100 Pander wrote: > Hi all, > > A new custom splash screen with lots of logos can be found here: > http://wiki.openmoko.org/wiki/Configuring_the_boot_splash_screens#Alternative_instructions > > Regards, > > Pander Personally I try to avoid splash screens, as they often end up consuming resources I would prefer were dedicated to booting up a hair faster. ;) I use the attached /etc/init.d/banner, which works wonderfully with Qi - it just displays a full-screen-sized OM logo formed from orange ASCII characters. I have a variation of it that prints a random Lao Tzu quote at the bottom, but I haven't figured out how to make it random using /bin/sh, while using my preferred /bin/bash (of course not installed by default) has problems displaying the logo. When used with Qi, this logo may be the only thing displayed until X starts, and it will reappear briefly when suspending/resuming/rebooting, which I like. Just drop it in /etc/init.d (there might be a "please wait, booting" version of banner there already) and "ln -s /etc/init.d/banner /etc/rcS.d/S03banner" if no previous banner existed. j banner Description: Binary data ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Troubles with opkg.org repo
On Sun, 8 Mar 2009 15:05:10 +0100 Thomas Zimmermann wrote: > Additionally there are a lot of Packages that don't contain a control > file with depencies and so on: > > Form the packages in debian format are these: > > 0_tshark_1.0.99+svnr26030_armv4t.ipk > 0_wireshark-common_1.0.99+svnr26030_armv4t.ipk > 1_wireshark_1.0.99+svnr26030_armv4t.ipk > > So if the authors of these packages can update them that would be > great. I contributed those three - sorry, I hadn't noticed they were in your list. Strange, I downloaded the files from opkg.org and they are significantly smaller than the originals I uploaded. They contain the beginnings of the correct data, specifically debian-binary and the start of data.tar.gz, but are truncated. The tshark package, eg, is supposed to be 106606 bytes, but the download from opkg.org is only 65536 bytes. In fact, all three are truncated to powers of two as well - multiples of 4096 bytes... I'll try resubmitting them yet again, though not until tomorrow. (I'm off to bed) Clearly the files are getting broken at some point in the process, I'll try to rule out upload. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Troubles with opkg.org repo
On Thu, 12 Mar 2009 08:05:03 +0200 "Risto H. Kurppa" wrote: > On Thu, Mar 12, 2009 at 7:51 AM, Joel Newkirk > wrote: > > On Sun, 8 Mar 2009 15:05:10 +0100 > > Thomas Zimmermann wrote: > > > >> Additionally there are a lot of Packages that don't contain a > >> control file with depencies and so on: > >> > >> Form the packages in debian format are these: > >> > >> 0_tshark_1.0.99+svnr26030_armv4t.ipk > >> 0_wireshark-common_1.0.99+svnr26030_armv4t.ipk > >> 1_wireshark_1.0.99+svnr26030_armv4t.ipk > > > >> > >> So if the authors of these packages can update them that would be > >> great. > > > > I contributed those three - sorry, I hadn't noticed they were in > > your list. Strange, I downloaded the files from opkg.org and they > > are significantly smaller than the originals I uploaded. They > > contain the beginnings of the correct data, specifically > > debian-binary and the start of data.tar.gz, but are truncated. The > > tshark package, eg, is supposed to be 106606 bytes, but the > > download from opkg.org is only 65536 bytes. In fact, all three are > > truncated to powers of two as well > > - multiples of 4096 bytes... > > > > I'll try resubmitting them yet again, though not until tomorrow. > > (I'm off to bed) Clearly the files are getting broken at some > > point in the process, I'll try to rule out upload. > > Thanks for uploading them again. > I hope uploading again helps but I have the feeling that it's > something else, or that the upload fails for some reason and it has > really nothing to do with your actions.. Let's see if the scripts get > fixed at some point. > > > r > I uploaded them again, then re-downloaded to be sure - as of right now, at least, they are correct and intact. Uploaded the same files, from the same desktop machine, with the same browser. :( j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Illume - Command line to display the keyboard ?
On Thu, 12 Mar 2009 10:17:13 +0100 kimaidou wrote: > Hi! > > Hum, I tried again, and from SSH too, but the command > /home/root/Scripts/illume-kbd-show does not hide the keyboard if it > is here. Could you send by email the binary you are using ? (if > possible) NB: I am using this file : > http://3v1n0.tuxfamily.org/openmoko/illume-kbd-show > > Thanks > Kimaidou "--help" gives some hints... ;) Try "illume-kbd-show -t" (for 'toggle') j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [SHR-Testing] Slow response when answering a call
On Fri, 13 Mar 2009 11:59:50 +0900
"W.Kenworthy" wrote:
> Just once, I do it straight after boot - lasts through suspend etc.
>
> The original called for move and I wrote copy - sorry - not sure if
> copy works as well, or it has to be move ("mv" vs "cp")
>
> BillK
What works is anything that causes E/Illume to reload. Any change to
a .desktop file will do it.
What I've used for a while now when editing edje themes and icons is
"killall -HUP enlightenment". Save the lines below as Relite.desktop
and it will work. (the icon is extracted from default.edj, it's also at
http://newkirk.us/om/icon_applications_restart.png ) I think I'd find
it irritating, but I suspect this can be 'automated' by sticking the
killall command in /etc/X11/Xsession.d/99relite - maybe with a 20
second delay.
j
[Desktop Entry]
Name=Relite
Comment=Elightenment reload
Exec=killall -HUP enlightenment
Icon=icon_applications_restart.png
Terminal=false
Type=Application
Categories=Application;Utility;
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: openvpn?
What are you having trouble with? I was able to cross-compile http://openvpn.net/release/openvpn-2.1_rc9.tar.gz on the first attempt with ". /usr/local/openmoko/arm/setup-env;./configure --host=arm-angstrom-linux;make" after easing http://people.openmoko.org/~zecke/om2008.8-dev/armv4t/openssl-dev_0.9.7g-r7_armv4t.opk into /usr/local/openmoko/arm/arm-angstrom-linux-gnueabi/usr/lib. I've not tested it yet. j - Original Message - From: Nick Van Fossen <[EMAIL PROTECTED]> To: Subject: openvpn? Date: Tue, 19 Aug 2008 14:42:49 -0700 > I'm trying to compile openvpn for my FreeRunner, which is > running 2008.8, but so far I haven't been successful. Has > anyone else tried to do this yet? If so I'd appreciate > any info on how you were able to. > > -Nick > __ > ___ Talk to your Yahoo! Friends via Windows Live > Messenger. Find out how. > http://www.windowslive.com/explore/messenger?ocid=TXT_TAGLM_WL_messenger_yahoo_082008 > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB keyboards "silent" in terminal (FSO)
I've gotten a few keyboards to work successfully in 2008.08, and a few would not work. One that would not work was no surprise to me - a Dell keyboard with a builtin USB hub - the keyboard and volume knob appear as individual USB devices, requiring an internal hub in the keyboard. That particular keyboard, unlike most, actually tells me on the bottom label that it requires 1.5A, which is why I was unsurprised. I don't know what current the Freerunner is capable of supplying to the USB port, but I'd be surprised if it could pump out 1.5A... (and even if it could it seems unlikely to last more than an hour or so ;) With 2008.08 when I plug in any simple (IE single USB device, no extraneous lighted keys) keyboard it is functional, and the onscreen keyboard silently goes away as long as it's plugged in. I'm awaiting a powerable USB hub I ordered to test if that permits the more complex keyboard to work as well. j - Original Message - From: Benito Torres <[EMAIL PROTECTED]> To: community@lists.openmoko.org Subject: Re: USB keyboards "silent" in terminal (FSO) Date: Tue, 19 Aug 2008 21:44:05 +0200 > On Tue, Aug 19, 2008 at 20:57 (+0200), Benito wrote: > > I dare to jump in as I have the same problem. > > Just found this: http://docs.openmoko.org/trac/ticket/1796 > (see Message-ID: <[EMAIL PROTECTED]> on this list). > > And this: > http://downloads.openmoko.org/framework/milestone2/updates/armv4t/xserver-kdrive-glamo_1.3.0.0+git4147f80c0bfedaff7749bccbd1e1d566dee0e04c-r7_armv4t.ipk > On my FSO-system there's > 1:1.3.0.0+gita51364e2f23d4b6331c5ed613ce3f7e15f8e540f-r6 > installed, which according to the git-commit-dates is > older than the afore mentioned ipk. For some reason the > ipk is not included in the fso-arm4vt-feed. > > After installation of the newer package and a reboot I had > to find that this sadly doesn't help with my > keyboard-problem. I'll try a real usb-keyboard (without > that usb<->ps/2-converter) before complaining further, but > as both mice (usb and ps/2) now work (funny thing without > a cursor :) I suppose the additional converter is not the > problem. > > So for now there's more light shed on the issue but the > story continues... > > Cheers, > /Ben > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Tester please?
I've packaged up the dnscache and daemontools programs by Dan J Bernstein (djbdns) and would like someone to test the ipk for me. I've got the full thing installed on my Freerunner now, including unnecessary files and additional files used for testing and unused services, so installing the package on my Freerunner would really only be a valid test if I reflashed first. daemontools is simply a daemon/service management system, and dnscache is the caching-only DNS server binary from the djbdns package. (which includes full DNS server, email RBL blackhole server, client and testing tools, etc) The effect of having them installed and running on the Freerunner is that any DNS lookups (like the tangoGPS lookup for every single map tile downloaded!!) are cached locally, so that repeat lookups are nearly instant - a big difference when using GPRS in particular. It starts several tasks, which normally consume about 1% of RAM and 0% of CPU, and about 300k on disk. I've been running it on 2008.08 (regularly 'opkg update'd) for a week now with no ill effects, only benefits. (well, apart from having to putz with /etc/resolv.conf frequently, which I've overcome personally but not yet in the ipk) I'm hoping someone can try to install it and let me know if all works well. There are a few things as yet incomplete: if uninstalled, the package removes the files, folders, and created user, but doesn't presently remove the startup line from /etc/inittab. Also, it's currently up to the user to cause their /etc/resolv.conf to always read 'nameserver 127.0.0.1' - more properly the installation should short-circuit ifup and so on so that they do not overwrite resolv.conf. I'm working out the best approach for that right now. (again it works on my Freerunner, but I want it simpler and more 'proper') It would be easier if all Freerunner networking utilized resolvconf... So the package is NOT ready for public consumption, though the programs it installs work as expected on my system. It is presently configured largely as it would be on a server in a network center, with modified paths, including logging all queries. (though logging is severely limited, I plan to eliminate it entirely for final release ipk) Any takers can download the ipk at http://newkirk.us/om/dnscache_2.1.5_armv4t.ipk and let me know of any problems encountered. NOTE: the installer does modify /etc/inittab, by adding the line "SV:123456:respawn:/usr/local/djbdns/package/admin/daemontools/command/svscanboot" at the end, and it does NOT yet remove that line when uninstalled. Once installed, make sure you have a working internet connection from the Freerunner, then check "cat /etc/resolv.conf" and invoke "echo 'nameserver 127.0.0.1' >/etc/resolv.conf", then try hitting a FQDN, like "ping google.com" and see if it resolves the IP from the FQDN. If you remove it you will need to manually remove the svcscanboot startup/respawn line from /etc/inittab, then reboot, then uninstall. (polish on the install/uninstall scripts will eliminate the need for that when released, but for now uninstall just uninstalls and removes created folders, links, and user 'dnscache') Separately, I had a query: Since this is technically two programs, from two separate packages of sourcecode, should it be distributed as two ipks? dnscache depends on daemontools, and packaging separately would be pretty simple, but I don't presently see the usefulness. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Substitute batteries
My guess is that it incorporates some constants that depend on the actual battery pack. Note that I don't actually KNOW that it's in the battery, but previous ML posts by folks at Openmoko refer to it as 'in the battery', and with a bl-5c installed almost all battery-related entries in sysfs are unpopulated. j On Thu, 21 Aug 2008 00:13:00 -0400, "Charles Pax" <[EMAIL PROTECTED]> wrote: > On Wed, Aug 20, 2008 at 11:29 PM, <[EMAIL PROTECTED]> wrote: > >> OK, I have a few Nokia batteries laying around, BL-5c and BL-6c, and >> desktop chargers. They power the Freerunner fine, and seem to charge > when >> they're in the FR plugged in. But the Freerunner doesn't know what to > do >> with them apart from that, presumably since they lack the Openmoko > coulomb >> counter circuit. > > > The coulomb counter circuitry is in the battery? Why isn't this on the > board > itself? > > -Charles ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Substitute batteries
On Wed, 20 Aug 2008 22:49:05 -0700, Michael Shiloh <[EMAIL PROTECTED]> wrote: > > > Joel Newkirk wrote: >> My guess is that it incorporates some constants that depend on the > actual >> battery pack. Note that I don't actually KNOW that it's in the battery, >> but previous ML posts by folks at Openmoko refer to it as 'in the > battery', >> and with a bl-5c installed almost all battery-related entries in sysfs > are >> unpopulated. > > This is to confirm that the coulomb counter is in the battery pack. Why? > Because it stores the count, and if you swap batteries, you need to know > the count for the battery in the device, not for the one you removed. Makes perfect sense. Is it a custom IC, something off-the-shelf, or what? Could it be cheaply retrofitted by a hobbyist into a BL-5c clone battery? I'm quite happy to crack one open, just not my Openmoko battery... ;) Barring the workability of that idea, I'm hoping we can adapt the battery management/monitoring software to fallback to the 'dumb cellphone' approach and let us safely charge, and have some idea of remaining capacity of, non-Openmoko batteries if it detects a battery present without the coulomb counter circuit. j >> >> j >> >> On Thu, 21 Aug 2008 00:13:00 -0400, "Charles Pax" > <[EMAIL PROTECTED]> >> wrote: >>> On Wed, Aug 20, 2008 at 11:29 PM, <[EMAIL PROTECTED]> wrote: >>> >>>> OK, I have a few Nokia batteries laying around, BL-5c and BL-6c, and >>>> desktop chargers. They power the Freerunner fine, and seem to charge >>> when >>>> they're in the FR plugged in. But the Freerunner doesn't know what to >>> do >>>> with them apart from that, presumably since they lack the Openmoko >>> coulomb >>>> counter circuit. >>> >>> The coulomb counter circuitry is in the battery? Why isn't this on the >>> board >>> itself? >>> >>> -Charles >> >> >> ___ >> Openmoko community mailing list >> community@lists.openmoko.org >> http://lists.openmoko.org/mailman/listinfo/community > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Tester please?
It's a pain, and I suspect massaging it into an external build framework would be more trouble than it's worth. DJB has some frankly weird build and install procedures for some of his software. I've built Qmail server and djbdns&co many times in the past for servers, so was able to work my way through it, having at various times on various platforms coerced it into building. The packages and original build instructions are at http://cr.yp.to/djbdns.html, but they're expecting to be in the destination server environment, and the build doesn't accommodate cross-compiling. I ended up working up a script that builds the full djbdns, tcpserver, and daemontools packages. Most of daemontools (well, obviously not the source and scripts) and the merest fraction of djbdns (just the dnscache binary, actually, plus the svcscan control structure created during install) are all that are needed, tcpserver is only needed for a true DNS server, since localhost recursive lookups will always utilize UDP. (TCP is only used for DNS when the response is too large for a single UDP packet, which is pretty rare outside of zone transfers between DNS servers) The script I wrote to build them is at http://newkirk.us/om/mokodnscache.sh - commented somewhat, it presumes that http://cr.yp.to/djbdns/djbdns-1.05.tar.gz and http://cr.yp.to/daemontools/daemontools-0.76.tar.gz are in the same directory as the script. (http://cr.yp.to/ucspi-tcp/ucspi-tcp-0.88.tar.gz as well, if you uncomment that section to build tcpserver as well) The result of the build will be a crapload of binaries you don't need, a handful you need, and install scripts. The install scripts /will/ work on the Freerunner, but will install everything of djbdns with symbolic links everywhere, while daemontools' installer creates links all over the place and leaves the binaries inside the build directory. The ipk I posted contains just dnscache and the necessary support, and as much as possible places everything under /usr/local, whereas the 'standard' install scripts require three root folders (/service, /commands, /admin) to contain everything, with more symbolic links than lines of code. (it creates links to links to the binaries!) http://newkirk.us/om/mokodnscache.sh is not elegant, isn't proper, but produces working binaries, which was my sole goal. (nevertheless, it's fairly thoroughly commented to explain what's going on when it short-circuits one of the build-time tests or alters a config file) The second half of the job I set out on is still only half finished, as I need to test the ipk and work on eliminating a few more unneeded hunks. (I think that by disabling logging of queries, I can cut the size in half by eliminating all the special DJB logging and timestamp support binaries - he's very careful in his code, required his own logging and timestamp support because he found syslog lacking, and wanted IIRC microsecond-accurate timestamps) When all is said and done, I suspect the better solution will be a standalone single-binary dns cache, but I looked at dnsmasq and decided to stick with what I was familiar with to start, and it turned out to be less resource-hungry than I'd feared so I may just stick with it if I can pare it down and polish it a bit further. j On Thu, 21 Aug 2008 16:25:42 +0930, Rod Whitby <[EMAIL PROTECTED]> wrote: > Timo Juhani Lindfors wrote: >> Joel Newkirk <[EMAIL PROTECTED]> writes: >>> http://newkirk.us/om/dnscache_2.1.5_armv4t.ipk and let me know of any >> >> Hmm, how do I build this from source? > > Indeed - I'd be happy to test bitbake recipes for these ... > > -- Rod > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Tester please?
On Fri, 22 Aug 2008 09:02:06 +0200, Joachim Steiger <[EMAIL PROTECTED]> wrote: > Joel Newkirk wrote: > [...] >> When all is said and done, I suspect the better solution will be a >> standalone single-binary dns cache, but I looked at dnsmasq and decided > to >> stick with what I was familiar with to start, and it turned out to be > less >> resource-hungry than I'd feared so I may just stick with it if I can > pare >> it down and polish it a bit further. > > nice work, i like djbdns and dnscache and use them myself a lot (on > servers, routers etc) > > i hope somebody can add a nice recipe to oe/om-oe including some fhs > compliance patches etc.. > > first i wanted to ask how you plan to solve the 'djbs licensing > madness'-issue, but i see he placed all that stuff under public domain > last december.. nice surprise ;) > with djbdns you may not notice this for years, i still have > installations with 4 year old binaries of it running Yes, that was a concern of mine at first as well, but I decided to pursue it at least for my own use. And then when I went to retrieve the sourcecode tarballs I discovered his license change affecting both djbdns and daemontools. Fortuitous, sort of a belated Christmas present. (for those unaware, DJBernstein's software was always available under a 'non-license' that permitted source-code redistribution only, but on 28Dec07 he placed djbdns and daemontools, and other packages on other days, into the public domain, so that distribution of a modified binary is now permissible) j > > kind regards > > > -- > > Joachim Steiger > Openmoko Central Services > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
USB host successes
I've been playing around with USB host mode for a couple weeks now, and
wanted to itemize my results. (previously touched on some of this in one
of the USB keyboard threads and on my blog) Any questions just ask, but I
mostly envision this as a thread where we can post results of various USB
hostmode tests. (to hopefully make their way onto the wiki after some
consolidation, cross-checking, and editing)
The needed cable arrived just before I flashed to 2008.08 - 2007 didn't
acknowledge keyboard, and ethernet wasn't in my hands then, other results
here apply to both 2008.08 (currently Raster's image upgraded via Zecke-dev
feeds) and 2007.02 versions.
2gb thumbdrive - automounts, reads & writes fine. ('Plain' Sony
thumbdrive, no encryption or compression, VFAT)
keyboards - most work flawlessly when plugged in (3 out of 4 tested),
including automatically disabling software keyboard. (Thanks Raster!!)
The one tested that did not work is a Dell multimedia keyboard with 2-port
USB hub and volume knob as a second USB device - it also states that it
draws 1.5A, so I was unsurprised when it didn't power up.
Canon S410 Elph digital camera - (self-powered) lsusb shows the device, no
drivers installed. (I'd like to be able to transfer pix from the camera to
the Freerunner)
Epson printer - (self powered) device is visible in lsusb, no driver
support installed/configured on my Freerunner yet...
Logitech webcams (Quickcam Communicate STX and Quickcam Chat [IIRC]) -
device visible in lsusb, but gspca and usb-video kernel modules are
unavailable ATM.
Netgear wifi adapter - powers device, lsusb shows it, no support enabled in
kernel AFAIK. Not really necessary on GTA02, though there are some
scenarios to utilize dual wifi, especially if at least one supports Wifi
Master mode. (iwconfig refuses on built-in wifi so presumably driver
doesn't support, at least - I'd love to get madwifi going, but suspect the
SDIO interface leaves us unsupported there since they won't even support
USB wifi devices with the same chipset as a supported PCI wifi card)
Ethernet 10/100 adapter - arrived today, works perfectly. Plugged into
ethernet then USB, and Freerunner powers device, establishes ethernet link
on eth1, and picks up IPs and routing. Perfect, beautiful, and painfree.
Pinging Google less than 7 seconds after plugging in.
That last one, ethernet, just made my Freerunner an indispensable network
diagnostic tool. :) (now I just need tshark/wireshark, ettercap, handy
things like that) I'm likely going to order a second, see if there's any
way I can position the Freerunner mid-cable with the two ethernets bridged,
and sniff traffic. ;) (obviously would be limited by USB 1.1 bandwidth but
probably still useful - the question is really whether the Freerunner can
power a hub and two adapters)
j
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: USB host successes
David Samblas wrote: > can you post model and where did you buy the ethernet adapter? > thanks a lot > El lun, 25-08-2008 a las 20:01 -0400, Joel Newkirk escribió: >> I've been playing around with USB host mode for a couple weeks now, and >> wanted to itemize my results. (previously touched on some of this in one >> of the USB keyboard threads and on my blog) Any questions just ask, but I >> mostly envision this as a thread where we can post results of various USB >> hostmode tests. (to hopefully make their way onto the wiki after some >> consolidation, cross-checking, and editing) The ethernet adapter is http://www.amazon.com/gp/redirect.html?ie=UTF8&location=http%3A%2F%2Fwww.amazon.com%2Fgp%2Fproduct%2FB000TUMBMQ&tag=joelnewkirk&linkCode=ur2&camp=1789&creative=9325 but while it works fine so far and is quite inexpensive, I'm not thrilled with the construction quality. (although I expect to be making alterations anyway, the housing wouldn't stay together, and once it's removed the ethernet jack is poorly anchored - clear housing plus clear tape makes almost invisible fix until I get at it with the dremel tool ;) j ps - sorry if I mess up anybody's mail threading, but mailman only wants to deliver from support list today, not community, so I'm reading via nabble and replying via email :( ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB host successes
Unless I'm very mistaken, those are both the same USB->Ethernet adapter. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: WLAN troubles
Derick Rethans wrote: > I get a DHCP lease, with nameservers and all, the route is correct too. > However, I can't do any sort of connection - nor ping. I have a similar problem when I fire up wifi, and it's caused by having two default routes. Try "ip r" and see if it lists two default routes - one out usb0 and one out eth0. If it does, try "ip r d default via 192.168.0.200" to delete the route pointing out usb0, and see if that resolves it for you. (Linux supports multiple default routes, but there should be a 'metric' specifying priority, or advanced custom routing tables to support multiple simultaneous uplinks - like load-balancing traffic on two DSL lines from one network) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Duped Messages (was Re: Please split this list!)
Rui Miguel Silva Seabra wrote: > Since we're talking about the mailing lists, I still receive (randomly), > three repeated mails, two repeated mails, etc... > > This mail from Vasco... I received it three times already! :) > > Rui > > On Mon, Sep 01, 2008 at 12:24:03PM +0100, [EMAIL PROTECTED] wrote: > If you examine the email headers you can see that the 'received' header documenting where the mail server handling the list (sita.openmoko.org running exim 4.63) received the message from the sender's mailserver (IE, their ISP's mailserver) differs from one copy to the next - this means that either the sending server is failing to recognize that the message has been delivered and resends, possibly the receiving server is failing to send the acknowledgement of receipt at the end of the SMTP transaction (at least in a timely fashion), so the sending server automatically retries. Received: from relay1.ptmail.sapo.pt ([212.55.154.21] helo=sapo.pt) by sita.openmoko.org with smtp (Exim 4.63) (envelope-from <[EMAIL PROTECTED]>) id 1Ka7yh-0002Ow-Ha for community@lists.openmoko.org; Mon, 01 Sep 2008 13:55:37 +0200 Received: from relay1.ptmail.sapo.pt ([212.55.154.21] helo=sapo.pt) by sita.openmoko.org with smtp (Exim 4.63) (envelope-from <[EMAIL PROTECTED]>) id 1Ka7fK-0005dL-Pf for community@lists.openmoko.org; Mon, 01 Sep 2008 13:55:37 +0200 Received: from relay1.ptmail.sapo.pt ([212.55.154.21] helo=sapo.pt) by sita.openmoko.org with smtp (Exim 4.63) (envelope-from <[EMAIL PROTECTED]>) id 1Ka7ZO-0003bU-8Q for community@lists.openmoko.org; Mon, 01 Sep 2008 13:55:21 +0200 Notice that the message ID differs between the three copies. This tells us that this is the point (when sapo.pt mailserver delivers to sita.openmoko.org mailserver) where the failure occurs. If it were the mailinglist server sending dupes, this header would be identical among all copies. Also, it doesn't depend on sending mailserver software, I've noted it happening with qmail as sender, exim, gmail.com, and others. (even mail.openmoko.org sometimes, such as Andy Green's reply to the '3G modem' thread) Based on past experience as admin of a cluster of mailservers that sometimes exceeded 1 million incoming SMTP connections per day, I suspect that either spam filtering or some testing for ML (IE, 'is this sender permitted to post?') takes place AFTER receiving the message but BEFORE telling the sending server that message was received, and is periodically taking longer than the sending server's SMTP timeout, so the sender gives up on the connection and tries again - meanwhile the exim server handling the ML eventually accepts the message. My suspicion is that the load on the (virtual?) server hosting the ML is getting to a level where processing messages sometimes takes longer than some sending servers are willing to wait. Properly, in such a situation, the sending server is supposed to resend, and the receiving server is supposed to discard the message it failed to fully receive before the connection was broken. Unfortunately my mailserver experience is with surgemail, qmail, sendmail, and some exchange (ick), but I've never worked with Exim, so I can't suggest anything specific to check in the server config. (When I've seen this caused by receiving mailserver it was most often qmail, and was caused by improperly configured/limited spawning that exceeded available RAM instead of deferring excess inbound connections - once dipping into swap, all bets are off regarding timely responses) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: 2008.8 update: Wrench doesn't appear even with illume-configinstalled
[EMAIL PROTECTED] wrote: >> On Mon, 1 Sep 2008 14:08:13 +0200 "Yorick Moko" <[EMAIL PROTECTED]> >> babbled: >> >> >>> On Mon, Sep 1, 2008 at 2:03 PM, <[EMAIL PROTECTED]> wrote: >>> Since the first update of 2008.8, I cannot get the wrench for >>> configuring >>> illume. I have installed illume-config, as the wiki said, but the >>> wrench >>> does not appear. I have searched and all articles say that installing illume-config is all that I have to do. I installed 080830 update, and got the same problem. Am I missing >>> anything? >>> ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community >>> https://docs.openmoko.org/trac/ticket/1863 >>> >> rm -rf ~/.e/e/config/illume (wipe your old one - if u had one) >> >> > > I did that, but wrench didn't appear even after rebooting :-( > > No offense intended, but are you looking for it in the right place? ;) By default (unless you've used illume-config itself to change it) it appears in the top-left corner only once the tray has been opened, is hidden normally. (this threw me at first, when I'd upgraded from 2007 and enabled it per the wiki) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Duped Messages (OffTopic but important to lists.openmoko.org operations)
Sarton O'Brien wrote:
> On Monday 01 September 2008 23:23:21 Joel Newkirk wrote:
>
>> Rui Miguel Silva Seabra wrote:
>>
>>> Since we're talking about the mailing lists, I still receive (randomly),
>>> three repeated mails, two repeated mails, etc...
>>>
>>>
>> per day, I suspect that either spam filtering or some testing for ML
>> (IE, 'is this sender permitted to post?') takes place AFTER receiving
>> the message but BEFORE telling the sending server that message was
>> received, and is periodically taking longer than the sending server's
>> SMTP timeout, so the sender gives up on the connection and tries again -
>> meanwhile the exim server handling the ML eventually accepts the
>> message. My suspicion is that the load on the (virtual?) server hosting
>> the ML is getting to a level where processing messages sometimes takes
>> longer than some sending servers are willing to wait. Properly, in such
>>
>
> Sounds like a decent theory. I've experienced abnormalities when pre-
> processing facilities (greylisting/virus scanning/spam filtering) are
> configured incorrectly or terminate abruptly or incorrectly.
>
> I imagine just looking at the logs around the time of duplication would give
> a
> fair indication as to where the issue is. Otherwise system load sounds
> reasonable, swap in anything time critical is not ideal.
>
> Sarton
>
Tested, insofar as I am able to test it being on the 'outside'. I
telnetted to sita.openmoko.org port 25 and manually sent a few
help-request emails to [EMAIL PROTECTED]
After that final period (a period by itself on a line ends an SMTP
message transfer) it sits there thinking for 12-15 seconds in silence,
then responds - slightly more delay that I'm used to in such tests, but
not unusual, (particularly if filtering is performed at that stage,
which would allow unacceptable emails to be rejected instead of having
to bounce - better for spam control that way) and one test it took only
7.5 seconds.
Then there was the test (same procedure, same entries, second test)
where after the period-enter I waited 4 minutes with no '250 OK'
response. Lacking that response code, the sending mailserver eventually
presumes the communication to have failed, and queues the message for
redelivery attempt. I finally killed my telnet session, and sure enough
the message "The results of your email commands" came back to me from
Mailman.
***HOWEVER: RFC-2821 specifies the timeout period while awaiting '250
OK' response is 10 minutes, so strictly speaking this is a
misconfiguration or bug in the SENDING mailserver, NOT Exim 4.63 on
sita.openmoko.org.***
But I wish you luck if you try to convince Google that their mailservers
are broken... This RFC was drafted in 2001, expanding (ESMTP) on the
1982 RCF-821 (SMTP), at which time waiting 10 minutes for a
500-character data transfer to be acknowledged may have seemed
reasonable. But with today's common end-user bandwidth, computer
speeds, and user expectations, that's just not sensible. (particularly
the last - imagine users waiting up to 10 minutes for Outlook Express to
finish sending a one-line email... if their outbound SMTP server did
this there'd be a lot of phone calls)
So the gist of it is that the mailserver sita.openmoko.org is confirmed
as the source of the dupes - though not strictly its fault - and whether
through resource scarcity on itself or a problem elsewhere (eg: DNS
server it uses for RBL lookups could be barfing) I can't determine.
Misconfiguration is a possibility, if it leads to resource
overcommitment. Strictly speaking the problem is at the sending
servers, but realistically it's unlikely to be fixed on most of them, so
if I were postmaster I'd be checking logs and looking for a local fix.
j
{successful dialog pasted here, failed was identical except for lack of
'250 OK' response - my entries prepended here with '>'}
[EMAIL PROTECTED]:/usr/share/misc$ telnet sita.openmoko.org 25
Trying 88.198.124.203...
Connected to sita.openmoko.org.
Escape character is '^]'.
220 sita.openmoko.org ESMTP Exim 4.63 Tue, 02 Sep 2008 03:20:42 +0200
>HELO newkirk.us
250 sita.openmoko.org Hello rrcs-70-62-125-137.midsouth.biz.rr.com
[70.62.125.137]
>MAIL FROM: [EMAIL PROTECTED]
250 OK
>RCPT TO: [EMAIL PROTECTED]
250 Accepted
>DATA
354 Enter message, ending with "." on a line by itself
>SUBJECT: Help
>
>
>help
>testing
>123
>
>
>.
250 OK id=1KaKZp-00035l-IE
>quit
221 sita.openmoko.org closing connection
Connection closed by foreign host.
[EMAIL PROTECTED]:/usr/share/misc$
{excerpt from RFC 2821 http://www.faq.org/rfcs/rfc2
geomagnetic sensor?
Wondering if anyone has explored this yet. The wiki has the page http://wiki.openmoko.org/wiki/Wishlist_-_Hardware:_Digital_compass on the idea. The AKM device is apparently no longer produced (link is dead, search of their page produces no results), and the Honeywell is significantly larger (and 2-axis instead of 3-axis) than the Yamaha. Personally, I'd be thrilled to be able to tuck a YAS529 inside my Freerunner... ~$10 each samples, I2C serial interface, about 4mm^3, it's just so appealing... With that plus the accelerometers and the GPS we'd have position and orientation nailed now pretty well. (well, at least within the vicinity of the earth's surface ;) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: A little zenity-gui for flashing images to the neo
On Sun, 7 Sep 2008 14:50:20 -0400, "Charles Pax" <[EMAIL PROTECTED]>
wrote:
> On Sun, Sep 7, 2008 at 1:19 PM, Dale Maggee
> <[EMAIL PROTECTED]>wrote:
>
>> I was more bored and wanted to exercise my scripting skillz, so I made
>> some changes to this script.
>>
>> I saved your code into a file called "flasher" and make it executable. I
> get the following error.
>
> ./flasher: line 67: syntax error near unexpected token `('
> ./flasher: line 67: `the USB-Device and booted into the Nor (for kernel
> and
> rootfs) or Nand'
>
> -Charles
Sounds like you did a cut'n'paste from email. The problem is that long
lines are wrapped, which breaks programs/scripts. There's at least three
places in the script that suffer from word-wrap, including several lines
that start at line 66.
Something like a bash script really MUST be sent as an attachment, to
ensure this breakage doesn't happen. Meanwhile you can go through the
script and removed the extra newlines, so that echo statements and comments
(the things I noted glancing at the script) are not wrapped.
j
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: A little zenity-gui for flashing images to the neo
On Sun, 7 Sep 2008 14:50:20 -0400, "Charles Pax" <[EMAIL PROTECTED]>
wrote:
> On Sun, Sep 7, 2008 at 1:19 PM, Dale Maggee
> <[EMAIL PROTECTED]>wrote:
>
>> I was more bored and wanted to exercise my scripting skillz, so I made
>> some changes to this script.
>>
>> I saved your code into a file called "flasher" and make it executable. I
> get the following error.
>
> ./flasher: line 67: syntax error near unexpected token `('
> ./flasher: line 67: `the USB-Device and booted into the Nor (for kernel
> and
> rootfs) or Nand'
>
> -Charles
Woops, please replace 'echo' with 'zenity' in my previous post - I was
thinking 'output' and just type 'echo' without thinking further.
j
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: USB host successes (good to have)
On Tue, 09 Sep 2008 19:36:55 +0200, Michele Renda <[EMAIL PROTECTED]> wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Hello > > I want to share something that I found in a shop: > > http://www.conectica.ro/display/1-63-19-Adaptoare/Adaptor_USB_A___USB_A___.html > > With it I could use the shipped cable (Usb to Micro Usb) to attach a > pendrive to my FR and to get my FR running. > > It cost around 1.50 euro, and it seem to be cheap to me. The good think > is that it permit to use the provided cable and to don't risk to brak > the micro usb part inside the FR. > > What do you think about it? > Should work just fine. I picked up a few of http://www.cross-mark.com/female-female-coupler-adapter-p-922.html (and http://www.cross-mark.com/pins-cable-players-p-734.html) and they work fine. Lets me couple the provided cable (and the replacements noted in parens) to the standard USB cable provided with most peripherals. j > Michele Renda > -BEGIN PGP SIGNATURE- > Version: GnuPG v1.4.6 (GNU/Linux) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org > > iD8DBQFIxrQ3SIAU/I6SkT0RAgYjAKCaB0IOXFcwVET99O8HKyq16edZoQCfQZAq > 5n0cnpOSMtdUf5tx8mZHbm8= > =ghWp > -END PGP SIGNATURE- > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Stylus Ink Replacement Cartridge?
On Sat, 6 Sep 2008 21:13:15 -0400, "Charles Pax" <[EMAIL PROTECTED]> wrote: > Does anyone know how to replace the ink cartridge in the stylus that comes > with the Freerunner? It seems like I can just pull it out, but I don't > want > to break it. Also, does anyone have a Staples of OfficeMax item number for > a > good replacement cartridge? Thanks. > > -Charles I picked up a refill at Staples for $1.99, a fine-point black Fisher Universal pressurized refill for 'multi-action pens', Staples item code 521860, works great. Fisher's model is #SU4F, UPC 7-47609-17141-9. It was kinda depressing how quickly the included ink runs out, I probably wrote a total of 1/2 page with it before needing this refill. They had another that should fit, medium point Staedtler 'multi-action' pen refills, staples item code 396404, but that one is a three-pack of black+blue+red for $3.29. The three-pack didn't seem like a great idea for the stylus (I'm sure it's intended for the 'select-a-color' style pens) - I can't imagine stopping to swap them out when I wanted red instead of black, or whatever, and I wouldn't want my stylus pen to always write red. (or blue for that matter - finepoint black is IT for me ;) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB host successes
On Tue, 9 Sep 2008 18:23:43 +0200, Maciej Delmanowski <[EMAIL PROTECTED]>
wrote:
> On Aug 25, Joel Newkirk wrote:
>> probably still useful - the question is really whether the Freerunner
> can
>> power a hub and two adapters)
>
> I've thought about different idea - can FreeRunner be powered from USB
> power
> adaper connected to the USB hub? Like this:
>
>
> Power plug
> |
> |
>USB Hub---Ethernet, thumb drive, other devices
> |
> |
> FreeRunner
>
> That would probably solved power requirements... But it's still 220V (or
> 110V)
> attached indirectly to the USB network, is it safe ('cause I have no
> idea...)?
> For example, can I connect another computer through USB or Ethernet card
> with
> that configuration?
>
> harnir
[NOTE: It's NOT - or better not be - 220V or 110V direct to the hub, it's
likely to be 5Vdc]
You should be able to do this with the Freerunner, but you have to set the
mode in sysfs correctly, AND you should have 15k resistors between the D+
and D- data lines and ground. (see
http://wiki.openmoko.org/wiki/GTA02_sysfs#USB_Host_.2F_Device) I'd ordered
a combo USB hub plus card reader that I wanted to try to utilize this way,
but when it arrived the hub was intermittent and the card reader was dead,
so I've not been able to try yet. (awaiting RMA) I've got a 5V 2A cord
from my lost Zaurus that I was going to power it all with, figured that'd
be adequate to charge the freerunner at 1000mA while powering the hub/card
reader and whatever I've plugged into it. Just needs a bit of rewiring
inside the hub.
As far as connecting a desktop PC and the Freerunner to the hub at the same
time, I'm not sure WHAT would happen if the Freerunner were in Host mode -
you'd get contention as to which was master of the bus, probably neither
would be able to work in that configuration. (I'm not even sure if the
design of a USB hub would allow it - I think it treats the 'peripheral'
connections differently than the 'host' connection, even apart from
expecting power to be present on the host port)
j
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: Get spare parts in NYC (batteries)
On Tue, 9 Sep 2008 12:49:50 -0400, Daniel Benoy <[EMAIL PROTECTED]> wrote: > Have you considered a spares pack? > http://us.direct.openmoko.com/products/spares-pack > > On Tuesday 09 September 2008 04:12:09 Stefan Fröbe wrote: >> Yes, I know as I also own quite a few BC-5s and remakes ;-) >> But I'm looking specifically for the original ones to have the coulomb >> counter and a working charge meter... >> Personally, I don't want to spend $50 when all I really want is one spare Freerunner battery. Don't want/need the headset or pouch, would use the extra batt if I had it but don't need it so would rather save the $$. (I've got two BL-5Cs and a BL-6C from my nokia 6600, so in a pinch I can use one of them - I make sure one is always charged and tucked in the armrest in my car - but as with Stefan I'd like to be able to see at least an estimate of remaining charge) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB host successes
On Tue, 09 Sep 2008 12:35:44 -0700, Michael Shiloh <[EMAIL PROTECTED]> wrote: > I've tried to answer many of these questions here: > http://wiki.openmoko.org/wiki/USB_host > > If you feel this information is unclear, lacking, or wrong, please > correct or please let me know. > > In particular, I'd love people to add any specific setups they have that > are working. > > Michael Thank you - exactly the information needed. The 47kohm resistor (between ID and GND) is needed to signal 1000mA charging permitted (that's what I was thinking of when I wrote about the 15k pair in my previous post), per both http://wiki.openmoko.org/wiki/USB_host and http://wiki.openmoko.org/wiki/Specialized_USB_cables, but unless I'm misreading they seem to contradict one another regarding the 15k pulldowns on D+ and D-. What is the correct info regarding those? j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Kernel 2.6.27?
Has anyone tried to move up to kernel 2.6.27? (for ASU/2008 or FSO) The latest I find at git.openmoko.org is 2.6.26. I'm interested because that's the first release with GSPCA webcam driver rolled in. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Fantastic Experience
On Fri, 12 Sep 2008 14:20:13 +0200, Matthias Apitz <[EMAIL PROTECTED]> wrote: > one question to the community: normally you can't switch easy between > normal USB mode, i.e. a host is talking to a USB device, and > host-to-host mode where two computers are talking to each other; the USB > FAQ even warns not to connecting two computers without any kind of USB > bridge, because this could damage the computers; see here: > http://www.usb.org/about/faq/ans5/ > > how is this electrically organised in the GTA02 that I could just use a > gender > changer to switch between host-to-host mode or host-to-device mode? > > thx > > matthias You cannot "just" use a gender changer. As you already perceive, that's just the physical connection being adapted. Take a few minutes and read through http://wiki.openmoko.org/wiki/USB_host and http://wiki.openmoko.org/wiki/Specialized_USB_cables - the gist of it is that you use a coupler or non-standard cable for the physical connection, and settings in the kernel (configurable on-the-fly, of course) tell the Freerunner whether it is USB Device or USB Host, and whether to expect 5V over the USB to charge & power itself, or to provide 5V over USB to power devices. There are also some clever things that can be done on this front that aren't exploited yet, like sensing a resistor's presence or absence and automatically switching things around. I wrote a shell script to let me choose (manually) among the various configurations more easily - when asked to go to 'power-providing' host mode, it checks for 5V presence on the USB port and refuses to send out 5V while it is so. I expect similar functionality will appear soon in the form of an applet. (context menu from the USB tray icon, perhaps?) As an aside: It'd be helpful if we could use a short standard term to differentiate: Host mode providing power vs Host mode permitting charging. Given the non-standard nature of Host mode communication on top of device-mode power, I figure there's no existing standard. "powered host mode" is descriptive but potentially confusing - what is 'powered': the Freerunner or the devices... Any suggestions, anyone, or someone know of a (short) existing term in use? j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
update-alternatives question
Quick question, hopefully someone can instantly see my error: I've built the full iproute2 toolset, and am working on an ipk/opk for ip and tc. The default om2008 at least, has /bin/ip linked to /bin/busybox. So that should be changed. For inspiration, I looked at the postinst script for grep, since it must perform the same changes. (point at full bin instead of busybox in postinst, restore in prerm) So I took the update-alternatives line from grep's postinst and simply replaced 'grep' with 'ip' everywhere. [EMAIL PROTECTED]:~# update-alternatives --install /usr/bin/ip ip ip.ip 100 update-alternatives: Error: cannot register alternative ip to /usr/bin/ip since it is already registered to /bin/ip Isn't this the point of update-alternatives - to allow multiple possible binaries, and 'register' the one with highest priority? What have I done wrong? And why is the component grep.grep (or in my case ip.ip)?? I checked and /usr/lib/opkg/alternatives/ip and /usr/lib/opkg/alternatives/grep are the same - busybox 50. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: update-alternatives question
That did it, thanks. j On Sat, 13 Sep 2008 21:09:11 -0700, Charles-Henri Gros <[EMAIL PROTECTED]> wrote: > Joel Newkirk wrote: >> Quick question, hopefully someone can instantly see my error: >> [EMAIL PROTECTED]:~# update-alternatives --install /usr/bin/ip ip ip.ip 100 >> update-alternatives: Error: cannot register alternative ip to >> /usr/bin/ip since it is already registered to /bin/ip >> >> Isn't this the point of update-alternatives - to allow multiple possible >> binaries, and 'register' the one with highest priority? What have I > done >> wrong? And why is the component grep.grep (or in my case > ip.ip)?? > > grep.grep is the binary that contains what the "grep" package installs. > You should probably install your "ip" as "ip.ip", or use /usr/bin/ip (if > that's where you install it) > > So the command would look like (I think): > update-alternatives --install /bin/ip ip /usr/bin/ip 100 > > -- > Charles-Henri > > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
/space/fic/openmoko-daily/...?
Odd. I'm building (cross-compile toolchain) wireshark/tshark, and encountered the error: /bin/sed: can't read /space/fic/openmoko-daily/openmoko/work/armv4t-angstrom-linux-gnueabi/glib-2.0-2.16.4-r0/glib-2.16.4/glib/libglib-2.0.la: No such file or directory Where is this reference coming from? I see it in config.log when I grep around for "/space/" - the following line to be specific: config.log:Configured with: /space/fic/openmoko-daily/openmoko/work/i686-armv4t-sdk-angstrom-linux-gnueabi/gcc-cross-sdk-4.1.2-r7/gcc-4.1.2/configure --build=i686-linux --host=i686-linux --target=arm-angstrom-linux-gnueabi --prefix=/usr/local/openmoko/arm --exec_prefix=/usr/local/openmoko/arm --bindir=/usr/local/openmoko/arm/bin --sbindir=/usr/local/openmoko/arm/bin --libexecdir=/usr/local/openmoko/arm/libexec --datadir=/usr/local/openmoko/arm/share --sysconfdir=/usr/local/openmoko/arm/etc --sharedstatedir=/usr/local/openmoko/arm/share/com --localstatedir=/usr/local/openmoko/arm/var --libdir=/usr/local/openmoko/arm/lib --includedir=/usr/local/openmoko/arm/include --oldincludedir=/usr/local/openmoko/arm/include --infodir=/usr/local/openmoko/arm/share/info --mandir=/usr/local/openmoko/arm/share/man --with-gnu-ld --enable-shared --enable-target-optspace --enable-languages=c,c++ --enable-threads=posix --enable-multilib --enable-c99 --enable-long-long --enable-symvers=gnu --enable-libstdcxx-pch --program-prefix=arm-angstrom-linux-gnueabi- --disable-libssp --disable-libmudflap --with-float=soft --with-local-prefix=/space/fic/openmoko-daily/openmoko/staging/arm-angstrom-linux-gnueabi/usr --with-gxx-include-dir=/space/fic/openmoko-daily/openmoko/staging/arm-angstrom-linux-gnueabi//usr/include/c++ --with-sysroot=/usr/local/openmoko/arm/arm-angstrom-linux-gnueabi --with-build-sysroot=/space/fic/openmoko-daily/openmoko/staging/arm-angstrom-linux-gnueabi --disable-libunwind-exceptions --with-mpfr=/space/fic/openmoko-daily/openmoko/staging/i686-linux/usr --enable-__cxa_atexit then again in libwsutil.la: wsutil/libwsutil.la:dependency_libs=' -L/usr/local/openmoko/arm/arm-angstrom-linux-gnueabi/lib -L/usr/local/lib /usr/local/openmoko/arm/arm-angstrom-linux-gnueabi/usr/lib/libgmodule-2.0.la -L/usr/local/openmoko/arm/arm-angstrom-linux-gnueabi/usr/lib /space/fic/openmoko-daily/openmoko/work/armv4t-angstrom-linux-gnueabi/glib-2.0-2.16.4-r0/glib-2.16.4/glib/libglib-2.0.la -ldl /usr/local/openmoko/arm/arm-angstrom-linux-gnueabi/usr/lib/libglib-2.0.la /usr/local/openmoko/arm/arm-angstrom-linux-gnueabi/usr/lib/libz.la -L/space/fic/openmoko-daily/openmoko/staging/arm-angstrom-linux-gnueabi/usr/lib' I've done ./configure --host=arm-angstrom-linux then hand-compiled rdps, then make. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
USB host/power script (was Re: Fantastic Experience)
On Sat, 13 Sep 2008 13:52:55 +0200, Christian Adams <[EMAIL PROTECTED]> wrote: > Am 13.09.2008 um 01:29 schrieb Joel Newkirk: >> automatically switching things around. I wrote a shell script to >> let me >> choose (manually) among the various configurations more easily - >> when asked >> to go to 'power-providing' host mode, it checks for 5V presence on >> the USB >> port and refuses to send out 5V while it is so. I expect similar >> functionality will appear soon in the form of an applet. (context >> menu >> from the USB tray icon, perhaps?) > > i was planning such an applet .. is your script somewhere available? > > ciao, morlac Very sorry, somehow I'd overlooked this post. (and somehow I found it) The script as it stands is attached. It was lost to an SD corruption (I was keeping /usr/local mapped to the card, and keeping mapfiles and temporary storage of things I was working on there) so I've been recreating it. j ggg Description: Binary data ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: (OM2008.8-update) (audio/GSM) alsamixer names
On Tue, 16 Sep 2008 22:01:51 +1000, Matt <[EMAIL PROTECTED]> wrote: > [EMAIL PROTECTED] wrote: >>> I'd like to suggest different names for the Alsa controls, because > quite >>> frankly I cannot relate to the ones I see in alsamixer. But first I'd >>> like to ask you guys if you feel the same need as me. Do you feel >>> comfortable with the present alsa channel names, or would you like them >>> to be clearer? >>> >> >> That would be a great deed. The present alsa channel names stink so much >> that only the bravest of the brave dare to get close to that subsystem. >> That goes a long way to explain why so far we still do not have the >> automagical configuration we all would like. Even on my desktop box, by >> the way, I had to rewire it to my stereo using the analog jack output >> because I lost the configuration for the coax digital. >> >> Minh >> >> >> > Renaming them might be useful, but understanding their effects would be > better. Agreed. Particularly in the context of writing a GUI interface that would simply let us set levels for speaker/mic for each use-case. :) That doesn't require changing anything in ALSA, just abstracting to more real-world-meaningful names. (EG 'speaker volume - ringer' or 'handset mic level - in-call') j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Freerunner Firewall
For anyone interested, I've posted two articles on my blog (I hate blogs)
regarding firewalling on the Freerunner -
http://jthinks.com/freerunner-simple-firewall and
http://jthinks.com/freerunner-advanced-firewall - that link to the required
package(s) and include the scripts and support to firewall your Freerunner,
as well as explaining some concepts and options including NAT & FORWARD to
use the Freerunner as a gateway for another computer.
Note that by default, only three ports apparently are open on OM2008 -
TCP22 (SSH), TCP111 (rpcbind/portmap) and possibly TCP6000 (remote X11 -
I've seen it open before but it's not right now, on OM2008.8-update on my
Freerunner), so firewalling of the default system doesn't gain much.
('netstat -ln'
to see what ports YOUR Freerunner is listening on)
The gist of the 'freerunner-simple-firewall' article is a script that lives
in /etc/init.d and on startup (or manual invocation) installs a set of
firewall rules using the netfilter firewalling support in the Linux 2.6.x
kernel. The 'simple' firewall requires only the iptables binary, which can
be obtained from either
http://www.angstrom-distribution.org/feeds/2008/ipk/glibc/armv4t/base/iptables_1.3.8-r4_armv4t.ipk
or http://newkirk.us/om/iptables_1.4.2-rc1_armv4t.ipk - the ipk on my site
excludes ip6tables (ipv6 firewall control - packaged separately by me but
bundled with iptables in Angstrom feed) but includes the iptables-save and
iptables-restore binaries (actually symlinks to the iptables-multipurpose
binary, whereas those two binaries are a separate package in the Angstrom
feed called 'iptables-utils') plus includes the script and ruleset
incorporated in the 'freerunner-advanced-firewall' article.
For anyone who wants to firewall their Freerunner but isn't interested in
reading the articles, you can either install my ipk, or install the
Angstrom ipk plus the script attached to this message, which should be
placed in /etc/init.d/iptables and then configured to run just after
networking is started with 'update-rc.d iptables defaults 42'. (note that
I do NOT currently have a feed set up, you can install with 'opkg install
http://newkirk.us/om/iptables_1.4.2-rc1_armv4t.ipk' or download and install
locally) Using my ipk results in the 'advanced firewall' while the
attached script plus the Angstrom ipk results in the 'simple firewall' -
rulesets are essentially the same, but management (IE, adding rules) is
easier with the 'advanced' solution.
Enjoy your pocket protector.
j
PS - the scripts and rulesets are released free of copyright - do with them
as you will. The two articles themselves are copyrighted, but I hereby
declare an explicit exception for any portions of them to be incorporated
into the Openmoko wiki if that is desirable, understanding that they will
then fall under the license (or non-license) applicable to the wiki
contents.
iptables
Description: Binary data
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: USB Networking vs. iptables
I notice that you list the DNS server as 212.6.108.140 (resolver0.ewetel.de), but have the DNAT rules pointing at 212.6.181.140 (an unnamed IP that seems to be owned by 'claranet')... Checking from the 'outside' (IE I'm not on your ISP's network, and I presume you are within the ewetel.de network) 212.6.108.140 is a DNS server which won't let me do recursive lookups, which is normal, but 212.6.181.140 seems unoccupied at this time, or 100% firewalled. If that doesn't resolve it, what's in your FORWARD and INPUT chains? Can you post the output of "iptables -vnL"? (the -'v' for verbose means the output will include counts of packets/bytes that matched each rule - useful for debugging sometimes when unexpected zeros appear) "iptables -vnL" shows all the filter chains, INPUT/OUTPUT/FORWARD. (plus any custom chains) INPUT would affect packets from the Freerunner to the FC6 box (IE, when resolv.conf points at 192.168.0.200) while FORWARD would affect packets when you have the outside DNS server in resolv.conf. j On Thu, 18 Sep 2008 17:22:29 +, Christian Weßel <[EMAIL PROTECTED]> wrote: > Hello mokos, > > I just have a DNS problem, I try to configure my FC6 following the guide > http://wiki.openmoko.org/wiki/USB_Networking#Proxying_with_iptables > because I have a simple static environment for my FR. > > FR.usb.ip = 192.168.0.202 > server.usb.ip = 192.168.0.200 > server.eth.ip = 192.168.1.10 > router.eth.ip = 192.168.1.254 > DNS.ip = 212.6.108.140 > > on server: > [EMAIL PROTECTED] ~]# cat /etc/resolv.conf > search home > nameserver 212.6.108.140 > nameserver 212.6.108.141 > > [EMAIL PROTECTED] ~]# iptables -L -t nat --line-numbers -n > Chain PREROUTING (policy ACCEPT) > num target prot opt source destination > 1DNAT tcp -- 192.168.0.202192.168.0.200 tcp > dpt:53 to:212.6.181.140 > 2DNAT udp -- 192.168.0.202192.168.0.200 udp > dpt:53 to:212.6.181.140 > > Chain POSTROUTING (policy ACCEPT) > num target prot opt source destination > 1MASQUERADE all -- 192.168.0.0/24 0.0.0.0/0 > > Chain OUTPUT (policy ACCEPT) > num target prot opt source destination > > on FR: > [EMAIL PROTECTED]:~# cat /etc/resolv.conf > nameserver 192.168.0.200 > > [EMAIL PROTECTED]:~# ping 74.125.19.147 -c 1 > PING 74.125.19.147 (74.125.19.147): 56 data bytes > 64 bytes from 74.125.19.147: seq=0 ttl=236 time=182.480 ms > > --- 74.125.19.147 ping statistics --- > 1 packets transmitted, 1 packets received, 0% packet loss > round-trip min/avg/max = 182.480/182.480/182.480 ms > > [EMAIL PROTECTED]:~# nslookup www.google.com > Server:192.168.0.200 > Address 1: 192.168.0.200 > > nslookup: can't resolve 'www.google.com' > > For me the masqueration seems to be fine, just something with DNAT is > wrong. > If I change the FR.resolv.conf to 'nameserver 212.6.181.140' it also not > working. > > But what's wrong? > > BTW: I got no SElinux security alerts, neither in secure nor in > messages. > ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Multimedia messaging
Has any progress been made on this front? (I'm running OM 2008.8-update ATM) Being able to receive messages with pix/audio attachements is very important to me - my asterisk server is set up to forward voicemails to my cell, and occasional still images (when motion is detected) from security webcams. The Freerunner is the first 'cellphone' I've owned that had sufficient resolution and screensize for the webcam stills to be really useful. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB Networking vs. iptables
Try "iptables -I RH-Firewall-1-INPUT -s 192.168.0.202 -j ACCEPT", or the same rule inserted at the top of INPUT and FORWARD chains. Your FORWARD chain simply jumps to RH-Firewall-1-INPUT, the same as the INPUT chain. RH-Firewall-1-INPUT blocks SSH from various specific IPs, then accepts only very limited specific connections, including ICMP,http,https,ssh,CUPS and ipsec but NOT including DNS... Lack of a rule accepting DNS in INPUT keeps you from doing DNS lookups at 192.168.0.201, lack of a rule accepting DNS in FORWARD keeps you from doing DNS lookups at any other host. If you want to keep it locked down tight on the Freerunner's traffic you can amend the rule above with '-p udp --dport 53', but other things (like email, FTP, VOIP, chat, and other things in the future) are probably desirable as well, and not permitted through. j > Chain INPUT (policy ACCEPT 0 packets, 0 bytes) > num pkts bytes target prot opt in out source > destination > 1 592K 375M RH-Firewall-1-INPUT all -- * * 0.0.0.0/0 > 0.0.0.0/0 > > Chain FORWARD (policy ACCEPT 0 packets, 0 bytes) > num pkts bytes target prot opt in out source > destination > 1 701 45828 RH-Firewall-1-INPUT all -- * * 0.0.0.0/0 > 0.0.0.0/0 > > Chain OUTPUT (policy ACCEPT 613K packets, 261M bytes) > num pkts bytes target prot opt in out source > destination > > Chain RH-Firewall-1-INPUT (2 references) > num pkts bytes target prot opt in out source > destination > 21246K 210M ACCEPT all -- lo * 0.0.0.0/0 > 0.0.0.0/0 > 22 898 78034 ACCEPT icmp -- * * 0.0.0.0/0 > 0.0.0.0/0 icmp type 255 > 23 0 0 ACCEPT esp -- * * 0.0.0.0/0 > 0.0.0.0/0 > 24 0 0 ACCEPT ah -- * * 0.0.0.0/0 > 0.0.0.0/0 > 25 72 20607 ACCEPT udp -- * * 0.0.0.0/0 > 224.0.0.251 udp dpt:5353 > 26 0 0 ACCEPT udp -- * * 0.0.0.0/0 > 0.0.0.0/0 udp dpt:631 > 27 0 0 ACCEPT tcp -- * * 0.0.0.0/0 > 0.0.0.0/0 tcp dpt:631 > 28330K 164M ACCEPT all -- * * 0.0.0.0/0 > 0.0.0.0/0 state RELATED,ESTABLISHED > 29 180 10764 ACCEPT tcp -- * * 0.0.0.0/0 > 0.0.0.0/0 state NEW tcp dpt:22 > 30 0 0 ACCEPT tcp -- * * 0.0.0.0/0 > 0.0.0.0/0 state NEW tcp dpt:443 > 314155 244K ACCEPT tcp -- * * 0.0.0.0/0 > 0.0.0.0/0 state NEW tcp dpt:80 > 329849 611K REJECT all -- * * 0.0.0.0/0 > 0.0.0.0/0 reject-with icmp-host-prohibited > > Due to the masquerade I checked, if it would helpful to change the > FR.resolv.conf to the same DNS (212.6.108.140), but I got just the known > result: > [EMAIL PROTECTED]:~# nslookup www.google.com > Server:212.6.108.140 > Address 1: 212.6.108.140 > > nslookup: can't resolve 'www.google.com' > > If I ping from FR to this IP I got a good result: > > [EMAIL PROTECTED]:~# ping 212.6.108.140 > PING 212.6.108.140 (212.6.108.140): 56 data bytes > 64 bytes from 212.6.108.140: seq=0 ttl=248 time=21.264 ms > 64 bytes from 212.6.108.140: seq=1 ttl=248 time=22.464 ms > 64 bytes from 212.6.108.140: seq=2 ttl=248 time=23.561 ms > > --- 212.6.108.140 ping statistics --- > 3 packets transmitted, 3 packets received, 0% packet loss > round-trip min/avg/max = 21.264/22.429/23.561 ms > > BTW, my router has no DNS function, it is just a router. ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: USB Networking vs. iptables
On Fri, 19 Sep 2008 16:21:13 +, Christian Weßel <[EMAIL PROTECTED]> wrote: > iptables -I RH-Firewall-1-INPUT -s 192.168.0.202 -j ACCEPT > > That's it. Now I am able to install Debian by following wiki guide > <http://wiki.debian.org/DebianOnFreeRunner> > > Thanx a lot. > > Am Freitag, den 19.09.2008, 07:35 -0400 schrieb Joel Newkirk: >> Try "iptables -I RH-Firewall-1-INPUT -s 192.168.0.202 -j ACCEPT", or the >> same rule inserted at the top of INPUT and FORWARD chains. You're most welcome. The one problem with your reasoning regarding the default policy of ACCEPT is that the last rule in the RH-Firewall-1-INPUT chain is a 'drop all' rule... Every RedHat/Fedora/CentOS box I've ever set up nearly the first thing I do is delete the default firewall and construct my own - I don't like the way they structure theirs. IMHO best practice (and clearest logic) is to enable a DROP policy on INPUT and FORWARD chains, and add explicit ACCEPT rules for desired traffic. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
accelerometer jitter
I wrote a short shell script to repeatedly take the raw output from one of the accelerometers (/sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/dump) and massage the data to output X,Y,and Z-axis readings in decimal, -128 to 128. With my Freerunner sitting flat, face up, on the concrete slab of my house, the 'Z' reading (positive Z is toward the back of the handset, BTW) fluctuates between 51 and 56. (X and Y range from -1 to 2) That's a variance of +/- 5%, and I'm fairly confident gravity isn't fluctuating significantly... ;) Is that to be expected, is my Freerunner defective, or am I doing something wrong? I'm trying to develop a simple tool that will determine 'down' based on output from the accelerometer, and indicate deviations from various 90-degree orientations - will I need to massage the data further to ignore variations, should I be averaging several readings, or what? j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: accelerometer jitter
On Sat, 20 Sep 2008 17:36:00 -0400, "Michael Fisher" <[EMAIL PROTECTED]> wrote: > On Sat, Sep 20, 2008 at 4:07 PM, Joel Newkirk <[EMAIL PROTECTED]> > wrote: > >> I wrote a short shell script to repeatedly take the raw output from one > of >> the accelerometers > (/sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/dump) >> and massage the data to output X,Y,and Z-axis readings in decimal, -128 > to >> 128. > As a bit OT but since you have gone to the trouble of developing the > script, > could you run a test while driving/riding in the car and post what the > readings might be? I was thinking of the possibility of having the FR > determine if it was in a moving car or not. > > Mike > > -- > Michael Fisher > [EMAIL PROTECTED] No problem, I'll alter the script to intersperse timestamps and take it for a ride tomorrow. It's dumping samples about once per second ATM - I'd initially included a 'sleep 1' command then commented it out since the loop execution ended up taking about 1 sec anyway. (chains of cat/grep/sed/cut piped) The one caveat is that I don't have a good mount for it in the car - for the requested test I'll just lay it on the floor on the passenger side and note the physical orientation. (more OT than this request, I found out that the rubberized surface on the back of the Freerunner is essentially non-stick if you try to lay the Freerunner on a Handstands StickyPad - I've had one of those on my dashboard for years, always kept my cellphone and Zaurus there even on a sloped dashboard, but the Freerunner won't stick to it even on the flat dash in my new car) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: accelerometer jitter
On Sat, 20 Sep 2008 23:08:42 +0100, Al Johnson <[EMAIL PROTECTED]> wrote: > On Saturday 20 September 2008, Joel Newkirk wrote: >> With my Freerunner sitting flat, face up, on the concrete slab of my > house, >> the 'Z' reading (positive Z is toward the back of the handset, BTW) >> fluctuates between 51 and 56. (X and Y range from -1 to 2) That's a >> variance of +/- 5%, and I'm fairly confident gravity isn't fluctuating >> significantly... ;) >> >> Is that to be expected, is my Freerunner defective, or am I doing > something >> wrong? > > What does the datasheet say about noise levels? You're looking at +/-5% of > reading but +/-1% of full scale which isn't implausible for a cheap > accelerometer. I can't find the relevant answer on the datasheet. It has some bar charts of "percent of parts" vs "sensitivity [mg/digits]" at 2.5v on page 37, but I can't understand what exactly they're plotting. :( >> I'm trying to develop a simple tool that will determine 'down' based on >> output from the accelerometer, and indicate deviations from various >> 90-degree orientations - will I need to massage the data further to > ignore >> variations, should I be averaging several readings, or what? > > Underspecified ;-) How accurate does it need to be? Is the phone expected > to > be stationary or moving? Is power consumption important? Sorry. ;) Near-stationary, power consumption immaterial, as accurate as possible. As a project to get my feet wet doing ground-up development (probably Python, which I need to learn) for the Freerunner (instead of just cross-compiling and fixing issues therein) I wanted to write a software 'bubble-level' tool. Laid flat on a surface you see a bubble, off-centered as a 'real' bubble-level would be to indicate pitch. More generally, held with the side or back against a vertical or horizontal surface to display an indication of how close to level (or vertical) the surface is. Lots of incremental additions and changes possible, like adding audio indication when level is achieved, interacting with DBus or what-have-you instead of reading sysfs, etc, making it an educational process on the specific environment of the Freerunner. So for the initial basic functionality averaging several samples over the course of a couple seconds works fine, but later refinements would benefit from more accuracy. (also actual implementation could attend to all three axes, not just look at one, which might help or hinder in this application, not sure) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: accelerometer jitter
Thanks - I was just reading through http://www.freescale.com/files/sensors/doc/app_note/AN3107.pdf about tilt measurements, which is linked from the wiki. (I'd missed it before) The one you linked is mostly about position calculations, double integration of the acceleration data, although it discusses noise and some management thereof. (which is much more important with the double-integral) In this usage, any decrease on one axis should be accompanied by a matching increase on a second axis, since I'm dealing specifically with vertical or horizontal orientations. The third axis only becomes significant in the "lying flat, face up" scenario, where decreases in 'z' will have matching increases in 'x+y' - noisier summing them, but probably workable. (since the common use-cases wouldn't demand as frequent updates) Which is actually the one that got me thinking about this in the first place - I was leveling my clothes washer and while looking for one of my levels my eyes drifted toward the Freerunner, and my brain said "hm - if I just laid it on top..." Which doesn't even consider the second sensor - I'm just looking at the one with all three axes aligned with the housing, not the one rotated 45 degrees about the Z-axis. Utilizing both would let me be more certain of small orientation changes. (I'm assuming noise isn't induced from outside, and so they wouldn't be in sync) This is turning out more calculationally involved than I'd first anticipated, but still looks workable. j On Sat, 20 Sep 2008 20:05:19 -0700, Steve Mosher <[EMAIL PROTECTED]> wrote: > Have a look at this might help. > > Note, we dont use a freescale accel, so some things may be different but > you'll get the basic idea. > > http://www.freescale.com/files/sensors/doc/app_note/AN3397.pdf?fpsp=1&WT_TYPE=Application%20Notes&WT_VENDOR=FREESCALE&WT_FILE_FORMAT=pdf&WT_ASSET=Documentation > > > Joel Newkirk wrote: >> On Sat, 20 Sep 2008 17:36:00 -0400, "Michael Fisher" > <[EMAIL PROTECTED]> >> wrote: >>> On Sat, Sep 20, 2008 at 4:07 PM, Joel Newkirk <[EMAIL PROTECTED]> >>> wrote: >>> >>>> I wrote a short shell script to repeatedly take the raw output from > one >>> of >>>> the accelerometers >>> (/sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/dump) >>>> and massage the data to output X,Y,and Z-axis readings in decimal, > -128 >>> to >>>> 128. >> >>> As a bit OT but since you have gone to the trouble of developing the >>> script, >>> could you run a test while driving/riding in the car and post what the >>> readings might be? I was thinking of the possibility of having the FR >>> determine if it was in a moving car or not. >>> >>> Mike >>> >>> -- >>> Michael Fisher >>> [EMAIL PROTECTED] >> >> No problem, I'll alter the script to intersperse timestamps and take it > for >> a ride tomorrow. It's dumping samples about once per second ATM - I'd >> initially included a 'sleep 1' command then commented it out since the > loop >> execution ended up taking about 1 sec anyway. (chains of > cat/grep/sed/cut >> piped) >> >> The one caveat is that I don't have a good mount for it in the car - for >> the requested test I'll just lay it on the floor on the passenger side > and >> note the physical orientation. (more OT than this request, I found out >> that the rubberized surface on the back of the Freerunner is essentially >> non-stick if you try to lay the Freerunner on a Handstands StickyPad - > I've >> had one of those on my dashboard for years, always kept my cellphone and >> Zaurus there even on a sloped dashboard, but the Freerunner won't stick > to >> it even on the flat dash in my new car) >> >> j >> >> >> ___ >> Openmoko community mailing list >> community@lists.openmoko.org >> http://lists.openmoko.org/mailman/listinfo/community > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Dead Accelerometers? (was Re: accelerometer jitte r)
On Sat, 20 Sep 2008 19:44:34 -0400, Joel Newkirk <[EMAIL PROTECTED]> wrote: > On Sat, 20 Sep 2008 17:36:00 -0400, "Michael Fisher" <[EMAIL PROTECTED]> > wrote: >> On Sat, Sep 20, 2008 at 4:07 PM, Joel Newkirk <[EMAIL PROTECTED]> >> wrote: >> could you run a test while driving/riding in the car and post what the >> readings might be? I was thinking of the possibility of having the FR >> determine if it was in a moving car or not. > No problem, I'll alter the script to intersperse timestamps and take it Ummm, turns out it IS a problem: I can't get any data from either accelerometer now. I now see: [EMAIL PROTECTED]:~# cat /sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/dump 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 3b 93 af b8 11 12 03 03 84 16 ee 0b 40 00 00 00 00 07 00 e4 00 00 00 00 00 00 fa 00 03 00 0b 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 regardless of any orientation changes or movement of the Freerunner - completely unchanging output. The third line used to look something like this (faceup): 47 00 e4 00 00 00 00 00 00 01 00 01 00 35 00 00 The first '01' is X, the second '01' is Y, the '35' is Z. Those three values are all that changed. I've tried rebooting, I've tried powerdown and restart, both without and with battery removal, I've tried explicitly activating the sensor (hadn't needed to before) at /sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/power/wakeup. Same with spi0.0 sensor - unchanging near-zero outputs on all three axes. (Values wrap from 00 to FF with FF representing '-1' - IE the face-down reading would have something around C5 for the Z reading) I'm about to reflash and reconstruct my environment, (I'd added celtune repository to get linphone running, forgot to remove it and opkg upgrade pulled a bunch of packages from there - after a second opkg upgrade enlightenment is broken and won't start on reboot) will update afterwards, and hopefully be able to post the requested data tomorrow... j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: DeadAccelerometers?
well, I flashed as follows, both rootfs and kernel from each: testing-om-gta02-20080921, 'base/empty' - seems a bit TOO empty, as I was unable to establish any USBnet communications with it. So I flashed to 2008.9 'stable' - accelerometers dead. Tried 2008.8-update - accelerometers dead. Tried 2008.08.08 - same deadness. Getting worried now - this one surely would have worked if it were a software matter... Tried 2008.4 - turns out it didn't offer sysfs access to accelerometers. Back to 2008.9 - still no joy. each reboot, it seems, the output from the accelerometers is slightly different, although each time each axis registers within the range of -5 to +12, and remains unchanged until the Freerunner is powered down. Then I checked again under 2008.9 (fresh flash) at ten minutes uptime, after composing this message, and suddenly both are reading dynamic data again!! I'm happy, but now I'm thoroughly confused, and somewhat worried as to what could have caused this in the first place... (as well as irritated that I went through 6 complete reflashes before it reached a working state again, something like 6 hours wasted and nothing learned) j On Sun, 21 Sep 2008 18:27:57 -0400, Joel Newkirk <[EMAIL PROTECTED]> wrote: > On Sat, 20 Sep 2008 19:44:34 -0400, Joel Newkirk <[EMAIL PROTECTED]> > wrote: >> On Sat, 20 Sep 2008 17:36:00 -0400, "Michael Fisher" > <[EMAIL PROTECTED]> >> wrote: >>> On Sat, Sep 20, 2008 at 4:07 PM, Joel Newkirk <[EMAIL PROTECTED]> >>> wrote: > >>> could you run a test while driving/riding in the car and post what the >>> readings might be? I was thinking of the possibility of having the FR >>> determine if it was in a moving car or not. > >> No problem, I'll alter the script to intersperse timestamps and take it > > Ummm, turns out it IS a problem: I can't get any data from either > accelerometer now. > > I now see: > [EMAIL PROTECTED]:~# cat /sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/dump > 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 3b > 93 af b8 11 12 03 03 84 16 ee 0b 40 00 00 00 00 > 07 00 e4 00 00 00 00 00 00 fa 00 03 00 0b 00 00 > 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 > > regardless of any orientation changes or movement of the Freerunner - > completely unchanging output. > > The third line used to look something like this (faceup): > 47 00 e4 00 00 00 00 00 00 01 00 01 00 35 00 00 > > The first '01' is X, the second '01' is Y, the '35' is Z. Those three > values are all that changed. > > I've tried rebooting, I've tried powerdown and restart, both without and > with battery removal, I've tried explicitly activating the sensor (hadn't > needed to before) at > /sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/power/wakeup. Same with > spi0.0 sensor - unchanging near-zero outputs on all three axes. (Values > wrap from 00 to FF with FF representing '-1' - IE the face-down reading > would have something around C5 for the Z reading) > > I'm about to reflash and reconstruct my environment, (I'd added celtune > repository to get linphone running, forgot to remove it and opkg upgrade > pulled a bunch of packages from there - after a second opkg upgrade > enlightenment is broken and won't start on reboot) will update afterwards, > and hopefully be able to post the requested data tomorrow... > > j > > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: accelerometer jitter
On Sat, 20 Sep 2008 17:36:00 -0400, "Michael Fisher" <[EMAIL PROTECTED]> wrote: > On Sat, Sep 20, 2008 at 4:07 PM, Joel Newkirk <[EMAIL PROTECTED]> > wrote: > >> I wrote a short shell script to repeatedly take the raw output from one > of >> the accelerometers > (/sys/devices/platform/spi_s3c24xx_gpio.1/spi0.1/dump) >> and massage the data to output X,Y,and Z-axis readings in decimal, -128 > to >> 128. > As a bit OT but since you have gone to the trouble of developing the > script, > could you run a test while driving/riding in the car and post what the > readings might be? I was thinking of the possibility of having the FR > determine if it was in a moving car or not. > > Mike OK, I started in the street near my house. Turned a 270 degree left turn (around a ~30-ft circle) then 90 degrees right and went down the street to a stop sign, left turn and on a couple hundred feet to a larger circle, which I circled 1.25 times at a near-steady 20mph, then went on to the convenience store. The entire trip was about 4 minutes, about 1.5 miles. Top speed was just over 30 mph (25mph limit throughout - don't send my GPS log to the cops ;) The Freerunner was sitting on the floor on the passenger side of the car, about 10" right of centerline, with the Y axis (top to bottom of screen in portrait mode) aligned to the direction of travel. So Positive-Y indicates vehicle deceleration, Positive-X indicates left turn, Negative-Z means the car probably bounced upward. Below is the output at roughly 1-second intervals from the accelerometer (#1, not #0, so axes aligned to housing) followed by the trip log from TangoGPS for the same time period. Top speed was just over 30 mph (25mph limit throughout - don't send my GPS log to the cops ;) There is at most a 5-second difference between start and stop points of the two datasets, and the right-hand turn off the small circle in front of my house should provide a good sync point (accel data goes negative first time), as will the ~30 second stop at a traffic light, ~20 seconds before data stops. j atest.sh output: X:14 - Y:-6 - Z:40 X:1 - Y:6 - Z:53 X:2 - Y:2 - Z:50 X:11 - Y:1 - Z:51 X:9 - Y:4 - Z:48 X:14 - Y:7 - Z:58 X:15 - Y:12 - Z:50 X:21 - Y:5 - Z:53 X:23 - Y:13 - Z:51 X:17 - Y:9 - Z:56 X:-8 - Y:5 - Z:52 X:-6 - Y:5 - Z:46 X:-4 - Y:5 - Z:49 X:3 - Y:1 - Z:54 X:6 - Y:6 - Z:56 X:6 - Y:7 - Z:54 X:6 - Y:6 - Z:58 X:4 - Y:5 - Z:59 X:5 - Y:8 - Z:49 X:6 - Y:6 - Z:53 X:8 - Y:5 - Z:50 X:13 - Y:3 - Z:57 X:15 - Y:10 - Z:52 X:8 - Y:9 - Z:49 X:9 - Y:14 - Z:47 X:4 - Y:17 - Z:53 X:5 - Y:6 - Z:53 X:5 - Y:2 - Z:55 X:9 - Y:-5 - Z:58 X:18 - Y:-6 - Z:53 X:15 - Y:-6 - Z:51 X:5 - Y:-2 - Z:56 X:1 - Y:7 - Z:49 X:6 - Y:5 - Z:52 X:4 - Y:8 - Z:54 X:2 - Y:11 - Z:54 X:-4 - Y:7 - Z:54 X:-1 - Y:7 - Z:51 X:7 - Y:-4 - Z:57 X:22 - Y:-1 - Z:53 X:25 - Y:7 - Z:53 X:18 - Y:4 - Z:47 X:19 - Y:7 - Z:53 X:15 - Y:5 - Z:51 X:19 - Y:9 - Z:48 X:17 - Y:8 - Z:52 X:21 - Y:8 - Z:48 X:27 - Y:9 - Z:53 X:27 - Y:12 - Z:50 X:24 - Y:9 - Z:47 X:20 - Y:8 - Z:52 X:7 - Y:7 - Z:52 X:-9 - Y:6 - Z:54 X:-9 - Y:8 - Z:49 X:3 - Y:2 - Z:54 X:4 - Y:-1 - Z:49 X:0 - Y:2 - Z:55 X:2 - Y:3 - Z:56 X:0 - Y:4 - Z:54 X:3 - Y:2 - Z:53 X:2 - Y:10 - Z:53 X:3 - Y:9 - Z:49 X:2 - Y:6 - Z:60 X:8 - Y:13 - Z:54 X:8 - Y:7 - Z:46 X:12 - Y:4 - Z:49 X:10 - Y:1 - Z:54 X:6 - Y:1 - Z:54 X:4 - Y:2 - Z:51 X:4 - Y:4 - Z:56 X:1 - Y:3 - Z:56 X:4 - Y:5 - Z:58 X:9 - Y:8 - Z:48 X:9 - Y:3 - Z:57 X:6 - Y:5 - Z:48 X:13 - Y:5 - Z:57 X:8 - Y:3 - Z:56 X:6 - Y:3 - Z:56 X:7 - Y:3 - Z:50 X:6 - Y:4 - Z:57 X:6 - Y:3 - Z:51 X:6 - Y:6 - Z:53 X:5 - Y:6 - Z:53 X:4 - Y:5 - Z:50 X:7 - Y:6 - Z:50 X:8 - Y:7 - Z:57 X:7 - Y:6 - Z:51 X:8 - Y:7 - Z:51 X:4 - Y:8 - Z:55 X:3 - Y:7 - Z:47 X:4 - Y:8 - Z:53 X:3 - Y:7 - Z:48 X:1 - Y:8 - Z:51 X:-4 - Y:8 - Z:55 X:-1 - Y:7 - Z:50 X:-1 - Y:4 - Z:54 X:-3 - Y:4 - Z:51 X:2 - Y:4 - Z:53 X:4 - Y:4 - Z:49 X:2 - Y:3 - Z:53 X:8 - Y:10 - Z:52 X:6 - Y:8 - Z:51 X:4 - Y:12 - Z:50 X:7 - Y:10 - Z:54 X:1 - Y:9 - Z:58 X:0 - Y:8 - Z:52 X:2 - Y:10 - Z:55 X:0 - Y:11 - Z:51 X:0 - Y:7 - Z:56 X:-1 - Y:6 - Z:53 X:-2 - Y:2 - Z:53 X:0 - Y:3 - Z:54 X:1 - Y:1 - Z:55 X:1 - Y:4 - Z:56 X:1 - Y:4 - Z:49 X:5 - Y:7 - Z:44 X:9 - Y:7 - Z:52 X:11 - Y:8 - Z:52 X:7 - Y:10 - Z:51 X:3 - Y:10 - Z:57 X:5 - Y:15 - Z:46 X:3 - Y:15 - Z:55 X:5 - Y:-4 - Z:55 X:11 - Y:-3 - Z:53 X:15 - Y:6 - Z:50 X:10 - Y:1 - Z:56 X:6 - Y:2 - Z:52 X:9 - Y:10 - Z:52 X:3 - Y:4 - Z:53 X:3 - Y:-1 - Z:55 X:5 - Y:1 - Z:54 X:6 - Y:5 - Z:50 X:4 - Y:2 - Z:51 X:2 - Y:4 - Z:53 X:2 - Y:5 - Z:53 X:4 - Y:5 - Z:53 X:5 - Y:5 - Z:54 X:3 - Y:3 - Z:47 X:6 - Y:3 - Z:50 X:2 - Y:4 - Z:56 X:5 - Y:6 - Z:48 X:4 - Y:5 - Z:55 X:4 - Y:2 - Z:55 X:5 - Y:3 - Z:54 X:2 - Y:2 - Z:55 X:6 - Y:1 - Z:57 X:2 - Y:0 - Z:53 X:4 - Y:3 - Z:51 X:2 - Y:4 - Z:56 X:1 - Y:1 - Z:56 X:6 - Y:4 - Z:51 X:3 - Y:3 - Z:52 X:6 - Y:2 - Z:55 X:3 - Y:7 - Z:49 X:5 - Y:5 - Z:51 X:3 - Y:7 - Z:55 X:4 - Y:7 - Z:47 X:4 - Y:7 - Z:49 X:5 - Y:3 - Z:55 X:5 - Y:7 - Z:48 X:5 - Y:12 - Z:4
Re: Nokia BL-5C replacement battery
On Tue, 23 Sep 2008 05:04:02 +0200, Joerg Reisenweber <[EMAIL PROTECTED]> wrote: > If you bought a Nokia BL-5C for a spare battery, you might want to check > this > site: > http://batteryreplacement.nokia.com/batteryreplacement/en/ > > cheers > jOERG Wow. After reading this I went to check the BL-5c from my old Nokia 6600, and couldn't find it. I also have a BL-6c and a new BL-5c clone I ordered online a couple months ago. That last one was on the desktop charger, and was almost twice as thick in the middle as it should be... extremely pillow-shaped. Needless to say I relocated it to the middle of the concrete garage floor away from anything flammable. (6-ft radius :) Thanks for the heads up. Apparently whoever made this clone battery, it has the same or a similar flaw to the genuine Nokia ones made by Matsushita in 2006. (always assuming it's not just someone selling off the portion of the 46 million made which never reached consumers, which actually seems pretty believable) But the timing is just remarkable. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Measure of the file /dev/input/event3 (Accelerometer data)
On Tue, 23 Sep 2008 03:59:00 -0700 (PDT), daniel103 <[EMAIL PROTECTED]> wrote: > > Hello again, > > > in /dev/input/event3 (or /dev/input/event2) you can see something like > that: > > > 00030e0 c157 48d7 82df 000c > 00030f0 c157 48d7 a770 000c 0002 0012 > 0003100 c157 48d7 a7ea 000c 0002 0001 0012 > > the last two columns are the accelerometer values. my question is ¿what > is > the measure of this values? ¿how i can obtain the decimal value? > > thanks so much http://wiki.openmoko.org/wiki/Accelerometer_data_retrieval Conversion from hex to decimal depends on programming language used. The Python script embedded near the bottom of the above-linked page is the basis for my V2 testing after some alteration. (I used 'bc' console calc to convert hex to dec in the bash-script V1 tests which read sysfs instead of /dev/input) As far as the 'measure', lay the phone flat on a table and check the Z-axis measurement, which is probably somewhere around 53-55 decimal - that's measuring 9.8 m/s^2, or 32 ft/s^2, or 1G. ;) The default mode for these guys is about 2G fullscale, though you can set it into a mode where fullscale is ~8G. (which, I presume, means an accompanying loss of detail) The counters represent -127 to 128, with FF==-1. (at least in 2G mode, which is all I've used) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: inotify-tools on openmoko
On Wed, 24 Sep 2008 15:44:34 +0100, Al Johnson
<[EMAIL PROTECTED]> wrote:
> On Wednesday 24 September 2008, jitendersingh wrote:
>> hi,
>>Do openmoko has support for inotify-tools.
>>(more information on inotify tools on
>> http://inotify-tools.sourceforge.net/)
>>if not, how can I port inotify-tools to openmoko.
>
> OpenEmbedded has inotify-tools 3.12 so there is a fair chance it will
> build
> and run on openmoko. Under mokomakefile it would be:
>
> make build-package-inotify-tools
>
If not, I successfully built 3.13 (untested though, my Freerunner's not
here ATM) with just a slight alteration to the configure script -
circumventing the inotify test that's not possible when cross-compiling
anyway. (seems kinda shortsighted to actually check if you're
cross-compiling but then fail instead of warn and attempt)
j
/usr/local/openmoko/source/inotify-tools-3.13$ diff configure.old
configure.new
19678,19682c19678,19684
< { { echo "$as_me:$LINENO: error: cannot run test program while cross
compiling
< See \`config.log' for more details." >&5
< echo "$as_me: error: cannot run test program while cross compiling
< See \`config.log' for more details." >&2;}
<{ (exit 1); exit 1; }; }
---
> { echo "Cannot run test program while cross compiling, just cross our
fingers..."
> # { { echo "$as_me:$LINENO: error: cannot run test program while cross
compiling
> #See \`config.log' for more details." >&5
> #echo "$as_me: error: cannot run test program while cross compiling
> #See \`config.log' for more details." >&2;}
> # { (exit 1); exit 1; }; }
> }
j
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community
Re: [2008.09] resolv.conf
On Thu, 25 Sep 2008 02:01:24 +0300, Flyin_bbb8 <[EMAIL PROTECTED]> wrote: >> In 2008.8 resolvconf didn't populate it >> with the nameserver supplied by dhcp when the wifi interface was brought >> up. >> I don't know if this is still the case in 2008.9. >> > > hmm actually it always filled it in for me from DHCP when connectin to > wifi > using "ifdown eth0 && ifup eth0" both in 2008.8 and in 2008.9 Both correct, actually, as it DOES fill it in from DHCP (if used) connecting wifi with 'ifup eth0', but IIRC resolvconf isn't used to do it, it just overwrites resolv.conf. The idea behind resolvconf (the bin) is that IT manages changes to the resolv.conf file, doing things like removing eth0-related DNS nameservers when eth0 goes down, going back to usb0-related nameservers automatically. ifup should feed DNS info to resolvconf and let it manage. Right now, I think DNS and default route changes when interfaces change up/down need some attention from the 'just-works' dept. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [2008.09] resolv.conf
On Thu, 25 Sep 2008 11:21:08 +1000, "Sarton O'Brien" <[EMAIL PROTECTED]> wrote: > On Thursday 25 September 2008 10:32:00 Joel Newkirk wrote: >> On Thu, 25 Sep 2008 02:01:24 +0300, Flyin_bbb8 <[EMAIL PROTECTED]> >> >> wrote: >> >> In 2008.8 resolvconf didn't populate it >> >> with the nameserver supplied by dhcp when the wifi interface was > brought >> >> up. >> >> I don't know if this is still the case in 2008.9. >> > >> > hmm actually it always filled it in for me from DHCP when connectin to >> > wifi >> > using "ifdown eth0 && ifup eth0" both in 2008.8 and in 2008.9 >> >> Both correct, actually, as it DOES fill it in from DHCP (if used) >> connecting wifi with 'ifup eth0', but IIRC resolvconf isn't used to do > it, >> it just overwrites resolv.conf. The idea behind resolvconf (the bin) is >> that IT manages changes to the resolv.conf file, doing things like > removing >> eth0-related DNS nameservers when eth0 goes down, going back to >> usb0-related nameservers automatically. ifup should feed DNS info to >> resolvconf and let it manage. >> >> Right now, I think DNS and default route changes when interfaces change >> up/down need some attention from the 'just-works' dept. > > > I still don't quite understand this confusion. > > udhcpc modifies /var/run/resolv.conf, as do I when needed. Any continually > changing value/file should reside in volatile, in my opinion. > > /etc/resolv.conf has, for as long as I've been playing, symlinked to > /var/run/resolv.conf. > > Except for the backend stuff not agreeing on this location, I have no > issues, > so long as udhcpc is used and an 'ip r d default' is issued prior to > invocation. > > If the plan is that these values will be written to the filesystem, I'd > say > I'll be modifying mine to the contrary. > > Sarton The program "resolvconf" is intended to work as follows: When an interface comes up, resolvconf should be called to "-a" add new nameserver(s) associated with the network connection on that interface. When an interface goes down resolvconf should get a "-d" delete nameserver(s) for the specified interface. The program resolvconf itself is supposed to make any/all changes to resolv.conf (wherever the actual file resides - the debian standard location is /etc/resolvconf/run/resolv.conf, actually) NOT the script or subsystem or network manager or whatever handles the up/down. Based on its configuration (notably /etc/resolvconf/interface-order - debian location) it uses whatever nameserver settings are needed, but remembers the others. So for example, usb0 coming up stuffs 192.168.0.201 in as nameserver, but does it by executing resolvconf. Then eth0 comes up and tells resolvconf its nameserver(s) - resolvconf decides which to use, likely eth0 will be set a higher priority than usb0. Later eth0 goes down. Now resolvconf automatically restores the nameserver settings from usb0. (presuming it is still up) If usb0 went down already, but ppp0 over gprs came up in the meantime, nameservers for ppp0 go live. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [2008.09] resolv.conf
On Thu, 25 Sep 2008 18:45:32 +1000, [EMAIL PROTECTED] wrote: > Joel Newkirk wrote: >> So for example, usb0 coming up stuffs 192.168.0.201 in as nameserver, > but >> does it by executing resolvconf. Then eth0 comes up and tells > resolvconf >> its nameserver(s) - resolvconf decides which to use, likely eth0 will be >> set a higher priority than usb0. Later eth0 goes down. Now resolvconf >> automatically restores the nameserver settings from usb0. (presuming it > is >> still up) If usb0 went down already, but ppp0 over gprs came up in the >> meantime, nameservers for ppp0 go live. >> > > OK, so this resolvconf thing is inherited from debian and has been > implented obviously without much thought to general functionality. I say > this in the context of normal (non-debian) services functioning as > expected when not massaged by resolvconf. > > I understand the intention, don't get me wrong, and it is indeed a nice > idea. But when all people want to do is gain connectivity with their > chosen method, this resolvconf facility seems to be more of a hinderence > and is unlike a lot of linux distros. Shouldn't this, at least > currently, be optional? > > If something as fundamental as _core_ network services are going to be > overlayed with something like this, it should really be working first, > rather thaan people editing files left and right without understanding > why. > > This is just what I think, take it as you will. In the end I am the > master of my device ;) > > I've actually just done a bit of research before posting and it > _appears_ this was implemented with good intention, just without the > required software support. It seems to be very much of debian origins > and a little bit of overkill for managing the facilities on the > freerunner, but that's just my opinion. If I was currently using usb0, > eth0 and a tun/tap interface simultaneously, that all flapped or had > dynamic dns requirments (sounds more like a router) it would make sense > to me. I imagine the end goal is to consolidate gprs and wifi dns but > currently no resolvconf is going to teach people how to get that going. > > When om2008 is mature and these kind of connections can be made easily, > without much knowledge, this utility would be a godsend ... as it stands > you currently need basic linux knowledge to understand why things don't > work and then you need to know how to manipulate resolvconf. > > Maybe it's my BSD upbringing, thanks for the clarification however. :) > > Sarton > IMHO the "simplest effective solution" should always head the list. In this instance I think that is to run dnscache or a similar service on the Freerunner itself, pointing somewhere like opendns.com's servers. (ok, maybe not simplest, though the local cache has various benefits. Pointing resolv.conf at opendns.com or wherever is simplest, assuming no network connection actually requires use of it's own DNS - corporate intranet wifi might, to reach 'private' records for example) If needed, newly-connected network connections that come equipped with new DNS servers (IE, wifi with DHCP) can insert/remove their DNS servers from the list used by dnscache, while resolv.conf permanently points at 127.0.0.1. This is what I've had running for a few weeks, but I hadn't yet attempted to get ifup/ifdown manipulating /etc/dnscache/root/servers/@. (which will essentially require a resolvconf-like manager ;) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [2008.09] resolv.conf
On Fri, 26 Sep 2008 10:24:18 +1000, "Sarton O'Brien" <[EMAIL PROTECTED]> wrote: > On Thursday 25 September 2008 22:06:29 Joel Newkirk wrote: >> IMHO the "simplest effective solution" should always head the list. In >> this instance I think that is to run dnscache or a similar service on > the >> Freerunner itself, pointing somewhere like opendns.com's servers. (ok, >> maybe not simplest, though the local cache has various benefits. > Pointing >> resolv.conf at opendns.com or wherever is simplest, assuming no network >> connection actually requires use of it's own DNS - corporate intranet > wifi >> might, to reach 'private' records for example) If needed, > newly-connected >> network connections that come equipped with new DNS servers (IE, wifi > with >> DHCP) can insert/remove their DNS servers from the list used by > dnscache, >> while resolv.conf permanently points at 127.0.0.1. This is what I've > had >> running for a few weeks, but I hadn't yet attempted to get ifup/ifdown >> manipulating /etc/dnscache/root/servers/@. (which will essentially > require >> a resolvconf-like manager ;) > > That all makes sense. Typically I've seen dnsmasq fulfill this function. > > My point was merely that the current solution is added complexity to a > rather > neat base filesystem. If this was a package I wouldn't have the same > gripe. > > Using dnsmasq, pointing at localhost, including opendns and using helper > scripts to insert and remove dns servers sounds less complex and less > confusing, with the added benefit of caching and less modification to the > base > system. It also means not having to incorporate something that doesn't > appear > to have had the modifications needed _to_ incorporate it. > > I don't like the requirement that all the software modifying network > settings > needs to be resolvconf aware. If we had a generic central solution people > could use whatever they want and not incur breakage ... > > Worst case, something like a resolv.conf pipe to a daemon that updates > resolv.conf.auto for dnsmasq would be better. Something central! > > Looking at the way resolvconf works, it almost seems like we should be > patching libc :S ...so we have a generic interface. > > Anyway ... I can feel fork coming on :P > > Sarton hmmm. I've been thinking on the situation, and have a proposal, as yet incomplete but IMHO promising - and it's nowhere near as drastic as forking. I see basically two approaches: One is an overarching network manager directly handling all the details, the other (my preferred) is to let DNS take care of itself independently, and address the multiple-default-route issue - probably with route metrics, though more advanced routing setup (requiring full "ip" binary instead of busybox) with multiple routing tables and rules is probably the "better" solution there. It starts with dnscache. Postinst in dnscache ipk removes the symlink /etc/resolv.conf and replaces it with a real file, pointing always at nameserver 127.0.0.1. (what I had going already) On my Ubuntu desktop, there's a possibly-useful file: /etc/resolvconf/update.d/dnscache - its purpose is to alter /etc/dsncachce/root/servers/@ (the list of servers dnscache is to query for anything uncached) automatically whenever resolvconf is triggered. This would be the way to hook in if resolvconf were working on the Freerunner as intended, which currently appears not to be the case. Alternately (or additionally), /etc/network/if-up.d/ and /etc/network/if-post-down.d/ allow us to perform the same changes automatically whenever an interface is brought up or taken down using ifup and ifdown. I've altered my /etc/init.d/dnscache script to include a 'refresh' feature, that first wipes /etc/dnscache/root/servers/@, then stuffs in any non-localhost nameservers presently listed in /etc/resolv.conf, then adds any nameservers found in /etc/resolvconf/run/resolv.conf, finishing up with any IPs listed in /etc/default/dnscache (default contents opendns.com IPs), wipes /etc/resolv.conf back to localhost only, then finally kicks dnscache so it uses the new servers/@ contents. Once my freerunner is back in front of me I'll test the init script changes - if it works as I hope, I can add a script to the ipk in each of /etc/network/if-up.d/ and /etc/network/if-post-down.d/ that simply call "/etc/init.d/dnscache refresh", and/or in /etc/resolvconf/update.d/. That makes the whole thing "just work" as a single ipk installing dnscache and all the support to override resolvconf AND 'manual' nameserver updates implemented by ifup/ifdown. If resolvconf is discarded it will work fine,
Re: New to Freerunner!!
On Fri, 26 Sep 2008 09:33:20 +0530, Kishore <[EMAIL PROTECTED]> wrote: (mercilessly snipped and interspersed with responses) > I am new to this community. I just got myself the freerunner a couple of > days back. I am a fairly experienced user of Linux/Debian/Kubuntu on > the desktop and this is my first experience with something like the > freerunner. Welcome! Roll up your sleeves and have some fun. :) > 1) Get the wifi to connect. It currently invariably always says something > like "Error: Unable to connect to the network" http://wiki.openmoko.org/wiki/Wifi cover this. It states that the GUI "should work fairly well" in updated 2008. I never had much success with it, but honestly didn't try much in the past few weeks. I've had best success with "ifup eth0" in terminal, once I'd set up a good wpa_supplicant.conf covering my WPA home AP, WPA work AP, unsecured work AP, and falling back to any unsecured AP that would talk to it. > 2) Get my GPS working. TangoGPS says no GPS found and "locations" > says unable to get a fix. http://wiki.openmoko.org/wiki/TangoGPS#Running_tangoGPS_on_Om_2008.8 should do the job. You need gpsd installed and config altered (GPS_DEV=/dev/ttySAC0), and you need to start gpsd, which you can do in Settings. > 3) How do i exchange data between my PC and the FR so that i could > push my MP3's and documents etc. Simplest is probably sftp or scp. http://wiki.openmoko.org/wiki/File_transfer - although it finishes up by stating that use as a USB storage device for the desktop is unsupported, while that is in fact possible and fairly easy to achieve now. (though it only presents the microSD as a usb drive to the PC, or more precisely any single partition - which should not be mounted on the Freerunner's filesystem while shared) > 4) A good music player, web browser and PDF reader. I can't help on that one. I toyed with the available media players and web browsers several weeks ago and that was all. > > Thanks to everyone who made this product what it is. Over the next > couple of months i wish to have my Qt apps i've written for the desktop > ported to this platform. In that regard i have not read anywhere about it > but i hope the USB port operates in host mode and i would need it to > interface with certain USB (1.1) hardware. > -- > Cheers! > Kishore > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Development environment suggestion ?
On Sun, 28 Sep 2008 18:42:54 +0200, Michael Tansella <[EMAIL PROTECTED]> wrote: >> Hello community lists ! >> This may be virtualbox images.. Nothing fancy, just ubuntu with >> toolchain, maybe even a "ready" eclipse version. Publish this by >> torrents, Maybe even with "good" to go software used during development. >> I spent lots of time before i got a working environment to develop/debug >> my openmoko in. > > +1 > This would be very nice I have made such an image. (a few times, in fact - first one got bloated before I thought about cloning it, second one got corrupted, third time's the charm, right? ;) I'm working on it right now. I started with Ubuntu Jeos 8.04.1. (JEOS is Just Enough OS - stripped-down to the bare essentials to get a working X-less console install in a virtualized machine) On top of that I selected (with aptitude) to install 'xubuntu-desktop', which is a metapackage that pulls in everything from Xubuntu. (Xubuntu is Ubuntu but with XFCE4 instead of Gnome or KDE) Don't really want 'everything' though, so before committing the installation I then go through and manually unselect a couple hundred packages, like X drivers for every video chipset (VM xserver only is required), office apps, etc. Bring in everything needed to build and install e17 (using easy_e17.sh) and install the Openmoko toolchain and VMware Tools. Then flush out all the development packages and i686 build tools not needed for cross-development, and remove the e17 build dir (~1gb). Then go through again with synaptic and purge another hundred or so packages, though more can still go. GDM offers e17 session at login as well, if desired, but default is xfce4. Right now, the image boots up and logs in user 'omdev' automatically (stop this in Settings->Login Window) with an xfce desktop. /usr/local/openmoko links in desktop and /home/omdev, desktop shortcuts for xfce4-terminal and xterm, and edje_editor. (the main reason I build e17 in the VM - for those wanting to change wallpapers, init screen, desktop layout, anything else "theme-ish" current Openmoko OS releases) It has Firefox3 and Gimp but not much else preinstalled in the way of applications. CVS, git, SVN, wget all installed and ready to run. A few other added console-based utils (mc, ipcalc, others) - vim as well. (though at about 24mb and me not using it, it was tempting to purge it) Set up to share USB devices - tested with memory cards, but my Freerunner isn't in hand at the moment. From VMware Infrastructure Web Access (latest VMware server) you can select any connected USB devices and they appear to the VM. In this instance, I plugged in a USB thumbdrive, after the desktop recognized it I told VMware to connect it to the VM, and Thunar file manager popped up on the VM desktop displaying drive contents. Painless, but I don't know how easy it would be to redirect the Freerunner uboot to it for dfu-util, nor if it's stable/reliable enough to flash the FreeRunner through this arrangement. (I suspect it will be workable, though maybe awkward needing to keep uboot idling while redirecting USB) usb0 is set up to communicate with the default FreeRunner usbnet config. I've tested the toolchain to be sure everything works as it should. (I added ". /usr/local/openmoko/arm/setup-env" to .bashrc so the toolchain is always ready in new terminals) The image is preconfigured for one CPU, an 8gb partition (sorry, no more available on my workstation when created) and 1gb RAM, (change in VMware before booting if needed) with desktop set to 1280x1024. (change in SettingsManager->Display) It currently uses 2.3gb of that partition, while a 7zip archive of the entire VM weighs in at about 1.4gb. What else should be here? As it stands it serves my needs, but others' will of course differ. I thought about bitbake (which I don't use) but it seems to me that the added weight in the VM wouldn't be justified in the distributable form - thoughts? I intend to rebuild it with a significantly larger partition as soon as I can clear some space on this workstation's HD, but that'll probably be several days from now. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Development environment suggestion ?
On Mon, 29 Sep 2008 18:19:04 +0200, "Olivier Migeot" <[EMAIL PROTECTED]> wrote: > On Sun, Sep 28, 2008 at 10:08 PM, Joel Newkirk <[EMAIL PROTECTED]> > wrote: > >> It currently uses 2.3gb of that partition, while a 7zip archive of the > entire VM weighs in at about 1.4gb. > > Did you try to "zero-fill" the image before compressing it? Not that > your ratio is anything bad, but it might get better (a lot of the > "used space" on a Linux partition is already filled of > easily-shrinkable text files, for example). > > A simple "dd if=/dev/zero of=/tmp/fillme ; rm /tmp/fillme" usually > does the trick, and I managed to compress some of my images (well, > those were more LAMP appliances than dev workstation, I have to say) > to higher ratios, IIRC. > > Good work anyway. > > -- > Olivier > M. Is that procedure applicable when VMware is set up to autosize the partition? I've got it set to 8gb partition size as a max. (stupid really that it won't let you select a max size larger than is presently available, when it's told not to actually allocate the space until used) Also, since posting that message I realized I'd forgotten to flush out apt archive again, which cleared up a good-size chunk of space. (IIRC, about 400mb) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: dead battery
A desktop charger for a Nokia BL-5c or BL-6c battery will do the job, at least the two non-Nokia ones I've tried. I've done it a few times. I've never let it sit on the charger until full, so I don't know if the charger can recognize full charge on the FreeRunner battery or not, but left on the charger for 15-20 minutes it will boot the FreeRunner. The most readily-accessible source is likely to be a friend with a Nokia phone using one of those batteries - several models over the past six or eight years. (mine is a 6600) If you know anyone with a nokia phone (now or the past several years) it's worth asking them what battery it uses, if they have a desktop charger, and if you can borrow it. (or keep it - if they're not using the phone ask for the charger and battery ;) j On Tue, 30 Sep 2008 13:04:07 -0600 (MDT), "Vince M. Clark" <[EMAIL PROTECTED]> wrote: > My battery is completely drained so I cannot boot to charge it. > > The following thread offered three possible solutions: > http://n2.nabble.com/Can%27t-boot-Freerunner-without-battery---Was:-Re:-unable-to-start-up-freerunner-after-batterie-was-full-down-td526783.html > > Does anyone know where I can get an external charger that will work with > the Freerunner battery? > > Vince Clark > [EMAIL PROTECTED] > (303) 493-6723 ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: dead battery
On Tue, 30 Sep 2008 13:35:56 -0600 (MDT), "Vince M. Clark" <[EMAIL PROTECTED]> wrote: > One more question. Maybe it will be obvious once I get my hands on a > charger, but how do you actually connect the battery? Will it fit in an old > Nokia phone? Or do you somehow connect the battery directly to the charger? The battery usually lays in just like it does in the phone. The battery will fit into the Nokia, but I'd be surprised if the Nokia agreed to charge it, since it won't recognize it. j > - Original Message - > From: "Joel Newkirk" <[EMAIL PROTECTED]> > To: "List for Openmoko community discussion" > > Sent: Tuesday, September 30, 2008 1:25:36 PM (GMT-0700) America/Denver > Subject: Re: dead battery > > A desktop charger for a Nokia BL-5c or BL-6c battery will do the job, at > least the two non-Nokia ones I've tried. I've done it a few times. I've > never let it sit on the charger until full, so I don't know if the charger > can recognize full charge on the FreeRunner battery or not, but left on > the > charger for 15-20 minutes it will boot the FreeRunner. > > The most readily-accessible source is likely to be a friend with a Nokia > phone using one of those batteries - several models over the past six or > eight years. (mine is a 6600) If you know anyone with a nokia phone (now > or the past several years) it's worth asking them what battery it uses, if > they have a desktop charger, and if you can borrow it. (or keep it - if > they're not using the phone ask for the charger and battery ;) > > j > > > On Tue, 30 Sep 2008 13:04:07 -0600 (MDT), "Vince M. Clark" > <[EMAIL PROTECTED]> wrote: >> My battery is completely drained so I cannot boot to charge it. >> >> The following thread offered three possible solutions: >> > http://n2.nabble.com/Can%27t-boot-Freerunner-without-battery---Was:-Re:-unable-to-start-up-freerunner-after-batterie-was-full-down-td526783.html >> >> Does anyone know where I can get an external charger that will work with >> the Freerunner battery? >> >> Vince Clark >> [EMAIL PROTECTED] >> (303) 493-6723 > > > ___ > Openmoko community mailing list > community@lists.openmoko.org > http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
qemu
In the course of working on my development VM, I installed qemu using Mokomakefile. It seems it installs a Qtopia image. Can someone lead me by the hand, or point out a howto/faq I missed, that clearly tells me how to set up other images with it, particularly FSO and 2008.9? I know that I need to tell qemu it's "gta02fake" instead of "gta01", but so far I've failed to get any images running with it apart from the one Mokomakefile pulls in. (they timeout while flashing) And is there any simple way to support having multiple images installed, choosing among them at launch time? All I can see for sure is how to 'flash' it with different images. (I've not used qemu much yet, so I still haven't fully wrapped my brain around the various symlinks and env settings affecting it) I'm hoping I can create desktop shortcuts to simple scripts that fire up qemu with various images, without having to fire up Mokomakefile and reflash each time - ideal would be fully independent, each available image having its own SD etc. Thanks for any info or advice. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Om2008.9] launching GPRS by hand
On Wed, 1 Oct 2008 15:32:23 +0200, Matthias Apitz <[EMAIL PROTECTED]> wrote: > El día Wednesday, October 01, 2008 a las 01:37:52PM +0200, Matthias Apitz > escribió: > >> >> Hello, >> >> The Wiki page http://wiki.openmoko.org/wiki/GPRS describes how to make >> GPRS working by hand in Om2007.2 and most of the stuff for chat-scripts >> and PPP options seems to be correct (compared with what I was using with >> my cellphone connected via V.24 to my laptop); >> >> in 2007.2 there was a gsmd daemon controlling the serial port >> /dev/ttySAC0 which seems now to be Qtopia's 'qpe', but I don't see any > method to >> stop and start this via scripts in /etc/init.d; how is this supposed to >> work now? > Maybe there is some way to just say to 'qpe': Go away for a while and > don't read or touch the modem /dev/ttySAC0? The rest of PPPD is fine > now, it just works as supposed to do; > > matthias > What you need is referred to as Multiplexing - the wiki page you referenced talks about it. The multiplexer allows multiple bindings to the GSM device, IE one for voice GSM and one for data GPRS, without the need to kill off QPE or whatever is tied to GSM in order to use GPRS. j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [2008.09][FSO] Programming interface to wifi Icon
On Wed, 01 Oct 2008 14:39:08 +0100, Arigead <[EMAIL PROTECTED]> wrote: > Sorry Yes I would like to use the Icon on the top of the illume bar. I > was thinking interfacing this to my daemon so regardless of app you can > see the wifi status the same as you can see the GSM status at present. > > If anybody knows about "somewhat different" :-) Would this presently be > set by Kernel code the Window Manager? It's part of the edje setup for the main screen, in illume.edj. (well, that can differ - I've always installed illume-config and illume-config-illume) You can think of the .edj file as a theme package of sorts, although it abstracts more UI handling out of the program than traditional theming. In OO fashion it separates all the user interaction into the edje file, with the main application interacting via signals to change UI or respond to user actions and triggers generated by the UI itself. You can set up in the .edj file to support changing icons triggered by signals from 'outside', and emit signals when some UI interaction (IE, click the icon) takes place. I'm interested in doing something similar with USB, showing whether it's in host or device mode, networking vs mass storage if device mode, sending or expecting power, forcing 500mA or 1000mA charging, and ability to change all the above. The program edje_editor is great for taking a look at such features. It's part of the E17 desktop setup, I've been playing with it in the developer VM I'm working on. It takes the intact .edj file and extracts all the configs, images, etc from it and gives you a nice WYSYWIG GUI editor. BTW - I printed out the actual text config from illume.edj, it's about 100 pages long... A simpler starting point to learn your way around .edj and edje_editor is illume_init.edj, the startup theme. (second view of the boots, with the green scanning back and forth instead of indicating percent progress) It's quite easy to drop in a different background image, redefine the animation, etc this way. (Now what we really need is a tool running ON the FR that decompiles the main .edj file and lets us replace wallpaper, etc - basic customization, not full-on UI redesign work) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: [Om2008.9] launching GPRS by hand
On Wed, 1 Oct 2008 16:18:33 +0200, Matthias Apitz <[EMAIL PROTECTED]> wrote: > El día Wednesday, October 01, 2008 a las 09:55:31AM -0400, Joel Newkirk > escribió: >> What you need is referred to as Multiplexing - the wiki page you > referenced >> talks about it. The multiplexer allows multiple bindings to the GSM >> device, IE one for voice GSM and one for data GPRS, without the need to >> kill off QPE or whatever is tied to GSM in order to use GPRS. > > Hi Joel, > > I've read this before, but all this information is outdated because it > talks about starting gsmd (which is not there anymore in Om2008.9) and I > don't even know if the Qtopia's qpe is aware of handling this device of > Multiplexing; > > there is not even a script in /etc/init.d to launch qpe; > > is there a working example of Multiplexing with qpe? > > at the moment I do it with this shell script ~/pppd.sh: The mokoscripts archive includes a replacement for /etc/X11/XSession.d/89qtopia. I don't see anything about gsmd, just gpsd (the same archive has agps stuff in it) and gsm0710muxd, the multiplexing gsm service. Also, the instructions specifically state that he's not sure about 2007.2 and FSO but has tested it with ASU. I successfully used the instructions (altered for t-mobile's internet3.voicestream.com service) under both 2007.2 and 2008.8, though to be honest I never tried it under 2008-update or 2008.9. (yet - once my FR is back in my hands I intend to set it up again) j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
