Re: USB host

2007-08-13 Thread Ian Stirling

Marcin Domanski wrote:

Yep it shouldn't be a problem - let's take an usb hub with external
power supply - connect it to neo and then connect the device to the
hub. I dont know how's the support for it but its an idea :)



That relies on the USB hub recognising a non-powered host.
Many do not.
(the first two I tested, a belkin tetrahub, and something else)


On 8/12/07, Ian Stirling [EMAIL PROTECTED] wrote:


assistivetechnology wrote:


Did anyone ever try to use the USB host with power supply, or build a HW
workaround to supply USB devices with power?


I'm contemplating the latter.
The kernel does not currently support USB host - it needs a patch.



___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Isn't this a little early?

2007-08-13 Thread andy selby
I sent the following to [EMAIL PROTECTED]

The domain freeopenmoko.com which is in your IP range is giving false
information.
The worst of it is that the testimonials claiming to have received a
Neo 1973 phone were up on the website before the phone was even
released.
The website also does not say that the device is not a fully
operational phone, it is a developer preview which is not suitable for
daily use, which is why its less expensive than comparably specced
models.
The site should also say it is not affiliated with the openmoko
project, the openmoko company or FIC.
Please ask them to correct it or kindly take down their site.

I'll post any follow up with them

___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Linux embedded dev board

2007-08-13 Thread Edwyn Stapel
2007/8/13, Edwyn Stapel [EMAIL PROTECTED]:

 nice dev kit.. but the neo is $150,- cheaper though and looks like it has
 more supported with their openmoko framework.

 Edwyn



___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: User Interface idea

2007-08-13 Thread D. Vicario
2007/8/12, Derek Pressnall [EMAIL PROTECTED]:

 One of the things that I've found on most pda-phones/smartphones is
 that they make the phone capabilities feel like just another
 application.  But when you want to use the device as a phone, it would
 be nice for the rest of the features to melt into the background.  The
 Motorola A780 does this nicely; it is a pda-phone when the flip is
 opened, but with the flip closed it becomes a regular feature-phone.

 So, that gives me an idea for UI layout.  First, have the interface
 support two virtual desktops, most likely via the window manager.  The
 primary desktop (PDA desktop) would remain how OpenMoko has it
 currently laid out.   The second desktop (PHONE desktop), however,
 should resemble the face of a normal phone.  The bottom half (or two
 thirds, whatever) of the screen would have a permanent phone keypad
 displayed (the keypad app), includuding directional buttons and
 several special-purpose buttons (answer/disconnect, function-A,
 function-B, Menu, OK, etc.).  Any application running in the Phone
 desktop would only be able to write to the window in the upper part of
 the screen, and they would receive their input through the keypad app
 runnin in the bottom.  The keypad app could have an API so that apps
 can request that certain keys be re-labled when that app is in the
 forground, but other than that the keypad would always display a
 similar layout for any running app.  This would enforce a consistant
 feel among the various apps.

 Now any app that wants to use the secondary Phone desktop would have
 to be specifically coded for it; I'm thinking that apps such as the
 Dialer would be running in the background, and have an active
 connection to the Phone desktop along with the PDA desktop.  The list
 of apps that should have Phone desktop capabilities would include the
 Dialer, SMS/Email, Media player, Calculator.


I'm in agreement with all what quoted above, and for obtain this I think
that must be implemented at least three things: Multitouch screen, Tactile
feedback via buzzer, and a good input method like these:
http://wiki.openmoko.org/wiki/Wishlist:Text_Input#New_input_methods (for me
finger splash could be very good). Substantially, IMHO the finger use of
principal functions of the phone joined with a good text input interface is
basically for the real usability of the terminal as a phone.
___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: User Interface idea

2007-08-13 Thread Jeff Andros
On 8/12/07, Derek Pressnall [EMAIL PROTECTED] wrote:

 snip

I'm thinking that apps such as the
 Dialer would be running in the background, and have an active
 connection to the Phone desktop along with the PDA desktop.

snip

um, do realize that we are working on a resource constrained system, and
probably will be for the forseeable future... I don't see this as TOO much
of a problem, as long as those apps unload most of themselves from memory
when they're not running, and they spend almost all of their time
sleeping... but watch out for that.

We'll have to see what people can come up with for this before I run this on
my system

-- 
Jeff
O|||O
___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: User Interface idea

2007-08-13 Thread Giles Jones


On 14 Aug 2007, at 00:02, Jeff Andros wrote:



um, do realize that we are working on a resource constrained  
system, and probably will be for the forseeable future... I don't  
see this as TOO much of a problem, as long as those apps unload  
most of themselves from memory when they're not running, and they  
spend almost all of their time sleeping... but watch out for that.


We'll have to see what people can come up with for this before I  
run this on my system




Exactly why the iPhone is very responsive and Windows Mobile phones  
aren't. Apple keeps things simple.


You need not worry about getting lost in applications if you have a  
button hardwired to the home screen. Windows Mobile has this on the  
smartphone edition, trouble is it's very fiddly to get back into any  
application you've come out of.


Being able to see a list of what is open and being able to switch to  
them is important for usability.



___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: User Interface idea

2007-08-13 Thread Derek Pressnall
On 8/13/07, Jeff Andros [EMAIL PROTECTED] wrote:

 um, do realize that we are working on a resource constrained system, and
 probably will be for the forseeable future... I don't see this as TOO much
 of a problem, as long as those apps unload most of themselves from memory
 when they're not running, and they spend almost all of their time
 sleeping... but watch out for that.

What I meant was that apps that would be in use would be split into
a piece that runs in the background, and a forground part that talks
to the display / input.  And the background part would only be running
when needed.
Now the phone app would always have a module running since it has to
listen for incomming phone calls.  The UI portion of the phone app
that uses the phone desktop would be the default / primary app that
is always available on it when nothing else is running (so that when
you switch to the phone desktop, you can start dialing out without
having to launch a phone app).  This UI would be smaller than the
pda-desktop version of the phone UI, since it doesn't have to worry
about drawing input buttons, etc; it talks to the already running
keypad app taking up the bottom half of the phone-desktop display.
However the UI frontend that is used from the pda-desktop side would
have a bit more code since it is responsible for it's own keypad, and
possibly takes care of more display items.

The Media player app could have a background piece that only run when
it is in use.  Again, it doesn't have any user interface components,
only an API for a frontend to talk to it.  The user can still exit
the app for both the frontend and backend to exit memory, but it will
go into pause mode if a higher priority app (such as the phone app)
needs a shared resource (the speaker).

This way, only the modules that are needed to be in memory are taking
up ram, and launching the (small) frontends seperately / in parallel
can give a snappier performance feel.

___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: User Interface idea

2007-08-13 Thread Giles Jones


On 14 Aug 2007, at 00:48, Derek Pressnall wrote:




What I meant was that apps that would be in use would be split into
a piece that runs in the background, and a forground part that talks
to the display / input.  And the background part would only be running
when needed.


It's sorta done like that already, we have GUI applications and  
runtime libraries.



Now the phone app would always have a module running since it has to
listen for incomming phone calls.


gsmd handles the calls, there's a library too.



The UI portion of the phone app
that uses the phone desktop would be the default / primary app that
is always available on it when nothing else is running


Sounds like you use the phone a lot? while that's right for you,  
other people use messaging or PDA functions more. You might find  
additional clicks to launch a dialer annoying, if you make dialer  
default someone else may think multiple clicks to browse their files  
annoying.


Lets not hardcode any form of functionality into the device, let  
people choose. Having a default desktop handler or a homescreen which  
can display plugins is the way to go about it.



___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Linux embedded dev board

2007-08-13 Thread Attila Csipa
On Monday 13 August 2007 22:22:24 Joe Friedrichsen wrote:
 It has more than a handful of similarities with the Neo (ARM9,
 touchscreen, USB, JTAG, but no GSM or GPS), and may help scratch some
 itches for new ideas and toys that have been floating on the list for

It's more of a PDA spec, if you can live without the screen, roughly the same 
spec can be had with a NSLU2 for well under 200$. If you add in a cheap GSM 
module via the serial port, that would be the poor mans Neo1973 (possibly 
could be even made to work with OpenMoko). But since it's not exactly pocket 
sized and needs HW hack to be battery powered it's not really much of a 
competition for the Neo :)


___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Login Manager

2007-08-13 Thread Jeff Andros
On 8/10/07, t3st3r [EMAIL PROTECTED] wrote:

 snip
 I see no effective way to combine these 2 different goals.One is
 prevents access to data but this will enforce bad guys to do full
 reflashing.Killing your (unusable) data but getting working (usable)
 phone.Another approach makes guys to believe phone is not defends itself
 and not secured.While it really silently tracks evildoers.
 /snip


simple... display contact info(email, friend's phone number, etc) to return
the phone at the login screen... I think my old(~2000 ish) WM PDA had an
option to do that... people can't get into the phone itself, but they can
figure out how to get ahold of you

-- 
Jeff
O|||O
___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Chinese input for OpenMoko

2007-08-13 Thread Jeff Andros
On 8/8/07, Shu Hung (Koala) [EMAIL PROTECTED] wrote:

 Hello,

 OpenMoko is developed and produced in Taiwan. And I live in HK.
 Both places uses Chinese as the common language.

 I'm just curious if there is any plan for Neo1973 to support Chinese
 input?
 If so, in what way? Hand-writing input? Or keyboard input method?


 Koala Yeung




 ___
 OpenMoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community


If you look real closely at the input panel, it shows a chinese mode
input... it doesn't work yet, but I'd guess it's planned

-- 
Jeff
O|||O
___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Debug Board JTAG Programming with openocd

2007-08-13 Thread Kyle Bassett
I have been unable to coerce my Neo to charge batteries, so I have a feeling
it has to do with the uboot image I flashed.  Should I manually charge the
batteries some other way? or continue trying to reflash the original uboot
revision?  (Can you even use the debug board if the battery is dead?)

My current uboot revision:
u-boot-gta01bv4-r11_e80955f07de03fef0196353e77534b2300193c1c_0_2348.bin

What I am trying to do right now is get up and running with openocd, but I
cannot get around this one error:

# openocd -f /etc/openocd/openocd.cfg
Info:openocd.c:84 main(): Open On-Chip Debugger (2007-01-31 12:00 CET)
Error:   jtag.c:1181 jtag_examine_chain(): JTAG communication failure, check
connection, JTAG interface, target power etc.

I know this means that it has found the debug board, but I am unsure what I
should do to initialize the phone.  (Simple as power?)

I have my Neo wired up to the debug board, (2.7v) dead battery inside.



Here are the steps I used to get where I am on Debian Etch:
-
moved 099_neo1973_debugboard.rules to /etc/udev/rules.d

compiled/installed libusb-0.1.12
compiled/installed libftdi-0.10

add file usr-lib.conf to /etc/ld.so.conf.d/
insert line /usr/local/lib
run ldconfig


compiled/installed openocd from svn:
svn co -r 137 http://svn.berlios.de/svnroot/repos/openocd/trunk openocd_137
./bootstrap
./configure --enable-parport_ppdev --enable-ft2232_libftdi
make; su; make install


made /etc/openocd/openocd.cfg with:
telnet_port 
gdb_port 
interface ft2232
jtag_speed 0
ft2232_vid_pid 0x1457 0x5118
ft2232_layout jtagkey
reset_config trst_and_srst
jtag_device 4 0x1 0xf 0xe
daemon_startup attach
target arm920t little reset_run 0 arm920t
working_area 0 0x20 0x4000 backup
run_and_halt_time 0 5000
ft2232_device_desc Debug Board for Neo1973

-

Some information I grabbed from here:
http://wiki.openmoko.org/wiki/Debug_Board
http://wiki.openmoko.org/wiki/OpenOCD#OpenOCD_and_Debug_Board
http://svn.openmoko.org/developers/werner/notes/openocd


Thanks,

Kyle
___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Debug Board JTAG Programming with openocd

2007-08-13 Thread Kyle Bassett
-Update:
I guess the debug board charges the batteries, because I checked the voltage
after typing the previous email and it's ~3.29v.  I still cannot get the Neo
to turn on though...

-Kyle

On 8/14/07, Kyle Bassett [EMAIL PROTECTED] wrote:

 I have been unable to coerce my Neo to charge batteries, so I have a
 feeling it has to do with the uboot image I flashed.  Should I manually
 charge the batteries some other way? or continue trying to reflash the
 original uboot revision?  (Can you even use the debug board if the battery
 is dead?)

 My current uboot revision:
 u-boot-gta01bv4-r11_e80955f07de03fef0196353e77534b2300193c1c_0_2348.bin

 What I am trying to do right now is get up and running with openocd, but I
 cannot get around this one error:

 # openocd -f /etc/openocd/openocd.cfg
 Info:openocd.c:84 main(): Open On-Chip Debugger (2007-01-31 12:00 CET)
 Error:   jtag.c:1181 jtag_examine_chain(): JTAG communication failure,
 check connection, JTAG interface, target power etc.

 I know this means that it has found the debug board, but I am unsure what
 I should do to initialize the phone.  (Simple as power?)

 I have my Neo wired up to the debug board, (2.7v) dead battery inside.



 Here are the steps I used to get where I am on Debian Etch:

 -
 moved 099_neo1973_debugboard.rules to /etc/udev/rules.d

 compiled/installed libusb-0.1.12
 compiled/installed libftdi-0.10

 add file usr-lib.conf to /etc/ld.so.conf.d/
 insert line /usr/local/lib
 run ldconfig


 compiled/installed openocd from svn:
 svn co -r 137 http://svn.berlios.de/svnroot/repos/openocd/trunkopenocd_137
 ./bootstrap
 ./configure --enable-parport_ppdev --enable-ft2232_libftdi
 make; su; make install


 made /etc/openocd/openocd.cfg with:
 telnet_port 
 gdb_port 
 interface ft2232
 jtag_speed 0
 ft2232_vid_pid 0x1457 0x5118
 ft2232_layout jtagkey
 reset_config trst_and_srst
 jtag_device 4 0x1 0xf 0xe
 daemon_startup attach
 target arm920t little reset_run 0 arm920t
 working_area 0 0x20 0x4000 backup
 run_and_halt_time 0 5000
 ft2232_device_desc Debug Board for Neo1973


 -

 Some information I grabbed from here:
 http://wiki.openmoko.org/wiki/Debug_Board
 http://wiki.openmoko.org/wiki/OpenOCD#OpenOCD_and_Debug_Board
 http://svn.openmoko.org/developers/werner/notes/openocd


 Thanks,

 Kyle


___
OpenMoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community