Re: [Wiki-gang] [Testing] Activity Set for 9.1.0
To clarify - the default image used in manufacturing contains Activities (the same ones shipped with G1G1 this year), right? My understanding is that (the image we ship) = (software release) + (customization key), so Chris's statement that our software releases don't ship with Activities is correct, but Greg's statement about starting Activity selection early for 9.1.0 makes perfect sense. -Mel Greg Smith wrote: Hi Chris, When this release is done we will update the default image used in manufacturing. All of our deployments except one currently take the default image. Many change it (usually adding more than subtracting) after they get the XOs in country but they start with our factory image. We will also offer people an option to clean install with a set of activities like we did in 8.2. Let me know if that's not clear or you need more info. Thanks, Greg S Chris Ball wrote: Hi Greg, Hi All, Learning from the last release, I want to start activity selection early for 9.1.0. Our software releases don't ship with activities, so I don't understand what's being selected here. We didn't choose a set of activities to ship with 8.2.0; we chose a set of activities to ship on a G1G1 image. Each deployment creates its own conjunction of {XO release+Activities}, with G1G1 being one such deployment when it's running. Could you elaborate on what the list of chosen activities would mean this time around? Thanks, - Chris. ___ Wiki-gang mailing list [EMAIL PROTECTED] http://wiki.laptop.org/go/Wiki-gang (_internal_name)s ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] One instance activity
On Wed, 10 Dec 2008, Bert Freudenberg wrote: On 10.12.2008, at 03:57, [EMAIL PROTECTED] wrote: On Tue, 9 Dec 2008, Bert Freudenberg wrote: On 09.12.2008, at 18:55, Eben Eliason wrote: Are you sure? Browse makes use of shared code, but still presents the user with the appearance of multiple instances. Right. The way to do it would be to create a unique D-Bus service in your activity. When the second instance tries to create that service it will notice that it already exists. It could then notify its first instance via said D-Bus service. you don't need to use D-Bus for this, it can be done by X without any other communication channels. I don't know the details for how to do this, but I've seen mozilla/firefox do this for a few years. to see this start firefox on one machine, connect to another machine and point the display back to the first one. then try and start firefox on that second machine. the end result will be a new window opening up, but running on the first machine (if you have trouble seeing the difference, make the two machines have different bookmarks, or give one network access that the other doesn't have) please don't develop new mechanisms to do things that already exist. It's not a new mechanism. The usage of a named D-Bus service to ensure unique program instances is documented and not my invention (though I cannot remember where I saw it first). I should have written One way to do this would be ..., I give you that. But you cannot know which way would be preferable for a given activity. And since it is private to the activity and does not affect other activities, no harm is done either way. E.g., twiddling X properties is hard in various high-level languages, in particular when using higher-level UI toolkits. Sugar currently requires two custom X properties and this is causing activity authors considerable pain. Even Sugar itself had to resort to C code, adding a custom native library to manipulate these properties, it was not easy in pure Python. This is in stark contrast to the nicely general and easy-to-use Python D-Bus bindings, which are similarly available in other high-level languages. So please consider that not all people like having to go down into the machine room to make new plumbing with a C compiler. Having done too much of that myself I can relate to them. my initial reaction to this is that this sounds like a gap in the python libraries that would be very useful to fill. getting someone to write a python library to better access the X properties would help many areas. any idea why nobody has written one yet? Python is much better than many languages at letting you write a library in C and then use it cleanly, so if doing this in python is as hard as you are indicating, why hasn't it been addressed? David Lang ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Way to tell if it is an XO
Thanks a lot for your answers. Yes, I think a shared credentials are the best way. Basically we want to offer a service just for the Xos and are working now on the authentication model. We are going to use webservices with WSShttp://www.oasis-open.org/committees/tc_home.php?wg_abbrev=wss and place a signed key on each XO that is going to use the service, to authenticate with the webservice provider. Regards, Marcel On Tue, Dec 9, 2008 at 8:14 PM, Carl-Daniel Hailfinger [EMAIL PROTECTED] wrote: Hi, On 09.12.2008 22:37, Chris Ball wrote: Hello, I need to know what is a good way to find out if the local machine running a python script is a XO. One way might be to cat /ofw/model. If the file doesn't exist, or doesn't return a letter and number, you probably aren't running on an XO. The goal is to check it the machine running the program is in fact an XO or some other machine, and to do it in a way that is very hard to fake from some other computer. I don't think making this hard to fake is possible, because any computer with root access can reply to your test in whatever way is necessary. I think you probably shouldn't try to accomplish this, or should try to use a different method such as a pre-shared credential. Yes, regardless of which test Marcel implements, it should be easy to fool the test in a few minutes. Regards, Carl-Daniel -- http://www.hailfinger.org/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Testing] Activity Set for 9.1.0
On Tue, Dec 9, 2008 at 7:57 PM, Greg Smith [EMAIL PROTECTED] wrote: All of our deployments except one currently take the default image. Many change it (usually adding more than subtracting) after they get the XOs in country but they start with our factory image. Where can this image be found? I'm curious as to what it contains. Thanks, Daniel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Bundling plugins with Browse
On Wed, Dec 10, 2008 at 3:47 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 22:29, Sayamindu Dasgupta wrote: On Wed, Dec 10, 2008 at 12:37 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 18:49, Sayamindu Dasgupta wrote: http://dev.laptop.org/~sayamindu/Browse-99.xo seems to be working for me. Please test. Attached is the patch which did the trick. Ugh. Why is it necessary to copy the plugin to $SUGAR_ACTIVITY_ROOT/ data ? That would only make sense if it is supposed to be modified at runtime. Do stuff placed in $SUGAR_ACTIVITY_ROOT remain persistent across reboots ? I seem to have troubles getting that (the file disappeared after reboot) You misunderstood. Why copy at all? Why not run it from the bundle? Copying from the bundle is only necessary for e.g. default configuration files that the activity needs to be able to change later (because the bundle itself is read-only). Ah, ok, I was blindly following cert8.db http://dev.laptop.org/~sayamindu/Browse-101.xo follows your solution and seems to work fine. Moreover, theoretically one can drop any plugin in the plugin directory of Browse, and expect it to work. Everyone, please test this as much as possible. Also, how do I include the source of the binaries (mozplugger, m4) in the Browse bundle ? Thanks, Sayamindu -- Sayamindu Dasgupta [http://sayamindu.randomink.org/ramblings] embed_pdf_in_browse.patch Description: Binary data ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] One instance activity
On 10.12.2008, at 11:56, [EMAIL PROTECTED] wrote: On Wed, 10 Dec 2008, Bert Freudenberg wrote: On 10.12.2008, at 03:57, [EMAIL PROTECTED] wrote: On Tue, 9 Dec 2008, Bert Freudenberg wrote: On 09.12.2008, at 18:55, Eben Eliason wrote: Are you sure? Browse makes use of shared code, but still presents the user with the appearance of multiple instances. Right. The way to do it would be to create a unique D-Bus service in your activity. When the second instance tries to create that service it will notice that it already exists. It could then notify its first instance via said D-Bus service. you don't need to use D-Bus for this, it can be done by X without any other communication channels. I don't know the details for how to do this, but I've seen mozilla/ firefox do this for a few years. to see this start firefox on one machine, connect to another machine and point the display back to the first one. then try and start firefox on that second machine. the end result will be a new window opening up, but running on the first machine (if you have trouble seeing the difference, make the two machines have different bookmarks, or give one network access that the other doesn't have) please don't develop new mechanisms to do things that already exist. It's not a new mechanism. The usage of a named D-Bus service to ensure unique program instances is documented and not my invention (though I cannot remember where I saw it first). I should have written One way to do this would be ..., I give you that. But you cannot know which way would be preferable for a given activity. And since it is private to the activity and does not affect other activities, no harm is done either way. E.g., twiddling X properties is hard in various high-level languages, in particular when using higher-level UI toolkits. Sugar currently requires two custom X properties and this is causing activity authors considerable pain. Even Sugar itself had to resort to C code, adding a custom native library to manipulate these properties, it was not easy in pure Python. This is in stark contrast to the nicely general and easy-to-use Python D-Bus bindings, which are similarly available in other high-level languages. So please consider that not all people like having to go down into the machine room to make new plumbing with a C compiler. Having done too much of that myself I can relate to them. my initial reaction to this is that this sounds like a gap in the python libraries that would be very useful to fill. getting someone to write a python library to better access the X properties would help many areas. any idea why nobody has written one yet? Python is much better than many languages at letting you write a library in C and then use it cleanly, so if doing this in python is as hard as you are indicating, why hasn't it been addressed? I don't know. And this is diverging from the subject at hand. Even if it was simple to set and read an X property then implementing single-instance apps via X properties still would be considerable more complex than using D-Bus. You would have to learn a whole lot about how X works first, and then design that mechanism on top of it. Having done a fair bit of X programming myself (the Squeak VM talks to Xlib directly, no toolkit) I would not exactly rate this as trivial. X can be abused to do IPC, sure, but it still is abuse. D-Bus is made to do that. - Bert - ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Testing] Activity Set for 9.1.0
Hi Dan, RTFM :-) http://wiki.laptop.org/go/Release_Notes/8.2.0#If_your_XO_is_not_connected_to_the_internet Thanks, Greg S Daniel Drake wrote: On Tue, Dec 9, 2008 at 7:57 PM, Greg Smith [EMAIL PROTECTED] wrote: All of our deployments except one currently take the default image. Many change it (usually adding more than subtracting) after they get the XOs in country but they start with our factory image. Where can this image be found? I'm curious as to what it contains. Thanks, Daniel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
9.1 Weekly Meeting on IRC at 2PM US ET Today
Hi All, Our weekly 9.1 meeting (was called feature roadmap) is today, Wed. 12/10 at 2PM US East Time on IRC Freenode.net #olpc-meeting channel. Agenda for this week is: 20 minutes - XO Camp update 20 minutes roadmap update and 9.1.0 page edits (especially schedule) 10 minutes finalize and approve trac usage 10 minutes agenda and action items for next week Minutes of last weeks meeting is here: http://lists.laptop.org/pipermail/devel/2008-December/021502.html Thanks, Greg S ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Resuming by default (was: One instance activity)
On Wed, Dec 10, 2008 at 6:09 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 20:39, Bert Freudenberg wrote: On 09.12.2008, at 18:55, Eben Eliason wrote: Pablo, for what reasons do you desire to prevent multiple instances from running? Hope there's a better reason than it's too easy to create unwanted instances ... (still waiting for the home view icons to resume by default) Circling back to the underlying problem ... why can't we indeed make the home view activity icon resume the latest journal entry? I think performance of the datastore query was the primary reason, apart from lack of time/resources. * this was the intention of the Journal all along (would make the idea of stopping/resuming activities much more obvious) The intention of the Journal was to /be/ the space from which one could resume previous work. This says nothing about the functionality of the Home screen. Experience has indicated that people (particularly kids) more often than not would prefer to resume the last thing they worked on, which led us to the proposed bi-modal home view. * it would solve Pablo's problem (this works today, resuming an already opened journal entry just switches to that activity) It solves his problem /if/ his problem was simply attempted prevention of accidental duplicate instances. If, on the other hand, his problem involved a technical constraint which prevents two instances from running simultaneously, it does not. We need to hear more about the root issue. * no more unwanted Journal entries (one would be reused all-over) This isn't quite the right idea. We wish to give preference to continuing work, but that's certainly not the same thing as saying that there is just one instance. No entries are reused. Instead, the most recent instance is simply shown by default, adjacent to a number of older instance, as well as a new instance. In any case, I think the better solution to reducing unwanted entries is to encourage naming, and provide a don't keep option in the naming prompt upon closing a previously unnamed unkept activity. * creating a new Journal entry would be an explicit action (need UI for that) Right. The current proposal is to provide the alternate functionality (start new) in the contextual menu (or by holding alt while clicking). That last item is the only problem I can see - in the activity one would need a start new entry button, just having this as optional menu item in the home view is not enough. The semantics would be just like the keep a copy button now, but also reset the activity to a blank slate. Hmmm, I'm not sure about this. I think it's important for the instance model to maintain that launching the most recent instance does indeed resume that specific instance. I think the in-place new document breaks this idea, harkening back to the old model of applications and files. If the upcoming datastore work focused on versioning as much as legacy file support, we would not even need that button (though it might still be convenient to have). Older instances would then readily be available in the Journal and could be resumed from there. Exactly. This is small part of the reason we didn't implement this yet, too. Versions would make it less problematic if a child resumed a recent activity, actually wanting a new document, and then destructively edited it. - Eben - Bert - ___ Sugar-devel mailing list [EMAIL PROTECTED] http://lists.sugarlabs.org/listinfo/sugar-devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Server-devel] Amateurish Workaround to Get Bonding to Work With eth1
Anna wrote: This is probably far from ideal, but it works for me and I'm putting it out there in case you're messing with XS 0.5 and need to have eth1 working now so you can test other stuff. I'm sure there's a more elegant solution. Put this in /etc/rc.local ifenslave lanbond0 eth1 That's odd, the network init scripts should be doing that ifenslave part, I think we're missing some small option in the ifcfg-eth1 file. Reboot. And there we go: [EMAIL PROTECTED] ~]# ifconfig eth1 eth1 Link encap:Ethernet HWaddr 00:0D:56:05:7C:DA UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1 RX packets:4348 errors:0 dropped:0 overruns:0 frame:0 TX packets:5381 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:368480 (359.8 KiB) TX bytes:6433735 (6.1 MiB) I've connected a few XOs to my test platform so far and haven't seen any unexpected issues. I don't know if you have to edit /etc/modprobe.d/xs_bonding with Jerry's suggestion below, but I've got that in there. options lanbond0 mode=active-backup miimon=100 options mshbond0 mode=active-backup miimon=100 options mshbond1 mode=active-backup miimon=100 options mshbond2 mode=active-backup miimon=100 Of course, you can just run ifenslave lanbond0 eth1, but it doesn't persist on reboot. Could you try adding HOTPLUG=yes to the ifcfg-eth1 file and disable your rc.local routine and retest please. That is only difference that I can spot between your non-working wired and working mesh lan. Jerry ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: Way to tell if it is an XO
On Wed, Dec 10, 2008 at 09:56:39AM -0200, Marcel Renaud wrote: Thanks a lot for your answers. Yes, I think a shared credentials are the best way. Basically we want to offer a service just for the Xos and are working now on the authentication model. We are going to use webservices with WSShttp://www.oasis-open.org/committees/tc_home.php?wg_abbrev=wss and place a signed key on each XO that is going to use the service, to authenticate with the webservice provider. Marcel, Is it important to keep the credential(s) secret? If so: * why? * for how long? * against what attack(s)? * how? * if (when) they leak, what next? Also, what are the incentives for keeping the credentials secret? for publishing them? Regards, Michael ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Multivnc for XOs(children) and Ubuntu box (teacher)
(Somewhat in continuation with the x11vnc and vncviewer discussion thread started here http://lists.laptop.org/pipermail/devel/2008-November/021281.html ) Has anybody tried/gotten multivnc to work ? I started the multivnc server on my Ubuntu box, and then I start multivnc on the client side (on an XO) and it recognizes a multivnc server running in the network and attempts to connect to it, but it exits with the following information displayed. Log attached below. Any help/pointers towards getting this running on XOs, and/or otherwise on a set of 2 desktops, as a first step towards getting this to run on the XOs - would be helpful. The instruction manual available online is in Chinese and I cant seem to find an English version. I will also be pinging to the RealVNC lists too. Thanks in advance, Arjun [EMAIL PROTECTED]:~$ multivnc_client -d 0 clientID(ok click): arjs vncdisplay=0 0 -1 -1 -1 -1 -1 -1 -1 -1 -1 0 recv-adv 40 type=8 length=40 serverID=teacher serverip=10.0.0.5 serverport=5 clientadv.client_ID: arjs type=10, flag=1 switch : rfbAllowAdv vncdisplay=0 vncconfig -display :0 -connect 10.0.0.5:5 /dev/null 21; echo -n $? vncconnect -display :0 10.0.0.5:5 /dev/null 21; echo -n $? vnc4config -display :0 -connect 10.0.0.5:5 /dev/null 21; echo -n $? print_popen : 127 print_popen : 127 print_popen : 127 print_popen : 127 print_popen : 127 print_popen : 127 print_popen : 127 print_popen : 127 print_popen : 127 3 [EMAIL PROTECTED]:~$ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
VGA-output on your XO?
http://arstechnica.com/journals/hardware.ars/2008/12/09/acer-releases-b223-displaylink-lcd-in-europe p. -- Polychronis Ypodimatopoulos Graduate student Viral Communications MIT Media Lab Tel: +1 (617) 459-6058 http://www.mit.edu/~ypod/ ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Way to tell if it is an XO
Hey Michael! I will give you guys some background info on the project The goal of the proyect is to provide some way of communication beteween the children who have an XO and their parents and family. Since most of the parents don't have a computer nor internet access, we determined telephone is the best way to reach them. Some of them don't have cellphones either, but line telephones, most of them do. My partner on this project and I work for a company that does text to speech and SMS2 to Phone Call services, so we are going to use the infrastructure for this project. So we are developing a simple client app for Sugar and the xo, in order to let children write a message and send it to their parents. The message will be translated to speech and, and the parents will recieve it in a phone call on their home line or mobile device. The message will be sent using web services and the transport will be SOAP over HTTP although the goal of the project was to base the authentication system on XMPP and also transport SOAP over xmpp. But for the prototype we aren't going to use XMPP. We plan to offer this service just for the children and the security and authentication concerns are not to let anybody else use the service. The only threat or attacks to the service is someone pretending to be a children with an XO to send messages since it is not free of charge for everyone, that's why we tought of WSS to probe authenticity, integrity and also that the message was originated by the sender. There is no threat on the XO's end to my knowledge since it is a one way only communication system ( Actually we did a reaserch for University and came up with XMPP using Jabber server to be the best way to achieve bidirectional communication) but this is a far bigger project because we need Jabber servers and authentication schemes to be used globbally here in Uruguay and that is far from happening I think.) Finnally, this project is only a prototype and it will be very difficult to deploy nation wide since there are commercial issues to settle. Anyways, hope I have been clear and thanks everyone for the support. Marcel Renaud On Wed, Dec 10, 2008 at 4:26 PM, Michael Stone [EMAIL PROTECTED] wrote: On Wed, Dec 10, 2008 at 09:56:39AM -0200, Marcel Renaud wrote: Thanks a lot for your answers. Yes, I think a shared credentials are the best way. Basically we want to offer a service just for the Xos and are working now on the authentication model. We are going to use webservices with WSShttp://www.oasis-open.org/committees/tc_home.php?wg_abbrev=wss and place a signed key on each XO that is going to use the service, to authenticate with the webservice provider. Marcel, Is it important to keep the credential(s) secret? If so: * why? * for how long? * against what attack(s)? * how? * if (when) they leak, what next? Also, what are the incentives for keeping the credentials secret? for publishing them? Regards, Michael ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: WPA-PEAP with MSCHAPv2
(to dig this back up) I have a similar situation here at my university - WPA2 enterprise w/ PEAP/MSCHAPv2. My Macbook Pro with F9 couldn't connect, neither can my XO (running F9, DebXO, or 8.2), but my Macbook Pro with F10 connects fine. I assume this is due to Network Manager .7 in F10. Any word on how NM .7 integration into sugar is coming, or what is left to be done? http://wiki.laptop.org/go/Network_manager_0.7 is pretty out of date it seems (or maybe nothing has really happened lately) yours, bobby On Tue, Sep 30, 2008 at 10:37 PM, Michael Stone [EMAIL PROTECTED] wrote: On Tue, Sep 30, 2008 at 01:01:11AM -0700, pato wrote: Pato, We are using a Cisco RADIUS server that implements WPA2 enterprise and WPA-PEAP with MSCHAPv2. The SSID is also hidden. Glad to hear from you -- PEAP/802.11i/802.1X and hidden SSIDs are not yet supported but will probably be supported for the next major release. If you're feeling impatient, then you can help get this feature working faster by helping to test Sugar as we port it to more recent versions of NetworkManager and as we implement support for negotiating network use authorization. (If you're a programmer, then you can also help with the implementation of these tasks.) Regards, Michael ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Server-devel] Amateurish Workaround to Get Bonding to Work With eth1
In my tests the following works on a fresh installation without modifying the xs_bonding: Put this in /etc/rc.local ifenslave lanbond0 eth1 Reboot. And there we go: Anna wrote: This is probably far from ideal, but it works for me and I'm putting it out there in case you're messing with XS 0.5 and need to have eth1 working now so you can test other stuff. I'm sure there's a more elegant solution. Put this in /etc/rc.local ifenslave lanbond0 eth1 Reboot. And there we go: [EMAIL PROTECTED] ~]# ifconfig eth1 eth1 Link encap:Ethernet HWaddr 00:0D:56:05:7C:DA UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1 RX packets:4348 errors:0 dropped:0 overruns:0 frame:0 TX packets:5381 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:368480 (359.8 KiB) TX bytes:6433735 (6.1 MiB) I've connected a few XOs to my test platform so far and haven't seen any unexpected issues. I don't know if you have to edit /etc/modprobe.d/xs_bonding with Jerry's suggestion below, but I've got that in there. options lanbond0 mode=active-backup miimon=100 options mshbond0 mode=active-backup miimon=100 options mshbond1 mode=active-backup miimon=100 options mshbond2 mode=active-backup miimon=100 Of course, you can just run ifenslave lanbond0 eth1, but it doesn't persist on reboot. Anna Schoolfield Birmingham ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
9.1.0 Weekly Meeting minutes
Ed, Michael, Marco, Chris, Paul, Joe, Greg and possible others met on IRC December 10th for 9.1 weekly meeting (formerly called roadmap meeting). - XO Camp We updated the goals for the camp. See: http://wiki.laptop.org/go/XOcamp_2 We discussed the need to focus on 9.1.0 features. Action item: Greg to edit the camp proposals to focus on top 9.1 priorities and remove most other items. - 9.1 Schedule and Feature Roadmap Greg noted that a new Semantic format for the feature roadmap will be put in place ASAP. See the new format at: http://wiki.laptop.org/go/Features-test Its locked for editing right now but once the new format is in place we can quickly tag the main features which must be in 9.1. Also noted that we will tag other features which people can work on and we will tie this page in to Trac and to the 9.1 page. The schedule was mentioned but it needs updating. http://wiki.laptop.org/go/9.1.0#Schedule Action item: Joe to read schedule and update as needed by next meeting: Wed. 12/17. Action item: Ed to read schedule and update as needed by next meeting: Wed. 12/17 - Trac bug scrub We discussed the need to agree on Trac conventions. Ed is the release manager and its up to him in the end. Action item - Ed to set a date for start of Trac scrub. He will tell us that date by Friday 12/12. - General There were some complaints that we don't have a plan or that we are spending too much time talking about stuff instead of working. There were also complaints that we don't reach consensus on a subject before moving to the next one. There was no consensus on the agenda but people wanted to make the meeting longer so that everyone's opinion could be heard. Just noting that for the record. My proposed agenda is below. Please comment and ask for changes or additions or subtractions as needed. Next meeting Wed 12/17 2PM US ET IRC Freenode.net #olpc-meeting Agenda 30 minutes - XO Camp review and status. 45 minutes - 9.1.0 status. Schedule review, status of major features, page updates and roadmap scrub status. 30 minutes - Trac scrub and conventions review. 15 minutes - set agenda for next week and review action items. Thanks, Greg S ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [PATCH] cafe_nand: remove busy-wait loop
So, the transfer rate dropped from about 9.6mb/sec to about 5.3mb/sec. I was afraid of this. Flash is pretty fast, and there's almost no parallelism available in it (e.g. you can't queue up a bunch of reads to occur when the data rotates under the heads, or start a seek early). Thus it's mostly worth sitting and waiting for the chip to finish, rather than making the CPU try to do something else. It all happens at chip speeds -- and OLPC designed the CaFE chip to run as fast as the flash chips and memory bus can run (because the Geode's native flash support, for some reason, didn't go that fast). [Note, however, that the Marvell spec for the chip doesn't give ANY speed specs. It does say it might be possible to run its PCI bus at 66 MHz rather than 33 -- which do we do?] I really think we need to separate the effects of filesystem compression from the effects of Flash storage. Other reports show systems being able to do many thousands of transactions per second to SSD (flash-based disks), but we don't come anywhere close to that. Why don't we? Nobody knows. There is a trivial way to make an uncompressed NAND image. Take an existing build's tar image, extract it to a directory, use the same mkfs.jffs2 command from the standard build log, but add --compression-mode none to build an uncompressed image. Burning that into NAND will let you test uncompressed reads. That will give you best-case reads, too, since all the data for a given file will be contiguous. Perhaps there is some sysfs knob that will also let you turn off compressed writes. It's unfortunate that there's no block device way to access Flash chips in Linux without using a filesystem at all. It's much easier to measure and tune I/O performance without a filesystem first, then see what (filesystem or driver) optimizations are required to make a particular filesystem fast on that device. John PS: During the long delay while the CaFe is erasing a block, it can do reads from other blocks (but not this one); do we do these in parallel? ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Bundling plugins with Browse
Sayamindu Dasgupta wrote: On Wed, Dec 10, 2008 at 3:47 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 22:29, Sayamindu Dasgupta wrote: On Wed, Dec 10, 2008 at 12:37 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 18:49, Sayamindu Dasgupta wrote: http://dev.laptop.org/~sayamindu/Browse-99.xo seems to be working for me. Please test. Attached is the patch which did the trick. Ugh. Why is it necessary to copy the plugin to $SUGAR_ACTIVITY_ROOT/ data ? That would only make sense if it is supposed to be modified at runtime. Do stuff placed in $SUGAR_ACTIVITY_ROOT remain persistent across reboots ? I seem to have troubles getting that (the file disappeared after reboot) You misunderstood. Why copy at all? Why not run it from the bundle? Copying from the bundle is only necessary for e.g. default configuration files that the activity needs to be able to change later (because the bundle itself is read-only). Ah, ok, I was blindly following cert8.db http://dev.laptop.org/~sayamindu/Browse-101.xo follows your solution and seems to work fine. Moreover, theoretically one can drop any plugin in the plugin directory of Browse, and expect it to work. Everyone, please test this as much as possible. Also, how do I include the source of the binaries (mozplugger, m4) in the Browse bundle ? Thanks, Sayamindu Does work fine. You can save the pdf once it is loaded. I guess we should add a palette (right mouse click on link) option to save the pdf directly - so you have both options. What do others think? Simon ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [PATCH] cafe_nand: remove busy-wait loop
Hi John, It's unfortunate that there's no block device way to access Flash chips in Linux without using a filesystem at all. It's much easier to measure and tune I/O performance without a filesystem first, then see what (filesystem or driver) optimizations are required to make a particular filesystem fast on that device. Does e.g. dd if=mtd0 of=somefile bs=1M count=20 accomplish this? (When a program expects a block device argument, you can give a raw mtd0 instead.) - Chris. -- Chris Ball [EMAIL PROTECTED] ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Bundling plugins with Browse
On Thu, Dec 11, 2008 at 2:08 AM, Simon Schampijer [EMAIL PROTECTED] wrote: Sayamindu Dasgupta wrote: On Wed, Dec 10, 2008 at 3:47 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 22:29, Sayamindu Dasgupta wrote: On Wed, Dec 10, 2008 at 12:37 AM, Bert Freudenberg [EMAIL PROTECTED] wrote: On 09.12.2008, at 18:49, Sayamindu Dasgupta wrote: http://dev.laptop.org/~sayamindu/Browse-99.xo seems to be working for me. Please test. Attached is the patch which did the trick. Ugh. Why is it necessary to copy the plugin to $SUGAR_ACTIVITY_ROOT/ data ? That would only make sense if it is supposed to be modified at runtime. Do stuff placed in $SUGAR_ACTIVITY_ROOT remain persistent across reboots ? I seem to have troubles getting that (the file disappeared after reboot) You misunderstood. Why copy at all? Why not run it from the bundle? Copying from the bundle is only necessary for e.g. default configuration files that the activity needs to be able to change later (because the bundle itself is read-only). Ah, ok, I was blindly following cert8.db http://dev.laptop.org/~sayamindu/Browse-101.xo follows your solution and seems to work fine. Moreover, theoretically one can drop any plugin in the plugin directory of Browse, and expect it to work. Everyone, please test this as much as possible. Also, how do I include the source of the binaries (mozplugger, m4) in the Browse bundle ? Thanks, Sayamindu Does work fine. You can save the pdf once it is loaded. I guess we should add a palette (right mouse click on link) option to save the pdf directly - so you have both options. Agreed - it would make sense to let users right click and save the PDF. -sdg- -- Sayamindu Dasgupta [http://sayamindu.randomink.org/ramblings] ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Wine activity
That's awesome work! I was able to install Wine and use it, including firefox and a win32 application I had previously build using mingw32 under Linux on another PC and uploaded to a webserver, and then downloaded using firefox inside wine. However, I did notice the following oddities: 1. When I later resumed the activity from the journal, the wallpaper was gone and nothing worked, although the start-menu items for firefox were still there. 2. It was not clear to me how to save wine's state to the journal. 3. At some point the usual 'leave full-screen mode' icon appeared in the upper-right corner, but clicking it seemed to have no effect other than to make it disappear, i.e. no sugar UI appeared and the desktop size did not change. 4. Wine crashed when I used Firefox's download manager to open the location of a downloaded file (winefile appeared briefly, then the whole activity crashed.) I have no idea why yet, but perhaps there is some information left in a log file somewhere I will find. On the bright side, this means it's fairly trivial to run at least some windows-only software on the OLPC now, which is great when there's not yet a Sugar or Linux version. -Ben On Tue, Dec 9, 2008 at 12:02 PM, Vincent Povirk [EMAIL PROTECTED] wrote: The Wine activity has advanced to the point where I think it's ready for testing by actual users. The current package, development history, and my todo list are at http://wiki.winehq.org/SugaredWine The intent of this project is to provide a shell that can be used to run Windows programs using Wine in the Sugar environment. It should be good enough that someone used to Windows can grab and install a Windows program without help, once the activity is installed. Ideally, the installer and software will both work fine in Wine and within the hardware limitations of an XO. In this ideal case, someone used to Windows should be able to operate it without help. If it does not live up to this ideal for platinum software (according to the Wine appdb) whose hardware requirements the XO meets, I want to know about it and hopefully fix it. Wine bugs and hardware limitations mean a lot of Windows programs won't work or won't work properly. On Linux, one can often push the compatibility much further than what works out of the box by looking at console messages (the log viewer works for this) and tweaking Wine. Don't expect everything to work perfectly, but don't give up if it doesn't. This is normal, even on Linux. Winehq.org has support channels for such cases (appdb, bugzilla, mailing lists, and the winehq irc channel). Most of the people there probably don't know anything about Sugared Wine, but collectively they should know more than I do about making Wine work in general. If a program doesn't work for you, you can go to any of those places for support. You can also email [EMAIL PROTECTED] That goes directly to me for now, but in the future (maybe the very near future) I may decide to send it somewhere public, like a mailing list, instead. Wine and the code that I developed for this project are licensed under the GNU LGPL. The entire package isn't quite LGPL because I included 7-zip. 7-zip is LGPL + unRAR restriction (you're not allowed to use the source code to create a RAR compressor). If you have a program that works well in this Wine package and would like to package it as a stand-alone .xo, please let me know. I already did most of the work for this so that I could include 7-zip and a firefox downloader/installer (and I could probably have included firefox itself if not for the fact that it would require uploading non-open-source code to repo.or.cz). Vincent Povirk ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Yay!, Bee, See (ABC) software
Hi, I've been playing around with this a bit, and I still can't figure out the xol files. When I download http://wiki.laptop.org/images/2/28/Yay-Bee-See-9.xol In Browse, it does get saved to the Journal, and when I start the xol file from the Journal it launches Browse with the main HTML file from the collection, and a subsequently launched vanilla Browse includes yay-bee-see in the images section of the Library. However, even after I keep both the .xol file and the Browse session, rebooting the machine causes yay-bee-see to disappear from the images section of the Library (and the kept Browse session to show a File Not Found message) until I open the .xol file again. Is this intended/expected behavior? Is there some way to keep user-installed Library Collections installed across reboots? Thanks, -Ben On Mon, Nov 24, 2008 at 9:08 PM, Samuel Klein [EMAIL PROTECTED] wrote: Yes, that's a fine baseline. As you point out, I had a hard time with the license field; enter what you like but please do include a full LICENSE file in the bundle that provides specific licenses (and attribution where required), image by image. If you download an xol file onto your xo from a webserver that has mimetypes set properly (such as w.l.o) it should automatically install itself into your Library/ directory. I don't know about that page not rendering properly on an XO; what version of Browse are you running? SJ On Mon, Nov 24, 2008 at 11:55 PM, Ben Wiley Sittler [EMAIL PROTECTED] wrote: Thanks! A few questions, though: 1. Is there any reason I shouldn't start with your version 2 .xol as my baseline? I'd like to update it to use the new lower-resolution, lower-quality images (which still look just fine on the XO-1 even in greyscale high-resolution mode zoomed out to the 1px = 1px scale.) 2. Is there some way to install the .xol more user-friendly than just unzipping it into the ~/Library directory? 3. I notice that in the description on the wiki for the bundle you wrote fdl text, pd, cc-by and cc-sa images. Some of the images are cc-by-sa and fdl, too. Also, the HTML text is actually pd (or at least it was in the version I released — of course you are welcome to license copyrighted derivative versions however you like.) 4. And finally, is there some reason the OLPC wiki does not work right when viewed from an XO-1? I had to go through URL-hacking contortions to open that page in Browse (it just said the page was empty otherwise.) Thanks, (and please pardon my ignorance!) -Ben On Mon, Nov 24, 2008 at 8:43 PM, Samuel Klein [EMAIL PROTECTED] wrote: Ben -- When you're zipping up the directory, if you add a metadata file in this subpath: library/library.info and give the resulting zip file the extension .xol, you'll have an XO library bundle. Here is a sample info file, with all required fields : http://wiki.laptop.org/go/Image:Yay-bee-see-library.info Note that the 'name' field in the info file should match the name of the root directory. Our standard is to increment the version # in the metadata every time you make a change; that allows tools like Sugar's software updater know when there are newer versions of packages available to install. SJ On Mon, Nov 24, 2008 at 7:31 PM, Ben Wiley Sittler [EMAIL PROTECTED] wrote: yeah, i added a 1200x900 version with more agressive JPEG compression which looks good both in color mode and in monochrome mode and is only 4 MiB or so: http://xent.com/~bsittler/yay-bee-see-olpc.zip hosted version: http://xent.com/~bsittler/yay-bee-see-olpc/index.html does that seem any faster? On Mon, Nov 24, 2008 at 3:26 PM, Gary C Martin [EMAIL PROTECTED] wrote: On 24 Nov 2008, at 17:21, Ben Wiley Sittler wrote: Hi, I have just joined this list and read through the archives, but could not find anything similar. I also didn't find mention of anything similar on the OLPC Wiki. I recently wrote some software for use by my daughter on her OLPC. It runs inside the Browse activity, either locally using a file: URI or over the network. I don't know whether it will be of interest to anyone else, but I have released the software to the public domain and packaged it along with scaled-down (1600x1200 or less) copies of some public-domain images and some copyrighted-but-free-to-redistribute images under GFDL, and various Creative Commons Attribution-Share Alike, Attribution, and Share Alike licenses. Individual attribution for each image is included in the application source code. Seems a great addition for the younger age range :-) I did notice that even on a high specced laptop (1.5Ghz, 2Gb ram, broadband connection) the background image was very slow to display (until it had been cached locally). One suggestion, 1600x1200 seems a bit large (even as a max size). For the XO, 800x600 (max!) would seem to be a fair max image size to save nand space and keep image quality. The XO screen is capable of 1200x900
Re: [PATCH] cafe_nand: remove busy-wait loop
On Wed, 2008-12-10 at 15:36 -0500, Chris Ball wrote: Hi John, It's unfortunate that there's no block device way to access Flash chips in Linux without using a filesystem at all. It's much easier to measure and tune I/O performance without a filesystem first, then see what (filesystem or driver) optimizations are required to make a particular filesystem fast on that device. What's wrong with just reading from the character device /dev/mtd0? Does e.g. dd if=mtd0 of=somefile bs=1M count=20 accomplish this? (When a program expects a block device argument, you can give a raw mtd0 instead.) No, that's only for mount. -- David WoodhouseOpen Source Technology Centre [EMAIL PROTECTED] Intel Corporation ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Way to tell if it is an XO
On Tue, 2008-12-09 at 16:37 -0500, Chris Ball wrote: Hello, I need to know what is a good way to find out if the local machine running a python script is a XO. One way might be to cat /ofw/model. If the file doesn't exist, or doesn't return a letter and number, you probably aren't running on an XO. Note that this is not an upstream'd kernel interface and thus really shouldn't be being depended on... Jeremy ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
New joyride build 2587
http://xs-dev.laptop.org/~cscott/olpc/streams/joyride/build2587 Changes in build 2587 from build: 2583 Size delta: 0.00M -dbus-x11 1.2.6-1.fc10 +dbus-x11 1.2.8-1.fc10 -cups 1:1.3.9-4.fc10 +cups 1:1.3.9-6.fc10 -cups-libs 1:1.3.9-4.fc10 +cups-libs 1:1.3.9-6.fc10 -dbus 1.2.6-1.fc10 +dbus 1.2.8-1.fc10 -dbus-libs 1.2.6-1.fc10 +dbus-libs 1.2.8-1.fc10 -libX11 1.1.4-5.fc10 +libX11 1.1.4-6.fc10 -libraw1394 2.0.0-4.fc10 +libraw1394 2.0.0-5.fc10 -- This mail was automatically generated See http://dev.laptop.org/~rwh/announcer/joyride-pkgs.html for aggregate logs See http://dev.laptop.org/~rwh/announcer/joyride_vs_update1.html for a comparison ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: New joyride build 2587
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Build Announcer v2 wrote: -libraw1394 2.0.0-4.fc10 +libraw1394 2.0.0-5.fc10 For all those XO's with firewire ports... (cups was also on the list. I don't know if that's deliberate or not.) - --Ben -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) iEYEARECAAYFAklAY8wACgkQUJT6e6HFtqTMogCcCLJ8EPQA9yZX9jccti3XBTsw 5uoAoIDrfoiXp0vTskBmKESF7dbnnlU2 =heEB -END PGP SIGNATURE- ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: VGA-output on your XO?
Polychronis Ypodimatopoulos wrote: http://arstechnica.com/journals/hardware.ars/2008/12/09/acer-releases-b223-displaylink-lcd-in-europe From http://en.wikipedia.org/wiki/DisplayLink and http://www.displaylink.com/downloads/publicfaq.shtml DisplayLink does not currently support the Linux operating system. DisplayLink is investigating support for other operating systems and will provide updates to this article when more information is available. You can use http://www.displaylink.com/contact.html to beg them to support their proprietary protocol in Linux. I added DisplayLink to http://wiki.laptop.org/go/Remote_display , with a disclaimer. -- =S ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Wireless activation improvements.
Ed, Kim, Emiliano, and Guadalupe, As described in http://dev.laptop.org/ticket/8976, I've succeeded in repairing wifi-activation to my satisfaction. However, in its present form, wifi-activation only makes use of the mesh and does not attempt to acquire leases over any 802.11b/g networks. Clearly, the wifi-activation technology would be more widely deployable if it could be used with access points as well as with the mesh. Therefore: do we have any interest in developing such a feature? Michael (Previous conversations with Emiliano and Guadalupe suggested to me that LATU would be pleased if we did so.) (Alternately, is LATU interested in developing such a feature with our aid, thereby expanding its skills in this area and potentially gaining access to the feature more quickly?) ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
New joyride build 2588
http://xs-dev.laptop.org/~cscott/olpc/streams/joyride/build2588 Changes in build 2588 from build: 2587 Size delta: 0.00M -kernel 2.6.27-20081201.1.olpc.672cde9409f412e +kernel 2.6.27-20081210.1.olpc.05aa2d840dc7b96 -- This mail was automatically generated See http://dev.laptop.org/~rwh/announcer/joyride-pkgs.html for aggregate logs See http://dev.laptop.org/~rwh/announcer/joyride_vs_update1.html for a comparison ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Wacom Bamboo with XO?
Wade Brainerd wrote: The only catch is that the drawing output lags the tablet strokes by quite a bit. My guess is that the event processing cannot keep up with the data rate. I seem to remember that there was a driver configuration related to that which might be used to throttle down the fire hydrant. The mouse events go at a higher priority than the expose events, so the screen update waits until you stop moving if it gets behind. But, at the end your entire stroke should appear. I would be concerned about reducing the update rate, it might lead to less accurate drawing. It is true that reducing the update rate would lower the pen position resolution of lines...but the Bamboo tablet has a linear resolution of 2400+ lines per inch which is about 12X the XO pixel resolution and 24X your 2X reduced drawing canvas. As a result, even with skipping pen deltas of 24 or less the pen output events will still be at the finest drawing precision. This is not even including the fact that the brush is more than a single pixel wide typically. If you add an entry under the first wacom Xinput entry in your xorg-dcon.conf file like: Option Suppress 60 you'll get much better drawing response with the existing infrastructure. I tried 10-72 and to keep up with faster sketching or larger brushes, the rate needs to be at 50 or greater. After the above edit and a 3-finger-salute to restart the X server on the XO, Colors! basically worked well with the tablet. There is a command line configuration program (xsetwacom) that could be used to configure Wacom tablet settings without editing the xorg-dcon.conf file. I don't know if a restart of X would then be required. It might be possible to base a control panel interface calling the underlying utility per Tomeu Vizoso's previous mention in this thread. To improve beyond the Suppress technique will require improved event handling optimization in colors.py. My guess is that the best approach will be to encapsulate tablet specific handling at a lower level for efficiency. I've got some optimizations planned to the C module which should make it better, and I think there are some bottenecks in the X server that slow things down and could be fixed, but my recommendation for now is to draw slower :) Cool! BTW, Colors! version 11 did not appear to have the photo snap capability. Was that removed? See Nirav's response. I'm also planning to add Paste support so you can just snap the photo in Record (or whatever) and paste it into the Reference image, then paint over that. I tried the earlier photo snap and it was awkward to toggle the underlying photo image. It would be more helpful if the photo could be drawn over as if on tracing paper. The toggle on/off would still be useful at different levels of refinement of the painting. There appeared to be a problem with the Zoom in and out function as zooming all the way out, and then back in results in the canvas offset by various amounts. I was not able to make it shift back without restarting the Activity. Finally, I'm not sure how to handle this but it was difficult to trigger the Frame exposure via the tablet since the drawing area seems to be exactly the entire screen. If it were to be extended a bit at the top then the Frame would be easier to get to from the tablet only. I've got an old serial Intuos that I'll try to get configured this weekend. I'll let you know how it goes... If we get this working, maybe Amazon could add a People who bought this item [the XO], also bought this [Wacom Bamboo]... :-) Best regards, Chris ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
2588 - Journal unusable
The Journal entries do not show the object icon - so they cannot be launched from. This SEVERE error has now lasted through five Joyride builds. I'm posting here, rather than write a ticket, as a way to get quickest to those who can do something about it. Or should users who have real XOs (rather than emulators) stop participating ? mikus ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Wacom Bamboo with XO?
On 9 Dec 2008, at 17:42, Wade Brainerd wrote: Hi Stan, Try a 'depmod' and then 'modprobe wacom' from the Terminal, and see if that helps. Feel free to grab me on IRC (as wadeb), I can help troubleshoot it. -Wade Hi Wade, this script worked well for me, and the tablet stays working after a reboot. I also made the xorg.conf changes as suggested by Chris (Option Suppress 60) and it's a very good improvement for XO hardware Before this extra step the canvas would only redraw once you stopped drawing, even with a small brush. --Gary On Tue, Dec 9, 2008 at 9:34 AM, Stanley Sokolow [EMAIL PROTECTED] wrote: I tried the new installation instructions on the wiki, first installing the Colors activity, then running the new shell script as per the wiki: wget http://dev.laptop.org/~wadeb/setupwacom.sh sudo sh setupwacom.sh It seemed to run ok without errors. I restarted the XO. Colors installed ok and works with the mouse, but the Wacom Bamboo tablet does not work, neither in Colors nor anywhere else. Looking at the /dev/input directory, I don't see the wacom device that should be there. When I rerun the shell script, it confirms that the linuxwacom package is already installed.The X11 configuration file contains the Wacom devices, so it's the right conf file. I'm stuck. Has anyone had success with the new instructions? By the way, when I tried the old instructions yesterday, the repomd.xml file was corrupted. (Actually, it was a perl script, not an xml file.) So, yum wouldn't run. Someone apparently fixed that problem.It's an xml file now and yum seems to run ok using it. Stan -- View this message in context: http://n2.nabble.com/Wacom-Bamboo-with-XO--tp1508107p1634938.html Sent from the Software development mailing list archive at Nabble.com. ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: 2588 - Journal unusable
On Thu, 11 Dec 2008, Mikus Grinbergs wrote: The Journal entries do not show the object icon - so they cannot be launched from. This SEVERE error has now lasted through five Joyride builds. I'm posting here, rather than write a ticket, as a way to get quickest to those who can do something about it. Or should users who have real XOs (rather than emulators) stop participating ? we were told to stop participating a dozen or so joyrides ago, they will tell us when we should start testing again. David Lang ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Wacom Bamboo with XO?
Hey Chris, thanks for the feedback! It's great to hear from people using Colors! on XO. On Wed, Dec 10, 2008 at 6:58 PM, Chris Marshall [EMAIL PROTECTED]wrote: It is true that reducing the update rate would lower the pen position resolution of lines...but the Bamboo tablet has a linear resolution of 2400+ lines per inch which is about 12X the XO pixel resolution and 24X your 2X reduced drawing canvas. As a result, even with skipping pen deltas of 24 or less the pen output events will still be at the finest drawing precision. This is not even including the fact that the brush is more than a single pixel wide typically. If you add an entry under the first wacom Xinput entry in your xorg-dcon.conf file like: Option Suppress 60 you'll get much better drawing response with the existing infrastructure. I tried 10-72 and to keep up with faster sketching or larger brushes, the rate needs to be at 50 or greater. After the above edit and a 3-finger-salute to restart the X server on the XO, Colors! basically worked well with the tablet. This is good reasoning, and I wasn't aware of the Supress option - I'm pretty new to Linux programming and X in particular. I had been considering implementing that kind of filter in the PyGTK mouse handler. I've updated the xorg-dcon.conf on dev.laptop.org to include this setting. There is a command line configuration program (xsetwacom) that could be used to configure Wacom tablet settings without editing the xorg-dcon.conf file. I don't know if a restart of X would then be required. It might be possible to base a control panel interface calling the underlying utility per Tomeu Vizoso's previous mention in this thread. Yep, 60 works well enough though that it's probably not worth a configuration option. I will still work on optimizing the painting performance though. I tried the earlier photo snap and it was awkward to toggle the underlying photo image. It would be more helpful if the photo could be drawn over as if on tracing paper. The toggle on/off would still be useful at different levels of refinement of the painting. That's a good idea. When I reimplement Photo Snap, I'll do it that way (so you are painting over it, e.g. the white background will show the image when it's enabled). There appeared to be a problem with the Zoom in and out function as zooming all the way out, and then back in results in the canvas offset by various amounts. I was not able to make it shift back without restarting the Activity. This should be documented better, or maybe have some visual feedback. Zoom in out keep the mouse cursor on the same point in the canvas before and after the zoom operation. This is handy for quick zooms, you just point at the area you want to detail and zoom in. In Git I have added a 'Center image' toolbar button though, since it can be hard to recenter. the image. You can also always scroll the image manually by holding the Hand button and dragging with the mouse. Finally, I'm not sure how to handle this but it was difficult to trigger the Frame exposure via the tablet since the drawing area seems to be exactly the entire screen. If it were to be extended a bit at the top then the Frame would be easier to get to from the tablet only. Maybe one of the Wacom buttons could be assigned to the Frame key event? Can this be done in xorg-dcon.conf? I've got an old serial Intuos that I'll try to get configured this weekend. I'll let you know how it goes... If we get this working, maybe Amazon could add a People who bought this item [the XO], also bought this [Wacom Bamboo]... :-) Great! Thanks again for the feedback, I'll try to release a new version before long with these suggestions. Cheers, Wade ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Wacom Bamboo with XO?
Hi Chris, On 11 Dec 2008, at 02:58, Chris Marshall wrote: There appeared to be a problem with the Zoom in and out function as zooming all the way out, and then back in results in the canvas offset by various amounts. I was not able to make it shift back without restarting the Activity. The hand keys actually do work in Colors! Yay – this is the first activity I know of that uses them, a big round of applause for Wade!! So, you can pan the canvas about like you can in Photoshop when holding the spacebar and dragging. Regards, --Gary P.S. thanks for the Option Suppress 60, makes the Wacom draw in realtime here on the XO :-) ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
New F series firmware release
I've just cut the first release of the F series firmware. New cool nifty OFW feature like NANDblaster, USB cdrom support, and other goodness. But the primary reason is the restructured EC code. Paul Fox did a ton of work re-organizing the EC code and cleaning up lots of cruft plus getting the system working under Wine. Many thanks to him. The only new EC feature is a quick blink before power off. New suspend code turns off the WLAN LEDs so it was really hard to tell if you had just suspended or if you were off. Now if you want to power off just press and hold the power button until you see several quick blinks and then the LED will go dark. You are now powered off. But since lots of code was moved around and various header files merged I expect to find some fallout. Hopefully its not too bad. -- Richard Smith [EMAIL PROTECTED] One Laptop Per Child ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Server-devel] Amateurish Workaround to Get Bonding to Work With eth1
On Wed, Dec 10, 2008 at 2:37 PM, Jerry Vonau [EMAIL PROTECTED] wrote: Anna wrote: ifenslave lanbond0 eth1 That's odd, the network init scripts should be doing that ifenslave part, I think we're missing some small option in the ifcfg-eth1 file. I am starting to agree -- we are either seeing an issue with timing _or_ a configuration issue. Perhaps our config for eth1 is broken. It'd be strange since I tested it quite a bit in the run up to the release, but it's also true that I was making subtle changes to the very same configs. Good news is: I can repro the problem on a machine I have here. Could you try adding HOTPLUG=yes to the ifcfg-eth1 file I spotted exactly the same difference and tested it -- does not seem to work. Hope to get to the bottom of it. (Apologies for the delay in getting into reproduce-and-debug -- I've had a hell of a time getting hardware I could play with for this, as I'm not @ my usual office, city or continent ;-) ) cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] Amateurish Workaround to Get Bonding to Work With eth1
On Wed, Dec 10, 2008 at 6:09 PM, Martin Langhoff I spotted exactly the same difference and tested it -- does not seem to work. Hope to get to the bottom of it. Alright, fixed. Credit to Anna and Jerry for narrowing down on the issue. The actual problem is laughably simple -- late in the dev cycle of 0.5 a typo sneaked in. A minor edit of ifcfg-eth1 fixes it, see: http://dev.laptop.org/git?p=projects/xs-config;a=commitdiff;h=acd64ab3d2342fbda08a944e31878db6b3b563f2 In any case, you can grab the rpm with the fix from http://xs-dev.laptop.org/xsrepos/testing/olpc/9/i386/xs-config-0.5.7.g11aaacf-1.noarch.rpm or perform yum --enablerepo=olpcxstesting install xs-config thanks everyone -- specially Anna -- for you help and patience. cheers, martin -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
2008/12/5 Reuben K. Caron [EMAIL PROTECTED]: After doing a clean install 4 times and dhcpd not starting Thanks for the dedication and the report - I managed to repro the issue as you describe below. ... When I cd /etc/sysconfig/olpc-scripts ./domain_config xs5.org It creates the xs_domain_name file in /etc/sysconfig/olpc-scripts/ that's the wrong landing place, as you recognised. The current working directory should not affect us. The patch is here: http://dev.laptop.org/git?p=projects/xs-config;a=commitdiff;h=ae030b75d85fa46962134424667dfa6fb7eba5cf and there's an xs-config 0.5.8-biglonghash.rpm on its way to the xs-dev server. Not quite in the same league as the bonding issue, but a good thing to spot and fix. cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel