FWD: Bonjour Problem
---BeginMessage--- -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi Guys, I tried to setup an Bonjour Chat with X0 Build 695 running in a VM on my Mac and my Bonjour account on my mac. The clients see each other but are not able to establish a communication via Chat. Do you have any clue why we see each other but are not abel to talk? Thanks a lot Kind Regards Daniel Taschik - -- mobil: +49 162 7604429 mail: [EMAIL PROTECTED] http: www.the-nana.de icq: 106018852 * msn: [EMAIL PROTECTED] aim: dtaschik * jabber: [EMAIL PROTECTED] skype: danieltaschik -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.8 (Darwin) iEYEARECAAYFAkfKdbUACgkQo9checNKopSgwwCfRe/J/UK6OfV906w+sZ6eFBky xqIAnilvDCAkRO4QExypMIrMlABY1hRZ =A0NE -END PGP SIGNATURE- ---End Message--- ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Bonjour Problem
Hi Daniel, I tried to setup an Bonjour Chat with X0 Build 695 running in a VM on my Mac and my Bonjour account on my mac. The clients see each other but are not able to establish a communication via Chat. Do you have any clue why we see each other but are not abel to talk? Yea, now I'm no expert, but zeroconf (Bonjour on the Mac) is used to advertise device names so both Mac and XO will see each other (very friendly), however the actual chat protocol/mechanics currently used by the XO are not the same as used by iChat – which is a pity but hopefully/maybe fixable at some point (sure I've seen some dev talk about this somewhere). I'm not sure if the actual chat protocol being used by either device is a formal standard, I hope it is, but I guess this particular compatibility would not be high priority other than for G1G1 owners who are likely to meet a Mac or two on their networks. ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: storing Activity parameters
I'm having some indecision about my own (very simple) activity state saving. Just had a though that may add another possibility to this thread. What if both journal and file-system were used to store the activity state, with the journal settings overriding the file-system settings. This way a new activity start-up could inherit the last used activity settings, and a specific journal start-up would inherit the settings used for that specific entry. So in the case of Speak I could have separate favourite journal entries for '3 eyed alien' and another for 'Mr square eyes', then if I also just started the activity from fresh I'd pick-up whatever the last used setting were (picked up from the FS). My Moon activity isn't complicated enough (just 2 key/value pairs so far) to make this much of a time saver, but for my planned Earth activity (a port of EarthGlobe, a Mac desktop app I wrote a few years back), I'd want to make use of shared location tagging (basically lon/ lat markers attached to some metadata) and some more detailed custom user settings that a user would not want to make each time. So the more detailed user settings (default location, some basic info/notes, viewing preferences) would be best going to the FS, and any tagged locations would go to the journal (so you could have different sessions for making different sets of geotags). I guess in this Earth case though there is a more clear separation between settings, and activity data. You'd (likely) want just one set of global settings for your details, while using the journal to store separate sets of tags - though I guess if you change location frequently you might want a journal entry for each location* Hm *probably a low usage case for our target audience. On 2 Mar 2008, at 01:29, Joshua Minor wrote: I implemented the save/load feature of Speak without fully understanding the other options. Now that I've seen the recent discussion about data vs instance vs the journal I think it would make more sense to have Speak save its state in a different way. On the other hand, the new frame redesign makes it much easier to resume an Activity, which would mean that parameters saved to the Journal, like Speak does now, would naturally be restored when you resume the Activity. A nice best practices document would be very handy. -josh ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Serious problems with rescan networks on wakeup feature.
Hi. I just upgraded to joyride-1738, and the new feature that rescans for networks when the laptop wakes up is causing a lot of trouble for me. I often leave long-running connections such as IMAP or IRC open. Before the change, the laptop would wake up, the card would reassociate, and there was a good chance that my connection would still be up. After the change, the connections are almost always dropped, even if I just turned the laptop off for a few seconds. What would be *great* is if it waited 5 or 10 seconds, tried the existing connection, and then restarted scanning *if* it wasn't working anymore. The lag isn't that bad, and it makes the experience where you just closed the laptop a minute ago and are still on the same network better. If there's anything I can do to help with this, please let me know! Cheers, Thomas Tuttle ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Today's mesh testing.
Hi, Daf and I got the school server jabberd/shared roster working today. We connected/registered 32 laptops to it with mesh TTL set to 1 for broadcast, and they were all able to see and join a shared chat session with each other. The workload on the spectrum analyzer increased from 18% (no-one connected) to 26% (all connected). The chat session is consistent -- no-one is dropping out and new messages are seen by each laptop, with a few seconds of lag. With the mass chat session still running, we shared a 500KiB PDF. First we joined the shared Read session with one laptop, and the download took 16 seconds to complete. We then joined two more laptops at once, the first download took 26 seconds and the second finished at 30 seconds. Five more at once: all finished around 1m00s. Ten more at once: the first finished at 2m18s, the last finished at 2m40s. There were no failures downloading the PDF. The sharing was unicast TCP, with mesh TTL set to 1, which explains the slightly worse than linear increase in download time for more laptops downloading at once. This is much more anecdotal than the full test plan, but we thought the testers currently in Peru would want to know what they can expect from the school server setup ASAP. We don't have more laptops upgraded and ready to join the network yet, but we don't have any reason to believe we've saturated the network -- with the PDFs downloaded and Chat still running, the duty cycle on the spectrum analyzer is now at 28%. (In general, wireless networks seem to start degrading around 40%.) - Chris and Daf. -- Chris Ball [EMAIL PROTECTED] ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Today's mesh testing.
Thanks for the info ! This is good news, as it means that schools up to a hundred students should work right now, given a school server and three active antennas... wad On Mar 2, 2008, at 7:05 PM, Chris Ball wrote: Hi, Daf and I got the school server jabberd/shared roster working today. We connected/registered 32 laptops to it with mesh TTL set to 1 for broadcast, and they were all able to see and join a shared chat session with each other. The workload on the spectrum analyzer increased from 18% (no-one connected) to 26% (all connected). The chat session is consistent -- no-one is dropping out and new messages are seen by each laptop, with a few seconds of lag. With the mass chat session still running, we shared a 500KiB PDF. First we joined the shared Read session with one laptop, and the download took 16 seconds to complete. We then joined two more laptops at once, the first download took 26 seconds and the second finished at 30 seconds. Five more at once: all finished around 1m00s. Ten more at once: the first finished at 2m18s, the last finished at 2m40s. There were no failures downloading the PDF. The sharing was unicast TCP, with mesh TTL set to 1, which explains the slightly worse than linear increase in download time for more laptops downloading at once. This is much more anecdotal than the full test plan, but we thought the testers currently in Peru would want to know what they can expect from the school server setup ASAP. We don't have more laptops upgraded and ready to join the network yet, but we don't have any reason to believe we've saturated the network -- with the PDFs downloaded and Chat still running, the duty cycle on the spectrum analyzer is now at 28%. (In general, wireless networks seem to start degrading around 40%.) - Chris and Daf. -- Chris Ball [EMAIL PROTECTED] ___ 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: questions about using Wireshark to monitor the Mesh
Thanks john, I will try out the commands you sent. where can I download the newest build of the XS? I can't find it at xs-dev.laptop.org or dev.laptop.org We're hoping to get a working version of the XS by the end of this week. Our pilot starts soon (April 13) and we want to get the XS working sooner rather than later. May I suggest bundling XFCE4 w/ the XS. I have installed it on my XS and it eases working w/ wireshark and other aspects of testing. At least for folks like myself that are less than expert w/ the command line. thanks again for your help John Bryan Kathmandu On Sun, 2008-03-02 at 10:52 -0500, John Watlington wrote: Those RPMs are already patched. What the patches allow are: support for mesh link layer messages (RREQ,PREQ, RREPLY, etc.) and decoding our new non-standard mesh multicast packets. That version doesn't dig into the telepathy packets. I have a patch from collabora that should do that, but haven't applied and tested it yet. I'll get it out ASAP (the patch is attached). IPv6 is turned off on recent school server builds.It breaks installations with more than a single school server --- see the trac ticket for details (sorry no number, I'm offline). mDNS is shown fine by the patched version, but should be turned off in a school server environment. In order to see all frames (and not just those containing IP packets), you have to bring up a special interface on the mesh driver (bringing down the regular one.) On servers with one wired ethernet interface, type: ifconfig eth1 down ifconfig msh0 down On servers with two wired ethernet interfaces, type: ifconfig eth2 down ifconfig msh0 down Then, on all types of servers, type: echo 7 /sys/class/net/eth2/lbs_rtap ifconfig rtap0 up Now point wireshark at rtap0 instead of msh0 to see more packets. The number echoed into lbs_rtap is a bit field indicating which frame types you want to see. I believe this is documented at http://wiki.laptop.org/go/Wireless Cheers, wad ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Today's mesh testing.
Really is good news. Something we can work from. -walter On Sun, Mar 2, 2008 at 10:13 PM, John Watlington [EMAIL PROTECTED] wrote: Thanks for the info ! This is good news, as it means that schools up to a hundred students should work right now, given a school server and three active antennas... wad On Mar 2, 2008, at 7:05 PM, Chris Ball wrote: Hi, Daf and I got the school server jabberd/shared roster working today. We connected/registered 32 laptops to it with mesh TTL set to 1 for broadcast, and they were all able to see and join a shared chat session with each other. The workload on the spectrum analyzer increased from 18% (no-one connected) to 26% (all connected). The chat session is consistent -- no-one is dropping out and new messages are seen by each laptop, with a few seconds of lag. With the mass chat session still running, we shared a 500KiB PDF. First we joined the shared Read session with one laptop, and the download took 16 seconds to complete. We then joined two more laptops at once, the first download took 26 seconds and the second finished at 30 seconds. Five more at once: all finished around 1m00s. Ten more at once: the first finished at 2m18s, the last finished at 2m40s. There were no failures downloading the PDF. The sharing was unicast TCP, with mesh TTL set to 1, which explains the slightly worse than linear increase in download time for more laptops downloading at once. This is much more anecdotal than the full test plan, but we thought the testers currently in Peru would want to know what they can expect from the school server setup ASAP. We don't have more laptops upgraded and ready to join the network yet, but we don't have any reason to believe we've saturated the network -- with the PDFs downloaded and Chat still running, the duty cycle on the spectrum analyzer is now at 28%. (In general, wireless networks seem to start degrading around 40%.) - Chris and Daf. -- Chris Ball [EMAIL PROTECTED] ___ 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 -- Walter Bender One Laptop per Child http://laptop.org ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Serious problems with rescan networks on wakeup feature.
On Mar 2, 2008, at 7:49 PM, Thomas Tuttle wrote: Hi. I just upgraded to joyride-1738, and the new feature that rescans for networks when the laptop wakes up is causing a lot of trouble for me. I often leave long-running connections such as IMAP or IRC open. Before the change, the laptop would wake up, the card would reassociate, and there was a good chance that my connection would still be up. After the change, the connections are almost always dropped, even if I just turned the laptop off for a few seconds. What would be *great* is if it waited 5 or 10 seconds, tried the existing connection, and then restarted scanning *if* it wasn't working anymore. The lag isn't that bad, and it makes the experience where you just closed the laptop a minute ago and are still on the same network better. The problem we have is the following: A student is using the laptop away from school/infrastructure, and is in simple mesh mode. In this mode, all service discovery and collboration is multicast.The student puts their computer to sleep (by closing the lid) and goes to school. Once they arrive at school, the last thing we want is for their laptop to try to use simple mesh --- it trashes spectrum and makes the school network not work, plus they won't see any of their friends that are (properly) connected through the school presence service. If we naively followed your suggestion above, the laptop would of course discover that the previous network state (simple mesh) was fine, and it would never discover that there were centralized services available. This reply is intended to spark discussion about better fixes. For example, how about only rescan when the laptop was in simple mesh mode when put to sleep, or if an attempt to reestablish the existing connection fails ? wad ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Serious problems with rescan networks on wakeup feature.
One thought, to add to the mix... How about designing the best solution for the 'deployment / school' situation, then offering a method for users to optionally 'tweak' the setting for their preference / situation. Like a sugar-control-panel option to 'show mesh enable/disable scan option'... then a enable scan or disable scan will appear in 'Neighborhood View'... as selectable items under Mesh 1, Mesh 6, and Mesh 11 icons. In a school/deployment situation, does the laptop always want to search for Meshes... even when the student is at home? Oh, sorry.. there were several ideas there melded into one :-) -Ixo On 3/2/08, John Watlington [EMAIL PROTECTED] wrote: On Mar 2, 2008, at 7:49 PM, Thomas Tuttle wrote: Hi. I just upgraded to joyride-1738, and the new feature that rescans for networks when the laptop wakes up is causing a lot of trouble for me. I often leave long-running connections such as IMAP or IRC open. Before the change, the laptop would wake up, the card would reassociate, and there was a good chance that my connection would still be up. After the change, the connections are almost always dropped, even if I just turned the laptop off for a few seconds. What would be *great* is if it waited 5 or 10 seconds, tried the existing connection, and then restarted scanning *if* it wasn't working anymore. The lag isn't that bad, and it makes the experience where you just closed the laptop a minute ago and are still on the same network better. The problem we have is the following: A student is using the laptop away from school/infrastructure, and is in simple mesh mode. In this mode, all service discovery and collboration is multicast.The student puts their computer to sleep (by closing the lid) and goes to school. Once they arrive at school, the last thing we want is for their laptop to try to use simple mesh --- it trashes spectrum and makes the school network not work, plus they won't see any of their friends that are (properly) connected through the school presence service. If we naively followed your suggestion above, the laptop would of course discover that the previous network state (simple mesh) was fine, and it would never discover that there were centralized services available. This reply is intended to spark discussion about better fixes. For example, how about only rescan when the laptop was in simple mesh mode when put to sleep, or if an attempt to reestablish the existing connection fails ? wad ___ 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: Serious problems with rescan networks on wakeup feature.
On Sun, 2 Mar 2008 22:37:22 -0500, John Watlington [EMAIL PROTECTED] said: On Mar 2, 2008, at 7:49 PM, Thomas Tuttle wrote: The problem we have is the following: A student is using the laptop away from school/infrastructure, and is in simple mesh mode. In this mode, all service discovery and collboration is multicast.The student puts their computer to sleep (by closing the lid) and goes to school. Once they arrive at school, the last thing we want is for their laptop to try to use simple mesh --- it trashes spectrum and makes the school network not work, plus they won't see any of their friends that are (properly) connected through the school presence service. If we naively followed your suggestion above, the laptop would of course discover that the previous network state (simple mesh) was fine, and it would never discover that there were centralized services available. This reply is intended to spark discussion about better fixes. For example, how about only rescan when the laptop was in simple mesh mode when put to sleep, or if an attempt to reestablish the existing connection fails ? That sounds good, actually. Basically, if there's a better connection available, use it, but if we have a known-best one (i.e., an AP), stick with it. I think this should be configurable, so deployments could decide the best way to do it based on their coverage, and G1G1 users could put their favorite (i.e. home/work/school) APs first. Ooh, there's an idea -- could there be a preferred ESSID list that is tried before mesh? My XO knows about my school network, but never uses it until it's tried a mesh first. Cheers, Thomas Tuttle ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Multiple Questions from a SysAdmin
Hey guys, I have multiple questions that maybe quite noobish. I have had them for a while and been looking for answers in the wiki, w/out success. 1. How can I cryptographically sign my own custom NAND image? We will need to do custom images for our pilot. For example, we need to include the Nepali activities we have built, Gcompris, and SocialCalc, that are not part of the standard build. The 6th graders at our pilot schools will require a different image than the 2nd graders. I have looked at the wiki page on building custom images and it looks fairly difficult. http://wiki.laptop.org/go/Building_custom_images Is there a way to cheat hard labor by using nandwrite or nanddump to simply copy customized image from an existing XO to a blank USB key? Also, have no clue how to crypto sign the custom image 2. Does Avahi work together w/ Telepathy (local mesh) and EJabber (multiple meshes) for the presence service? If so, how? I would appreciate if someone could point me to a good resource to learn more about how avahi works. I have only found a little bit of info about it on the web. Or does it simply handle network interface configuration? 3. What aspects of the mesh does Jabber enable that aren't there by default? 4. When I turn on an XO it always shows Mesh Networks 1, 6, and 11 even if there are no other XO's around. Why is this? I have always presumed that Network 1 was the XO's own instance of the mesh in case it didn't see a school server. But what about the other two? Thanks in advance for your help. I very much appreciate it. -- Bryan W. Berry Systems Engineer OLE Nepal, http://www.olenepal.org ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Emulating the School Server hardware (first things first)
Hello Chris, Hi Daf, Great news. I am located in Peru, tomorrow (today!) On Monday, we (a group of volunteers in Peru) will mount a small lab for helping the global OLPC efforts. I have check the info about the School Server, I understand that the info on the wiki can be a little old or not upgraded... no problem. We would thank if you update us with the configuration that we must set up for the school servers, because I have found this contradictory info: a) Networking devices: there is instructions (written) that the school server must have 4 network connections (UTP) to avoid the use of external switches. But there is a diagram (image) that shows an external switch to develop a network. b) we can use wi.fi access point (linksys) to emulate the OLPC access point... but we need confirmation about what trademark are you using for the access points... in this way we can use the same. c) Are the School Servers that will be installed in Peru without fans? (wich motherboard) What is the intended energy that the School Server must consume? This includes only the CPU or includes the switches/monitor/modem and other equipment related to the CPU? What about the access point? Finally, we have strong interest in connecting this Linux (XOs or regular native installions over x86 machines) to the normal Windows systems. The most possible scenary that the OLPC project will find here in Peru is that it will be needed to send a USB mule to the nearest point/town with regular internet access (internet cafe or cabina). We are studying the Wizzy and all the UUCP related issues. We have detected the need to particularize some aspects due to the nature of Internet in Peru (mostly related to the dominant position of some telecoms here and its reluctant participation in something that will not be under they control... more than 10 years of this position will not change over the night). If you have solutions studied for this USB mule transport then we will stop our worries and studies and turn to other issues, please let us know. Thanks and we hope to collaborate with our small forces here. Best regards, Javier Rodriguez Lima, Peru Chris Ball wrote: Hi, Daf and I got the school server jabberd/shared roster working today. We connected/registered 32 laptops to it with mesh TTL set to 1 for ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [OLPC Networking] Issues with the wireless mesh devices
Forwarding due to the quietness over on [EMAIL PROTECTED] =) Kim Hawtin wrote: Is this the correct forum to post questions around the wireless mesh devices? I took an XO to a community wireless[1] monthly meeting this week. We had a number of problems with other wireless devices, we believe directly related, to the XO being turned on, then stopped when the XO was turned off. I purchased some kit the same as in the APs that we use and hope to either confirm or discount the XO as the culprit. The network admin guys are quiet concerned about the DDOS possibilities. The access point is a Alix router board with a pair Atheros wireless minipci NICs. Here is a snippet from the log on the host; ath1: device timeout ath1: hardware error; resetting ath1: 0x0020 0x 0x, 0x4800 0x 0x ath1: ath_reset: unable to reset hardware; hal status 3 ath1: device timeout ath1: hardware error; resetting ath1: 0x0020 0x 0x, 0x6000 0x 0x ath1: ath_reset: unable to reset hardware; hal status 3 It appears that both Atheros NICs went awol at the same time and the AP/router needed to be hard reset/power cycled. Are there any tools that I can use to determine whats going on here? I noticed there was a wireshark patch is that for the XO itself? I've asked the netadmins for the APs for as much info as they can give us. So hopefully we can resolve this sooner rather than later =) regards, Kim -- [1] www.air-stream.org ___ Networking mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/networking ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
[ANNOUNCE] Compressed RAM swap device (compcache) 0.2 released
Hi All, I am pleased to announce compcache 0.2 - Compressed RAM based swap device for Linux. - Project Home: http://code.google.com/p/compcache/ - compcache-0.2: http://compcache.googlecode.com/files/compcache-0.2.tar.bz2 * Introduction compcache is virtual RAM based block device which acts as swap disk. Pages compressed to this disk are compressed and stored in memory itself. This helps a lot in performance under memory pressure as it avoids/delays swapping to slower hard-disks (Desktops) or using flash as swapping device which suffers from wear-leveling issues (Embedded). * Use Cases - Embedded Devices - Low Memory Desktops (Virtual Machines!) Some performance numbers can be found on Project Home. Testing on Linux VMs with typical workload (KDE, Firefox, Openoffice, Amarok etc.) shows 5% wastage by underlying allocator together with significant performance gain under memory pressure. README file included contains usage details. * Changelog: version 0.2 vs 0.1 - Fixed bug on systems with highmem - Better filtering-out of non-swap requests - Export statistics through proc nodes: - /proc/compcache - /proc/tlsfinfo - Debug and Statistics support for allocator and compcache can now be individually turned on/off by setting DEBUG, STATS to 0/1 in respective header files - Swap device now renamed to /dev/compcache0 - Added scripts: use_compcache.sh and unuse_compcache.sh See README for usage - Default compcache size set to 25% of RAM - Lots of code cleanups - Updated README - Created Changelog :) All suggestions welcome. Cheers! Nitin ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Testing 200 XO's in two weeks time for Nepal's pilot
Hey guys, Nepal should receive its shipment of 200 XO's in roughly 14 days from today. Here is the Rough Test Plan I have in mind 1. Boot into firmware and run test-all 2. Load customized image based on 656 build 3. Test localization on each XO (read Nepali, type in Nepali) 4. Test that basic activities like browse, E-Paati, EToys work 5. Associate XO's w/ school server via active antenna, test basic school services, cache, moodle, file downloads, chat Questions: How do I get developer keys for all 200 XO's and then how do I deactivate the developer keys after I no longer need access to the firmware? Which anti-theft features of Bitfrost have been implemented on the XO's we will receive? I distinctly recall there was some kind of mechanism where the XO would phone home periodically to a central database to see if it matched a list of stolen XO's. How many XO's can a single active antenna support? We only have two active antennas at the moment. Should we buy extra regular access points to back up our active antennas? Again, would love to know if particular AP is preferred and how many XO's one can support. I read in the devel list today that the WRT54G is not preferred. We will have two pilot schools. One w/ 110 students and the other w/ 50 students. I appreciate any other ideas on testing the XO's, particularly testing the batteries and the network. The Test_Config_Notes page in the wiki refers to this script to connectivity http://wiki.laptop.org/images/8/8a/Status.sh Can anyone tell me if it is up-to-date? Notes: These XO's may have been tested thoroughly at the factory but I would prefer to at least run the firmware's test-all command after the XO's arrive in Nepal. I have been a sysadmin for about 10 years and I have always run some kind of diagnostics on new hardware before I put it into production. Thanks in advance for everyone's help. -- Bryan W. Berry Systems Engineer OLE Nepal, http://www.olenepal.org ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel