**** BEGIN LOGGING AT Mon May 11 02:59:57 2009 May 11 06:37:11 Hi! Buongiorno! May 11 07:15:37 good morning! May 11 07:54:36 morning May 11 07:54:46 moin May 11 07:57:59 moin May 11 08:08:24 und gibt schon wieder neue features bzw. fixes? May 11 08:08:57 am freitag gabs ne neue firmware May 11 08:09:03 blog.fonosfera.org May 11 08:10:57 hört sich ja gut an May 11 08:12:36 werd ich evtl. heute mal testen May 11 08:32:27 good morning May 11 08:33:35 hi May 11 09:32:56 hi all May 11 09:35:48 hi May 11 10:21:34 Hello! My fonera2 2.2.6.0 rc1 Pause the torrent after little time... :( May 11 10:21:55 you have suggestion? May 11 10:22:47 nope... May 11 10:22:59 uhm... May 11 10:23:14 i try another dl... May 11 10:25:24 Danicasti: any other additional info? May 11 10:26:24 why? May 11 10:27:03 It's first time i try torrent in the 2.2.6 May 11 10:27:19 ...bad english :( May 11 10:28:07 iurgi: another torrent works... May 11 10:30:10 Iurgi: Any progress on the feedback I sent over the weekend? May 11 10:32:38 I mean, any clue of what is working wrong inside... May 11 10:36:24 Danicasti: good May 11 10:36:38 Decodecoding: well, all I can say is it looks like a DNS problem May 11 10:36:52 we really need to see what happens if you connect that fonera to a different connection May 11 10:38:19 Well, I couldn't bring it today to the office, but I hope I can tomorrow... Anyway, did you see the same report on the fonosfera trac? It was exactly the same in otther countri, the "JAVI IP cached" thing... May 11 10:38:36 the IP cached thing is normal May 11 10:38:46 it's the whitelisting systme May 11 10:38:52 the fonera has a watchdog May 11 10:39:10 that checks whether chilli/coova is alive eveyr 5min or so May 11 10:39:20 and this might be generating those logs May 11 10:39:39 but it shouldn't since it should not be using a whitelisted domain but a 'blacklisted' one instead May 11 10:39:52 But, who is JAVI? It can't be a neighbor of mine since the same appears in other reports... May 11 10:40:58 Anyway, if there is any test that can be done, just tell. I have ssh access to the box. May 11 10:50:50 lol May 11 10:50:53 JAVI is the developer May 11 10:50:54 XD May 11 10:51:07 he was quite bad at writing logs... as you can see May 11 10:51:33 Decodecoding: yes, you can do some test... May 11 10:57:58 May 11 10:55:33 Fonera user.err kernel: Killed process 8158 (unworkable) May 11 10:58:09 That's when the torrent is paused May 11 10:59:12 that's the oom killer imo May 11 11:10:53 i'm back May 11 11:11:21 ok May 11 11:22:15 is it normal, that the dl speed from rs.com is only 2x ~75kb/s May 11 11:22:39 browser dl speed is: ~750kb/s May 11 11:23:04 maybe i should try the new rc image May 11 11:23:35 c_up: hmmm May 11 11:23:37 the fonera is not a PC May 11 11:23:43 you can't expect the same speed May 11 11:23:54 also, the speed from RS is variable, even for the same file May 11 11:24:17 I would say, in optimal conditions, with the fonera you can reach up to 500KB/s in total as maximum May 11 11:24:24 and that's what I get usually May 11 11:25:31 yeah, with orig flipper i got about 500kb/s May 11 11:26:10 yes, that's the expected speed May 11 11:49:06 wil that be increased in future or is it limited by hardware? May 11 11:59:35 imo, that is a quite nice speed for such a small device... May 11 12:01:43 yes, but a faster dl would be a better ;) May 11 12:03:57 yeah XD May 11 14:04:15 iurgi: any news about torrent ? May 11 14:05:30 blogic: not yet May 11 14:23:13 hmmm May 11 14:23:19 blogic: I think I got something May 11 14:23:27 the torrent folder needs to be empty, right? May 11 14:23:33 I think that's the reason May 11 14:23:41 empty May 11 14:23:43 ? May 11 14:23:48 I had a previously downloeded file and torrent May 11 14:23:51 ah ok May 11 14:23:55 delete it May 11 14:23:56 yes May 11 14:24:00 will delete it and retry May 11 14:24:03 I think that's the cause May 11 14:24:20 trying again May 11 14:26:16 hmmm May 11 14:26:26 same error... maybe it's in an uncertain state May 11 14:26:28 will reboot May 11 14:26:43 well, I deleted the torrent folder May 11 14:26:53 it has been created and the .torrent is there May 11 14:26:58 but it says paused May 11 14:27:00 will reboot now May 11 14:27:12 or want me to test something for debug, before I do? May 11 18:22:06 Hey guys, im online Witz iPhone because im habe no Internet Donnestion anymore May 11 18:23:16 Updated top rc1, ping From fonera works. But not from pc, any idea? May 11 18:24:44 ok, rebooted twice, waited 10 minutes and back on line! May 11 18:30:43 uh? May 11 18:30:45 strange May 11 18:31:16 maybe communication between router an fonera!? May 11 18:31:40 yeah May 11 18:31:45 maybe it was the dsl line? May 11 18:32:19 no, ping from fonera to google (putty) worked fine May 11 18:32:25 only http access May 11 19:18:35 *waiting for the usb sound device* May 11 20:34:56 hi all May 11 21:00:46 Hi all. Finally made it to get home... May 11 21:22:10 hi May 11 21:23:24 Iurgi, the mail Pablo sent to the list, is related to torrents pausing or torrent causing system crashes, or both? (H) May 11 21:23:35 mmmm May 11 21:23:49 I think the ones that pause May 11 21:23:55 I see May 11 21:24:16 looks like azureus creates torrents not following the bendoce format May 11 21:24:22 and they are corrupted May 11 21:24:33 most high end torrent downloaders don't mind May 11 21:24:40 but unworkable coomplains about it May 11 21:24:48 we're thinking of doing a dirty hack... May 11 21:24:49 sounds reasonable May 11 21:24:55 but then it will need a loooooooooooot of testing May 11 21:32:57 bendoce? May 11 21:33:38 the thing is that the unworkable parser is not as robust as many other torrent clients May 11 21:33:49 and that can be the cause of many "failed torrents" reported May 11 21:33:54 Pablo, I've started with the basics...: Ubuntu and Debian iso torrents :-d May 11 21:34:09 by robust i mean not accepting valid .torrent files May 11 21:34:25 with torrent files, after a bit of research i've found that happen as with HTML files May 11 21:34:48 most are valid, because they are generated by other programs May 11 21:35:01 but many of them are not valid, but May 11 21:35:15 since the client (firefox/explorer azureus/whatever) accepts it May 11 21:35:22 they pass May 11 21:35:49 the unworkable parser works with well formed .torrents, and i figure that debian and ubuntu torrents are certainly May 11 21:36:11 but .torrents made by users around the world using azureus or whatever can or can not be valid May 11 21:36:57 that can also be the reason why when making a "canonical test" this does not seem to happen, because to test upload/download speed we all use ubuntu and debian torrents, since they have the most avaliability May 11 21:37:17 "illegal" torrents, wich are the ones users are going to download, are not always going to be valid May 11 21:38:03 azureus ctorrent and the reference torrent client written in python can afford having many tricks in the parsing code for those non-valid .torrents May 11 21:39:08 just like firefox/explorer/safari can show you a page that have only

title

paragrah and not the whole
....<body>...</body> and everything to be HTML X.Y compliant May 11 21:39:18 <Decodecoding> Well, the thing is that the developers switched to unworkable searching for more system stability but seems that there is some background issue still undiscovered... May 11 21:39:33 <pablo_bz> no, actually they are good news May 11 21:39:38 <Decodecoding> Maybe if that problem is found... May 11 21:39:54 <pablo_bz> unworkable seems stable downloading and naive parsing May 11 21:40:15 <Decodecoding> ..switching back to ctorrent could be a possible... May 11 21:40:21 <pablo_bz> its simplicity is both the cause of stability and rejecting real-world dirty torrents May 11 21:41:07 <pablo_bz> no, imho between bad-good and good-bad, the better will be getting good-good May 11 21:41:19 <Decodecoding> But, are those dirty torrents related to the crashes...? May 11 21:41:35 <pablo_bz> i do not know May 11 21:41:49 <pablo_bz> that's why i asked for samples. May 11 21:41:55 <pablo_bz> there are three factors here May 11 21:41:56 <Decodecoding> I see May 11 21:42:11 <pablo_bz> the .torrent, the client, and _the network_ May 11 21:42:12 <Decodecoding> well, so let's try and test.... May 11 21:42:57 <pablo_bz> change one of the three factos and the results will vary May 11 21:42:58 <Decodecoding> Holy crap, 180 kb downloading a ubintu iso... May 11 21:43:25 <pablo_bz> what we strive is getting the best possible client for most .torrents and most networks May 11 21:44:08 <Decodecoding> and then there's the chilli restarts I always see in the serial console May 11 21:44:23 <pablo_bz> chilli restarts??? May 11 21:44:40 <pablo_bz> arent you confused by chilli_radconfig calls made by cron? May 11 21:44:52 <pablo_bz> or chilli forks() when a client connects? May 11 21:45:24 <Decodecoding> http://pastebin.com/m436c616 May 11 21:46:04 <Decodecoding> from line 120 May 11 21:48:55 <pablo_bz> but Decodecoding May 11 21:49:21 <pablo_bz> # May 11 21:49:21 <pablo_bz> May 11 21:38:35 Fonera daemon.info hostapd: ath1: STA 00:19:d2:36:46:66 WPA: group key handshake completed (WPA) May 11 21:49:24 <pablo_bz> # May 11 21:49:27 <pablo_bz> May 11 21:42:23 Fonera user.alert syslog: redirector returned invalid data May 11 21:49:29 <pablo_bz> # May 11 21:49:32 <pablo_bz> May 11 21:42:23 Fonera user.crit syslog: enqueue --> RestartChilli May 11 21:49:58 <pablo_bz> oh May 11 21:50:07 <pablo_bz> you are referring to the restart in line 131 May 11 21:51:40 <pablo_bz> yes May 11 21:51:46 <pablo_bz> it seems arbitrary May 11 21:52:13 <pablo_bz> it would be worth investigating when chilli says this "May 11 21:43:43 Fonera user.alert syslog: redirector returned invalid data" May 11 21:52:29 <pablo_bz> because that are it's last words May 11 21:53:05 <Decodecoding> Sorry, my fonera crashed: http://pastebin.com/m7af795f4 May 11 21:53:43 <Decodecoding> The Debian iso was downloading with no problem and full speed (180 Kb) May 11 21:58:39 <pablo_bz> OOM May 11 21:58:50 <pablo_bz> :( May 11 21:59:17 <pablo_bz> out of memory... May 11 22:03:56 <Decodecoding> Well, I reported this to blogic and Iurgi last weekend so they know May 11 22:09:07 <iurgi> yes May 11 22:09:17 <iurgi> but I think your oom is due to teh dns problem May 11 22:09:34 <iurgi> pablo his chillispot restarts every little time and the foneras is always in offline mode May 11 22:09:57 <Decodecoding> It appears on in the fonero profile May 11 22:10:17 <iurgi> uh?yes, because that is only refreshed every hour May 11 22:10:25 <iurgi> but the power led is orange, right? May 11 22:11:02 <Decodecoding> no, that only happens connected to a linksys router. If connected to my cable-modem, all lights are green and ok May 11 22:11:06 <iurgi> ah May 11 22:11:12 <iurgi> ok May 11 22:11:23 <iurgi> and the oom also happened in this situation? May 11 22:11:36 <Decodecoding> yes May 11 22:11:48 <iurgi> hmmm May 11 22:11:50 <Decodecoding> and with different internet access too May 11 22:11:54 <iurgi> that's too bad May 11 22:13:07 <iurgi> ok guys May 11 22:13:10 <iurgi> I'm going to bed May 11 22:13:11 <Decodecoding> I've seen some people reporting this crashes but I guess I'm the only one with the serial cable watching these logs... :-S May 11 22:13:19 <iurgi> yes May 11 22:13:29 <iurgi> also, you're one of the few with the latest firmware too May 11 22:13:31 <Decodecoding> well, let's call it a day May 11 22:13:35 <iurgi> yes May 11 22:13:47 <iurgi> good night May 11 22:13:55 <Decodecoding> C-Ya May 11 22:20:12 <pablo_bz> ah May 11 23:26:41 <burton-ng> hello May 11 23:28:12 <burton-ng> has anybody tested webam in new firmware release= May 11 23:28:12 <burton-ng> ? May 11 23:28:22 <burton-ng> mine is no longer working **** ENDING LOGGING AT Tue May 12 00:15:12 2009 **** BEGIN LOGGING AT Tue May 12 00:15:31 2009 **** ENDING LOGGING AT Tue May 12 02:59:57 2009