**** BEGIN LOGGING AT Sun Jul 14 03:00:00 2013 Jul 14 08:54:10 ((sudo requires user passwd not root passwd)) unless you fix that nonsense: Jul 14 08:54:13 ~jrtools Jul 14 08:54:14 from memory, jrtools is http://wiki.maemo.org/User:Joerg_rw/tools Jul 14 08:55:03 ((Which rsync switches should i use)) Jul 14 08:55:06 ~jrtools Jul 14 08:55:07 somebody said jrtools was http://wiki.maemo.org/User:Joerg_rw/tools Jul 14 08:57:47 and good (UGT) morning Jul 14 08:59:11 http://desktop-zero.blogspot.nl/2010/03/nokia-n900-guide-sync-folders-between.html Jul 14 08:59:16 good morning Jul 14 08:59:17 also still handy for rsync and n900 Jul 14 08:59:43 well, still i have problem with find -printf options. Jul 14 09:00:14 i think i should use | xargs .... but don't know how exactly. Jul 14 09:12:19 ok Jul 14 09:12:24 i give up with this pan config Jul 14 09:12:27 it's impossible heh Jul 14 09:12:41 maybe someone can make an easier way to do it because for the life of me i can't seem to figure it out Jul 14 10:17:45 whats the current status of the autobuilder ? Jul 14 10:19:39 M4rtinK2, sixwheeledbeast is working on that, but afaik it's not "auto" but "mano"... Jul 14 10:20:00 well, the packages seem to build fine Jul 14 10:20:13 but it doesn't look like they are imported to the repository Jul 14 10:21:15 No, hence the word "mano"... Jul 14 11:15:19 M4rtinK2: repo-management is fsckedup Jul 14 11:15:47 M4rtinK2: ping merlin1991 who is importing packages to repo manually Jul 14 11:19:30 also, to all developers: PLEASE check your http://maemo.org/packages/view/ if there are recent multiple ""Old version cleaned by repository management"", and if so, IMMEDIATELY HOLLER here, best by highlighting sixwheeledbeat or me Jul 14 11:20:45 DO NOT upload multiple times same version number of your package, when it doesn't build instantly (it may take a while), or if it already built without problems Jul 14 11:22:17 generally autobuilder is NOT meant to do development there! you're NOT supposed to upload multiple versions per day! Jul 14 11:22:53 to develop and test your app, you use SCRATCHBOX on your local PC Jul 14 11:23:32 We might implement an auto-lock to your account when you uppload multiple times on same day Jul 14 11:27:01 please be aware that each upload that successfully builds (if it were promoted to extras-devel automatically) causes a new version of packages.gz that has to get build across *all* packages in repo, and will cause *all* devices to download the new packages.gz to check if there are updates available Jul 14 11:28:10 so, don't consider two new versions of your package per day to be a particularly appreciated thing to do Jul 14 11:29:34 it's also embarrassing for you as a developer - it makes you look foolish and not having a plan about where to head to with your package Jul 14 11:31:40 well, that looks kinda extreme Jul 14 11:31:55 sorry? Jul 14 11:32:02 IIRC, N900 don't check for updates that often Jul 14 11:32:24 that's not the point Jul 14 11:32:47 and N900 HAM checks for updates just so often Jul 14 11:32:52 also, Devel is imho expected to change often, but yeah, most people have it enabled due to the QA not really working that well Jul 14 11:32:54 default once every 24h Jul 14 11:33:04 and on every time you open it Jul 14 11:33:28 but - as mentioned - that's ot the point Jul 14 11:33:29 then at least for the N900s it should not be an issue Jul 14 11:33:39 ORLY? Jul 14 11:34:18 even if someone uploaded a package every 10 minutes, or once in 24 hourse, it should "look" the same for the N900 Jul 14 11:34:32 not an issue to download a new 15MB packages.gz multiple times a day, because some fool confused autobuilder with his personal development IDE? Jul 14 11:34:32 if I understand it correctly Jul 14 11:35:44 shouldn't happen if it really checks only every 24 hours Jul 14 11:36:00 if it autoupdates every time it is opened, that would be an issue Jul 14 11:36:10 HAM checks for date/version of packages.gz every time you start it (actually even after every action you did in HAM) Jul 14 11:36:14 but still, how often does one open HAM ? Jul 14 11:36:30 oh, after every action, that's quite bad Jul 14 11:36:31 honestly, you're goinfg to argue with me about that? Jul 14 11:37:08 I just think this should be solvable server side Jul 14 11:37:24 autobuilder is a means to propagate new *tested* working apckage version to repo, *nothing* else Jul 14 11:37:46 sure Jul 14 11:38:05 and yes, I'm going to solve this server side: multiple uploads to same package per day -> account locked Jul 14 11:38:22 I don't have issues with that & nothing what can go wrong with my packages as they are basically just compressed Jul 14 11:38:34 well, that what I think is too extreme Jul 14 11:39:03 what about: no upload in 4 hours - upload package db at once Jul 14 11:39:05 just dont allow it to happen and tell the user why Jul 14 11:39:07 I think uploading 3 versions of a package per day is too extreme Jul 14 11:39:29 tell them to try again in 24h and to read the manual orso Jul 14 11:39:30 package uploaded in last 4 hours - queue an update in 4 hours Jul 14 11:40:32 so basically, if there are multiple versions in a given timeslot, the db would be updated for the whole batch Jul 14 11:40:47 but sure, could be problematic due to the mess autobuilder is Jul 14 11:40:48 more and more "developers" ask for upload rights to autobuilder/repo without even having a garage project to upload. And it seems they all think the standard way is to check your edits in sourcecode on autobuilder Jul 14 11:41:57 even if they are "developers" i think it is good that people are still interested in Fremantle Jul 14 11:42:06 well, everyone started as a "developer" Jul 14 11:42:17 sure, but we at techstaff can't handle that Jul 14 11:42:43 maemo.org not been designed to handle that Jul 14 11:43:00 it's meant to get used in a different way than that Jul 14 11:43:09 well, if the autobuilder indeed needs manual attention, high usage could be a problem Jul 14 11:43:28 well, the problem is that everyone has devel enabled Jul 14 11:43:42 actually the fact that it doesn't promote to repo automatically is a bless currently Jul 14 11:44:14 well, I have users waiting for an update, so its kinda of an issue for me presently :) Jul 14 11:44:16 no, the problem is that "devels" think it's leete to upload 5 new versions per day Jul 14 11:44:30 so what about just setting a cronjob twice a day ? Jul 14 11:44:51 see my plea above Jul 14 11:45:15 multiple uploads while previos version got not processed seem to fsckup repo-management Jul 14 11:45:44 well, that sure is an issue Jul 14 11:46:25 and I guess not easy to fix due the usual not-documented-at-all state of the infra Jul 14 11:47:16 edit, upload to autobuilder, bitch about compile takes >30min, read error logs, edit [...] --- NOT the way to use this infra Jul 14 11:47:41 well, that indeed is bad Jul 14 11:47:53 the >30min compile time especially Jul 14 11:48:16 unless trying to debug a autobuilder specific build issue Jul 14 11:48:18 add to that brainless messing with version numbers Jul 14 11:48:20 but those should be rare Jul 14 11:49:30 yes, this is a procedure ONLY allowable to make your perfectly locally compilable package match any weirdness in autobuilder Jul 14 11:50:54 I ponder to add a requirement to add scratchbox compiler log with timestamp and signature to the package, otherwise it gets rejected Jul 14 11:51:24 will teach "developers" how the SOP looks like Jul 14 11:52:40 ::: edit, compile LOCALLY, read LOCAL error logs. no: goto start. ONLY THEN upload to autobuilder Jul 14 11:52:46 well, development & packaging for the N900 is already hard as is, so adding additional barriers like this is IMHO not really a good idea... Jul 14 11:53:06 another issue might be that they might not know about this ? Jul 14 11:54:00 and I'm pushing to not grant autobuilder/repo upload permissions to users that don't have their project in garage Jul 14 11:54:32 hmm, check number 1 here: http://extras-cauldron.garage.maemo.org/HOWTO.html Jul 14 11:54:58 it explicitly tells users to upload until it works Jul 14 11:55:10 could be part of the confusion Jul 14 11:56:37 adding a "build locally first, don't needlessly strain the infrastructure until you package works" here & to the main autobuilder Web UI might help Jul 14 11:57:26 yes, it probably been addressed to experienced devels that would feel ashamed to upload stuff to autobuilder and have their error logs exposed to general public, when they didn't check the whole package locally before Jul 14 11:58:16 and yes, such a remark on that page might help. also helping might be to add fremantle there Jul 14 11:58:55 as you can tell by careful reading, this page is terribly out of date Jul 14 11:59:53 it origins from a time where every devel dropped by in this very channel and talked to peers to learn how to do things right Jul 14 12:00:17 (pre N900 times) Jul 14 12:04:27 anyway, autobuilder/repo-management is fubar atm. So please help to keep it alive during the times we're searching for somebody capable to fix it! Jul 14 12:06:23 sure, but I'm not really that experienced with admining stuff Jul 14 12:06:34 but I can at least donate & help with testing Jul 14 12:06:55 and develop apps so people keep using their N900s :) Jul 14 12:07:01 I posted 3 *simple* rules above for every devel to notice Jul 14 12:08:31 1) don not upload multiple times a day 2) do take good care about your version numbers, or your package will get messed up by repo-management 3) check your http://maemo.org/packages/view/ if there are recent multiple ""Old version cleaned by repository management"" Jul 14 12:09:59 ( 1. point is actually not that strict, if you wait for error logs of your failed build before you upload next version) Jul 14 12:10:33 so add them at the top of the autobuilder Web UI or somewhere where they will be actually seen ? Jul 14 12:10:46 (however do NOT upload multiple compiling versions, possibly even with same version number. EVER!) Jul 14 12:11:02 btw, regarding packages/view/* Jul 14 12:11:15 http://maemo.org/packages/view/modrana/ doesn't seem to load ever Jul 14 12:11:36 then it already exceeded point of no return with that cruft Jul 14 12:12:23 makes apache2/php5/midgard process OOM abort Jul 14 12:12:26 well, that's indeed a horrendous input validation oversight for the original autobuilder authors Jul 14 12:12:38 no Jul 14 12:12:53 err, well, yes Jul 14 12:12:56 sorry Jul 14 12:13:08 I kinda wonder if setting up an OBS instance with Fremantle SDK target would be easier than fixing this mess Jul 14 12:13:30 maybe, but how to integrate that with existing stuff? Jul 14 12:13:41 I wonder if there has been any work done on allowing one to choose between N900 as a host or N900 as a guest via USB networking: https://wiki.maemo.org/USB_networking Jul 14 12:14:31 DocScrutinizer05: it could run in parallel with autobuilder first (if there are enough resources) Jul 14 12:14:35 see http://maemo.org/packages/view/findmine/ which not *yet* tripped the line between OMG and OOM Jul 14 12:14:38 psycho_oreos: what do you mean? Jul 14 12:14:48 host for what? Jul 14 12:15:10 everyone would build in their own repo and only push to a shared one once the package builds correctly Jul 14 12:15:24 the shared repo could then be synced by a cronjob to Devel Jul 14 12:15:26 Macer: :( Jul 14 12:15:30 what's not working? Jul 14 12:15:49 just a quick uninformed idea how it might work :) Jul 14 12:16:00 kerio, bit complicated. I meant like USB networking "host" is where N900 becomes a router and I use N900's connection [wlan/3G]. Though I also want to switch between that mode and making N900 as a "guest" where it connects through to PC and uses PC's network instead. Jul 14 12:16:27 psycho_oreos: libicd-network-usb to be a "guest" Jul 14 12:16:33 and mobilehotspot to be a host Jul 14 12:16:54 there are your UIs Jul 14 12:17:24 kerio, eww mobilehotspot. I prefer wired networking though, besides I really don't like to use mobile hotspot in public places. Jul 14 12:17:40 mobilehotspot supports sharing a connection via usb Jul 14 12:18:42 Can it also be strictly for USB though? i.e. choose USB as the option and it does not turn wl1251 as ad-hoc device? Jul 14 12:18:58 i think so? Jul 14 12:19:06 idk, go ahead and check Jul 14 12:19:15 iirc you can't do both at the same time, actually Jul 14 12:19:17 Hmm I'll go have a poke at it, thanks :) Jul 14 12:19:26 or just make yourself some script Jul 14 12:20:07 Yeah that would be ideal, sometimes if I want to use N900's onboard wifi as network dongle then its nifty (in a sense). Jul 14 12:20:18 [Sun Jul 14 10:50:12 2013] [error] [client 151.77.230.21] PHP Fatal error: Maximum execution time of 30 seconds exceeded in /usr/share/php/midcom/lib/midcom/helper/_dbfactory.php on line 501, referer: http://www.mobileblog.it/post/14920/modrana-un-navigatore-gps-open-source-per-n900 Jul 14 12:21:42 wow, that looks quite weird Jul 14 12:21:53 *nods* was in the process of writing up a BASH script but I hit a snag. I asked in #bash and one of them suggested that I might want to try asking here for an abstraction tool. Jul 14 12:22:28 30 seconds ? why ? Jul 14 12:23:10 Probably it acts like a watchdog so to speak, looking for hung commands/process Jul 14 12:24:05 i.e. you don't want a nasty script to go around in indefinite loop without exitting, it may cause the daemon to crash. Jul 14 12:25:16 well, sure, but what takes 30 seconds on a package page to render Jul 14 12:25:22 that's the issue Jul 14 12:25:51 Whatever is on line 501 in /usr/share/php/midcom/lib/midcom/helper/_dbfactory.php I suppose :) Jul 14 12:27:10 referer: http://maemo.org/packages/repository/builds/fremantle_extras-devel_free_source/all/ referer: http://maemo.org/packages/ referer: http://maemo.org/packages/search/?org_maemo_packages_search%5B1%5D%5Bproperty%5D=name&org_maemo_pac...bla referer: http://maemo.org/packages/view/qdl-plugins-dailymotion/ referer: http://maemo.org/packages/package_instance/view/fremantle_sdk_free_armel/libc6-dev/2.5.1-1eglibc27+0m5/ Jul 14 12:27:49 a referer loop ? Jul 14 12:28:01 or something like that Jul 14 12:28:15 in /usr/share/php/midcom/lib/net/nemein/wiki/wikipage.php on line 818 idcom/core/privilege.php on line 459 Jul 14 12:29:00 no, our db simply cluttered with shit like in http://maemo.org/packages/view/findmine/ Jul 14 12:30:17 ouch Jul 14 12:30:20 we also seen OOM aborts, where apache2/php/midcom simply allocated more than the granted ~400MB, for a single page Jul 14 12:30:30 well at least the package name is quite fitting Jul 14 12:30:37 ugh Jul 14 12:31:36 Too much info in changelog I suppose? Jul 14 12:31:51 [2013-07-13 23:29:28] (pc-connectivity-manager) dunno if I said it before, but look at the mess @ http://maemo.org/packages/view/pc-connectivity-manager/ Jul 14 12:31:52 [2013-07-13 23:31:50] particularly 2011-12-02 22:47 and 2012-12-29 04:23 Jul 14 12:32:23 [2013-07-10 00:12:12] http://maemo.org/packages/view/libqtm-experimental/ dead Jul 14 12:33:22 [Sun Jul 14 12:33:08 2013] [error] [client 88.217.63.147] PHP Fatal error: Allowed memory size of 367001600 bytes exhausted (tried to allocate 93 bytes) in /usr/share/php/midcom/lib/midcom/core/querybuilder.php on line 326 Jul 14 12:33:35 Whoa, looks like there needs to be some sort of truncating. Jul 14 12:33:49 ^^^^ that's been me, trying to surf to ^^^ URL Jul 14 12:35:19 DocScrutinizer05: 367mb for a single webpage? Jul 14 12:35:23 so maybe you now understand why I say "in a week or so I *have* to take down all that mess since it runs into errors caused by errors caused by errors. Reaching criticality" Jul 14 12:36:55 repo management cluttering db, thus making apache abort OOM, thus no more midgard admin interface to that package to even delete it or do whatever to it, to fix the mess Jul 14 12:37:26 we're rapidly going over the cliff Jul 14 12:37:36 stuff like this is a nightmare Jul 14 12:37:43 indeed Jul 14 12:37:46 fixing a running production system Jul 14 12:38:10 + facing people who ask is the site down, etc. Jul 14 12:38:32 well ,you can just tell them "X is broken" Jul 14 12:38:35 basically libqtm-experimental/ is zombie state and non-recoverable Jul 14 12:38:51 other packages as well Jul 14 12:38:58 others following soon Jul 14 12:39:17 Yeah you can tell them its broken but you have to say that for everyone that comes in and asks the same question. Jul 14 12:39:26 zombie state? :x Jul 14 12:39:53 well, I can't do anything to that package to fix it (except on very low sql level maybe) Jul 14 12:40:10 or by increasing apache memlinit to 2GB Jul 14 12:40:41 so I at least could reach the midgard admin menu on that webpage to delete or edit the package Jul 14 12:40:53 if you have the source package, could you just purge the original an resubmit it ? Jul 14 12:41:02 oh Jul 14 12:41:10 I *cannot* Jul 14 12:41:30 yeah "oh" :-O Jul 14 12:41:39 panic mode getting popular Jul 14 12:41:59 so i guess that leaves only the SQL stuff (purging the records that cause this) Jul 14 12:42:10 but I do remeber some insane queries Midgard does Jul 14 12:42:21 that you've posted during the migration Jul 14 12:42:34 that's however prone to a global midgard db messup, since that are ~50 tables that all need to be consistent Jul 14 12:42:39 so that would probably also be quite complicated Jul 14 12:43:10 Sounds like in either way its one massive brain******. Jul 14 12:43:32 well, at least make db backups, if possible :) Jul 14 12:43:36 needs thorough understanding how midgard works, to do the right thing to the right tables (plural!) Jul 14 12:43:56 we have db backups, luckily Jul 14 12:44:01 good! Jul 14 12:44:16 but that might mean we have to rollback maemo.org to ~april Jul 14 12:44:47 and that still leaves repo in a messed up state Jul 14 12:45:09 at least all packages since april would be orphaned then Jul 14 12:45:37 how big is the db ? Jul 14 12:45:57 & does it support making snapshots while running ? Jul 14 12:46:01 a humble 50GB Jul 14 12:46:19 no Jul 14 12:46:24 not really Jul 14 12:46:41 well, that's indeed an issue Jul 14 12:46:50 well, it's mysql and our sysops have installed some means to do proper dumps for backup Jul 14 12:46:56 having say daily incremental backups would be a good safety net Jul 14 12:47:14 we have those Jul 14 12:47:58 every time blade-a load goes up significantly it's probably backupPC doing backups of the VMs Jul 14 12:48:05 incl db VM Jul 14 12:48:29 happens once / day, ~00:00 CET Jul 14 12:48:50 that's indeed good to have Jul 14 12:49:08 yeah, but we don't keep those backups forever Jul 14 12:49:15 in case it gets completely borked for some reason Jul 14 12:49:45 is it full or incremental ? Jul 14 12:50:12 I asked for keeping 4 weeks worth of incremental backups, plus one incremetal backup per month against original state Jul 14 12:50:24 dunno if warfare implemented it Jul 14 12:50:55 every Sunday we do a full backup, during week is incremental Jul 14 12:51:19 we keep 8 weeks of sunday backups iirc Jul 14 12:51:55 well, that should hopefully have the worst case covered for the time being Jul 14 12:52:13 nothing much more that can be done when fixing a live production system Jul 14 12:52:58 well, actually a cluster, making things even worse Jul 14 12:53:24 anyway, good job taking things this far already! :) Jul 14 12:53:57 considerin Maemo being such an undocumented & complicated beast Jul 14 13:00:10 seen anthonie Jul 14 13:00:15 ~seen anthonie Jul 14 13:00:17 anthonie <~anthonie@ip26-214-210-87.adsl2.static.versatel.nl> was last seen on IRC in channel #maemo, 2h 39m 2s ago, saying: 'No, hence the word "mano"... '. Jul 14 13:01:38 ~seen anthonie_ Jul 14 13:01:38 anthonie_ <~anthonie@ip26-214-210-87.adsl2.static.versatel.nl> was last seen on IRC in channel #maemo, 3d 19h 44m 56s ago, saying: 'yes, I think that's pretty well possible'. Jul 14 13:03:41 kerio, thanks, just checked Qt Mobile Hotspot and the latest under extras-devel does have an option to choose either USB or WLAN. Jul 14 13:35:31 M4rtinK2: http://xkcd.com/705/ Jul 14 13:39:20 sure, uptime is important :) Jul 14 13:39:50 that reminds me - what distro is Maemo actually running on ? Jul 14 13:39:55 well, the VMs Jul 14 13:40:02 Debian I guess ? Jul 14 14:23:19 vw 27 Jul 14 14:23:25 fail sorry Jul 14 14:28:31 Maemo uses Debian package management, it does not necessarily mean that it is running on Debian :) Jul 14 14:36:14 psycho_oreos: of course, but I mean the infrastructure Jul 14 15:16:50 alas ubuntu Jul 14 15:34:19 http://privatepaste.com/20e2d92083 Jul 14 15:34:46 /etc/php5/apache2/php.ini s/max_execution_time = 30/max_execution_time = 99/ Jul 14 15:36:04 * DocScrutinizer05 frowns at php5 causing apache2 processes to stall rather than quit when max_execution_time expires Jul 14 16:03:11 PID USER PRI NI VIRT RES SHR S CPU% MEM% TIME+ Command Jul 14 16:03:12 19959 www-data 20 0 1454M 1195M 20412 S 17.0 20.2 3:16.90 /usr/sbin/apache2 -k start Jul 14 16:11:24 16:07:30 write(19, "\303\0\0\0\3UPDATE repligard SET changed=NULL,action='update',typename='midgard_parameter', object_action=4 WHERE guid='0bab5c186a3111dd961d834f021af853f853' AND realm='record_extension' AND sitegroup = 1", 199) = 199 Jul 14 17:06:33 M4rtinK2: http://maemo.org/packages/view/modrana/ loads now Jul 14 17:07:35 there are still pages that fail to load though, even with 99s timeout Jul 14 17:10:55 Where is the storage location of 'contacts data'? Like rtcom-eventlogger for messaging. Jul 14 17:11:37 somebody (I think merlin1991) already tracked down the reason why ""Old version cleaned by repository management"" fails and thus gets rescheduled Jul 14 17:12:44 amin007110: I'd say .osso-abook Jul 14 17:13:31 in /opt ? Jul 14 17:14:05 in ~user/.osso-abook/db/addressbook.db Jul 14 17:14:13 hmmm Jul 14 17:14:41 why? Jul 14 17:14:50 just to know Jul 14 17:14:59 hmmm Jul 14 17:15:07 sqlite3 ? Jul 14 17:15:16 nah, berkley iirc Jul 14 17:16:39 http://maemo.cloud-7.de/maemo5/usr/local/bin/contact Jul 14 17:33:13 DocScrutinizer05, you know when in settings>accounts we touch an account to modify, it becomes highlighted. Well, a few years ago, when i touched a jabber account, battery becomes empty and device turned off. Since next restart, that row in account settings has remained highlighted regardless of which account fit there. Jul 14 17:34:56 Deleting and recreating does not make sense, unless I totally forgot about restoring account backup. Jul 14 17:35:37 make any difference* i mean Jul 14 17:50:32 for me the highlighting is just fraction of a second, until that account opens in detail view Jul 14 17:51:34 I gather you got a garbled account file/whatever that is making this accounts list blow chunks Jul 14 17:52:27 anyway (IM/VoIP)accounts there are not related to addressbook Jul 14 17:52:37 i know Jul 14 17:53:09 These are separate questiobns Jul 14 17:53:59 Well, How can I find this account file/whatever ? Jul 14 17:54:32 I'm not sure. Start at ~user/.rtcom-accounts Jul 14 17:55:11 gconftool -R /|less / Jul 14 17:56:52 accounts.cfg in there has a good part of stuff related to those accounts. Don't ask me though where the passwords are stored Jul 14 17:57:53 already saw account.cfg. Nothing related to highlighting rows. Jul 14 17:58:20 there's *nothing* anywhere related to highlighting Jul 14 17:58:33 that's a symptom of a bug Jul 14 17:58:43 probably in accounts.cfg Jul 14 17:59:07 or in (not) attached avatar image Jul 14 17:59:34 some syntax error or inconsistency or illegal char Jul 14 17:59:41 anyone here from the us who want to help me a bit? Jul 14 18:00:31 opened in vi. just 13 lines. General settings. like auto online, enabled, protocol, bla bla bla Jul 14 18:00:35 amin007110: I didn't assume there's a "highlight=permanent" in there Jul 14 18:02:00 amin007110: maybe an important line is missing. I dunno Jul 14 18:05:17 amin007110: AIUI you can create new accounts? then create a second account exactly like the messed up one, and compare their entries one by one Jul 14 18:07:19 I think it's not related to any account, it's related to the row number the account stand on. Jul 14 18:07:40 Imagine i have 12 accounts Jul 14 18:07:55 and the 7th one is highlighted Jul 14 18:08:18 when i delete the 7th, the previous 8the one take its place. Jul 14 18:12:09 mhm, now that's extremely weird Jul 14 18:13:04 I'd honestly suggest to reflash Jul 14 18:14:18 well, then if restore accounts backup, we will the row there again :D Jul 14 18:14:26 if I restore * Jul 14 18:14:38 we will see * Jul 14 18:17:47 It`s super mega giga hyper super DIFFICULTLY to understand, but i don't know is it super difficulty made or I am hyper idi0t!!! Jul 14 18:18:23 might even be a hardware defect, in digitizer Jul 14 18:18:32 nope Jul 14 18:18:51 digitaizer has been changed 2 times during these years Jul 14 18:18:53 then it shouldn't show up in a fresh reflashed device Jul 14 18:19:08 not even when you create 12 accounts Jul 14 18:19:10 in a fresh yes Jul 14 18:19:27 the problem is backup! Jul 14 18:19:42 restoring backup in fact Jul 14 18:19:42 and when it returns on restoring from osso-backup, then it's clearly related to what's in there Jul 14 18:20:11 and in backup the content is pretty much plaintext Jul 14 18:20:25 once you untargz'ed it Jul 14 18:20:56 and search for what ? Jul 14 18:20:56 mc is awesome to browse a backup and view the content of every single file in that tarballs in tarballs Jul 14 18:21:18 search for stuff that seems related to accounts Jul 14 18:22:20 like I already said, it might be an illegal char somewhere, or a garbled syntactical structure, or whatever Jul 14 18:23:13 I still don't get it how it could jump from one account to next, on deletion or insertion of an account Jul 14 18:23:24 and then as well survive reboots Jul 14 18:23:57 you say it's on 7th account? so what happens when you delete all accounts but 6 Jul 14 18:24:24 where is the highlight then? Jul 14 18:24:51 empty area become highlighted Jul 14 18:25:15 maybe it's a bug that existed since beginning of times, and just nobody ever noticed it since... 7 accounts, duh! Jul 14 18:25:46 and you say you have only 13 lines in your accounts.cfg? how's that possible? Jul 14 18:26:02 whou you mind to see ? Jul 14 18:26:07 whould Jul 14 18:26:19 I mean I have only 2 or 3 accounts and some 100 lines in there Jul 14 18:27:10 lol, I told you once a time, I, myself, am a bug :D Jul 14 18:27:38 # wc accounts.cfg Jul 14 18:27:40 78 71 1732 accounts.cfg Jul 14 18:28:18 nah, you're just dreaming that, I'm sure. Did you notice you were a bug when you awoke one morning? Jul 14 18:28:37 Nokia-N900:/home/user/.rtcom-accounts# wc accounts.cfg Jul 14 18:28:38 13 14 239 accounts.cfg Jul 14 18:29:03 ugt? Jul 14 18:29:38 Kafka Jul 14 18:30:28 and one of his worst stories at that! Jul 14 18:30:34 :) Jul 14 18:31:00 but im not gregor samsa :D Jul 14 18:31:04 :) Jul 14 18:31:12 you are no travelling salesman? Jul 14 18:31:13 :< Jul 14 18:33:32 Of course the answer depends on what you mean by reasonable. Jul 14 18:48:16 ok, 239 is a virgin accounts.cfg Jul 14 18:48:48 for any jabber or sip account there should be sth in this file Jul 14 18:49:22 btw in backup you find it in comm_and_cal.zip/uzip://Root/tmp/.rtcom-accounts Jul 14 18:52:42 you might want to cp -a /tmp /tmp-backup; rm -rf /tmp/* Jul 14 18:53:38 Well, I guess when i disable an account in settings, Maemo removes its related configs from the file. Jul 14 18:53:55 though tmpfs on /tmp type tmpfs (rw,noatime,size=1024k) Jul 14 18:54:53 amin007110: so what? you say you have 13 accounts? Jul 14 18:54:58 Or maybe i am wrong, Now comparing some differenct backups Jul 14 18:56:22 Well, found one with 10528 lines ! hehe :D Jul 14 18:56:36 sounds more like it Jul 14 18:57:25 still I don't get it how you can have 12 accounts showing in GUI when your accounts.cfg is virgin Jul 14 18:58:06 I'd thik an account not configured in accounts.cfg can't show up in GUI whatsoever Jul 14 18:58:13 think even Jul 14 18:58:51 hmmm Jul 14 18:59:16 i found that 7th account Jul 14 18:59:32 and there is not any noticeable difference with others Jul 14 18:59:35 you're sure you didn't do leete personalization with new unix user accounts, or switching user to root, or changing fs on /home or whatever? Jul 14 19:00:18 Yes Jul 14 19:01:23 sorry, no clue Jul 14 19:01:50 sth seems to be severely mystical on your system Jul 14 19:02:43 accounts that show up in settings/accounts should as well show up in ~user * * accounts.cfg Jul 14 19:03:47 well, i found them in previous 2011 made backups Jul 14 19:04:05 but not in current file and half of them in 2012 Jul 14 19:04:08 you say your file has 239 chars like the one in my backup labeled "virgin", yet you claim to see 12 accounts in GUI, and 7th of them is highlighted permanenly Jul 14 19:04:11 hello guys, right now im trying to reflash my phone (because of no uboot menu - just options like uboot console) but uboot is faster > flashing didn't start. Some ideas how to solve that? (by the way 'attached kernel' option is not working, srsly dont know why).. the problem with uboot is probably coused by 'formated' MyDocs partition, uboot is trying to find there file named boot.scr or bootmenu.scr... but how i said my MyDocs is clean aft Jul 14 19:04:19 although i had all of them Jul 14 19:05:29 lexxik: charge your battery, follow instructions in : Jul 14 19:05:33 ~flashing Jul 14 19:05:34 from memory, maemo-flashing is http://wiki.maemo.org/Updating_the_tablet_firmware Jul 14 19:07:05 lexxik: attached kernel not working probably caused by you having PK kernel modules installed Jul 14 19:08:03 DocScrutinizer05, is it possible that fsck restores a virgin accounts.cfg instead of the original? Jul 14 19:08:22 anyway uboot isn't "faster", it's just the next think after NOLO when NOLO refuses to flash Jul 14 19:08:34 amin007110: hardly Jul 14 19:09:09 unless the original got garbled on fs level and fschk deleted or moved it to lost&found Jul 14 19:09:28 then the GUI would create a virgin new file I guess Jul 14 19:10:33 since that's obviouslym SOP on maemo: when a config file is missing, the process trying to open it creates it anew Jul 14 19:11:15 but then that should result in all your accounts vanishing Jul 14 19:11:21 AIUI Jul 14 19:11:28 there is no accounts.cfg in lost+found at the moment. Jul 14 19:11:51 unless that particular GUI looks at some friggin telepathy crap config files as well Jul 14 19:12:17 in lost&found it wouldn't be called anything resembling original name Jul 14 19:12:56 any files rescued to lost&found are called file0000 .. file9999 iirc Jul 14 19:13:08 #385478 Jul 14 19:13:13 inode name Jul 14 19:13:46 but if it was a directory, files in it represent real names Jul 14 19:13:52 and there's for sure an option to fsck to rather delete/purge those orphaned "files" rather than moving stuff to lost&found Jul 14 19:14:09 is 'green light' battery status indication important when flashing? Jul 14 19:14:45 err yes Jul 14 19:15:38 battery check: switch off device, press&hold "U" key, attach NOKIA wallcharger aka fastcharger, watch screen Jul 14 19:15:52 it should show NOKIA and USB symbol upper right Jul 14 19:16:18 if it doesn't stay there until you remove charger, then your battery is (too) empty Jul 14 19:18:01 rationale: when NOLO decides battery is too low, it tries to skip flashing requests and to boot into system to start charging - however sensible that might be Jul 14 19:18:26 so when that happens, for you next thing you see is uBoot Jul 14 19:18:48 alas afaik uBoot can't charge battery Jul 14 19:19:29 so if that's your problem, then an external charger or rescueOS are your options to recover Jul 14 19:21:11 the problem: when i plugg in the wallcharger, the diod is orange and after a while, uBoot starts. Does it mean i need to charge the battery in another phone? (eg. my 5800XM?) Jul 14 19:22:36 yes Jul 14 19:22:56 your battery is completely depleted Jul 14 19:23:02 ~flatbatrecover Jul 14 19:23:02 Remove battery for 1 minute. Insert battery. Plug powered ***NOKIA WALLCHARGER*** to device. Watch steady amber. Let sit and charge. Do NOT try to boot. After 30 min, you got either a) a booted up N900, b) flashing amber which means you can boot, c) steady amber going off - in this case start over again with ~flatbatrecover Jul 14 19:23:27 ^^^ only applies to non-fubared rootfs devices Jul 14 19:25:01 anyway steady amber means the device can't boot up at all, not even to NOLO, due to battery way too low Jul 14 19:25:26 steady amber is emergency charging until NOLO can boot Jul 14 19:25:55 you can't charge battery to a level where NOLO would allow flashing, by emergency charging Jul 14 19:27:14 so what to do now: charge battery in my historical 5800XM, put it in N900 and? flash? or is there option how to save my half-bricked phone? Jul 14 19:27:51 you can boot to rescueOS and restore the uboot.scr (or whatever) file Jul 14 19:28:02 or you reflash Jul 14 19:28:07 rootfs Jul 14 19:29:03 anyway you should first charge the battery in your ancient phone, if that works Jul 14 19:29:11 by the way, when the battery is out of the phone and the charger is plugged in, the uBoot is still running. Interesting.. Jul 14 19:29:11 *completely* Jul 14 19:29:34 hmm, that's indeed interesting Jul 14 19:30:06 ooh, you mean when you remove battery while uBoot is displayed? that's normal Jul 14 19:30:14 ~hotswap Jul 14 19:30:14 hmm... hotswap is http://talk.maemo.org/showthread.php?t=57734 Jul 14 19:30:42 NOLO checks for battery presence. uBoot doesn't Jul 14 19:34:02 cool.. yet another thing i didnt know about n900 :) Jul 14 19:35:39 hah, when you wanna hear about things yiu didn't know yet of N900, you're talking to the right one ;-P Jul 14 19:36:51 I think i created a few more info bits scattered oall over the place than just http://talk.maemo.org/showthread.php?t=57734 and ~jrtools Jul 14 19:37:14 and Jul 14 19:37:22 ~factinfo flatbatrecover Jul 14 19:37:23 flatbatrecover -- created by DocScrutinizer <~halley@openmoko/engineers/joerg> at Mon Jul 26 02:02:58 2010 (1084 days); last modified at Wed Mar 27 16:02:38 2013 by kerio!~kerio@Maemo/community/contributor/kerio; it has been requested 63 times, last by DocScrutinizer05, 14m 20s ago. Jul 14 19:37:55 hai Jul 14 19:38:13 however the question is if you *really* wanna know about all that ;-) Jul 14 19:43:11 actually Jul 14 19:43:14 ~jrtools Jul 14 19:43:14 extra, extra, read all about it, jrtools is http://wiki.maemo.org/User:Joerg_rw/tools Jul 14 19:43:21 might be interesting Jul 14 19:43:54 as well as some of the other stuff in this junkjard Jul 14 19:44:05 err oops Jul 14 19:44:23 sorry I meant maemo.cloud-7.de Jul 14 19:44:37 (junkjard) Jul 14 21:19:13 so... now i have my battery charged so the phone is reflashable.. i dont have any important datas in it, but tons of applications and a few "spezial" configs&tweaks (u know.. but nothing important), so now, with the rescueOS, i need to create boot.scr on the MyDocs? If yes, what is its content? All the time before this 'issue' i thought that there is no uBoot related stuff on MyDocs.. :) Jul 14 22:18:44 ask pali, or any other uBoot user Jul 14 22:19:14 I guess providing a boot.scr via pastebin is not asking for much Jul 14 22:23:04 another possibility: you flash kernel only - matching to what your rootfs installed kernel modules say is the right one for you (you probably know) - and then install uBoot again Jul 14 22:23:39 ...from booted maemo, as usual Jul 14 22:24:12 this will re-create the boot.scr and the uBoot + attached stock kernel, and anything else needed Jul 14 22:25:21 see flasher --help or wiki page or man page, how to flash kernel-only Jul 14 22:28:06 sorry, my uBoot is really ancient, I guess it won't help at all for solving your problem Jul 14 22:32:24 tbh I'm not sure if I have uBoot installed at all. Last time I booted the device in question is like 12 months ago Jul 14 22:32:47 probably not, since I wouldn't need it Jul 14 22:33:43 anyway I have no boot.scr Jul 14 22:34:28 and I guess yur boot.scr is proprietary to your system config, as set up in bootmenu/* config files Jul 14 22:35:03 so probably the kernel-flashing is the better alternative Jul 14 22:38:03 thanks... i will try --flash-only=kernel option... however, thanks :) Jul 14 22:38:12 yw Jul 14 22:38:40 my n900 is stuck in USB mass storage mode, even though it isn't even connected to a computer Jul 14 22:38:52 consequently, it won't turn off. what should i do? Jul 14 22:39:03 it won't turn off? Jul 14 22:39:39 DocScrutinizer05, "device is being used via usb. disconnect device from the pc and try switching device off again" Jul 14 22:39:57 WTF? Jul 14 22:40:04 (i've never seen that before because i've never tried to switch it off while connected to USB) Jul 14 22:40:05 never seen or even heard of this Jul 14 22:40:22 and it says mass storage mode in the status area and all that Jul 14 22:40:38 simply connect device to PC or charger, then disconnect again Jul 14 22:40:52 i plugged it in, put it in mass storage mode, and noticed it wasn't appearing on the PC Jul 14 22:41:00 it's a known minor inconvenience Jul 14 22:41:06 then i accidentally dismissed a popup message while confirming it was in usb mass storage mode Jul 14 22:41:09 then i unplugged it... Jul 14 22:42:08 that fixed it, thanks Jul 14 22:42:13 yw Jul 14 22:43:10 happens to me rarely even on unplugging fastcharger Jul 14 22:43:17 it's a race condition Jul 14 22:43:19 (hehe :)..what about #maemo-techsup or #maemo-help huh? :-) Jul 14 22:43:35 indeed Jul 14 22:43:47 sure, if I can ask for entrance fees ;-P Jul 14 22:44:13 sell PWs... Jul 14 22:44:21 PWs? Jul 14 22:44:32 passwords Jul 14 22:44:38 jkshsdjkh Jul 14 22:44:44 how much? Jul 14 22:44:48 ;-P Jul 14 22:44:59 sfdjwklrjh Jul 14 22:45:01 eqzhwgcrcwn Jul 14 22:45:07 take 3, pay 2 Jul 14 22:45:23 how much do you want for them? Jul 14 22:45:24 by minutes Jul 14 22:45:53 lexxik: buy me a lunch Jul 14 22:45:54 i will set up the third one as my root passwd, it is rly strong Jul 14 22:46:51 DocScrutinizer05, in this case the usb hub is clearly a bit weird anyway Jul 14 22:47:19 possibly Jul 14 22:47:53 well, second attempt neither phone nor hub noticed the connection. 3rd attempt, without the hub in the way, went normally Jul 14 22:48:04 hubs - particularly unpowered ones - may cause some trouble, since they don't alloow even 500mA charging Jul 14 22:48:29 DocScrutinizer05, why should that matter? Jul 14 22:48:51 because ENUM will fail in weird ways then, and finally hub may shut down the port Jul 14 22:49:19 so ive tried the --flash-only=kernel option, the uBoot didnt load but after Nokia logo is... nothing.. for a while it seemed promising (screen powered) but than my beloved brick shutted down... so the 'total reflash' is the only option now? Jul 14 22:49:35 N900 isn't prepared for a hub that doesn't allow 500mA charging Jul 14 22:50:07 lexxik: no, you just flashed the wrong kernel Jul 14 22:50:40 uBoot can't load since it got replaced by the kernel you flashed Jul 14 22:51:10 but the kernel needs to match the kernel modules in /lib/modules/* Jul 14 22:51:49 IroN900:~# ll /lib/modules/ Jul 14 22:51:51 insgesamt 0 Jul 14 22:51:52 drwxr-xr-x 2 root root 7600 2011-03-06 21:52 2.6.28-omap1 Jul 14 22:51:54 lrwxrwxrwx 1 root root 12 2009-10-30 12:32 current -> 2.6.28-omap1 Jul 14 22:52:12 yours might look pretty different, and the kernel you need to flash depends on that Jul 14 22:53:34 if the kernel and modules don't match, the kernel will panic after loading Jul 14 22:53:54 not much will happen after that, and nothing will happen before Jul 14 22:54:38 usually device reboots on a kernel panic Jul 14 22:56:12 regarding 500mA charging... Is there a good alternative to stock N9 wall-charger (1A), that provides around 500mA or even less? Jul 14 22:56:19 you said "attached kernel" didn't load on uBoot, so you evidently didn't have stock kernel on your root Jul 14 22:56:40 huh, why would you want 500mA or less? Jul 14 22:57:13 and no, no usb charger provides less than 500mA, that would be rather insane Jul 14 22:57:31 to not stress the battery Jul 14 22:57:43 lol, it doesn't work like that Jul 14 22:57:44 not USB, but wallcarger Jul 14 22:58:34 so what i need now? the same 'stock' kernel that was in my n900 before uBoot issue? Jul 14 22:58:54 no, you evidently had something non-stock Jul 14 22:59:15 maybe you had powerkernel Jul 14 22:59:16 eg. kernel power? Jul 14 22:59:32 I almost would bet you had KP Jul 14 23:00:05 ~kp Jul 14 23:00:06 rumour has it, kp is http://talk.maemo.org/showthread.php?t=85665 Jul 14 23:00:40 somewhere in this thread I'd expect a mini-howto flash KP via flasher --flash-only=kernel Jul 14 23:01:18 i will try to find it there, thx again. [hitting thx button] Jul 14 23:01:24 anyway you could try KP Jul 14 23:01:32 :-D Jul 14 23:01:47 it can't be worse than stock kernel Jul 14 23:02:04 for your recovery effort Jul 14 23:02:53 and most recent one (kp52) would even work if you had thumbified system Jul 14 23:03:05 i will give you some free and strong passwords for that ;) Jul 14 23:03:07 dcnjwecvjwcvbujvbwujvbujv Jul 14 23:03:16 but keep it secret! Jul 14 23:03:36 and I'm not all that sure but seem to recall KPisn't picky with version of its modules, i.e. KP52 might work with KP50 modules Jul 14 23:04:10 sorry we don't buy, just sell. Also no return of used ones Jul 14 23:04:24 LOL... Jul 14 23:04:32 But anyway, the higher the charging rate (mA), the less charging cycles the battery will take, in theory Jul 14 23:05:00 well, only in theory, and even then not entirely correct Jul 14 23:05:10 not linear Jul 14 23:05:28 depending on type Jul 14 23:05:29 actually LiIon suffers much more from deep discharge than from speed charging Jul 14 23:07:11 ShadowJK did some tests, and it seems the cell is happy with the standard 800mA charge it gets for a short while during initial charging of deep discharged cell Jul 14 23:07:40 increasing that to 950mA though seems to significantly reduce battery life Jul 14 23:08:03 very usual recommendation is 0.7C Jul 14 23:08:07 see... Jul 14 23:08:10 http://www.androidpolice.com/2010/12/14/your-battery-gauge-is-lying-to-you-everything-you-need-to-know-about-bump-charging-and-inconsistent-battery-drain/ Jul 14 23:08:13 ~0.7*1320 Jul 14 23:08:13 924 Jul 14 23:08:34 batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries Jul 14 23:08:49 now the last one is scientific and worth aread Jul 14 23:08:52 meh, I#ve seen *all* that years ago Jul 14 23:09:10 ~batteryfaq Jul 14 23:09:10 somebody said batteryfaq was http://wiki.openmoko.org/wiki/Battery_Questions_and_Answers Jul 14 23:11:05 > >Table 2: Cycle life as a function o fdepth of discharg eA partial discharge reduces stress and prolongs battery life.<< Jul 14 23:11:51 300@100%disch vs 1500@50%disch Jul 14 23:14:57 the whole artice doesn't even mention charge-current Jul 14 23:17:01 the question now.. is where to get kernel-power v52 image compatible with flasher, im right? is it the same .bin format as 'full' firmware images? Jul 14 23:18:25 I'm not sure about that Jul 14 23:18:50 but I guess it is Jul 14 23:19:22 since basically on device flashing is also done via flasher-3.5(fremantle/ARM-build) Jul 14 23:20:03 Commercial chargers do not allow changing the charge voltage limit. Adding this feature would have advantages, especially for laptops as a means to prolong battery life. When running on extended AC mode, the user could select the “long life” mode and the battery would charge to 4.00V/cell for a standby capacity of about 70 percent. Before traveling, the user would apply the “full charge Jul 14 23:20:03 mode” to bring the charge to 100%. Some laptop manufacturers may offer this feature but often only computer geeks discover them. Jul 14 23:20:08 so basically whatever is used to "install" KP on device should also be fine for flashing via PC Jul 14 23:20:49 LOL, seems this article is rather old Jul 14 23:21:02 seversal laptops now offer that feature Jul 14 23:21:14 2010 Jul 14 23:22:07 but ok, so you would not recommend going 500mA? Jul 14 23:23:34 if i know, or KP installation is used .deb package.. so i will download it and 'dpkg -x' it and see what the package contains... are there any chances that flashable image is included? Jul 14 23:24:37 --for Jul 14 23:26:03 win7mac: if you're worried about your battery life, use ShadowJK's charge21.sh tweaked to a lower constant-voltage (sth like 4050 instead 4200) Jul 14 23:26:43 lexxik: sure, it will be included Jul 14 23:29:57 thanks, will look into that charge21.sh Jul 14 23:33:16 so.. ive found .fiasco image :) i will give it a try :P Jul 14 23:33:54 I wonder where from you got KP52.deb Jul 14 23:34:45 pali has an annoying habit to hide his tarbal and .deb attachements *somewhere* arbitrary random post number in his threads Jul 14 23:35:29 or maybe you're better off with KP50, dunno what you had installed Jul 14 23:36:00 http://repository.maemo.org/extras-devel/pool/fremantle/free/k/kernel-power/ Jul 14 23:36:03 kernel-power_2.6.28-10power52_armel.deb Jul 14 23:36:35 im sure it was kp52 :) Jul 14 23:50:29 * lexxik is hapy Jul 14 23:50:45 it worked! :DDDDD Jul 14 23:51:08 DocScrutinizer05: lunch is yours... Jul 14 23:53:11 sleep is mine... gn o/ Jul 15 00:01:27 lexxik: great! Jul 15 00:06:50 lexxik: so now just install KP52.zimage and uBoot again, and you're set Jul 15 00:07:34 I guess there must be a KP52-for-uBoot package Jul 15 00:08:48 yeah, its named 'kernel-power-bootimg' Jul 15 00:09:32 kernel-power-bootimg_2.6.28-10power52_armel.deb Jul 15 00:09:37 just checked Jul 15 00:09:56 if i want uBoot now, i must install that and 'u-boot-update-bootmenu' ...? Jul 15 00:10:13 is uBoot still installed i guess Jul 15 00:10:47 contans a zimage-2.6.28.10-power52 Jul 15 00:11:10 you nuked uboot by flashing powerjernel Jul 15 00:11:41 so first install zimage-2.6.28.10-power52, then install uBoot again Jul 15 00:11:53 uBoot installation will pick up zimage-2.6.28.10-power52 Jul 15 00:12:29 well, actually your zimage-2.6.28.10-power52 might still be there (I guess in /boot/*) Jul 15 00:13:06 so firstly the kernel-power-bootimg, than u-boot-flasher and finaly the u-boot-update-bootmenu Jul 15 00:13:36 and yes, u-boot-update-bootmenu should pick up the config files for that zimage from bootmenu/* and create a new boot.scr Jul 15 00:14:09 but you still need to flash uBoot to kernel partition Jul 15 00:14:27 so yep, what you said Jul 15 00:15:31 or rather, I guess u-boot-flasher will automagically invoke u-boot-update-bootmenu Jul 15 00:16:37 also my nitdroid partition is on its place, and the .item's files are still in /etc/bootmenu.d... so just u-boot-updateblahblah and uBoot will work as before? Jul 15 00:16:57 ...in PREinstall script Jul 15 00:18:00 look, you _first_ want to make sure there's a proper boot.scr file, only _then_ flash uBoot to kernel partition in NAND Jul 15 00:19:11 this is why I guess that installing uBoot the recommended way will auto-invoke u-boot-update-bootmenu _prior_ to flashing uBoot Jul 15 00:19:40 is it problem problem? Jul 15 00:20:02 since _after_ flashing uBoot you're in a limbo state at best, until you get a proper boot.scr Jul 15 00:21:10 to be utterly clear on this: when you flash uBoot now without taking care to first create proper boot.scr, then you are *exactly* where you started several hours ago Jul 15 00:22:20 ...hm.. i dont understand.. when i will install uboot, it doesn't create it? Jul 15 00:24:15 anyway.. what to do now? Jul 15 00:27:59 run u-boot-update-bootmenu Jul 15 00:28:14 check that you got a boot.scr where you think it shall be Jul 15 00:29:05 err, first check if your zimage-2.6.28.10-power52 is inb /boot and the according files in bootmenu.d/* Jul 15 00:29:24 then run u-boot-update-bootmenu Jul 15 00:29:34 then check boot.scr Jul 15 00:29:51 then install uBoot exactly the way it's recommended to get installed Jul 15 00:29:59 done. bootmenu.scr is Jul 15 00:30:06 located on myDocs Jul 15 00:30:26 and zImage of kp52 is also on its place Jul 15 00:30:51 so now the u-boot-flasher? Jul 15 00:30:55 yes Jul 15 00:31:05 well, the uBoot package Jul 15 00:53:55 u-boot-flasher is already in newest version.. reinstall it? or let i be? Jul 15 00:57:11 you need to reimstall it Jul 15 00:57:58 this info about newest version is obsolete incorrect info from before you flashed powerkernel via PC Jul 15 01:02:43 ok.. everything is working... and just for info, the 'attached kernel' is now the power-kernel v52? Jul 15 02:21:03 robbiethe1st ! Jul 15 02:21:06 you're alive :D Jul 15 02:21:43 you were gone for lik 4,5 days Jul 15 02:22:38 Um, more like three Jul 15 02:22:49 friday/staturday... Jul 15 02:26:38 OK, Jul 15 02:26:39 um Jul 15 02:26:50 address is 141 willms rd, elk, WA 99009 Jul 15 02:27:15 ... of course, this is the wrong window, but meh Jul 15 02:52:14 :D **** ENDING LOGGING AT Mon Jul 15 02:59:58 2013