**** BEGIN LOGGING AT Fri Jul 12 02:59:58 2013 Jul 12 08:48:05 https://fbcdn-sphotos-f-a.akamaihd.net/hphotos-ak-ash4/255018_10152027777808272_2089311588_n.jpg Jul 12 08:49:12 hi all Jul 12 08:49:48 it has been a while since I logged into this channel, glad to see everybody are still here Jul 12 09:03:54 ~usbfix Jul 12 09:03:54 rumour has it, usbfix is http://talk.maemo.org/showthread.php?t=75920 - and **NEVER** use epoxy (unless you want to seal your device for underwater) Jul 12 10:55:39 * DocScrutinizer05 moos at evil daystar Jul 12 11:10:36 werecow Jul 12 11:15:39 hehe Jul 12 11:17:28 >> No Real Programmer works 9 to 5. (Unless it's the ones at night.) << ( http://www-personal.edfac.usyd.edu.au/staff/souters/Humour/Real.Programmer.Stories.html ) Jul 12 11:18:13 so 80s Jul 12 11:19:59 so hard to get right. The text always jumps between sarcasm, cynicism, and damn true statements Jul 12 11:20:15 Yeah, but the romantic idea of a' Jul 12 11:20:28 er of a 'programmer' is actually not so romantic and its 20years old Jul 12 11:20:38 and its overly disgustingly male Jul 12 11:20:46 to the point of saturation Jul 12 11:21:02 except in "Real Programmers don't wear high heeled shoes." Jul 12 11:21:03 probably also the reason why just about any piece of software is shit Jul 12 11:21:08 :) Jul 12 11:22:54 probably it lacks a "Real Programmers don't participate in emancipation and gender discussions." Jul 12 11:23:50 though implicitly included, we have "A Real Programmer goes to discos to watch the light shows." Jul 12 11:24:34 :) Jul 12 11:24:57 DocScrutinizer05: I was once told by a lead programmer to *avoid* multithread code Jul 12 11:25:03 that real enough yet? Jul 12 11:25:21 she says it makes code hard to read and understand Jul 12 11:25:26 it's a pity that natural languages have a ubiquitous concept of gender Jul 12 11:25:49 it's a pity that was is called a 'real programmer' is always this bro-tastic male person Jul 12 11:25:59 also, make sure to hang the main UI so that our users know our app is busy Jul 12 11:27:04 definitely not bro for rejecting multithreaded and responsive code Jul 12 11:27:53 from a code efficiency perspective it's just insane to bother about the 2% of female programmers Jul 12 11:28:22 that is implying code is effecient :) Jul 12 11:28:23 I guess wemon are too smart to act like real programers Jul 12 11:28:38 healthy dose of female improves anything that is overly male Jul 12 11:28:56 code is efficient inside a gzip archive Jul 12 11:29:36 yes, alas we're far away from anything like a healthy dose of female in IT Jul 12 11:29:45 instead of a hung UI, why not make a modal progress bar window? Jul 12 11:29:53 Nope, multithreaded code is ugly Jul 12 11:30:06 * amin007110 is thinking about: Does restoring those lost+found stuffs have better consequent than reflashing?! according to http://blog.windfluechter.net/content/blog/2011/03/30/1095-updated-automatically-restore-files-lostfound Jul 12 11:30:37 amin007110: very likely has some corruption in them? Jul 12 11:30:59 women* btw Jul 12 11:31:37 jon_y, maybe. Jul 12 11:31:55 (code) I consider natural language a code here Jul 12 11:32:43 btw, anybody know if int * a = (int[2]) {1,2}; legal in C? Jul 12 11:33:01 kind of funky looking but seems to work Jul 12 11:33:19 not sure what kind of search term to put in google Jul 12 11:33:28 coding the article http://www-personal.edfac.usyd.edu.au/staff/souters/Humour/Real.Programmer.Stories.html in a gender neutral way would cause massive drop in code efficiency, that's just not justified given we talk about maybe 0.1% of female subjects that are tr´argeted by that text Jul 12 11:34:07 and those 0.1% would probably feel insulted Jul 12 11:34:25 since they never forget the name of their husband Jul 12 11:34:55 "A Real Programmer might or might not know his wife's name. He does, however, know the entire ASCII (or EBCDIC) code table." Jul 12 11:36:13 does a real programmer do VB? Jul 12 11:36:26 LOL Jul 12 11:36:43 lol Jul 12 11:37:05 probably a troll, I saw one at thedailywtf, he said C should inherit single "=" from VB in if/else comparisons Jul 12 11:37:23 read the article, a Real Programmer does FORTRAN exclusively (except the few assembler things he uses to tackle the problems that can't get done in FORTRAN) Jul 12 11:37:38 I should learn Fortran Jul 12 11:37:56 with C bindings, cause that's where the party's at Jul 12 11:39:12 also, real programmers don't do Agile Jul 12 11:40:06 I present to your a custom tree structure in java, with no tree seaarch, nor a way to remove elements after it's added Jul 12 11:40:29 since no one thought it through Jul 12 11:40:49 in fact, it doesn't even have any public methods Jul 12 11:41:02 methods? Jul 12 11:41:14 public interfaces Jul 12 11:41:15 structures? Jul 12 11:41:28 usable functions, oop wise Jul 12 11:41:33 not sth a Real Programmer would ever bother with Jul 12 11:41:36 everything is inherited Jul 12 11:41:59 that kind of bullshit convinced me agile is bullshit Jul 12 11:42:36 the reason for such a crappy class was because "customer never asked for anything more that requires it!" Jul 12 11:43:03 I gave up trying to refactor the code after a few weeks Jul 12 11:43:29 jon_y: you honestly need to read http://www-personal.edfac.usyd.edu.au/staff/souters/Humour/Real.Programmer.Stories.html Jul 12 11:43:44 sure Jul 12 11:44:03 also, everything that uses the tree struct uses typeof instead of proper inheritance Jul 12 11:44:21 completly neglecting the purose of OOP to begin with Jul 12 11:44:58 >>What of the future? It is a matter of some concern to Real Programmers that the latest generation of computer programmers are not being brought up with the same outlook on life as their elders. Many of them have never seen a computer with a front panel. Hardly anyone graduating from school these days can do hex arithmetic without a calculator. College graduates these days are soft -- protected from the realities of programming by source Jul 12 11:45:00 level debuggers, text editors that count parentheses, and "user friendly" operating systems. Worst of all, some of these alleged "computer scientists" manage to get degrees without ever learning FORTRAN! Are we destined to become an industry of Unix hackers and Pascal programmers?<< Jul 12 11:45:10 >>Real Programmers aren't afraid to use GOTOs. Jul 12 11:45:15 So TRUE Jul 12 11:45:32 java doesn't have gotos because it is pretentious Jul 12 11:45:40 neither does it have unsigned integers Jul 12 11:46:35 using for(;;)/break as fake goto in java gets tired after awhile Jul 12 11:47:12 lol Jul 12 11:47:35 so does dealing with programmers that write shell scripts masquerading as java code Jul 12 11:47:52 hehehehehehehe muhahaha Jul 12 11:48:06 eg using map instead of set for unique objects Jul 12 11:48:45 in FORTRAN there's no for(;;) Jul 12 11:48:59 well, it's java Jul 12 11:49:05 there are calculated goto Jul 12 11:49:25 I should get promoted to code review nazi at work Jul 12 11:49:28 but nothing you couldn't do with smart self-modifying code Jul 12 11:50:18 job should come with an armband and a whip, along with stylish designer uniform Jul 12 11:52:07 you know how COBOL (and so probably FORTAN as well) do subroutine calls? There's no stack. On entering the the subroutine, the return address gets written to the GOTO at end of the subroutine. Funny when you later on run into same code segment by some other means than a subroutine call Jul 12 11:53:17 DocScrutinizer05: I bet it's real fun when it gets ported to harvard archs Jul 12 11:54:08 COBOL is per definitionem not multithreaded Jul 12 11:54:58 what about fortran? Jul 12 11:55:06 In FORTRAN(-90; not before) you just prepend the word "recursive" in front of subroutine/function, and then the implementation will deal with it however it likes. Jul 12 11:55:24 recursive? Jul 12 11:55:35 it has to be recursive? Jul 12 11:55:52 I guess they meant reentrant Jul 12 11:56:15 does fortran has any concept of static variables? Jul 12 11:56:37 but probably that term wasn't minted yet, when FORTRAN90 got defined Jul 12 11:57:06 is there a fortran 2003? Jul 12 11:57:28 even http://en.wikipedia.org/wiki/Fortran#Fortran_2008 Jul 12 11:57:37 ok, 2008 Jul 12 11:58:00 http://en.wikipedia.org/wiki/Fortran#Fortran_2003 as well Jul 12 12:01:29 DAMN! FORTRAN2003 introduces function pointers :-D Jul 12 12:01:44 And there's a "save" attribute for persistent local variables (close to that meaning of C's 'static') already in FORTRAN-77. Jul 12 12:01:47 err procedure pointers, ok ok Jul 12 12:02:16 function pointers in my fortran! Jul 12 12:02:20 At least they release new versions more often than the C committee. Jul 12 12:02:37 I'm waiting for my C11 Jul 12 12:02:51 C99 is pretty complete Jul 12 12:03:06 so I'm fine even if it was never updated Jul 12 12:03:37 you want functors? go back to c++! Jul 12 12:04:16 doing c++ style pure virtuals just fine in C too Jul 12 12:04:49 >>Support for international usage [...] and choice of decimal or comma in numeric formatted input/output.<< LOL! COBOL has "DECIMAL POINT IS COMMA" since dunno 1968 or sth Jul 12 12:05:42 #define , . Jul 12 12:05:47 sure is fun Jul 12 12:06:11 is that valid C? Jul 12 12:06:44 Shamefully, no. Jul 12 12:06:55 (Macro names need to be identifiers.) Jul 12 12:11:35 speaking of identifiers, C++ can overload punctuation marks Jul 12 12:11:44 I'm still wondering how to use them though Jul 12 12:12:26 overloaded operator$ or operator->, operator, Jul 12 12:12:41 golden: http://en.wikipedia.org/wiki/Fortran#Humor Jul 12 12:14:35 * DocScrutinizer05 remarks that it's evidently Friday, judging by the topics discussed Jul 12 12:15:19 judging by shit that wont work its friday indeed :( Jul 12 12:15:46 casual friday Jul 12 12:17:06 ok Jul 12 12:17:08 since its friday Jul 12 12:17:14 i have the best movie trailer of the year for you Jul 12 12:17:15 http://www.youtube.com/watch?feature=player_embedded&v=iwsqFR5bh6Q Jul 12 12:24:14 awesome words: http://harmful.cat-v.org/cat-v/ Jul 12 12:24:36 >>Unfortunately their advice has been completely ignored, and today Unix has become overcome by exactly the kind of mistakes they warned against.<< Jul 12 12:25:16 * DocScrutinizer05 mumbles "busybox" Jul 12 12:33:25 DocScrutinizer05: windows 8 Jul 12 12:33:29 taking up 20GB Jul 12 12:34:04 I didn't even know cat(1) has arguments Jul 12 12:36:54 POSIX cat(1) only has one; no -v there. ("-u: Write bytes from the input file to the standard output without delay as each is read.") Jul 12 12:42:56 man rev|tac|rev Jul 12 12:43:03 madly useful ;-P Jul 12 12:45:00 C11 is a lie Jul 12 12:46:14 also, i have a fully updated MAME romset again! yay! Jul 12 12:46:15 ls Jul 12 12:46:21 whoops, wrong window Jul 12 12:47:19 cat - - is a tiny bit counter-intuitive in how it works Jul 12 12:47:51 why? Jul 12 12:47:59 ls Jul 12 12:48:03 gaaah, wrong window again Jul 12 12:48:15 i need to install an eye tracker to switch the window focus Jul 12 12:49:34 well, scrap that, my mistake Jul 12 12:49:55 i didn't know stdin could be reopened, from a terminal Jul 12 12:49:59 so maybe that's kinda strange Jul 12 12:50:03 kerio: (eyetracker) brilliant hack Jul 12 12:50:55 kerio: (re-opened) it probably isn't Jul 12 12:51:04 but i typed ^d Jul 12 12:51:50 >> Jul 12 12:51:52 If a file is '-', the cat utility shall read from the standard input at Jul 12 12:51:53 that point in the sequence. The cat utility shall not close and reopen standard input when it is referenced in this way, but shall accept multiple occurrences of '-' as a Jul 12 12:51:55 file operand. Jul 12 12:52:00 << (man 1p cat) Jul 12 12:52:02 i see Jul 12 12:53:09 I think ^D is EOF but not automatically closes the filehandle Jul 12 12:53:32 hm, makes sense Jul 12 12:53:46 after all a file(handle) gets closed by the process that opened it, not by the read() Jul 12 12:55:42 it's just common practice that a process does read() until EOF, then close() Jul 12 12:56:21 not all do, though. see tail -f Jul 12 13:04:34 yfw a random Swedish email that was mistaken for spam praise your product after passing through google translate Jul 12 13:05:16 can't read Swedish, still is kind of random on a English ML Jul 12 14:17:13 eh? Jul 12 14:17:24 whatever Jul 12 14:17:56 * DocScrutinizer05 ponders to port OpenSUSE bash excellence to fremantle bash Jul 12 14:19:41 suse: wc -l /etc/profile 384 /etc/profile Jul 12 14:20:24 IroN900:~# wc -l /etc/profile 16 /etc/profile Jul 12 14:30:12 http://susefaq.sourceforge.net/howto/bash.html though this one is ancient Jul 12 14:47:06 should turn your tweaks into one huge script to use by DEUs Jul 12 14:54:06 hm? Jul 12 14:54:21 like this? -> Jul 12 14:54:24 ~jrtools Jul 12 14:54:25 it has been said that jrtools is http://wiki.maemo.org/User:Joerg_rw/tools Jul 12 14:55:38 sure I'd adjust the "script" on this wiki page to import a adapted suse /etc/profile (and other related files, particularly for bash-completion) Jul 12 14:56:49 but also for keybindings atc. E.G it sucks donkeyballs that fremantle bash doesn't know about history-completion by page-up Jul 12 14:57:07 s/completion/search/ Jul 12 14:57:07 DocScrutinizer05 meant: but also for keybindings atc. E.G it sucks donkeyballs that fremantle bash doesn't know about history-search by page-up Jul 12 14:58:30 I think I need to create an even more general approach, since I see same problem (missing personalization of shell) on other platforms/boxes as well Jul 12 14:59:30 something to not only deploy my ssh pubkey but same time also my shell aliases and whatnot Jul 12 15:52:35 * amin007110 is thinking about Volleyball match result, Iran vs Germany, 21:30 IRST. Jul 12 15:53:28 <_nicolai_> hi, is someone a maintainer for autobuilder? autobuilder does not import new packages. Jul 12 16:08:12 http://slashdot.org/topic/datacenter/hp-keeps-installing-secret-backdoors-in-enterprise-storage/ - Epic Fail Jul 12 16:11:20 _nicolai_: the problem is known, but we don't have a maintainer Jul 12 16:14:22 http://www.theatlanticwire.com/national/2013/07/science-has-average-penis-size-thing-nailed-down/67105/ Jul 12 16:30:30 Pali: ping Jul 12 16:31:44 freemangordon: pong Jul 12 16:32:24 Pali: do you expect me to do anything else re omap_smc3 patch? Jul 12 16:32:32 BTW Dave did one more comment Jul 12 16:33:22 freemangordon: it is needed something to do with patch? Jul 12 16:33:51 Pali: hmm, I'd say yes :) Jul 12 16:34:18 Dave wrote something only about comments, right? Jul 12 16:34:49 Pali: remove r12 from stmfd/ldmfd and maybe remove dmb. Maybe I don;t read it correctly Jul 12 16:35:04 ok Jul 12 16:35:13 Pali: I guess you should ask Jul 12 16:35:45 Pali: honestly, I have no idea what comment to put for dsb/dmb Jul 12 16:36:27 Pali: besides the already existing comments in omap_smc2() Jul 12 16:47:15 http://xkcd.com/1237/ Jul 12 17:31:18 <_nicolai_> DocScrutinizer05, I would like to help, but I don't know anything about autobilder, is there anything I can do? Jul 12 17:56:49 _nicolai_: the problem actually is midgard Jul 12 17:56:58 _nicolai_: I can manually trigger the import when you ping me Jul 12 18:05:09 <_nicolai_> merlin1991, please do so. It is for my "ledpulse" package. maemo package interface says: "old version cleaned ..." every 4 hourse. The newest version to import is ledpulse 0.3 Jul 12 18:10:19 _nicolai_: done Jul 12 18:19:20 is the extras assistant web interface working? i always get "no repository selected" just after upload Jul 12 18:19:35 and scp/dput not working for me either Jul 12 18:49:07 merlin1991: did you already conntact Jonne Nauha because of my maintainer request for dropn900? Jul 12 19:33:14 merlin1991: no, the problem is autobuilder, which is only 10% midgard and 90% friggin scripts Jul 12 19:34:20 merlin1991: and our real problem is autobuilder / repo-management spamming our db with bogus records, until everything explodes Jul 12 19:36:38 see ^^^ <_nicolai_> merlin1991, please do so. It is for my "ledpulse" package. maemo package interface says: "old version cleaned ..." every 4 hours Jul 12 19:36:51 other packages do that every 4 minutes Jul 12 19:38:28 and I _have_ _to_ shut down this whole stuff when that continues, we already have packages that can't get inspected via package interface since they are simply too long Jul 12 19:41:19 I several times explained we have no maintainer for autobuilder, Eero is busy, and x-fade missing in action completely since end of last year Jul 12 19:42:01 sure you're free to ask teh guys we asked 100 times before, maybe you're more lucky Jul 12 19:43:27 and I definitely will shut down stuff in a few days, if nothing changes in that regard Jul 12 19:44:33 junk piles up in our db, and no simple means to remove it from there Jul 12 19:44:41 stuff like autobilder or stuff like m.o? Jul 12 19:45:19 and things approach a degree of fubar where everything comes to a grinding halt anyway, so I hope I will shut it down before ZILCH works anymore Jul 12 19:45:54 win7mac: that depends what's fsckdup when I shut it down Jul 12 19:46:28 we already see fatal failures in www.m.o due to oversized package histories Jul 12 19:46:48 I won't disable single packages/view/* pages Jul 12 19:47:11 if that gets worse that what it already is I will have to take down www.m.o as well Jul 12 19:47:57 are you 1000% sure thats not because the raid is 99,9% stuffed? Jul 12 19:48:36 fun detail: when a www.m.o/packages/view/ page doesn't work anymore, then there's not even a midgard way to delete that whole package, since... the page doesn't work anymore Jul 12 19:49:19 no, I'm 10000000000000000000000% sure Jul 12 19:49:23 :-S Jul 12 19:49:35 goooooooooooooooood ;-) Jul 12 19:49:57 I still just can't believe it... ;) Jul 12 19:51:09 you just can't believe what? Jul 12 19:52:45 that a 4 or 6 "removing old package version from repository" every 4 minutes, since January, do eventually result in insanely large web pages on ~packages that result in apache processes stalling or getting killed due to OOM? Jul 12 19:53:29 or that we have no means to remove those 500 events per day for a single package from our 946 midgard datatables Jul 12 19:54:21 or that I *WILL* shut down everything that is acting self-destructive, in a week the latest Jul 12 19:56:20 http://maemo.org/packages/view/findmine/ Jul 12 19:56:42 and that's one of the still working ones Jul 12 19:56:59 still loads... Jul 12 19:57:10 at least it been a few days ago Jul 12 19:58:59 we have other ones I don't want to post here since every time somebody clicks them, a apache/php process on www dies from out-of-memory Jul 12 19:59:19 I mean, I could think of having so limited and probably heavily fragmented disc-space may add to the trouble already there Jul 12 19:59:31 nonsense Jul 12 19:59:36 ok Jul 12 20:00:20 please don't explain how it works, I trust you ;) Jul 12 20:00:57 I wish I could explain how it works, actually I can't even explain how it doesn't work Jul 12 20:01:50 hm... have no faith in council setting up a mail like discussed in #meeting 3 weeks ago? Jul 12 20:02:09 but I can point you to http://monitor.maemo.org/ganglia/?r=hour&cs=&ce=&m=load_one&s=by+name&c=maemo&h=&host_regex=&max_graphs=0&tab=m&vn=&sh=2&z=small&hc=4 Jul 12 20:02:40 ^^ I mean that it will have some kind of positive answer? Jul 12 20:03:09 you're constatly urging me to c&p what I wrote 15 min or hours or sometimes days before Jul 12 20:03:31 I several times explained we have no maintainer for autobuilder, Eero is busy, and x-fade missing in action completely since end of last year Jul 12 20:03:49 yeah and there was an idea Jul 12 20:03:59 [2013-07-12 21:42:01] sure you're free to ask teh guys we asked 100 times before, maybe you're more lucky Jul 12 20:05:04 http://mg.pov.lt/maemo-meeting-irclog/%23maemo-meeting.2013-06-21.log.html#t2013-06-21T22:35:04 Jul 12 20:05:27 not my business, ther are other 4 council members Jul 12 20:06:10 I outright refuse to take track of everything Jul 12 20:06:34 I announced burnout some 3 months ago Jul 12 20:07:05 what makes you think things changed to the better for me? Jul 12 20:08:39 don't you see I understand that and try to help out? Jul 12 20:10:08 I'll ping thedead Jul 12 20:14:56 http://mg.pov.lt/maemo-meeting-irclog/%23maemo-meeting.2013-06-21.log.html#t2013-06-21T22:43:25 Jul 12 20:15:56 I know. and indeed there are some more councilors over here, sopposedly... Jul 12 20:25:00 ~ seen chem|st Jul 12 20:25:07 chem|st is currently on #maemo #harmattan #maemo-ssu, last said: 'pong'. Jul 12 20:38:21 what happened to that guy (jacewoski?) that said he'd help out with autobuilder? Jul 12 20:38:52 he thought it's unbearable to coordinate own actions with sysops Jul 12 20:39:16 DocScrutinizer05: the problem though stopping everything from running is 1 script that does not run thanks to a problem with midgard Jul 12 20:39:26 so yes I can pinpoint it down to midgard Jul 12 20:40:08 *shrug* Jul 12 20:40:31 to me it seems there's too much running Jul 12 20:43:31 ah, so there was some process he didn't want to follow. Are all the scripts well understood? with all the smart guys around here, it seems impossible that it can't be understood and hopefully tweaked to work better. Ideally it would be nice to have a dev/test environment for all this... they are virtual right? Jul 12 20:44:07 that's not feasible Jul 12 20:44:36 not feasible due to money? resources? Jul 12 20:44:44 resources Jul 12 20:45:03 we need ~4..6 VM running concurrently Jul 12 20:45:26 if not more Jul 12 20:45:29 for dev/test someone could run that on their workstation Jul 12 20:45:37 good luck Jul 12 20:48:09 anthonie tried that, right? - progress so far? Jul 12 20:48:40 anthonie tried to install ONE vm with midgard1 Jul 12 20:49:24 midgard is mysql based however Jul 12 20:50:02 and we can't clone that db in one complete chunk, since it has sensitive data as well Jul 12 20:50:16 and it is the 2nd vm in that picture Jul 12 20:50:30 not install, clone. it's how it would be handled in an enterprise. i understand there are constraints here though. without a "playground" it's impossible to make significant changes or test things like upgrades Jul 12 20:50:30 garage with postgresql is 3rd Jul 12 20:51:22 autbuilder 4th Jul 12 20:51:36 repo 5th Jul 12 20:51:49 sensitive data, I assumed so. you'd have to clone and then scrub the sensitive data. a big job for sure, but worth it imo Jul 12 20:52:33 sure, if we find somebody to do that, he can en passant fix the problems we face right now Jul 12 20:53:33 for sure the competence any such person needs for doing that clonig&scrubbing is sufficient to fix the simple little bugs we have right now easily Jul 12 20:53:50 I imagine that would take a small team Jul 12 20:54:21 for example we have scripts with hardcoded mysql account credentials Jul 12 20:54:46 ugly but not surprising Jul 12 20:54:47 and sure, same credentials are in db VM in mysql as well Jul 12 20:55:28 we have passwd files with 30k user accounts, same for groups Jul 12 20:55:53 at some point you need to make a balance between operations and security and some folks need to be trusted Jul 12 20:56:19 yes, but I won't hand out clones of our sensitive data Jul 12 20:57:10 would the invest in a 3rd blade make sense to have those vm's needed for testing? They could later be used for other purposes Jul 12 20:57:29 it would help Jul 12 20:57:37 I wouldn't either (security is my day job). what would we need to have our own dev environment at iphh, under our (your) control? Jul 12 20:57:49 a new blade? some storage? Jul 12 20:58:03 yes Jul 12 20:58:35 we've already set up a very limited test env for www VM Jul 12 20:58:49 with a local mysql to emulate the central one Jul 12 20:59:04 to test midgard upgrade etc Jul 12 20:59:05 to me it sounds like that is quickly becoming a priority, especially with all the interactions between m.o. systems Jul 12 20:59:10 *new blade, some storage and some testers! Jul 12 20:59:20 testers and admins Jul 12 20:59:27 no manpower to pursue it further Jul 12 20:59:41 and some oversight/planning Jul 12 21:00:01 with no HR there's nothing to plan Jul 12 21:02:15 yes, that's the big problem... time/people. Jul 12 21:02:23 Time for the banner and a SOS tmo thread Jul 12 21:02:32 afk for a bit Jul 12 21:03:12 ...give picture of the situation and ask for help Jul 12 21:03:41 BTW, why am I not banner-admin already?! Jul 12 21:05:20 ...I promise to ask Doc + chemist for approval of new banners Jul 12 21:06:15 I'd like to help, but I'm nol wizard.... I do some of this stuff in my day job and figuring out things work is something I do day to day. The big problem is time... it's hard to commit to being "the guy" Jul 12 21:06:55 correction: no wizard & how things work Jul 12 21:07:28 midgard I know nothing about for instance Jul 12 21:07:33 we'd need a team Jul 12 21:07:43 barely even heard of it Jul 12 21:07:55 antonie already offered his assistance Jul 12 21:08:16 *anthonie Jul 12 21:11:26 ok, I just don't want to complain and not offer to help somehow. :) I'd really like to be able to upload to the Autobuilder. Jul 12 21:13:02 shawn, we'll come back on you if we get such a project going Jul 12 21:13:57 anthonie suggested noise some on tmo as well, BTW Jul 12 21:14:25 *some noise Jul 12 21:16:06 I know it's kind of stabbin in the dark, but what can we loose? Jul 12 21:17:38 how much $$ would another blade inkl. HDDs be roughly? Jul 12 21:18:50 everybody interested can contribute to augment the knowledge in http://wiki.maemo.org/Maemo.org_Infrastructure/Autobuilder Jul 12 21:19:03 maybe 800 Jul 12 21:19:15 first rough guess Jul 12 21:20:39 well, if that's the last thing we can do for maemo, lets use donations for that Jul 12 21:20:58 (contribute) we allow "guest" accounts on several maemo VM, so you can investigate what the scripts do. I also provide a "bio-proxy" to root cmdline Jul 12 21:21:15 no idea if it worth taking the risk though, if it remains fruitlesss and things g down anyway Jul 12 21:22:01 win7mac: we have no manpower to investigate an existing once working system that now has a few flaws Jul 12 21:22:21 ^^ so whoever really wants, actually *can* look into things? Jul 12 21:22:23 I doubt we can find the manpower for such a testbed project Jul 12 21:22:50 (look into things) sure Jul 12 21:23:01 ok Jul 12 21:24:11 I can provide cleaned out copies of every file you're interested in Jul 12 21:25:14 and you also have no further ideas how to proceed? Jul 12 21:25:49 I need to take care about any passwords, any user privacy, and about some statistically significant data. Other than that there's nothing secret in maemo.org Jul 12 21:27:12 well, for now I'd proceed with spotting why repo-management over and over tries to clean out obsolete packages Jul 12 21:27:26 then try to figure how to fix the problem Jul 12 21:28:00 and particularly: DOCUMENT IT Jul 12 21:28:13 right! Jul 12 21:28:17 so others can pick up where I dropped Jul 12 21:28:32 absolutely Jul 12 21:29:25 ...very underestimated in community-driven infra... Jul 12 21:29:50 it sounds like with autobuilder (and other stuff) a lot of scripts run in the root cron, so those would definitely need to be relayed via you Jul 12 21:30:09 ...like with board mails... Jul 12 21:30:19 I'm willing to take look and try to help document/understand Jul 12 21:32:20 sure thing Jul 12 21:33:58 shawnjefferson: could you please register with nickserv? Jul 12 21:34:15 /msg nickserv help register Jul 12 21:40:49 I have registered... just did an identify Jul 12 21:42:40 shawnjefferson: I just invited you to our admin channel Jul 12 21:43:26 great! i just joined that channel Jul 12 21:43:48 you're sure? Jul 12 23:07:06 Hello Jul 12 23:07:45 Had a question regarding a custom desktop Jul 12 23:07:52 I can do: gconftool-2 /apps/osso/hildon-desktop/applets --dump Jul 12 23:08:03 and it shows the position of all my icons on the desktop Jul 12 23:08:31 How can i take that, store it, and then load it for later if i make changes but want to change back to the original positions listed in the file without moving icons by hand? Jul 12 23:08:38 Any help would be greatly appreciated Jul 12 23:22:22 Nobody, ? Jul 12 23:40:18 Snafu777: gconftool --help Jul 12 23:40:37 Doc! Jul 12 23:40:38 Amen Jul 12 23:40:47 I've read the man page Jul 12 23:40:59 and I dont have a /root/.gconf or a /home/user/.gconf Jul 12 23:41:07 I think that might be why i cant do what im trying to do? Jul 12 23:41:12 I just don't know enuf about gconf Jul 12 23:41:17 then you know you can store values via gconftool as well as read them Jul 12 23:41:25 i tried Jul 12 23:41:27 nothing worked Jul 12 23:42:41 Pretty much I have a very customized way to re-create the N900 from the filesystem up Jul 12 23:42:45 repartitioning etc.. Jul 12 23:42:58 Last step for me is to set the icons via a script automatically Jul 12 23:43:15 you *might* need root permissions to do gconftool -s for some keys Jul 12 23:43:20 got it Jul 12 23:43:26 doing everything as root Jul 12 23:44:05 I own an N900. It is a phone with Maemo (linux-like) for the OS. I am trying to find a way via gconftool-2 to dump the current location of icons on the Desktop. Jul 12 23:44:06 From what i have been able to try so far: gconftool-2 /apps/osso/hildon-desktop/applets --dump Jul 12 23:44:06 That command will dump the current position of all icons on the desktop to the screen. How can I take that output and re-load it later if I move an icon? Jul 12 23:44:06 I am trying to avoid having to manually move icons should I reflash. Jul 12 23:44:08 Bleh Jul 12 23:44:14 NOT what i intended to paste, sorry Jul 12 23:44:41 Load/Save options Jul 12 23:44:42 --dump Dump to standard output an XML description of all entries under a directory, recursively. Jul 12 23:44:42 --load Load from the specified file an XML description of values and set them relative to a directory. Jul 12 23:44:42 --unload Unload a set of values described in an XML file. Jul 12 23:44:56 dude, I had that on my other window Jul 12 23:45:19 So i tried..... gconftool-2 /apps/osso/hildon-desktop/applets --dump > desktop.xml Jul 12 23:45:20 I deleted an icon from the desktop Jul 12 23:45:29 then i tried gconftool-2 --load desktop.xml Jul 12 23:45:32 with no effect Jul 12 23:45:51 *shrug* Jul 12 23:45:53 then i tried gconftool-2 --shutdown && gconftool-2 --spawn Jul 12 23:45:56 never tried that Jul 12 23:46:20 How would you accomplish the task DocScrutinizer05 ? Jul 12 23:46:48 I probably had used gconftool -R /apps/osso/hildon-desktop/applets Jul 12 23:47:12 and then RTFM Jul 12 23:47:27 to find about dump, load, unload Jul 12 23:47:45 anyway the inverse function to load seems to be unload, not dump Jul 12 23:48:48 but it should prolly make no difference Jul 12 23:50:29