**** BEGIN LOGGING AT Mon Jul 30 02:59:59 2012 Jul 30 03:02:00 Yes! Jul 30 03:02:05 YES! YES! YES! even! Jul 30 03:02:12 theora works too, mission accomplished Jul 30 03:02:31 oknow go sleep Jul 30 03:02:35 summing up - for full xiph foundation codes playback, one need ogg-support and gstreamer0.10-theora Jul 30 03:02:48 no freaking diea why it haven't worked before Jul 30 03:02:54 but using Nuke is good idea, sometimes Jul 30 03:02:55 that's mumbo jumbo for me Jul 30 03:03:05 thanks for trying to help, zeq, DocScrutinizer, qwazic and others Jul 30 03:03:05 but at least you're typing to someone Jul 30 03:03:18 no worrioes, other read backscrolls ;) Jul 30 03:03:21 others* Jul 30 03:03:28 but then it's too late :D Jul 30 03:03:31 but I appreciate Your nonwhitstanding dedication to be carying listener ;) Jul 30 03:03:54 that's what happens when someone is extremely bored Jul 30 03:05:08 go watch some porn, it's more honest than religion, anyway Jul 30 03:06:08 but after this astonishing victory over mafw forces of darkness, it's really time for some sleep Jul 30 03:06:09 that's why i was reading what you typed Jul 30 03:06:13 see ya! Jul 30 03:06:18 gives it a whole new meaning eh? Jul 30 03:06:24 hehe Jul 30 04:20:04 using my n900 over lte Jul 30 04:20:06 woop! Jul 30 04:20:23 LTE or just 3.5G? Jul 30 04:21:30 LTE :-P tethered to my lumia 900 Jul 30 04:21:32 heh Jul 30 04:21:52 love the N900 but not att support :-/ Jul 30 04:28:01 Macer, my N900 will be a step up for me, 3.5G on T-Mobile, instead of just rated 7.2MB. (Though that latter figure isn't even close to real-world.) Jul 30 04:28:31 well.... the lte tethering is ok i suppose Jul 30 04:28:40 my modem is broken in my n900 anyways Jul 30 04:28:47 but i love my n900 :( Jul 30 04:28:47 yeah, except for battery life. Jul 30 04:28:53 your modem got broke? Jul 30 04:28:57 yeah. that part sucks Jul 30 04:28:59 yes Jul 30 04:29:05 how, an accident? Jul 30 04:29:11 keeps disconnecting saying it cant read the sim Jul 30 04:29:15 no Jul 30 04:29:23 bad sim card reader then? Jul 30 04:29:24 just up and died on me one day Jul 30 04:29:28 maybe Jul 30 04:29:37 probably not the modem. Jul 30 04:29:38 doesnt matter. i dont have tmob anymore anyways Jul 30 04:29:44 tmob is awful in chicago Jul 30 04:29:58 I don't mind tmob, they're Ok in NYC, better at least than atandt. Jul 30 04:30:14 AT&T is awful in NYC, as everyone knows. Jul 30 04:30:31 combination of iPhone hipsters and no new towers. Jul 30 04:30:39 in chicago att is AERSOME ;) Jul 30 04:30:53 er Jul 30 04:30:56 awesome Jul 30 04:31:06 Macer, I don't envy you, because I save a lot of money with T-Mobile than if I had to have AT&T. Jul 30 04:31:26 i suppose.. but the monthly bill wasnt my concern Jul 30 04:31:31 the shit data was ;) Jul 30 04:31:33 I use T-Mobile prepaid. Jul 30 04:31:59 I figure since AT&T is shit in NYC, I don't want the roaming anyway. Jul 30 04:32:39 when i went cross country from chicago to washington state... Jul 30 04:32:47 and I get to have no contract, so if I move, no problem switching to someone else. Jul 30 04:32:51 us cellular was the only company in the sticks :) Jul 30 04:33:03 i was constantly roaming Jul 30 04:33:05 us cellular is CDMA, right? Jul 30 04:33:22 one thing i have to give nokia credit for.. their maps app for wp7 is awesome Jul 30 04:33:29 yah Jul 30 04:33:49 nokia maps downloads to your phone Jul 30 04:34:03 and isn't totally awful like ovi maps Jul 30 04:34:09 Macer, I'll have to try it if the XDA hackers have made it available for other WP7 devices. I have an HTC HD2, it can run every mobile phone OS known to man. Even Meego. ;) Jul 30 04:34:11 what an abortion that app was Jul 30 04:34:35 wow really? Jul 30 04:34:36 well, it can't run Maemo... Jul 30 04:34:36 Speaking of which, I-mobile has roming agreements with AT&T Jul 30 04:34:45 so, if you've got T-mobile... Jul 30 04:35:22 robbiethe1st: a little play after the buyout fell through? Jul 30 04:35:23 Macer, HTC HD2 can run Android, WinMo (that's what it runs native), WP7, Meego, Ubuntu, and some other stuff in emulation. Jul 30 04:35:34 in emulation?? Jul 30 04:35:36 lol Jul 30 04:35:38 Macer, I have it running Jellybean now. Jul 30 04:35:49 Macer, only for Windows 9x and stuff like that. Jul 30 04:35:50 Well, they had it before the buyout, and then they had to agree to like 5 more years of roaming agreement Jul 30 04:35:51 so... Jul 30 04:35:52 using what? vbox? lol Jul 30 04:36:01 Macer, no, QEMU-ARM Jul 30 04:36:12 robbiethe1st: trying to keep up with verizon? Jul 30 04:36:22 Macer, it doesn't need emulation to run Android or WP7 though. Jul 30 04:36:22 i still think the block was lame Jul 30 04:36:38 why would it need emu for maemo? :) Jul 30 04:36:48 isnt that arm as well? Jul 30 04:36:54 Macer, it doesn't run Maemo as far as I know. Jul 30 04:37:00 oh Jul 30 04:37:17 it does run Meego, but it's a closed beta, and I'm not in. Jul 30 04:37:35 Android and WP7 though are available from XDA. Jul 30 04:37:42 a closed beta of open software :) Jul 30 04:37:47 yeah, weird. Jul 30 04:37:51 high fives for the open movement! Jul 30 04:37:55 lol Jul 30 04:38:33 Meego is not really that open, which is why I went with an n900. I loved my n810, if it had had hardware acceleration and better performance, I probably would have kept tethering it. Jul 30 04:38:57 lol Jul 30 04:39:12 some might argue that maemo isn't that open as well :) Jul 30 04:39:36 well, then there's Openmoku, but that's not very functional. Jul 30 04:40:03 it is more a problem with hardware ip than anything Jul 30 04:40:10 if I have root, the hardware drivers aren't as big as a problem. Jul 30 04:40:28 i would love to see a company release a device that is wide open :) Jul 30 04:40:42 it'd be nice, but it isn't quite neccesary from a developer or user standpoint except for developing the OS. Jul 30 04:41:13 Macer, not going to happen, regulations require GSM be closed source, and I don't believe that CDMA carriers would be more partial to letting it on their networks either. Jul 30 04:41:25 well... i a Jul 30 04:41:32 m not a programmer Jul 30 04:41:50 but from what i can gather the closed hw bits of maemo really stunted its growth Jul 30 04:42:10 Not really; closed SW bits more. Jul 30 04:42:12 the N900 itself had huge hurdles Jul 30 04:42:24 as far as running different OSes on it Jul 30 04:42:34 ie: modem, gps, cams Jul 30 04:42:37 yeah, the HD2 is the king of the OS changing stuff. Jul 30 04:43:16 it has similar hardware to the HTC Desire, which has open source drivers. Jul 30 04:43:24 or the Nexus One. Jul 30 04:43:25 It would seem that, so long as your other OS's had a reasonably close kernel, you could just wholesale load the kernel modules and userland stuff in for that support Jul 30 04:43:53 robbiethe1st: at the time that was my arguement lol Jul 30 04:43:57 As far as GSM goes, I don't see why it can't be both open and closed: The radio itself is closed, and exposes a fully open API. Jul 30 04:44:11 nobody believes in backporting i guess Jul 30 04:44:17 robbiethe1st, it doesn't though, that's the thing. I run winmo, android, and wp7. Android definitely has a different kernel than the other two which use WinCE kernels. Jul 30 04:44:21 i dont see why a kernel has to be bleeding edge Jul 30 04:44:45 Actually, though, the real problem is that all kernel modules should come with source. Jul 30 04:44:48 robbiethe1st, it's mostly due to the fact that HTC kept making the same phone over and over again I guess. Jul 30 04:44:50 if all the hw functions then a kernel update is unnecessary unless there is a fundamental flaw Jul 30 04:45:04 The closed bits should come in userland blobs, like Nvidia doesa Jul 30 04:45:17 robbiethe1st: yeah.. that was the openworlder arguement heh Jul 30 04:45:53 had some great showings in here a long time ago about gpl vs lgpl lol Jul 30 04:46:01 robbiethe1st, that's what I meant by completely open, the radio has to be closed. Jul 30 04:46:23 ...Of course, I'd also like to see a 'closed' radio with a specific HW flaw that allowed someone to bypass all restrictions. Short a pair of pins, and hey, you're at your own risk. Jul 30 04:46:26 the rest can be whatever you want almost, as long as it doesn't abuse the network so much that they don't want it on there. Jul 30 04:46:35 And we(the hw company) did our reasonable best. Jul 30 04:47:19 robbiethe1st, when you say we, you're at Nokia? Jul 30 04:47:19 i think the biggest fear is phreaking cell towers Jul 30 04:47:32 No Jul 30 04:47:53 yeah, I thought most of the maemo people had either left, or were polishing up their resumes. Jul 30 04:48:19 EdLin: i honestly wonder how far maemo/meego developer gets you in this world :) Jul 30 04:48:36 the Symbian people were outright sent away. Jul 30 04:48:53 as close as it could be to being fired under Finland's labor laws. Jul 30 04:48:59 nokia could have been apple years before iphone Jul 30 04:49:10 that is just piss poor management Jul 30 04:49:23 Macer, I want a computer in my pocket again with a nokia internet tablet, how far that gets me is not that important. Jul 30 04:49:24 they dropped the ball bad on that one Jul 30 04:49:32 true. Jul 30 04:49:55 they were way ahead of the curve but didnt capitalize on it Jul 30 04:50:29 they got screwed by ms as well Jul 30 04:50:39 lumia 900 will not be getting win8 Jul 30 04:50:41 heh Jul 30 04:50:45 Nokia has been bought out by Microsoft essentially, MS didn't do the actual deed, but they replaced most of their key executives with 'softies, which is as far as you can get with Finalnd's foreign ownership laws. Jul 30 04:51:16 wait until the angry lumia owners get a taste of that :) i have a lumia but only because they were paying people to "buy" them Jul 30 04:51:20 Macer, I don't think the CEO or his minons from MS give a rat's behind about Nokia, they're there to promote Windows on phones. Jul 30 04:51:40 i got a $100 discount (the cost of the lumia) on my first bill Jul 30 04:51:49 and â$100 in accessories Jul 30 04:51:56 $100 Jul 30 04:52:24 that's pretty nice. Personally I'd hate to have a Lumina 900, the very thing is a mockery of the N9. Jul 30 04:52:38 if that were the case the lumia 900 would be their target phone for win8 Jul 30 04:52:49 since they sold so well Jul 30 04:53:00 Macer, they didn't sell well. They sold 300,000 of them. Jul 30 04:53:06 in the US Jul 30 04:53:12 $20 upgrade fee and they'd start to challenge the bigger OSes Jul 30 04:53:17 wp7 is not bad at all Jul 30 04:53:33 no way lol Jul 30 04:53:38 that cant be right Jul 30 04:53:39 overall Android flagships sold ten times as many. Jul 30 04:54:26 from what i understand the lumia 900 did very well Jul 30 04:54:35 Macer, 300,000 in the US, and about 700,000 in EU and Asia. Sounds like a lot total actually, but not compared to Samsung which sold 30,000,000 phones. Jul 30 04:55:00 lol Jul 30 04:55:59 trying to find figures Jul 30 04:57:02 Macer, there are actually twice as many Windows Mobile phones in use than Windows Phone 7. It's been a flop. Jul 30 04:57:10 of course they are only showing percentages heh Jul 30 04:57:25 Macer, look at those percentages. You do the math. Jul 30 04:57:44 3.7% of the mobile market share is wp Jul 30 04:58:02 and the mobile market is in the hundreds of millions. Jul 30 04:58:23 which means you get... about a million phones. Jul 30 04:58:26 like I said. Jul 30 04:58:41 i am still looking for numbers Jul 30 04:58:42 with 300,000 in the significant to Nokia now US market. Jul 30 04:58:47 maybe wikipedia has them Jul 30 04:59:51 Macer, if you really think WP7 is doing well, at a 3.7% market-share, be my guest. The N9 outsold all but the Lumina 900 and that's only because Nokia tried their best to make it fail. Jul 30 05:02:49 i honestly dont know Jul 30 05:03:06 i am curious... i am looking for sales numbers but cant seem to find any Jul 30 05:03:10 they really had a potential hit on their hands with the N9. Jul 30 05:03:15 so it most likely means it isnt heh Jul 30 05:03:30 i feel like getting an N9 but no hw keyboard Jul 30 05:03:44 Macer, there are no official sales figures from Nokia, these are good estimates. Apple and Samsung and HTC release sales figures, Nokia is not. I think that in itself says it all. Jul 30 05:03:44 N900 had the best hw qwerty ever made Jul 30 05:03:59 yeah.. probably Jul 30 05:04:06 Macer, debatable, the HTC Touch Pro 2 had a great qwerty. Horrid OS though. Jul 30 05:04:06 i want an N950 :-/ Jul 30 05:04:27 EdLin: well... i love my n900 kb ;) Jul 30 05:04:32 i'm using it now heh Jul 30 05:04:36 yep, it looks like a good keyboard. Jul 30 05:04:48 i can only assume the n950 has a good kb as well Jul 30 05:04:57 I'm getting the n900 for the software, not the hardware. Jul 30 05:05:10 too bad they ditched the kb with the n9 Jul 30 05:05:13 though other than some flaws with the USB port, the hardware is fine. Jul 30 05:05:13 :( Jul 30 05:05:31 well... my sim reader doesnt work Jul 30 05:05:36 yep. :( Jul 30 05:05:40 which is also a common issue it seems Jul 30 05:06:02 I hope my N900 is in good shape, I'm getting it from a chinese dealer in hong kong that I've been told is problematic. Jul 30 05:06:10 this is my 2nd n900 Jul 30 05:06:33 first one was usb port snapping off Jul 30 05:07:00 yeah, that's a known issue. Jul 30 05:07:12 i didnt even know that until after it snapped off Jul 30 05:07:24 and this one has the bad sim reader heh Jul 30 05:07:36 but i still use it all the time as a nit Jul 30 05:07:39 Macer, I won't laugh at you. I didn't know xbox 360s had hardware issues until after I got one. ;-) Jul 30 05:07:56 oh.. the solder issue? Jul 30 05:08:13 the USB port solder issue with the n900, yes. Jul 30 05:08:13 where you have to reflow it by forcing it to burn out? heh Jul 30 05:08:23 i meant the 360 :) Jul 30 05:08:31 no, the 360 is worse than that. Jul 30 05:08:45 i actually fixed one with the washers Jul 30 05:08:53 I finally got, as my third 360, a "short" model that should last. Jul 30 05:08:57 i didn't think that would actually work heh Jul 30 05:09:05 they finally, after over 5 years, fixed the problem. Jul 30 05:09:09 i still have one in my closet Jul 30 05:09:21 i was going to riceburn it witha lit case etc Jul 30 05:09:31 but didnt really ccare enough about it Jul 30 05:09:36 N900 wise, The way to solve the USB problem is to take it apart and put a blob of solder between the edges of the USB port and the surrounding metal can covers Jul 30 05:09:47 Macer, why bother when MS will replace the thing for free for 3 years or so? Jul 30 05:10:00 robbiethe1st: if i knew that i'd have done it ;) Jul 30 05:10:12 EdLin: i got it free Jul 30 05:10:16 oh. Jul 30 05:10:27 plus the warranty is voided after i ripped it apart heh Jul 30 05:10:30 Macer, if you're the registered owner tho, that won't matter. Jul 30 05:10:32 lol Jul 30 05:10:35 I didn't know it until I screwed my first one up Jul 30 05:10:40 Still have to go repair it... Jul 30 05:11:01 robbiethe1st: my 2nd one is fine except the sim reader.. but no tmob anyways Jul 30 05:11:16 i cant do 2G ;) Jul 30 05:11:28 Yeah Jul 30 05:11:42 Mine, well, the cellmodem's totally shot Jul 30 05:11:45 i was considering ebaying it Jul 30 05:11:58 someone might buy it for whatever reason Jul 30 05:12:04 Picked up a pair of broken N900's for like $150 back when Jul 30 05:12:12 i sold my first ones screen for $50 recently Jul 30 05:12:13 Macer, it'd be a good replacement for an iPod Touch. Jul 30 05:12:17 I'd probably be the ne Jul 30 05:13:01 wish there was something else like the n900 out there Jul 30 05:13:09 but nothing else compares Jul 30 05:13:23 its all the same trifecta stuff.. ios wp or android Jul 30 05:13:27 it's sad :-/ Jul 30 05:14:15 i need to try to get plasma active on my tf101 Jul 30 05:15:59 anyways.. speaking of tf101 ... i just swapped over to it :) going to watch cold case... ttyl Jul 30 05:16:18 OK, enjoy. Jul 30 05:55:58 Just put my libc6 build on my device. Feels generally a little quicker. It's annoying I can't get it to run in SB1 though. Jul 30 05:58:53 zeq: do you know what exactly new features libc uses from 2.6.28, besides accept4? Jul 30 06:07:03 pselect and ppoll for sure Jul 30 06:08:05 hmm, but if the other parts of the system are build without those being called, will it have any benefit? Jul 30 06:08:58 yes, because virtually everything links to libc, and that's how those syscalls get used. Jul 30 06:09:17 in fact that's generally how syscalls get used! :) Jul 30 06:10:44 libc will emulate them if kernel support is unavailable (low min kernel version) Jul 30 06:10:55 aah, i see Jul 30 06:11:34 zeq: what is the risk to hit a kernel bug undiscovered till now because of those syscalls emulation Jul 30 06:12:00 very very low Jul 30 06:12:40 zeq: sorry to play devils advocate, but changing libc is risky think todo. Thus my questions. Jul 30 06:12:49 in fact the emulation is known to be racy Jul 30 06:13:16 did you build it with thumb? or it is arm only Jul 30 06:14:22 zeq: BTW re LTO - you should pass -lto to both compiler and linker according to docs Jul 30 06:14:35 -flto that is Jul 30 06:14:52 yes, have not done so somewhere? Jul 30 06:15:02 have we* Jul 30 06:15:15 in fennec if I am not mistaken. at least your mozconfig i saw Jul 30 06:15:29 that would be a bug ;) Jul 30 06:16:07 I doubt it explains the memory usage Jul 30 06:16:27 i've build fennec with some ac_ options from microb, stil didn;t try it though Jul 30 06:16:32 sure Jul 30 06:16:41 which options? Jul 30 06:16:44 What's the best app for recorder on n900 do you think? recaller? Jul 30 06:17:30 zeq: http://pastebin.com/gsmT1GMQ Jul 30 06:19:14 --disable-js-binary < what does that do? Jul 30 06:19:34 NFC :D Jul 30 06:19:39 going to try it Jul 30 06:20:52 I think the updater is what checks for new extension versions when you update, is that right?? Jul 30 06:21:08 zeq: sounds like that Jul 30 06:21:29 zeq: some of the patches in microb are how sql commits Jul 30 06:21:44 that would explain slow page opening in fennec Jul 30 06:22:02 do they also do something about the idle cpu consumption? Jul 30 06:22:11 I thought I saw a patch that might hint at that Jul 30 06:22:16 ie sync/async commit Jul 30 06:22:24 ok Jul 30 06:23:08 well, i suspect idle cpu usage is because of idle support enabled for maemo, never checked though Jul 30 06:26:06 anyway, if we don't find a way to reduce memory usage, further tunnings are pointless Jul 30 06:26:28 I'm sure it's a bug that it's so high Jul 30 06:26:41 yeah Jul 30 06:26:48 I can't be supposed to use that much! Jul 30 06:27:14 maybe you should rebase Jul 30 06:27:28 yeah, I'll do that Jul 30 06:27:31 have you made any commits? Jul 30 06:27:40 no Jul 30 06:27:50 okay, I'll just rebase and push then Jul 30 06:31:57 zeq: great, anyway I gtg, won't be able to test it soon, so no hurry Jul 30 06:32:20 see you later freemangordon Jul 30 06:33:13 zeq: damn, it just does not make sense to use ~100MB no matter if opening TMO or FB Jul 30 06:33:19 zeq: bb Jul 30 06:33:44 no it doesn't, we'll need to track down where it's going and file a bug Jul 30 06:39:54 fennec rebased Jul 30 06:40:07 freemangordon: ^ Jul 30 06:44:46 zeq: ok Jul 30 07:34:35 --disable-js-binary = jazelle? Jul 30 07:41:01 DocScrutinizer05: If it is, we definitely want it. I'd e surprised though. Jul 30 07:44:33 freemangordon: re. pselect, ppoll. I'm feeling really dumb here, the ARM syscalls were MIA until fairly recently, missing in our kernel, glibc is in error assuming they are generally available @2.6.28. What's worse, when I ported Gentoo to ARM EABI, I actually created the ppoll/pselect patch for the ARM architecture, which didn't get applied! Jul 30 07:46:17 zeq: you want it, means? No-Jazelle (as I seem to recall it's obsolete cruft anyway), or Yes-Please-Jazelle? Jul 30 07:47:29 well, as I recall the problem with Jazelle was lack of open specification and difficulty to utilise, not performance per se. Jul 30 07:48:07 :shrug: Jul 30 07:48:48 So in theory it could be useful, but in practice for the above reason I doubt this is a Jazelle implementation! Jul 30 07:50:30 re. pselect/ppoll: It might actually be worth backporting support into the kernel, it was introduced for very good reason, fixes real hangs, especially with the racy emulation in glibc (pselect was used far too much given it's lack of kernel support) Jul 30 07:51:36 ummm what? no pselect support? Jul 30 07:51:42 as for ABI breakage, that isn't a problem as it's only glibc that requires the support, userspace already assumes (wrongly) that it works properly Jul 30 07:51:58 DocScrutinizer05: I know, nuts. Jul 30 07:52:14 what is pselect? Jul 30 07:52:26 but, but, how does the system work then? Jul 30 07:52:38 DocScrutinizer05: our glibc emulates it with a racy code Jul 30 07:52:57 https://bugs.launchpad.net/ubuntu/+source/linux/+bug/319729 Jul 30 07:54:00 I'd forgotten this, despite as I mentioned above having ported the support to ARM miyself! :S Jul 30 07:54:16 s/miyself/myself/ Jul 30 07:54:16 zeq meant: I'd forgotten this, despite as I mentioned above having ported the support to ARM myself! :S Jul 30 07:56:16 DAFAQ! Jul 30 07:57:02 indeed Jul 30 07:58:58 IMHO Nokia should have hooked it up in the Maemo kernel, there was no good reason not to since the syscalls were reserved anyway. Jul 30 08:01:06 hooked what up? Jul 30 08:02:42 the pselect systemcall Jul 30 08:02:49 jonwil: ^ Jul 30 08:03:02 ok Jul 30 08:03:09 s/systemcall/syscall/ Jul 30 08:03:09 and what does pselect do anyway? Jul 30 08:03:29 it's a non-racy select() Jul 30 08:03:37 (when it's not "emulated") Jul 30 08:03:41 oh ok Jul 30 08:04:26 I should probably get working on my wl1251-cal replacement :) Jul 30 08:04:31 http://lwn.net/Articles/176911/ Jul 30 08:04:52 jonwil: ^ good article about it Jul 30 08:06:49 I probably fail to wrap my head around it Jul 30 08:07:37 shouldn't any signal stash up anyway, until the (p)select would return immediately after call when signal is pending? Jul 30 08:10:12 or qould setting of sigprocmask automatically clear some signals? Jul 30 08:10:20 would* Jul 30 08:10:49 example program linked here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/319729/comments/6 Jul 30 08:11:00 was(!) Jul 30 08:11:05 oh well Jul 30 08:11:39 ah, it's still there, just clicked on the wrong link ;) Jul 30 08:12:11 https://launchpadlibrarian.net/22877243/childspin.c Jul 30 08:13:56 zeq: you ran this prog against our glibc? Jul 30 08:13:59 I didn't get enough sleep last night Jul 30 08:14:11 me neither ;-) Jul 30 08:15:00 no, I haven't. But I know what implementation we have. I could give it a spin though ;) Jul 30 08:15:31 I mean if we run into trouble here, we'd actually need a cssu-kernel aiui Jul 30 08:16:02 unless there's some nifty workaround Jul 30 08:16:29 I believe in normal usage it's fairly "unlikely" Jul 30 08:16:45 but then it depends... Jul 30 08:16:57 races are fun like that :) Jul 30 08:17:21 general murphy rule: a possible race *always* happens, you just notice it after kernel version freeze Jul 30 08:17:32 IMO it should never have been implemented in glibc until kernel support was in place Jul 30 08:18:21 if people knew they couldn't depend on it, they wouldn't have coded to rely on it. Jul 30 08:18:34 yep Jul 30 08:19:02 unless... glibc does some nifty tricks to catch the race Jul 30 08:19:08 they don't Jul 30 08:19:48 so they thought "duh! let's replace it by sigprocmask pselect ? Jul 30 08:19:54 " Jul 30 08:20:15 err sigprocmask; select Jul 30 08:20:30 The glibc implementation happened because it is part of POSIX.1 Jul 30 08:20:53 the kernel implementation happened because of the broken glibc implementation Jul 30 08:21:18 OUCH Jul 30 08:21:50 the bigger ouch is it getting left out of the ARM kernel for so long Jul 30 08:21:58 yep indeed Jul 30 08:22:00 inexplicably Jul 30 08:22:20 though [citation needed] - run the testcode! Jul 30 08:22:24 The real problem is that the ARM kernel has had so many forks over the years that its impossible to keep track of anything :P Jul 30 08:23:04 I bet fmg is happy to find another kernel fix he can implement Jul 30 08:23:36 :) Jul 30 08:23:51 jonwil: that's probably a big part of it Jul 30 08:23:54 and I'm feeling my refusal to think about cssu-kernel erode Jul 30 08:24:22 thumb :shrug: Jul 30 08:24:52 bug in kernel that potentially affects *all* apps/processes: FIXIT! Jul 30 08:25:07 I was pushing way back when to get it in, but then I got frustrated with my inability to code PXA SRAM MM support, and put my Zaurus on a shelf... Jul 30 08:25:49 I took on a project above my ability Jul 30 08:26:01 zeq: pretty please run the test code, tell us if a race happens or not Jul 30 08:26:18 ok Jul 30 08:26:29 What kernel version is our kernel based on? Jul 30 08:26:34 (both versions!) Jul 30 08:26:42 of testcode Jul 30 08:26:57 #ifdef USE_SELECT Jul 30 08:27:38 jonwil: uname ? Jul 30 08:33:09 DocScrutinizer51: technically it's a glibc bug, with a kernel fix Jul 30 08:33:20 yep, groked that Jul 30 08:33:53 please edit the perror() in that prog a bit, so no doublettes Jul 30 08:34:07 I've broken my language settings on my N900 :( Jul 30 08:34:24 if (pselect (0, NULL, NULL, NULL, NULL, &oldmask) < 0) { Jul 30 08:34:25 if (errno != EINTR) { Jul 30 08:34:27 perror ("select"); Jul 30 08:34:41 last line oughta be perror("pselect") Jul 30 08:35:29 also I'm not happy about the empty sigchld_handler (int signum) Jul 30 08:35:55 from my daily debugging I know what ARM compilers do to such code Jul 30 08:36:06 feel free to fix it up, I'm happy to test again with a new version if you like Jul 30 08:36:22 "bwahaha idiot coder, let's move that to /dev/null" Jul 30 08:37:24 :shrug: I wonder if a printf() would do Jul 30 08:37:52 or a bla++ Jul 30 08:38:31 while that's running, how do I fix my language support? Jul 30 08:40:48 umm, wats da problam Jul 30 08:43:49 * DocScrutinizer05 glares at 'as' on his N900, missing friggin gcc Jul 30 08:44:14 I installed my glibc and all my native language translations disappeared Jul 30 08:44:38 for a tiny cruft like that https://launchpadlibrarian.net/22877243/childspin.c I don't feel like starting SB Jul 30 08:44:41 I switched back to the repo version of libc6 and it's still broken Jul 30 08:45:12 all the device languages have disappeared Jul 30 08:45:13 uuuumpf Jul 30 08:45:49 maybe your 21MB locale blob got lost? Jul 30 08:45:59 probably Jul 30 08:46:07 but how can I regenerate it? Jul 30 08:46:14 or replace it? Jul 30 08:46:19 not at all :-S Jul 30 08:46:24 hmmm Jul 30 08:46:28 restore it from somewhere Jul 30 08:46:38 it's a nokia blob Jul 30 08:46:41 :-( Jul 30 08:47:01 I made a backupmenu backup a little while ago Jul 30 08:47:14 that should suffice Jul 30 08:49:37 DocScrutinizer05: rootfs? Jul 30 08:50:27 can't spot it now Jul 30 08:50:31 :-S Jul 30 08:51:26 The easy way to restore it is to just do apt-get install --reinstall posix-locales Jul 30 08:51:34 thx Jul 30 08:52:29 DocScrutinizer05: the test has hung Jul 30 08:55:17 Provides: Jul 30 08:55:18 2.7.2-1maemo7-binonly0+0m5 - Jul 30 08:55:21 ??? Jul 30 08:55:46 is that posix-locales? Jul 30 08:57:31 / # ls -l /usr/lib/locale/locale-archive Jul 30 08:57:32 -rw-r--r-- 1 root root 23911920 Feb 19 2010 /usr/lib/locale/locale-archive Jul 30 08:57:52 :nod: Jul 30 08:58:56 http://www.youtube.com/watch?v=gTcJ53uaQuU&feature=youtube_gdata_player Jul 30 08:59:11 nuke that and you get funny button labels like "tana-fi-homebutton" Jul 30 08:59:14 damn euros are bad lol Jul 30 09:01:06 ..."since that's the only way the Nokia translators can handle it" Jul 30 09:01:09 :-S Jul 30 09:02:23 stab_me_internet_connection :P Jul 30 09:03:07 literally? Jul 30 09:03:16 that's what it says :) Jul 30 09:03:21 :nod: Jul 30 09:03:26 you nuked that file Jul 30 09:03:39 23911920 Feb 19 2010 /usr/lib/locale/locale-archive Jul 30 09:03:50 must have. attempting to reinstall posix-locales now Jul 30 09:04:23 I'd probably not dare that Jul 30 09:04:30 rather restore it from backup Jul 30 09:04:41 or first do a ls ;-) Jul 30 09:05:42 still no languages :( Jul 30 09:05:55 ls -l /usr/lib/locale/locale-archive Jul 30 09:06:00 looks like I'll have to restore from backup :( Jul 30 09:06:34 file exists Jul 30 09:06:42 reboot? Jul 30 09:07:27 afk, have to run Jul 30 09:07:30 giving it a go Jul 30 09:07:49 see ya doc Jul 30 09:10:05 yay working! :D Jul 30 09:20:17 hehe: http://www.spinics.net/lists/arm-kernel/msg38114.html <- that was me, no wonder I forgot, it was 5 years ago! Jul 30 09:21:00 I'm going to go aheard and backport the official version Jul 30 09:21:59 impressive number of responses I generated there! Jul 30 09:22:04 hi qwazix Jul 30 09:25:21 zeq: hi Jul 30 09:27:55 zeq: we're all impressed, you're a kernel hacker, and We Are Not Worthy. Jul 30 09:29:27 ;-) Jul 30 09:29:30 I'm a poor excuse for a kernel hacker. I think I've got about 1 official contribution! Jul 30 09:30:21 zeq: I ran Linux 0.95 on a 386sx-16, didn't do much kernel hacking on it other than recompiling it without SCSI support for my PC, which took about 48 hours to compile. Jul 30 09:30:42 that's like scoring 1 goal in a football career! ;) Jul 30 09:30:45 lol Jul 30 09:31:28 zeq: some American football players never score a goal, but I know what you mean, the no-hands sport as opposed to the American steroid-contest. ;-) Jul 30 09:31:30 EdLin: that's pretty early on, I think my earliest kernel was 1.0 something Jul 30 09:31:49 yes, that one :) Jul 30 09:32:01 zeq: 1.0 wasn't for over a year after 0.95 I think, maybe even 2 years. Jul 30 09:32:53 zeq: European football will never succeed in American sports networks, because it's too hard to find time for commercials. American football manages to get some commercials in about every 3-10 minutes. Jul 30 09:33:53 :nod: those were the days, 0.9x *sigh* Jul 30 09:34:22 DocScrutinizer05: I remember seeing the first Slackware ad on Usenet and thinking "boy is that thing bloated". Jul 30 09:34:27 EdLin: thats a good explanation why "soccer" never made it bit in the US Jul 30 09:34:32 where you had to patch your kernel with some weird binary patcher, to get errr. wut? kernel cmdline right? Jul 30 09:35:22 I was a RISC OS dev before I found RiscBSD, then onto Linux Jul 30 09:35:27 DocScrutinizer, I don't recall debugging the early kernels, I just lived with whatever bugs they had more or less, though I did patch them sometimes to the latest patch level. Jul 30 09:36:00 nah, back then it was a immanent part of installing a kernel Jul 30 09:36:12 you had to patch it to tell it about rootfs or whatever Jul 30 09:36:18 DocScrutinizer, oh yes, that stuff. Jul 30 09:36:27 but I used a boot/root disk combo. Jul 30 09:36:40 that came later iirc Jul 30 09:36:48 nope, it was around at 0.95 Jul 30 09:36:49 anyone creative want to figure out what I could do with my old N900 that is no longer being used :p? Jul 30 09:37:28 EdLin: aaah yes, might be right: boot from floppy to HDD Jul 30 09:37:38 bindi: probably give it to me because mine comes from a Chinese eBay seller that I hear is not reliable. :( Jul 30 09:37:48 next! Jul 30 09:38:17 DocScrutinizer, it wasn't too hard to get it to boot on the hard drive, as long as you had the proper MS-DOS partitions for dual booting. Jul 30 09:38:39 loadlin, yayaya Jul 30 09:38:49 yeah, I think I remember that bit. Jul 30 09:39:09 bindi: I will give you $10 for it. Jul 30 09:39:28 nah, I could sell it for 60€ but I rather find a use for it Jul 30 09:40:01 bindi: I use Skype a lot, got a 100-voice-minute/5GB data plan. Does that count as a use for it? Jul 30 09:40:41 I got a better idea Jul 30 09:40:44 also use the Linux stuff on it too, want to use gpg-encrypted email on it using another email client. Jul 30 09:40:53 ipcam with face recognition Jul 30 09:40:56 :-D Jul 30 09:41:02 cool. :) Jul 30 09:41:15 i wonder if this is possible on the n900 alone Jul 30 09:41:34 I don't know, they manage to do face recognition on Android phones. Jul 30 09:41:55 there's a game/demo of face recognition for the n900 Jul 30 09:41:58 burgerface Jul 30 09:42:05 http://www.youtube.com/watch?v=A-Wh4NuRfeU Jul 30 09:42:43 cute Jul 30 09:44:50 is the stock email client open-source? I'm wondering if it would be possible to add gpg-encryption if it isn't a feature. Jul 30 09:45:52 I recall the n810 didn't have that feature in its stock email client, I used a third party client instead. Jul 30 09:46:12 forget which one. Jul 30 09:49:15 I didn't do much email with the n810, but plan to more with the n900 of course due to the always-on connection. Jul 30 09:49:41 so gpg encryption is going to be important, as one of my frequent email correspondants loves to use it. Jul 30 09:50:38 * EdLin is talking to himself, the best form of encryption Jul 30 09:52:16 EdLin: it's one of the features conspicuously missing from modest Jul 30 09:52:47 zeq: that's what I was afraid of. Is modest easily worked on? Jul 30 09:53:07 EdLin: I've not looked at it Jul 30 09:53:22 or maybe I'll just install a third party Linux email client and call it a day. Jul 30 09:53:22 not exactly easily Jul 30 09:53:24 codewise Jul 30 09:53:32 but some guys already did it Jul 30 09:53:45 also on tinymail - under the hood of modest Jul 30 09:53:49 aiui Jul 30 09:54:17 DocScrutinizer05: so there's gpg support in modest? Jul 30 09:54:21 DocScrutinizer, you have some links to their project? Jul 30 09:54:24 nope Jul 30 09:54:43 EdLin: check CSSU Jul 30 09:55:04 Didn't think I'd have failed to notice that appear as a new feature! Jul 30 09:55:22 DocScrutinizer05, which CSSU is it in, standard or testing? Jul 30 09:55:27 t Jul 30 09:56:01 cool, I guess I'll be joining the ranks of CSSU testers then, this is an important feature for me. Jul 30 09:56:02 portrait support and treeviews are buzzwords I recall Jul 30 09:56:10 as well as "delete mails on server" Jul 30 09:56:43 EdLin: there's NO gpg support in CSSU modest! Jul 30 09:57:06 sorry I've been fuzzy in my posts above Jul 30 09:57:15 DocScrutinizer05, OK, I misunderstood you. :) What exactly do you mean? Jul 30 09:57:32 but some guys already did it (edit/patch modest) Jul 30 09:57:48 so it's waiting to be included in CSSU? Jul 30 09:58:00 portrait support and treeviews are buzzwords I recall Jul 30 09:58:10 (modest) Jul 30 09:58:14 as well as "delete mails on server" Jul 30 09:58:18 (tinymail) Jul 30 09:58:35 yes, but what about gpg. Come out with it man! ;-) Jul 30 09:58:45 NFC Jul 30 09:58:54 afaik nobody dared to bother yet Jul 30 09:58:55 RFID Jul 30 10:00:20 DocScrutinizer05, it shouldn't be too hard if you have access to the message text buffer in a reasonable form, just send the stuff to gpg, and read back what it spits out. Maybe use a special program to handle this if neccesary like with APG in Android. Jul 30 10:02:16 I heard several users use claws Jul 30 10:02:27 yeah, that's the one I used on n810. Jul 30 10:02:31 I recall it now. Jul 30 10:03:04 but claws kind of has a lack of a nice interface, it's definitely a port of a desktop program. Jul 30 10:03:51 at least, on the n810 it was, don't see any reason why n900 would be different in that regard. Jul 30 10:04:38 I'll have to see how much I like Modest, if I like it a lot, I might persuade myself to hack on it. Jul 30 10:06:01 otherwise I'd use claws or mutt. Jul 30 10:10:38 modest lacks basically everything I need for maqnaging my mail Jul 30 10:11:58 honestly, any arbitrary webmailer interface is a feature monster compared to modest Jul 30 10:12:00 DocScrutinizer05, then I probably won't bother messing with it. Jul 30 10:12:43 DocScrutinizer05, too bad, the n900 is a sort of power-user's device. Jul 30 10:12:55 I was hoping they'd put in a better email client. Jul 30 10:12:57 I *hate* webmailer, but for N900 it'd be the alternative of choice for my mailing needs Jul 30 10:13:24 well, webmail doesn't have good gpg support, other than cutting and pasting it by hand. :) Jul 30 10:13:39 ;-P Jul 30 10:13:53 web.de once had Jul 30 10:14:13 modulo you had to share your private key of course Jul 30 10:14:17 eww Jul 30 10:14:26 I wouldn't share my private key... Jul 30 10:14:46 well, actually they generated a keypair for you Jul 30 10:14:59 same difference. Jul 30 10:15:05 yup Jul 30 10:15:25 webmeailer by design defeats the purpose of pgp Jul 30 10:15:43 I heard about a "secure" webmail site, one day the police came knocking and insisted they give up their keys, they caved. Jul 30 10:15:55 not web.de, a different one. Jul 30 10:18:45 I actually keep an encrypted (with gpg symetrical encryption option) copy of my key on another server, but it is not usable in that state by others of course. Jul 30 10:19:03 it's for backup. Jul 30 10:20:22 EdLin, hosting your own mail servers is so easy nowadays, it's a wonder people don't do it Jul 30 10:20:44 Hurrian, because they get their email rejected as spam if it comes from an ISP? Jul 30 10:21:10 yes, there's ways around that, but they're a big hassle. Jul 30 10:21:33 not to mention filtering the incoming spam. Jul 30 10:21:34 EdLin, that happens? I would think that spam filters scan for content Jul 30 10:21:55 * ArkanoiD_ got replacement new metawatch Jul 30 10:22:13 of course, there is PGP encryption for sending secure messages over public services, but you still don't keep your files on your server Jul 30 10:22:15 Hurrian, not only content, they care about who connects with them. Let me dig up an old article about what you have to do to avoid gmail and yahoo et. al. dumping your email on the floor. Jul 30 10:24:24 Hurrian, darn, it's an old article on codinghorror.com, I don't think I bookmarked it and site: in search engines isn't turning it up. Jul 30 10:25:04 Hurrian, it isn't as simple as just setting up postfix to send to others. Jul 30 10:26:17 http://www.codinghorror.com/blog/2010/04/so-youd-like-to-send-some-email-through-code.html Jul 30 10:26:19 here Jul 30 10:27:50 you have to get your ISP to set up a reverse PTR record for you. Jul 30 10:28:13 or the big email providers will summarily put it in /dev/null Jul 30 10:28:32 and there's a bit more after that too. Jul 30 10:28:56 Hi,nandtester here-will have to continue with recovering my N900 when back home in Germany...will return tomorrow, so might start on Wednesday...can't do it here,because can't charge my phone here...will let you know,when I continue on that issue Jul 30 10:30:20 EdLin, ouch, wow Jul 30 10:30:33 the sigmask syscalls patch from upstream git applied fairly cleanly :) Jul 30 10:31:05 Hurrian, now you know why people don't set up their own mail servers at home. ;-) Jul 30 10:31:20 other than as leaf nodes to a hub that has done all of that. Jul 30 10:33:14 you can thank malware for making this happen, and Microsoft for making malware have such a happy home on PCs. ;-) Jul 30 10:34:30 I run my own mailserver Jul 30 10:34:52 outgoing mail is routed through my isp smtp though Jul 30 10:35:06 zeq: well, that isn't what we were talking about. Jul 30 10:35:22 obviously anyone can do that part without a hassle. Jul 30 10:36:45 of course, there will be a bit of a hassle if you're messing with exim on Debian, but that's Debian's fault for never fixing it. ;-) Jul 30 10:37:26 but that's a different sort of hassle, of course. Jul 30 10:38:15 Well, I have SPF correctly configured for my domain etc, I have POP3 and IMAP access, and I'm about to add webmail for my gf. Jul 30 10:38:29 Using sendmail! Jul 30 10:38:35 zeq: OK, but it's outbound via your ISP. Jul 30 10:39:00 just because not everywhere is IPv6 connected Jul 30 10:39:04 zeq: I remember the old sendmail that ran as root.... :) Jul 30 10:39:06 IPv6 goes out directly Jul 30 10:39:19 IPv6 <3 Jul 30 10:39:23 IPv4 is via a smarthost Jul 30 10:39:57 my home router unfortunately is not IPv6 compatible. Jul 30 10:40:19 I've thought of upgrading it, but it's otherwise a very good and fast router. Jul 30 10:41:37 zeq, I think fennec memory usage isn't a bug Jul 30 10:41:45 check out http://play.qwazix.com/fennec_mem_stats.tar.gz Jul 30 10:42:08 fennec eats up no less than 150mb even on maemo4 Jul 30 10:42:24 (all stats with firefox start and tmo open) Jul 30 10:42:34 doesn't mean it isn't a bug Jul 30 10:43:00 versions vary from 1.1 -> 17.0 Jul 30 10:43:29 still doesn't mean it isn't a bug Jul 30 10:44:08 :nod: Jul 30 10:44:18 so it isn't a bug in the sense of something introduced in a version Jul 30 10:44:30 no, that would be a regression Jul 30 10:44:31 or maybe It's better to say it isnt a regression? Jul 30 10:44:53 you beat me to it :) Jul 30 10:45:11 anyway, check those stats, maybe there is something useful in there Jul 30 10:45:44 going to try latest fennec Jul 30 10:46:19 I tested fennec 1.1 on N810, official fennec from ovi (fremantle), your fennec, harmattan ovi version and linux desktop Jul 30 10:46:41 I can measure android too if that can be useful Jul 30 10:46:46 and? Jul 30 10:47:00 android check might be useful Jul 30 10:47:01 what are the conclusions? Jul 30 10:47:13 freemangordon, http://play.qwazix.com/fennec_mem_stats.tar.gz Jul 30 10:47:31 I don't have any conclusions, I just packed the data for you. Jul 30 10:47:33 the thing is most devices today have more memory than we do, certainly where desktop firefox runs! Jul 30 10:47:45 I just saw that it is a ram-hungry beast always Jul 30 10:47:51 yeah Jul 30 10:50:15 cannot check the stats right now, i am connected via rd from n900 to my desktop, will check later Jul 30 10:50:19 mozilla development is very feature oriented Jul 30 10:51:49 most developers are focussed on new developments, and probably release branch optimizations don't always get forward ported back into central Jul 30 10:52:09 will check now on android Jul 30 10:52:15 certainly vendor fixes and optimization seem not to Jul 30 10:52:40 Did I ever mention I hate android? Jul 30 10:52:54 maybe :) Jul 30 10:53:23 I'd give a lot of money to have any kind of linux run on the galaxy note Jul 30 10:56:30 AARGH, you need root to take screenshots?? Jul 30 10:57:34 Ah I forgot, it has a samsung app that does that... Jul 30 11:02:44 force close... how sweet.. Jul 30 11:06:51 omg!! fennec android 600mb RAM Jul 30 11:07:16 the native java version Jul 30 11:10:40 hmmm, latest m-c seems much better Jul 30 11:10:53 ~m-c Jul 30 11:10:55 aroun 20MB less usage Jul 30 11:11:03 vi__: mozilla-central Jul 30 11:11:49 so fennec 17.a1-20MB? Jul 30 11:11:57 zeq: but it does not resize correctly without your pathc Jul 30 11:12:04 vi__: less ;) Jul 30 11:12:15 sweet! Jul 30 11:12:28 uses around 80-85 to open TMO, which is pretty much acceptable Jul 30 11:13:28 freemangordon: did I miss a patch? Jul 30 11:13:51 zeq: no, yesterday we concluded that ! patch is not needed Jul 30 11:14:00 now it turns out we were wrong Jul 30 11:14:04 oh Jul 30 11:14:08 ? Jul 30 11:14:19 random failure? Jul 30 11:14:26 well, didn't check it with the patch, but without it it does not resize correctly Jul 30 11:14:37 weird Jul 30 11:15:01 it was resizing correctly for me without the patch Jul 30 11:15:14 no, fennec seems much faster, it seems the rendering code was optimized Jul 30 11:15:18 Currently the patch isn't in the tree, and hasn't been for a couple of weeks Jul 30 11:15:30 yes, and fennec was slow :) Jul 30 11:15:36 that's true Jul 30 11:15:39 now it is microbpfast again Jul 30 11:15:45 good! Jul 30 11:15:47 microb-fast Jul 30 11:15:54 but does not resize correctly Jul 30 11:15:58 I got that :) Jul 30 11:16:02 boo Jul 30 11:16:04 yeah, i know Jul 30 11:16:07 :P Jul 30 11:16:25 did you try it with the missing change? Jul 30 11:16:36 no, i've just tried it Jul 30 11:17:20 the xul version doesn't let me to install it... Jul 30 11:17:37 could you push the change, sorry for that, but it is a nightmare to edit via R-D via 3G Jul 30 11:17:50 I will build it Jul 30 11:18:00 yes I'll apply it and repush Jul 30 11:18:05 qwazix: could you check what exactly is failing? Jul 30 11:21:07 freemangordon, I was trying to install it on android Jul 30 11:21:29 I don't think any of us needs to know why it's failing... Jul 30 11:24:20 freemangordon: pushded Jul 30 11:24:27 pushed* Jul 30 11:27:37 qwazix: sorry, misread, i thought it is failing to UNinstall on MAEMO :D Jul 30 11:27:52 zeq: ok, will pull and rebuild Jul 30 11:28:28 freemangordon, it also did that and HAM crashed, but when I went to remove via apt-get it removed successfully Jul 30 11:28:39 I suppose HAM does autoremove? Jul 30 11:30:44 HAM is awful Jul 30 11:31:08 the only thing it does is reboot after you install packages marked as "reboot after installing" Jul 30 11:31:46 kerio, not only that Jul 30 11:32:02 oh yeah, it also sucks 12 minutes of my cpu's time every time there's a catalogue update Jul 30 11:32:04 or it starts Jul 30 11:32:32 it also closes connections and applications when you upgrade maemo Jul 30 11:32:45 that's what i said Jul 30 11:33:07 it reboots after it installs the upgraded metapackage (marked as sys,reboot) Jul 30 11:33:36 and puts device on offline mode while upgrading Jul 30 11:33:50 and closes all other applications Jul 30 11:33:58 and checks if there is enough battery Jul 30 11:34:05 and enough rootfs space Jul 30 11:34:18 (and spits out an entirely unhelpful message) Jul 30 11:34:20 yeah, i bet the 12 minutes of cpu grinding are to check those things Jul 30 11:35:52 with just a handful of repos enabled! Jul 30 11:36:34 by the way, is there a reason extras-devel doesn't use diffindex? Jul 30 11:41:05 kerio, because diffindex takes a lot of CPU power to crunch? Jul 30 11:41:20 it's painful to apt-get update E-D Jul 30 11:42:35 freemangordon: did you see any of my discussion with DocScrutinizer05 above re sigmask syscalls? Jul 30 11:42:39 fennec xul on android (1.5 dual core, 1GB ram) is slower than on maemo Jul 30 11:42:53 zeq: no Jul 30 11:43:26 qwazix: you're not surprised, are you Jul 30 11:44:23 freemangordon, I knew android sucks, but so much? Jul 30 11:45:23 you said it ;) Jul 30 11:45:38 qwazix, firefox is unholy slow even with their "native uI" Jul 30 11:45:44 our kernel actually lacks pselect, ppoll et al. which is a twofold problem. Our userspace relies quite a lot on pselect, and unfortunately we rely on the buggy implementation in glibc which a simple test program proves it causes hangs. Doc agreed with me we really should backport the syscalls and use them. Jul 30 11:46:21 I've already patched the kp51, attempting to build now Jul 30 11:46:47 what? Jul 30 11:46:54 hehe Jul 30 11:47:04 zeq: ok, awaiting patches Jul 30 11:47:34 zeq: btw you can clone kernel-cssu and work here, Pali will pull the patches for KP from kernel-cssu tree Jul 30 11:47:45 ARM somehow got forgotten despite *me* posting a patch 5 years ago! Jul 30 11:47:46 s/here/there/ Jul 30 11:47:46 freemangordon meant: zeq: btw you can clone kernel-cssu and work there, Pali will pull the patches for KP from kernel-cssu tree Jul 30 11:47:57 Hurrian, with native UI it's acceptable Jul 30 11:48:09 but still is lacking compared to chrome for android Jul 30 11:48:30 still slower than the native browser. it's useless unless you needed something in Gecko. Jul 30 11:48:36 freemangordon: ok. I just need to do a few thinks afk right now... Jul 30 11:48:42 s/thinks/things/ Jul 30 11:48:42 zeq meant: freemangordon: ok. I just need to do a few things afk right now... Jul 30 11:49:52 zeq: ok, when you're back the tree is here: https://gitorious.org/community-ssu/kernel-cssu Jul 30 11:50:04 I know you like gitorious :D Jul 30 11:50:05 freemangordon: http://www.spinics.net/lists/arm-kernel/msg38114.html <-- here's where I tried to fix last time! (and forgot all about it until this morning!) Jul 30 11:50:21 freemangordon: ok thx :) Jul 30 11:50:52 hello Pali, are you there? Jul 30 11:51:00 yes Jul 30 11:51:03 Hi! Jul 30 11:51:22 I have a question about kp51 an u-boot: http://talk.maemo.org/showpost.php?p=1244727&postcount=66 Jul 30 11:51:26 zeq: i'll read your discussion with doc later Jul 30 11:51:45 * zeq is now afk Jul 30 11:52:02 I'm not sure which steps to take Jul 30 11:52:42 install uboot from this thread: http://talk.maemo.org/showthread.php?t=81613 Jul 30 11:53:15 I already have it installed Jul 30 11:53:42 but version is 2012.04-1 Jul 30 11:54:01 how is it best to upgrade? Jul 30 11:54:13 zeq, I give up with fennec for android, the file I linked has htop screenshots but I couldn't save about:memory for xul version Jul 30 11:54:45 download version 2012.04-2 (DEB packages) and install them Jul 30 11:54:48 Pali I need to remove .04-1 first? Jul 30 11:54:58 no Jul 30 11:55:32 so just download, dpkg -i and done? Jul 30 11:55:45 and reboot Jul 30 11:55:53 ok Jul 30 11:56:11 also, if you want to boot kernel-power, you need to install package kernel-power-bootimg Jul 30 11:56:28 it's already installes Jul 30 11:56:32 *installed Jul 30 11:56:50 will I need to reinstall after u-boot? Jul 30 11:57:25 ~tmo is trolls, morons and oxen. Jul 30 11:57:26 vi__: cannot alter locked factoids Jul 30 11:59:57 freemangordon, fennec installs, removes, and autoremoves cleanly now. Jul 30 12:00:44 vi__ why are you mentioning that just now? Jul 30 12:00:46 bennypr0fane, if you installed kernel-power-flasher package, yes reinstall u-boot-flasher Jul 30 12:01:08 but if you did not isntalled kernel-poiwer-flasher package now, reinstalling u-boot AGAIN is not needed Jul 30 12:01:49 no, I meant, after I installed U-Boot, will I need to reinstall kernel-power-bootimg? Jul 30 12:02:59 no Jul 30 12:03:09 just run: u-boot-update-bootmenu Jul 30 12:03:20 this will update uboot bootmenu :-) Jul 30 12:03:40 ok, thanks Jul 30 12:04:46 freemangordon: ping Jul 30 12:10:45 yaaayyy, kp51 successfully booted! thank you Pali! Jul 30 12:11:01 ok Jul 30 12:12:34 is anyone in here using nemo on n900? Jul 30 12:14:23 I can't get Nemo to boot either :-( Jul 30 12:15:26 It should just boot when U-Boot is installed with the sdcard with Nemo image present - but it always boots Maemo. It used to work before though Jul 30 12:50:50 hm, the automounter kinda screwed me up Jul 30 12:51:26 it mounted my vfat partition as i was deleting it to make a separate swap partition :s Jul 30 12:57:05 does anyone use the swapset package? Jul 30 13:02:30 freemangordon: I fetched orig source with debian/rules get_orig_source, unpacked it, then after creating the patch with quilt, what is the correct dpkg-buildpackge command to get clean the tree and attempt to build? I'm obviously doing it wrong because the patches won't apply! Jul 30 13:04:06 (probably because they've already been applied) Jul 30 13:04:17 will removing the back cover also swapoff the swap *partitions* i have on the uSD? Jul 30 13:06:08 freemangordon: never mind, I've got it I think. Jul 30 13:14:13 thumb+compcache is pretty great Jul 30 13:14:18 but the back cover thing is kinda bad :s Jul 30 13:21:12 will removing the back cover also swapoff the swap *partitions* i have on the uSD? <-- to my knowledge, no. Jul 30 13:21:20 sweet :D Jul 30 13:23:14 freemangordon, Pali: I don't suppose anybody has tried building the kernel with the new toolchian? Jul 30 13:23:41 s/toolchian/toolchain/ Jul 30 13:23:42 zeq meant: freemangordon, Pali: I don't suppose anybody has tried building the kernel with the new toolchain? Jul 30 13:23:47 i don't suppose anybody has tried building the kernel with the thumb2 instruction set? :D Jul 30 13:24:03 kerio: not a good idea! Jul 30 13:24:08 why not? :c Jul 30 13:24:10 (due to the errata) Jul 30 13:24:16 well, you'd add the errata too Jul 30 13:24:22 the kernel handles the hw bug in userspace Jul 30 13:24:32 who's going to do it for the kernel? ;) Jul 30 13:24:39 zeq, I'm using toolchain in madde Jul 30 13:24:39 zeq, and more times I used ubuntu arm toolchain to build u-boot Jul 30 13:24:39 (Ubuntu/Linaro 4.6.3-1ubuntu5) Jul 30 13:26:06 now kernel + u-boot fit into kernel nand area Jul 30 13:26:11 ok, well, it's not working with linaro-4.7.2012.7 + binutils-2.22 Jul 30 13:26:21 so building kernel with thumb2 (to reduce size) is not needed Jul 30 13:26:24 Pali: k Jul 30 13:26:29 drivers/char/hw_random/omap3-rom-asm.S: Assembler messages: Jul 30 13:26:31 drivers/char/hw_random/omap3-rom-asm.S:20: Error: selected processor does not support ARM mode `smc #1' Jul 30 13:26:32 why do we need to embed the kernel in uboot again? Jul 30 13:26:50 kerio and where you want to add kernel? Jul 30 13:27:21 Pali: idk, the rootfs? Jul 30 13:27:26 maybe /boot Jul 30 13:27:29 kernels usually stay there Jul 30 13:27:36 don't they? Jul 30 13:27:39 kerio, sorry but ubifs driver is too big in uboot Jul 30 13:27:48 and also do not working correctly Jul 30 13:27:50 you'd have a lot of space if you didn't put the kernel in it :D Jul 30 13:27:52 k Jul 30 13:27:52 :c Jul 30 13:28:09 I was not able to read ubifs rootfs in qemu Jul 30 13:28:12 hm Jul 30 13:28:14 shame Jul 30 13:28:18 what about optfs? Jul 30 13:28:21 (where I deleted kernel and added only uboot binary) Jul 30 13:28:30 can uboot access the emmc? Jul 30 13:28:33 kerio, kernels are now read from MyDocs Jul 30 13:28:40 ew, vfat :s Jul 30 13:28:54 yes, fat driver is ok in uboot Jul 30 13:28:59 but what happen if you delete kernel? Jul 30 13:29:05 you will have your device unbootable Jul 30 13:29:10 what happens if you nandtest all the mtd partitions? Jul 30 13:29:19 and you will need to reflash full device Jul 30 13:29:20 that dude did that Jul 30 13:29:28 kerio, you must be root for nandtest Jul 30 13:29:30 people will always find a way to fuck things up Jul 30 13:29:43 Pali: that's why you put the kernel in optfs Jul 30 13:29:45 but deleting files exported to Windows computer is really easy Jul 30 13:29:54 read-only for root :) Jul 30 13:30:11 and if you broke optfs and you want to backup data? Jul 30 13:30:47 really, working kernel in kernel nand area is good idea Jul 30 13:30:53 good point Jul 30 13:30:59 you still have one working kernel Jul 30 13:31:07 vi__: pong Jul 30 13:31:29 and if you have installed bootmenu you can export eMMC to host computer Jul 30 13:31:42 zeq: ithere was an additional compilers swith in new compilers to support smc instructions Jul 30 13:31:59 freemangordon: I've patched both cssu and kp with the new syscalls Jul 30 13:32:02 or enable ssh via usb network to fix broked maemo parts (I used it more times and I did not need to reflash device) Jul 30 13:32:16 zeq: the rotate/painting issue in fennec might a bug, not related to ! Jul 30 13:32:32 zeq: kernel boots? Jul 30 13:32:35 freemangordon: so they've broken our fix? Jul 30 13:33:00 freemangordon: no, idea, build failed due to lack of your above mentioned switch Jul 30 13:33:04 zeq: well, looks like more general bug, as scrolling and zoom are broken too Jul 30 13:33:07 it should do though Jul 30 13:33:11 zeq: use 4.2.1 Jul 30 13:33:16 patch applied from git quite cleanly Jul 30 13:34:28 I don't think I happen to have the maemo sdk 4.2.1 compiler :P Jul 30 13:34:34 Pali: what's up with KPS 0.15, why x-fade removed it from repos? Jul 30 13:34:40 zeq: you have Jul 30 13:34:50 FREMANTLE_ARMEL target Jul 30 13:35:12 freemangordon, it was not in repos Jul 30 13:35:20 only in interface Jul 30 13:35:35 but I'm not able to push new version, it still reject it Jul 30 13:35:59 Pali: try with 0.16 Jul 30 13:36:03 it should work Jul 30 13:36:37 "Changelog: no changes, maemo.org is stupid" Jul 30 13:36:50 freemangordon: 0.15r1! Jul 30 13:37:17 kerio: leave the maintainers to decide, ok? Jul 30 13:37:27 I do not know if it will work... I pushed version 0.15, build successfull and maemo.org did not imported it into rpeository Jul 30 13:37:43 Pali: yes, because there is 0.15 Jul 30 13:37:57 does not matter if it is removed or not Jul 30 13:38:01 but there was no 0.15 before I pushed 0.15 first time Jul 30 13:38:03 it exists in DB Jul 30 13:38:17 Pali: well, autobuilder, you know :D Jul 30 13:38:29 also ivgalvez wrote that merlin will got maemo package permissions Jul 30 13:38:43 I will wait some days Jul 30 13:39:02 I know, but he still does not have, and noone knows when it will happen Jul 30 13:39:24 Pali: don't hold your breath Jul 30 13:39:29 better push 0.16 Jul 30 13:41:07 zeq: you have FREMANTLE_ARMEL target, aren't you? Jul 30 13:47:36 zeq: " mfinkle: but nightly has some viewport bustedness too" Jul 30 13:47:41 freemangordon: Yes, I have FREMANTLE_ARMEL target :) cs2007q3-sb8 Jul 30 13:48:06 well, that is 4.2.1, try gcc -v Jul 30 13:48:24 indeed it is, I don't know what I was thinking :) Jul 30 13:48:26 that is what mortals use to build stuff for Fremantle :P Jul 30 13:48:44 zeq: saw my paste from #mobile? Jul 30 13:49:08 for some reason I thought there was some maemo toolchain Jul 30 13:49:34 it is, in FREMANTLE_ARMEL :P Jul 30 13:49:36 yes, I see that, could well be what you are seeing Jul 30 13:49:45 i bet on it Jul 30 13:50:02 freemangordon: cs2007q3-sb8 is a scratchbox toolchain Jul 30 13:50:20 I know, I was kidding Jul 30 13:50:23 :) Jul 30 13:50:40 anyway, KP is perfectly compilable with it Jul 30 13:50:43 I'll try building with it now Jul 30 13:50:46 as is KCSSU Jul 30 13:51:01 will be afk while it compiles though Jul 30 13:51:26 zeq: it is fast, some 10 minutes Jul 30 13:52:14 zeq: " and still broken today" Jul 30 13:52:48 crap, I don't have my repos set up properly on this target apparently! Jul 30 13:56:16 that's better :) Jul 30 13:56:50 AaronMT: or inwards busted too Jul 30 13:56:52 zooming and panning just totally boned Jul 30 13:56:54 AaronMT: where is the bug Jul 30 13:56:59 zeq: ^^^ Jul 30 13:57:02 hehe Jul 30 13:57:07 LOL Jul 30 13:57:25 I hate to be right :( Jul 30 13:57:51 keep an eye out when they've fixed it, and I'll rebase again Jul 30 13:58:30 right now I must go do some housework! Jul 30 13:58:36 (kernel build is underway) Jul 30 13:59:27 ok Jul 30 14:02:47 freemangordon: If I have thumb repo enabled should I disable community-testing? Jul 30 14:03:21 vi__: no Jul 30 14:04:35 freemangordon: TY Jul 30 14:05:22 hm, what's the rationale for preferring external swap to internal again? Jul 30 14:05:28 internal is faster, isn't it Jul 30 14:06:54 kerio: I think some are worried that internal causes undue flash wear on a non-removable resource. Jul 30 14:07:09 i see Jul 30 14:07:12 good point Jul 30 14:07:25 also 2gb is bigger than 700mb, so swap fragments less Jul 30 14:35:01 kerio: It is to reduce IO collisions. Jul 30 14:35:55 i see Jul 30 14:35:55 If you are trying to R/W to eMMC it is just a pain in the ass to have to swao there too. Jul 30 14:38:09 zeq, ping Jul 30 14:38:13 fingers-crossed, going to reboot with new e Jul 30 14:38:17 kernel Jul 30 14:38:31 MohammadAG: pong Jul 30 14:38:32 I'll ask after the reboot then :P Jul 30 14:40:33 it seems to be working... :D Jul 30 14:40:58 zeq, openwrt again :) Jul 30 14:40:59 so MohammadAG, ask away :) Jul 30 14:41:04 ah Jul 30 14:41:09 ok Jul 30 14:41:27 but basically, this router isn't well known :p Jul 30 14:41:39 DSL-2760U Jul 30 14:42:15 did I miss the question? :) Jul 30 14:42:28 MohammadAG: zeq, openwrt again Jul 30 14:42:28 I don't know where to start :P Jul 30 14:42:30 zeq: ah Jul 30 14:42:32 zeq: ok Jul 30 14:42:34 MohammadAG: but basically, this router isn't well known Jul 30 14:42:36 MohammadAG: DSL-2760U Jul 30 14:42:40 I telnet'd to the stock firmware Jul 30 14:42:41 oh, right.... Jul 30 14:43:01 telnet says BCM96328 Broadband Router Jul 30 14:43:44 MohammadAG: go on Jul 30 14:44:10 that's what I got :P Jul 30 14:44:22 it's similar to 2765U Jul 30 14:44:43 it's the internals that matter Jul 30 14:45:29 as long as you build for the right chipset and you have the means to flash it you should be able to get something running... Jul 30 14:45:43 cpu model : Broadcom4350 V7.5 Jul 30 14:46:15 do I need some toolkit to build? Jul 30 14:46:31 there's an sdk for OpenWrt Jul 30 14:46:31 also I found that I have to use kernel 3.3 Jul 30 14:46:45 it's pretty straightfoward Jul 30 14:47:02 it's configured like the Linux kernel Jul 30 14:47:22 I found this http://translate.google.com/translate?hl=en&sl=zh-CN&u=http://hi.baidu.com/%25BE%25C5%25B6%25A6%25CB%25DC%25BD%25BA/blog/item/d464a924cd063b2dc99559fc.html&prev=/search%3Fq%3DOpenWRT%2B2760U%26hl%3Den%26client%3Dubuntu%26hs%3DbmX%26channel%3Dfs%26biw%3D1366%26bih%3D557%26prmd%3Dimvnsfd&sa=X&ei=-Z0WUKXFE-eQ4gTO5YC4DQ&ved=0CHQQ7gEwBA Jul 30 14:47:43 your best bet it to build from vcs Jul 30 14:48:34 if you build from the VCS you maximize the chances of your board being supported Jul 30 14:49:30 https://forum.openwrt.org/viewtopic.php?id=15201 Jul 30 14:49:43 and is there a way to make sure I don't brick? Jul 30 14:49:47 since I can't find the stock firmware anywhere Jul 30 14:50:05 can't you back it up from the router web UI? Jul 30 14:50:11 the firmware itself? Jul 30 14:50:25 no, the Web UI is as restricted as it gets Jul 30 14:50:29 there's usually an option to back up the current firmware Jul 30 14:50:34 :( Jul 30 14:50:37 that sucks Jul 30 14:50:37 this is D-Link FW Jul 30 14:50:51 under the hood it's a good router Jul 30 14:50:58 it has telnet, sshd, samba Jul 30 14:51:13 it connects with pppd Jul 30 14:51:36 are there any available tools in shell for firmware? Jul 30 14:51:43 # cat /proc/version Jul 30 14:51:43 Linux version 2.6.30 (root@localhost.localdomain) (gcc version 4.4.2 (Buildroot 2010.02-git) ) #1 Tue Nov 2 14:21:47 CST 2010 Jul 30 14:51:50 that's pretty recent Jul 30 14:52:08 nvram and nvramUpdate Jul 30 14:52:12 I'm surprised you can't find the firmware for it Jul 30 14:52:21 I can find the russian firmware Jul 30 14:52:26 but mine is BN, whatever that is Jul 30 14:53:02 actually this might be it Jul 30 14:53:04 http://www.bergfiles.com/i/bf4a51d40fh32i0 Jul 30 14:53:29 usually routers have a firmware recovery mode Jul 30 14:53:38 so you can reflash your bricked device Jul 30 14:53:38 DSL-2760U-BN_1.07_11022010_cfe_fs_kernel <-- filename Jul 30 14:53:46 this has JTAG I think Jul 30 14:54:32 I mean over ethernet :) Jul 30 14:54:44 unless you can use the JTAG? Jul 30 14:55:15 Your best bet would probably be the OpenWrt forum ^ Jul 30 14:55:31 I'll be back later, but I need to go again now... Jul 30 14:59:49 freemangordon, Pali: the new kernel works, I need to patch the linux-headers too, to match before rebuilding libc6. Jul 30 14:59:58 I really need to learn chinese Jul 30 15:00:08 * zeq is afk Jul 30 15:38:58 freemangordon, Pali: how should linux-kernel-headers be versioned? shall I just bump the date? Jul 30 15:39:50 DocScrutinizer05: ^ Jul 30 15:39:51 zeq, what is problem of linux-kernel-headers? Jul 30 15:40:52 Pali: I'm adding the missing TIF_RESTORE_SIGMASK syscalls Jul 30 15:41:02 ie pselect() etc Jul 30 15:41:11 zeq, that should be patched in kernel-power Jul 30 15:41:14 package: linux-kernel-power-headers Jul 30 15:41:32 glibc needs to be rebuild with the new syscalls Jul 30 15:41:47 that builds against linux-kernel-headers Jul 30 15:41:58 or am I missing something? Jul 30 15:42:14 please do not touch SDK kernel packages (like linux-kernel-headers) Jul 30 15:42:31 kernel-power/kernel-cssu is place to patch Jul 30 15:42:57 linux-kernel-power-headers provides same what linux-kernel-headers Jul 30 15:43:12 are you sure? Jul 30 15:43:28 I thought it provided headers in /usr/src Jul 30 15:43:37 with symlink via /lib/boot... Jul 30 15:43:46 zeq, no Jul 30 15:44:12 linux-kernel-power-headers comes from kernel-power; linux-kernel-headers comes from kernel Jul 30 15:44:36 both has same, files in /usr/include Jul 30 15:45:09 Pali: sorry, ok, but there are packages providing kernel headers as described abovee Jul 30 15:45:12 above* Jul 30 15:45:48 kernel headers are in kernel-power-headers (or kernel-headers) Jul 30 15:46:28 user include files (e.g. /usr/include/linux/something.h) are in linux-kernel-headers and linux-kernel-power-headers Jul 30 15:46:31 Pali: see dpkg -L kernel-power-headers Jul 30 15:46:43 zeq, linux-kernel-power-headers Jul 30 15:46:56 add *linux-* prefix Jul 30 15:47:07 oh, right Jul 30 15:47:43 and that gets built along with the kernel, yes? Jul 30 15:47:50 or no...? Jul 30 15:49:48 answer is yes Jul 30 15:49:58 So I've already updated the headers :) Jul 30 15:50:04 that makes like easier :) Jul 30 15:50:21 Now I just need to build glibc against the new headers Jul 30 15:50:49 with the only issue the lack of support in sb/qemu :( Jul 30 15:54:07 zeq, yes linux-kernel-power-headers are build with kernel-power Jul 30 15:54:09 the most recent available qemu devkit binary does have support for pselect() Jul 30 15:54:32 but not ppoll() for some odd reason Jul 30 15:54:37 zeq, use linaro qemu Jul 30 15:55:00 there's a devkit I can download? Jul 30 15:55:12 Pali: do you have a link? Jul 30 15:55:14 linaro qemu is included in ubuntu 12.04 Jul 30 15:55:23 I don't have ubuntu Jul 30 15:55:36 but I guess I can extract it from the deb Jul 30 15:55:42 if you need binaries, just download deb packages Jul 30 15:55:59 that works with scratchbox though? Jul 30 15:56:26 http://packages.ubuntu.com/precise/qemu-user-static Jul 30 15:56:29 I spent a good part of my weekend trying to get >1.0 qemu working with scratchbox Jul 30 15:56:31 this is static linked version Jul 30 15:56:51 static linked version does not depends on other libraries, so should work Jul 30 15:57:42 but I did not tested it in scratchbox... Jul 30 15:58:15 scratchbox requires qemu to be linked with libsb Jul 30 15:58:29 along with very strict glibc version Jul 30 15:58:36 it's a PITA Jul 30 15:59:07 life is much easier with SB2 Jul 30 15:59:42 why must be qemu linked with libsb? Jul 30 16:00:09 because scratchbox 1 is very hacky Jul 30 16:00:14 also dynamic linked version is here: http://packages.ubuntu.com/precise/qemu-system Jul 30 16:00:21 maybe you can try to recompile it Jul 30 16:00:44 I have qemu on my system but SB requires --sbox-call parameter Jul 30 16:00:45 so other reason why not ti use sb1 Jul 30 16:01:29 I'm going to switch to SB2, but I wanted to make sure nobody else *had* to just because I'm fixing the pselect() hang bug :/ Jul 30 16:04:25 I patched recent qemu from debian to have --sbox-call, and linked it with libsb, but I didn't have the exactly right glibc to link with because recent qemu won't build against the ancient version sb is built with. Jul 30 16:05:00 I was chasing my own tail for hours :( Jul 30 16:09:50 so how does one edit a file with no nano/vi Jul 30 16:09:53 or sed Jul 30 16:10:46 MohammadAG: creating a new file can be done with cat > file < do you have ed? Jul 30 16:11:32 zeq, I gave up on OpenWRT on the thing Jul 30 16:11:42 I just want to hand over all IPs to my router Jul 30 16:12:18 Creating a custom OpenWrt build is quite a task Jul 30 16:12:20 I edited the file, dropped http server, still doesn't work Jul 30 16:13:01 * zeq is afk again Jul 30 16:15:14 oh yay, it works outside LAN Jul 30 16:22:50 Here's a good WTF Jul 30 16:23:01 In a real-time monitoring application for our server Jul 30 16:23:38 if a user, connected and actively using said server, has an "f" as the third letter of his username, he may not be displayed in the real-time monitoring app. Jul 30 16:24:19 and by "he may" I mean, there's a very precise 4/10 chance he may not be displayed. Jul 30 16:25:16 so user ffuuuuu will always be displayed correctly, but user fffuuuu will not. Jul 30 17:16:49 lol Jul 30 17:21:12 zeq: I still fail to see why you need new lqemu Jul 30 17:21:19 qemu even Jul 30 17:21:46 we don't need it working in SB, but on the device Jul 30 17:21:53 AIUI Jul 30 17:21:56 that's true Jul 30 17:22:25 except for anything build related on sb Jul 30 17:22:32 binaries build in SB (with old libc and qemu) will work (hopefully) on the device Jul 30 17:22:33 I don't know how much that applies to Jul 30 17:22:48 but then it doesn't matter if sb has a different C lib Jul 30 17:22:57 (as long as it's the same version) Jul 30 17:23:06 it's ABI that matters afterall Jul 30 17:23:07 zeq: we'll not upgrade libc in SB Jul 30 17:23:21 only on device Jul 30 17:23:32 that way there is no need for new qemu Jul 30 17:23:37 yeah, it doesn't matter if one uses glibc emultation and the kernel syscall Jul 30 17:23:44 yep Jul 30 17:23:45 (and your time is saved) Jul 30 17:24:13 yep Jul 30 17:24:45 zeq: BTW https://bugzilla.mozilla.org/show_bug.cgi?id=778580 Jul 30 17:24:46 04Bug 778580: was not found. Jul 30 17:24:57 see the duplicates :D Jul 30 17:25:14 wow Jul 30 17:25:28 somebody really screwed up! Jul 30 17:25:31 :) Jul 30 17:31:28 building glibc against patched headers Jul 30 17:31:57 good for zeq :P Jul 30 17:32:23 will be if it works ;) Jul 30 18:26:59 apart from running out of space on my rootfs, it works! Jul 30 18:36:36 I don't suppose there's any way to compact the dpkg status and available databases? Jul 30 18:39:59 why? Jul 30 18:42:07 I've run out of rootfs, and they seem to be taking a lot of it Jul 30 18:42:28 right now though, I don't seem able to boot! O_o Jul 30 18:43:25 I had to delete bootimg to get the libc6 install to complete (run out of that much space) Jul 30 18:44:12 that shouldn't stop it booting though, or... Jul 30 18:44:17 zeq that is heavy. Jul 30 18:44:30 If you can sy Jul 30 18:44:35 oops Jul 30 18:44:56 if you can boot, you can symlink shitloads of stuff to opt. Jul 30 18:45:09 locales etc Jul 30 18:45:37 a lot of stuff in /usr/bin that should be on opt etc. Jul 30 18:45:38 I don't think it's going to :( Jul 30 18:45:40 what a pain Jul 30 18:45:46 :( Jul 30 18:45:48 ze Jul 30 18:46:05 the fuck is wrong with my keyboard? Jul 30 18:49:48 you know, it gets worse Jul 30 18:50:10 apparently my MyDocs is corrupt too Jul 30 18:53:30 BM terminal is working :) Jul 30 18:55:50 why would BM terminal work but normal boot not? Jul 30 18:56:22 I think something broke when it ran out of rootfs Jul 30 18:57:15 zeq: how big is libc? Jul 30 18:58:00 ~800kb Jul 30 18:58:30 funny thing is I've now got lots of rootfs space!?! Jul 30 18:58:41 24M Jul 30 18:59:00 freemangordon: is bootimg needed to boot? Jul 30 18:59:50 I thought it would work with just the image on NAND Jul 30 19:02:37 zeq: bootimg is not needed Jul 30 19:03:25 I think my ubifs went crazy Jul 30 19:03:43 after rebooting from BM I've rebooted without problems Jul 30 19:06:13 freemangordon: you saw me say it worked? Jul 30 19:06:39 kernel + libc I mean Jul 30 19:16:28 ze Jul 30 19:17:24 think my battery might be playing up Jul 30 19:17:47 not charging Jul 30 19:18:07 zeq: great Jul 30 19:18:19 libc i mean :) Jul 30 19:18:37 yeah, that went well, my battery not so much! Jul 30 19:18:48 I'll dig out a spare Jul 30 19:21:00 zeq: did you run the test program? Jul 30 19:21:10 I can't boot again Jul 30 19:21:30 all my batteries are flat and seem not to want to change! Jul 30 19:21:35 charge Jul 30 19:22:00 phone fault? Jul 30 19:25:36 seems to be charging now Jul 30 19:26:38 I'll let it charge to green Jul 30 19:33:10 I need to stop running my batteries down so low Jul 30 19:33:16 I'm up and running again Jul 30 19:33:21 yay Jul 30 19:33:22 call me docFool! When looking into calendar before elaving office, I noticed I am tagged 'holiday' for the whole week Jul 30 19:33:57 my locales are b0rken again! hehe Jul 30 19:34:21 too late, I already got appointments for wednesday and fri Jul 30 19:34:41 zeq: wtf are U doing? Jul 30 19:35:06 I'm running my new libc with pselect() support Jul 30 19:35:10 :D Jul 30 19:35:24 kernel pselect() support Jul 30 19:35:36 how's that killing the locale? Jul 30 19:35:42 goodness knows Jul 30 19:36:01 you fixed it this morning? Jul 30 19:36:12 yep Jul 30 19:36:45 no clue what been wrong? Jul 30 19:36:47 zeq, lol, own patches are hardest to back or forward-port;) Jul 30 19:37:14 zeq, don't worry, my .ogg and theora playback was also killed for no reason, only nuking everything gstreamer related and installing from scratch fixed it Jul 30 19:37:28 no reason why locale couldn't act the same ;) Jul 30 19:37:45 XP Jul 30 19:41:02 something wrong with the libc6 locale support I think :S Jul 30 19:41:09 weird Jul 30 19:41:18 X-Fade: Woody14619 ping Jul 30 19:41:28 X-Fade: Woody14619 apps4meego is down Jul 30 19:42:00 Is it? Jul 30 19:43:20 Woody14619: 502 Jul 30 19:43:23 locales are available, but changing locale triggers the watchdog! Jul 30 19:43:29 Is that even Nokia hosted? Jul 30 19:43:42 Woody14619: not only me getting it, as well client wouldn't work too Jul 30 19:43:53 that's how it seems Jul 30 19:44:41 Woody14619: i dunno, but last we had thiis proble X-Fade fixed it, as he is not much around here, you maybe know way to contact him? btw ask him who else can be contacted in emergency cases Jul 30 19:45:58 DocScrutinizer05: do I need to do something to update the watchdog watch list? Jul 30 19:46:04 X-Fade is off for the night. I can try a back channel, but good odds he's aware of the issue. When a server goes down, most IT groups have pagers going off... Since local time is well after work hours it may take sometime for them to get it fixed though. Jul 30 19:46:50 zeq, You realize that app4meego is doomed, anyway? Jul 30 19:46:59 i.e. no plans to support it after nokia plugs off? Jul 30 19:47:06 until i missed something for those 3 weeks Jul 30 19:47:12 s/for/due/ Jul 30 19:47:12 Estel_ meant: until i missed something due those 3 weeks Jul 30 19:47:23 zew, sorry for pinging You, i mean ZogG Jul 30 19:47:31 zeq even Jul 30 19:47:37 no worries Jul 30 19:48:43 Woody14619: it's ok, i just wonder if her knows Jul 30 19:58:09 freemangordon: ping Jul 30 19:59:29 zeq: watchdog watchlist? no such thing on maemo5, thanks to the holy flying spaghettimonster Jul 30 20:00:07 ok, well something definitely seems to be triggering the watchdog :S Jul 30 20:00:27 zeq: it's just fremantle doesn't handle locale in any sane or faintly familiar way Jul 30 20:01:29 I don't understand why rebuilding the same libc works fine but triggers reboot on setting locale Jul 30 20:01:36 and probably some process that thinks it can't live without *standard* (as in Nokia standard) locale infra will blow chunks and pull the whole system into orcus, via dsme Jul 30 20:01:57 I'm still using posix-locales Jul 30 20:02:07 fremantle doesn't Jul 30 20:02:13 even reinstalled it to make sure, and the locales are available in the list Jul 30 20:02:17 you can't rebuild locale-cache Jul 30 20:02:46 so wtf is happening? Jul 30 20:03:39 WTF did Nokia make the translations secret source? Jul 30 20:04:33 for the fun of it, and NIH Jul 30 20:05:50 look into modest (for example) .desktop file. there's the age old known way to have name[EN]=foo name[ZC]=bar. Nokia doesn't use that and invented their own shit Jul 30 20:06:42 can't we just run sed on all this and make it standard l10n? Jul 30 20:07:21 since we dunno how to extract the real names from locale-blob, we can't Jul 30 20:07:31 I still don't understand why rebuilding the same glibc version breaks the locale cache Jul 30 20:07:49 voodoo programing. Jul 30 20:07:51 ;) Jul 30 20:07:59 hmmm Jul 30 20:08:14 something has to be decoding it Jul 30 20:08:27 * RST38h moos at Doc Jul 30 20:08:42 ah, the locales... Jul 30 20:08:46 Name=mcen_ap_name Jul 30 20:08:47 Comment=mcen_ap_name_thumb Jul 30 20:08:49 Exec=/usr/bin/modest -t 30 Jul 30 20:08:50 WTF?! Jul 30 20:09:01 is there a patch missing in the libc6 src pkg? Jul 30 20:09:12 mcen_ap_name_THUMB???????? Jul 30 20:09:24 LOL Jul 30 20:10:09 aaah thumbnail Jul 30 20:10:22 ah Jul 30 20:10:31 anyway that's your modest.desktop Jul 30 20:10:39 if you haven't guessed Jul 30 20:10:50 I don't need to guess Doc Jul 30 20:11:02 I'm not getting any but that crap Jul 30 20:12:01 maybe the glibc patches aren't getting applied, I'm going to check carefully Jul 30 20:12:37 how many bloody patches did they need?!? Jul 30 20:13:28 I once planned to do a lil fancy, open a requester with all the available apps' names and user tags those who shall autostart on boot. On finding this BULLSHIT in modest.desktop Name=mcen_ap_name I gave up Jul 30 20:15:12 there's obviously ZERO way to transcode mcen_ap_name to "email" (or whatever it's in your locale) on cmdline Jul 30 20:15:37 unless you build your own RE'd lookup table Jul 30 20:15:53 thanks Nokia! RLY Jul 30 20:16:33 >:-( Jul 30 20:16:48 hi RST38h Jul 30 20:17:49 I mean, what weapon needs product manager to use to make devels code such BS? Jul 30 20:18:25 zeq: pong Jul 30 20:19:09 zeq: I suspect using HAM to upgrade libc won;t trigger the watchdog ;) Jul 30 20:19:28 freemangordon: it's not that installing libc triggers it Jul 30 20:19:34 but? Jul 30 20:19:38 but the locale gets cleared Jul 30 20:19:45 it breaks locales? Jul 30 20:19:46 then when setting it reboot Jul 30 20:19:54 yeah Jul 30 20:20:01 zeq: I suspect your device is FUBAR Jul 30 20:20:12 Doc reckons it's the crazy Nokia translations Jul 30 20:20:12 better falsh it Jul 30 20:20:20 flash even Jul 30 20:20:35 zeq: you have only ione n900? Jul 30 20:20:38 *one Jul 30 20:20:48 freemangordon: it did this, this morning, when I had a glibc with broken pselect() Jul 30 20:21:01 zeq: check /var/*/*/dsme/rebootcounters or whatever it's called Jul 30 20:21:13 when I reinstalled stock libc, _and_ then reinstalled posix-locales Jul 30 20:21:17 it was back to normal. Jul 30 20:21:35 vi__: zeq has only one n900, can we do something about it ;) Jul 30 20:21:58 Woody14619: ^^^ Jul 30 20:22:59 zeq: I have to go afk, GF needs my attention ;), but anyway I think you need a development device Jul 30 20:23:10 I'm confused doc... what are you wanting me to look at? Jul 30 20:23:28 freemangordon: maybe Estel_ can sell me one Jul 30 20:23:36 Woody14619: a development device for zeq Jul 30 20:23:36 see you later Jul 30 20:23:40 dd Jul 30 20:23:43 oops Jul 30 20:23:44 bb Jul 30 20:23:49 :) Jul 30 20:24:22 zeq: naah, it will cost you an arm, a leg and 30 years of slavery :P Jul 30 20:24:25 bb guys Jul 30 20:24:35 Given what's going on with Nokia right now, I'm not sure we can move something like that. :( Jul 30 20:24:36 DocScrutinizer05: reboot_count currently at 20 Jul 30 20:24:51 going to set locale and see where it ends up Jul 30 20:24:55 you found the friggin files? Jul 30 20:24:58 path? Jul 30 20:25:37 hmm, GF playing angry birds on n950 :D:D:D Jul 30 20:26:07 Woody14619: could you organize some donation from the community Jul 30 20:26:17 we already did that for DavyP Jul 30 20:26:25 zeq, sell what?:) Jul 30 20:26:31 n900 Jul 30 20:27:03 sure, I can Jul 30 20:27:39 For devoted developer like zeq I could even do it under costs, as a way of my donation (i'm to poor to donate enough in monetary way) Jul 30 20:27:46 where's the friggin dsme logs about reboots?? Jul 30 20:27:56 Estel_: lets see if community can donate one device like it did for DavyP Jul 30 20:28:03 vi__: where are you man Jul 30 20:28:05 zeq, if You're interested, I can find a device for 150 euro, of course reinforcing port myself Jul 30 20:28:06 DocScrutinizer05: what is the prompt after setting to save locale? Jul 30 20:28:10 zeq: ^^^ Jul 30 20:28:14 freemangordon, I see Jul 30 20:28:27 waiting for reboot Jul 30 20:28:30 sure, if Community can, why not :) Jul 30 20:28:47 that would be cool :D Jul 30 20:28:53 zeq: ? prompt after setting to save locale ?? Jul 30 20:29:03 please rephrase Jul 30 20:29:05 merlin1991 bought device the same way from me, and I think he is satisfied :) Jul 30 20:29:20 (cheapest I was able to find, exactly 150 euro) Jul 30 20:29:54 Estel_: not so long ago zeq had a problems with his washing machine, better spend those on a new one :P Jul 30 20:30:02 Estel_: if I can't get one from community I'll probably take you up Jul 30 20:30:15 freemangordon, so either vi__ and DavyP way, or ping me zeq or You :) Jul 30 20:30:17 freemangordon: yeah... Jul 30 20:30:26 priorities eh? Jul 30 20:30:29 washing machine, huh. dish, or clothes one? Jul 30 20:30:30 yeah Jul 30 20:30:36 GFs ;) Jul 30 20:30:51 ah, this kind of washing machine? Jul 30 20:31:01 dafaq, I can't find the dsme logs Jul 30 20:31:03 sounds quite sexist, though Jul 30 20:31:12 Estel_: it does not matter, they are not cheap :D Jul 30 20:31:15 zeq: where's your bootcount? Jul 30 20:31:27 zeq: path? Jul 30 20:31:33 freemangordon, sure, but I'm experienced on repairing both, and many times, problems are trivial to solve :D Jul 30 20:31:45 repairing GFs? Jul 30 20:31:46 it's now 0 Jul 30 20:31:52 unless You mean GF as washing machine, meh Jul 30 20:31:56 ? Jul 30 20:31:58 Estel_: :P Jul 30 20:32:02 freemangordon, this one too ;) Jul 30 20:32:14 photo aplications required, though Jul 30 20:32:21 erm I mean Jul 30 20:32:22 zeq: what command do you issue to get result ? Jul 30 20:32:24 photo of problem Jul 30 20:32:36 not necessary macro Jul 30 20:32:56 Estel_: anyway, I think vi__ will do the job wight Jul 30 20:32:58 freemangordon, now pray that Your Gf doesn't break into your irc logs ;) Jul 30 20:33:00 *right Jul 30 20:33:02 mee to Jul 30 20:33:05 me even Jul 30 20:33:15 and me too ;) and me! and me! Jul 30 20:33:17 I'm wondering if it might be I'm clicking on a reboot icon but can't read the label... Jul 30 20:33:19 all 3 of us Jul 30 20:33:28 hehe Jul 30 20:33:29 can somebody please try changing their language Jul 30 20:33:41 da, zeq, zdrastwujcie Jul 30 20:33:41 zeq: why? Jul 30 20:33:41 and tell me what prompt appears after saving Jul 30 20:33:51 hold on Jul 30 20:33:53 could it be "reboot now" Jul 30 20:33:57 ? Jul 30 20:34:11 zeq: for sure you need a reboot if you change the language Jul 30 20:34:23 I need help. Jul 30 20:34:25 lemme fire my devel device Jul 30 20:34:28 ... Jul 30 20:34:32 vi__: we need to Jul 30 20:34:36 vi__, with Gf? suuure Jul 30 20:34:38 too Jul 30 20:34:52 start with taking a screwdriver... Jul 30 20:34:58 hah, ms vi_ would roundhouse you to the head. Jul 30 20:35:05 * DocScrutinizer05 is off for dinner, since this thing drives him mad Jul 30 20:35:17 my keymap has gone insane. Jul 30 20:35:18 enjoy doc Jul 30 20:35:28 and I cannot get it back. Jul 30 20:35:36 vi__, setxkbmap ? Jul 30 20:35:40 vi__: could you please read the backscroll Jul 30 20:35:43 or rx-51? Jul 30 20:36:18 zeq, BTW, are you comfortable with soldering iron? Jul 30 20:36:31 zeq: "All applications will be closed and device will restart. Continue anyway?" Jul 30 20:36:58 * freemangordon is afk Jul 30 20:37:05 hw do I tell what keymap is loaded? Jul 30 20:37:36 ah! Jul 30 20:37:55 okay it's just not translating then Jul 30 20:38:11 I've rebuilt without -mthumb Jul 30 20:38:24 lets see what happens Jul 30 20:38:54 my keyboard has reverted back to stock layout. Jul 30 20:39:07 if that doesn't work, I'll set it back to softfloat Jul 30 20:39:12 And it wont go back to the way I want. Jul 30 20:39:23 I have setxkvd us Jul 30 20:39:32 maybe the Nokia translation mechanism uses FP?!? O_o Jul 30 20:39:34 but it does fuck all Jul 30 20:39:42 WTF is going on. Jul 30 20:40:01 I am about ready to launch my n900 of a long cliff. Jul 30 20:40:26 vi__: that will permanently fix the layout Jul 30 20:40:39 vi__: wait a little, i'm not yet there to catch it Jul 30 20:41:27 obviously L18N day Jul 30 20:41:37 err L10N Jul 30 20:41:52 :) Jul 30 20:41:59 I can set the keymap to it for example. Jul 30 20:42:07 and it works like the it map Jul 30 20:42:34 FUCK THIS PIECE OF SHIT. Jul 30 20:43:02 ooh, now your capslock mapping got borked Jul 30 20:43:37 freemangordon: HAM says update incompatible Jul 30 20:43:40 fyi Jul 30 20:43:54 wait Jul 30 20:43:56 ... Jul 30 20:44:20 what EXACTLY does that piece of fucking retarded hildon welcome shit fucking do? Jul 30 20:45:06 good question Jul 30 20:45:09 vi__: probably the app that appears to set the time sometimes when the battery is removed? Jul 30 20:45:29 or is that called something else? Jul 30 20:45:35 yep, I guess he meant that one Jul 30 20:45:49 I shouldn't have run HAM Jul 30 20:46:44 PAH! HAHH! Jul 30 20:46:51 locate dsme Jul 30 20:47:12 why did that take so long for me to think about locate? Jul 30 20:47:23 >> /var/lib/dsme/stats/lifeguard_restarts Jul 30 20:47:45 zeq: but of course, i meant lbc as a part of -mp thingie :) Jul 30 20:48:03 *libc Jul 30 20:48:27 Doc see above, I was unknowingly clicking on reboot device Jul 30 20:48:32 who T F F would look in var/lib for some statistics Jul 30 20:49:32 zeq: waaaait WUT? you rebooted, not some watchdog??? Jul 30 20:50:09 sonovvabitch. Jul 30 20:50:51 * DocScrutinizer05 heads out for dinner now, for good Jul 30 20:50:53 add 1 more thing to the list of really retarded things nokia made your N900 do. Jul 30 20:51:12 hildon welcome HOSES YOUR KEYMAP Jul 30 20:51:31 I apologise for my outburst. Jul 30 20:51:36 DocScrutinizer05: yeah, how embarrassing Jul 30 20:51:43 vi__: please check backscroll, zeq needs a devel devi9ce Jul 30 20:52:57 are you suggesting I give him mine or is this a serious request to try and make it happen? Jul 30 20:56:59 here is a question. Jul 30 20:57:25 where did hildon-welcome get a copy of the original unmodified keymap? Jul 30 20:57:53 to copy over the modified rx-51? Jul 30 20:58:10 has it got a secret backup stashed away? Jul 30 20:59:49 zeq: do you really need a dev device? Jul 30 21:00:15 also how does the new usb hostmode work? Jul 30 21:00:58 vi__: it would be really handy, I'm when I break things I'm without a primary device Jul 30 21:01:16 it would also let me test things out more easily Jul 30 21:02:12 like right now I'm without any translations from working on libc6, and the lables are rather cryptic! Jul 30 21:03:02 vi__: there has to be a whole bunch of secret phantom files. At very least for factory-reset Jul 30 21:03:10 well, we could start a donations drive on TMO. Jul 30 21:03:29 It worked for davyp. Jul 30 21:03:42 However the device he go was a real shitter. Jul 30 21:03:56 It worked, but it was quite fucked. Jul 30 21:04:21 zeq would you need one with a USB port? Jul 30 21:04:42 LOL Jul 30 21:04:45 although davyp was a lot more 'visible' on the forum. Jul 30 21:05:02 DocScrutinizer51: go for dinner already! Jul 30 21:05:24 I'm already eating :-P Jul 30 21:05:51 rx-51 you know ;-) Jul 30 21:06:02 vi__: usb port is rather handy for flashing, and charging if nothing else... Jul 30 21:06:47 I probably can get used N900 in rather good condition for less than 150 Jul 30 21:07:18 or find them yourself on ebay Jul 30 21:08:20 WITHworking usb Jul 30 21:11:22 DocScrutinizer05: I bought my last N900 on ebay,, supposedly new, but not really. I was really worried about the USB port (always am) Jul 30 21:11:41 I've had two devices fail that way :( Jul 30 21:12:01 u strengthened it with solder etc? not done mine yet, seems ok... Jul 30 21:12:27 well, some precausions will stop that problem Jul 30 21:12:29 Nokia stole the second one, they insisted on replacing it with an E7, which gf subsequently got stolen within a couple of weeks! Jul 30 21:12:59 DocScrutinizer51: euro? Jul 30 21:13:10 actually each monkey can solder down a working USB Jul 30 21:13:15 vi__: yep Jul 30 21:13:17 DocScrutinizer51: can you solder down the USB? Jul 30 21:13:29 yep Jul 30 21:13:33 DocScrutinizer51: do you have a paypal account? Jul 30 21:13:44 yep Jul 30 21:14:36 so if we raised 150 euro on TMO, you could get zeq a second n900? Jul 30 21:14:52 probably Jul 30 21:15:03 probably or can? Jul 30 21:15:27 make that 130, 20 are already in from me Jul 30 21:15:49 I see your 20 and match it. Jul 30 21:15:56 :) Jul 30 21:16:01 vi__: I forgot my crystal ball at home Jul 30 21:16:02 thanks guys! Jul 30 21:16:33 zeq: dont count your cows till your chickens come home yet. Jul 30 21:16:52 I appreciate the support :) Jul 30 21:17:14 zeq: have you and FMG 'announced' fennec yet? Jul 30 21:18:17 FMG put it in the thumb repo, but we're not happy to announce it until we've resolved some current breakage in mozilla-central Jul 30 21:19:00 am I allowed to talk about it? Jul 30 21:19:12 sure Jul 30 21:19:19 but with warnings Jul 30 21:19:28 it's not ready yet Jul 30 21:19:43 IE tomorrow when I make a post and I am listing why you NEED a device can I mention you and FMG are brewing up fennec. Jul 30 21:19:58 vi__: as soon as I find a good used one For reasonable pricemI can get it Jul 30 21:19:59 of course :) Jul 30 21:20:06 it is alpha etc, not really released. Jul 30 21:20:16 DocScrutinizer51: good. Jul 30 21:20:35 A lot of people want fennec. Jul 30 21:21:19 With Qwazix saying "it is better than fennec on dualcore 1.5MHz androids with 1GB RAM. Jul 30 21:21:28 hype shall be made. Jul 30 21:21:38 I don't care too much, but I want zeq in and activ Jul 30 21:22:06 vi__: it has potential to work well, but we need to get the memory consumption down Jul 30 21:22:14 DocScrutinizer51: I cocour, however to have a sucessful donation drive you have to ride a wave of hype. Jul 30 21:22:25 ^concour. Jul 30 21:22:44 ok :) Jul 30 21:22:48 zeq: And that is why it is still alpha and you need another one. Jul 30 21:22:53 ! Jul 30 21:23:50 I moved to a new house. I live on the bottom floor away from the street looking into the garden through big windows. There are 2 foxes who play here every night. Jul 30 21:23:55 and a badger too. Jul 30 21:24:11 ... Jul 30 21:24:13 nice :) Jul 30 21:24:25 that is at night. Jul 30 21:24:43 during the day there are 2 angry squirrels. Jul 30 21:25:04 not only do they fight at each other... Jul 30 21:25:15 they activley plot one anothers demise. Jul 30 21:25:26 We get foxes in our garden Jul 30 21:25:28 thatms love Jul 30 21:25:47 plenty of squirrels too! Jul 30 21:25:51 zeq: what's wrong with the current fennec, apart from the fact that's horrendously slow? Jul 30 21:26:04 also, why does linux use a swap partition with lower priority than another one? Jul 30 21:26:06 :s Jul 30 21:26:08 *with a Jul 30 21:26:24 kerio: you think it is slow now? Jul 30 21:26:32 current build (not in repo) has broken widget layout is is pretty much unusable Jul 30 21:26:49 the version in the repo has a really bad memory leak Jul 30 21:26:50 vi__: well, slower than microb Jul 30 21:27:22 kerio: that is an unfair comparison. Microb is loaded in the background ALL the time. Jul 30 21:27:32 It's been at least as fast a microb, in the last few weeks, but there's always something else not working. Jul 30 21:27:39 where do you think your ram goes? Jul 30 21:27:52 it feeds a family of ponies! Jul 30 21:27:58 we also need to change the keyboard handling to support the hildon input menthod properly Jul 30 21:28:16 and I still need to get the fullscreen toggle working Jul 30 21:28:18 zeq: oh, that's why fn d doesn't give me the same as fn+d? Jul 30 21:28:29 yes Jul 30 21:28:37 fullscreen? Jul 30 21:29:03 so you can see the hildon status bar Jul 30 21:29:21 toggle like in microb Jul 30 21:29:43 by the way, having an actual fullscreen toggle is really weird Jul 30 21:29:56 (as in, some way to run wmctrl -b toggle,fullscreen) Jul 30 21:30:05 the toolbars get mixed up all the time Jul 30 21:30:20 kerio: well it is at best a hack. Jul 30 21:30:30 why? Jul 30 21:30:34 hildon was not designed to be run in that way. Jul 30 21:30:46 well hildon was designed in the wrong way, isn't it Jul 30 21:30:52 *wasn't Jul 30 21:31:06 kerio: no, you were. Jul 30 21:31:21 aisleriot handles it well Jul 30 21:31:24 fuck yeah gnome Jul 30 21:32:02 ~tmo! is trolls, morons and ogres. Jul 30 21:32:03 vi__: okay Jul 30 21:32:03 kerio fucks gnomes? Jul 30 21:32:14 DocScrutinizer51: yes. Jul 30 21:33:17 kerio: http://www.catonmat.net/blog/top-ten-one-liners-from-commandlinefu-explained/ Jul 30 21:34:02 should i bother? Jul 30 21:34:06 even using all the normal libc6 build options, just with some gcc47 build fixes it still isn't working. I'm going to try a libc build with the 4.2 compiler Jul 30 21:34:45 zeq WTF? Jul 30 21:35:08 I'm trying to work out exactly what cause the locales to go wrong Jul 30 21:35:13 vi__: huh? Jul 30 21:35:31 kerio: it is some cool shit. Jul 30 21:35:48 If I can get a build working, I can try one thing at a time Jul 30 21:35:49 but no forkbomb roullette. Jul 30 21:35:56 hehe Jul 30 21:36:14 zeq: ls -laR / >before Jul 30 21:36:15 zeq: take you logical approach elsewhere please. Jul 30 21:36:38 gtg bb Jul 30 21:37:21 first thing I'm going to do is get a baseline working device Jul 30 21:37:28 installing stock libc6 Jul 30 21:41:01 ok. stock is working Jul 30 21:43:58 hm, swap on uSD + compcache makes things a little bit slower Jul 30 21:44:01 at least, it feels that way Jul 30 22:07:58 damn, PayPal is a bitch Jul 30 22:08:03 how I hate this service Jul 30 22:08:38 one guy who bought old computer form me *accidentaly* clicked to open claim (wanted to open claim about other purchase), and immediately, bot me and he informed paypal about mistake and requested closing case Jul 30 22:08:52 I even, just to be on safe side, provided scans of delivery documents Jul 30 22:09:04 and guess what, ScamPal decided today to close case... in favour of buyer! Jul 30 22:09:06 =) Jul 30 22:09:19 buyer immediately send me money back - it's good that he was honest one Jul 30 22:09:27 what the fuck Jul 30 22:09:52 but, the thing is, that PayPal got provision paid AGAIN, and also, converted my money from one currency to another, taking anothjer povision, to delete (-) net saldo Jul 30 22:09:55 so, they' Jul 30 22:10:01 have earned 4x provision, totally Jul 30 22:10:05 just by being a bitch Jul 30 22:10:42 Estel_: how do i fit two batteries in an otterbox? Jul 30 22:10:47 their custom center try to tell me, that it's ok for them to have such rule in agreement with customers, and of course they refuse to give it back Jul 30 22:11:03 kerio, no idea, ki don't use otterbox. Be creative Jul 30 22:11:19 Estel_ in the end they want your money Jul 30 22:11:19 it's not even here yet Jul 30 22:11:31 I contacted goverment's custom right service, #@$#@$/ Beware of PayPAl, it's becoming even more arrogant scam service than ever Jul 30 22:11:59 we will see how their #$@#$ scam rules apply to Polish law Jul 30 22:12:11 as they're offering their services here, under polish domain, not in @#%$#5 luxemburg Jul 30 22:12:34 and they can put their #@$#@$@# luxemburg contact mail, they want me to sent "complains", in their scamish ass Jul 30 22:13:09 btw funny fact - if buyer wouldn't send me money back, they could ask him to do so, but wouldn't force his money back to my account, or start legal action Jul 30 22:13:22 their "forcibly money transfer" is only in favor of buyer, never the other way Jul 30 22:13:38 paypal is great if you're a dishonest buyer and you want to screw the seller over Jul 30 22:13:41 i'm officially going to bounce off every paypal payment, that won't be amrked as transfer to friends/relatives Jul 30 22:13:45 kerio, exactly Jul 30 22:14:11 It's pity that aliexpress showed scampal middle finger, and deleted payments via paypal Jul 30 22:14:22 i just don't get what's wrong with wire transfers Jul 30 22:14:27 I would buy half if Cchiense country, then request refund Jul 30 22:14:31 they're *definetely* traced Jul 30 22:14:43 I would like to see how PayPal arguments their "rules" to chinesee special forces ;) Jul 30 22:14:58 kerio, international wire transfers are expensive Jul 30 22:15:09 for sxmall-to-mdoerate sums, they're too expensive Jul 30 22:15:15 and 4x commission from paypal is cheap? Jul 30 22:15:28 for huge sums, like money for black alu project, of course i convince every donoro/pre-order one to do so Jul 30 22:15:32 kerio, haha Jul 30 22:15:39 buyer don't care, as it's me who paid it :< Jul 30 22:16:05 but honestly, and unfortunatelyt, for such sums, even 4x commision from paypal is less expensive than wire transfer from US :( Jul 30 22:16:19 I really hope for bitmoney stabilizing on current value Jul 30 22:16:24 it's stable for more than year now Jul 30 22:16:28 bitcoins, I mean Jul 30 22:16:44 after initial speculation and 30 USd value dropping to 4 USd, it seems that 4-5 USd is real value for bitcoin Jul 30 22:16:54 I'll glady tell PayPal to go and sodomize itself Jul 30 22:16:57 with a stun baton Jul 30 22:17:06 electric one Jul 30 22:17:53 Estel_: hmm, the otterbox looks awfully tight Jul 30 22:17:57 ScapPal, eBayshit, and every sub-company of this scamming family Jul 30 22:18:16 kerio, just buy otterbox big enough to handle mugen-cover sized N900 Jul 30 22:18:18 job done Jul 30 22:18:28 it's not custom made Jul 30 22:19:00 by the way, how's the full aluminium cover going? Jul 30 22:19:12 stable and slowly to the target :) Jul 30 22:19:19 can you make one with three battery slots? Jul 30 22:19:26 they should fit sideways, don't they Jul 30 22:19:34 I'm starting vacation (2 weeks and little more) from day after tomorrow, though Jul 30 22:19:35 hm, maybe not Jul 30 22:19:49 kerio, no, 2x18650 slot is priority Jul 30 22:19:57 and they don't fit sideways, unfortunately (bl-5j) Jul 30 22:19:57 18650? :o Jul 30 22:20:08 yea, 2x 18650 for custom body replacement Jul 30 22:20:12 wtf is a 18650 Jul 30 22:20:13 for backcover only, no chances Jul 30 22:20:18 google it ;) Jul 30 22:20:20 is that mAhs? Jul 30 22:20:22 nope Jul 30 22:20:24 haha Jul 30 22:20:30 :( Jul 30 22:20:38 18650 is standard cylyndrical most popualr in the world LiIon cell Jul 30 22:20:45 best price Jul 30 22:20:47 best availability Jul 30 22:20:49 ew, cylindrical Jul 30 22:21:02 will never disappear, it's used in notebooks batteries and flashlights Jul 30 22:21:20 yea, but they have higher capacity density, even if You consider "wasted" space due to cylindrical Jul 30 22:21:23 do they fit in the battery slot, in length? Jul 30 22:21:38 phillips or panasonic (can't remember) made 3Ah one, lately, 3.8 mAh is also avaiulable Jul 30 22:21:42 no, they don't Jul 30 22:21:47 that's why You need custom slot Jul 30 22:21:55 and that's why they will be available for custom body Jul 30 22:21:56 yeah but it becomes insanely huge :( Jul 30 22:21:59 as slot will be re-made :) Jul 30 22:22:03 oh, i see Jul 30 22:22:03 nope Jul 30 22:22:12 I'll get rid of standard battery slot Jul 30 22:22:30 that's why it's no chance for backcover only, as it would be extremely fat Jul 30 22:22:37 18650 have their dimensions in name Jul 30 22:22:42 1,8 - diameter Jul 30 22:22:47 65 - lenth Jul 30 22:22:53 0 - ammount of problems with buying them ;) Jul 30 22:23:10 18mm diameter, 65 mm length Jul 30 22:23:13 how much space is there to reclaim, in the n900? Jul 30 22:23:20 2x is 6Ah easily, soon 6,8 Ah too Jul 30 22:23:28 some, with creative planning Jul 30 22:23:36 You can't cut off camera secondary board ;) Jul 30 22:23:38 i both like and don't like the sound of that Jul 30 22:24:02 well, 18650 are always first to have new awesome capacity, and are always cheaper to other types, comparing same capacity Jul 30 22:24:18 no worries, 18650 variant of full body replacement will be optional Jul 30 22:24:21 is there a way to fit two in the same space as a regular n900 with a regular back cover? Jul 30 22:24:29 full body replacement with standard bl-5j is also going to be available Jul 30 22:24:41 kerio, wtf? no, 18650 is only for mugen-cover sized one Jul 30 22:24:53 i hope that a mugen-style xl backcover will also be available Jul 30 22:24:55 for standard cover You get ONE battery, no matter what, and it's bl-5j Jul 30 22:24:56 er, i meant one Jul 30 22:25:01 of course Jul 30 22:25:16 if You read thread, it's going to be available as aluminum mugen-style backcover for normal N900 body... Jul 30 22:25:25 i see Jul 30 22:25:29 flat backcover alu replacement with better kickstand (mugen too) for normal body Jul 30 22:25:33 (with lens cover, i hope :) ) Jul 30 22:25:37 and, full body replacement in 2 variants (for now) Jul 30 22:25:41 including backcover Jul 30 22:25:44 of course, with lens cover Jul 30 22:25:58 but, backcover from ful lbody replacement *probably* won't be same as backcover for normal body Jul 30 22:26:03 due to locking mechanism Jul 30 22:26:15 hm Jul 30 22:26:18 that is why full body is sold with backcover only, not separately Jul 30 22:26:21 i see Jul 30 22:26:38 I thjknk about two varianst of full body rep, one normal, with either flat or mugen-sized backcover... Jul 30 22:26:39 18mm diameter? Jul 30 22:26:49 650mm length? Jul 30 22:26:53 and one extreme, for 18650, with mugen-sized backcover only (possibly, 2mm thicker) Jul 30 22:27:02 (I mean, bigger in thickness) Jul 30 22:27:06 DocScrutinizer, haha, noi Jul 30 22:27:08 * zeq is being called to bed Jul 30 22:27:10 18 mm diameter Jul 30 22:27:12 65 lenght Jul 30 22:27:14 night guys Jul 30 22:27:17 0 stands for whatever You want Jul 30 22:27:18 0 problems Jul 30 22:27:21 night zeq! Jul 30 22:27:35 so... old-style body with flat back-cover, old-style body with xl backcover and crazy 18650 body with OMGHUEG backcover? Jul 30 22:27:40 1.8 dia sounded like in inch Jul 30 22:27:42 DocScrutinizer, why You see sausages everywhere?:P Jul 30 22:27:46 in styluses, in batteries... Jul 30 22:27:59 kerio, kinda yes Jul 30 22:28:15 + 2 types of backcovers only, sadly, not compatible with new bodies (but compatible wioth normal bodies) Jul 30 22:28:20 sounds complicated, but it isn't ;) Jul 30 22:28:23 Estel_: what about the screen part? Jul 30 22:28:28 DocScrutinizer, it's 18 mm, for sure Jul 30 22:28:32 kerio, not touched Jul 30 22:28:35 hm Jul 30 22:28:39 whole screen module remains as-is, it's metallic anyway Jul 30 22:28:45 including sloiding mechanism Jul 30 22:28:49 too much hassle, nothing to gain Jul 30 22:28:52 i wanted a transparent aluminium front cover! :D Jul 30 22:29:01 aka glass?;) Jul 30 22:29:16 no, transparent aluminium Jul 30 22:29:17 projects for future, after ending those - same, but from exotic wood ;p Jul 30 22:29:37 de luxe, completely unprofitable for usage, glamour model ;P Jul 30 22:29:44 still kinda cool, wooden body Jul 30 22:30:06 I'm also thinking about N950 and N9 (with slot for keyboard and sliding) bodies, but this is for distant future Jul 30 22:30:12 you know, .45 is some 11mm Jul 30 22:30:21 .38 is ~9mm Jul 30 22:30:32 DocS, hm? Jul 30 22:30:40 sure, but? Jul 30 22:30:45 1.8 probably is a light ship gun Jul 30 22:30:46 n900-shaped gun! Jul 30 22:30:50 hehe Jul 30 22:30:56 with *two* bullets Jul 30 22:31:12 shot inject via i2c Jul 30 22:31:28 Estel_: will the xl backcovers also have the external power port? Jul 30 22:31:42 (what kind of port is it? miniusb?) Jul 30 22:31:54 this one is kinda problematic. Probably, I will settle with compromise proposed, to leave slot for any kind of socket one may want to put there Jul 30 22:32:04 You know, tiight aluminum in one slot place Jul 30 22:32:06 thin* Jul 30 22:32:13 as putting external port for everyone isn't best idea Jul 30 22:32:14 i don't like thin aluminium :( Jul 30 22:32:16 due to lack of fuse, etc Jul 30 22:32:45 thin = you can easily punch it off if You want, and insert port there, for charging battery directly like dr_frost_dk, or for mod to soolder to n900 mainboard +5V, or... Jul 30 22:32:58 anyway, I'm not going to have motherboard of people's devices Jul 30 22:33:07 how I'm suppose to make external port for charging via testpads Jul 30 22:33:13 I'd prefer a plastic patch with 2 gold pads and a magnetic iron behind it Jul 30 22:33:16 everyone would need to solder it on own Jul 30 22:33:28 DocScrutinizer, I also though about springs solution Jul 30 22:33:39 Estel_: external charging port! Jul 30 22:33:42 meh, magnet >> springs Jul 30 22:33:44 but it defeats space for dual-batteries or 18650's OTOH Jul 30 22:33:56 well, can't your port also be used for charging? Jul 30 22:34:03 what "my port"? Jul 30 22:34:14 isn't it *your* guide on TMO? Jul 30 22:34:18 no? Jul 30 22:34:24 hm, maybe i'm mistaken Jul 30 22:34:34 no one created 2nd usb port for N900 Jul 30 22:34:42 alternate way connection is possible, but it's same port Jul 30 22:34:47 - some safety mechanism Jul 30 22:34:50 like fuse for 5V Jul 30 22:34:55 no, i meant dr_frost_dk's direct battery connection Jul 30 22:34:59 (it was him) Jul 30 22:35:08 or external ESd protection for USB (only chips internal one remain - it's rated 8,8 kV, so it's Ok, but still) Jul 30 22:35:24 direct battery connection allow to charge only using dedicated LiIon charger Jul 30 22:35:35 no way to Put it under 5V without boom Jul 30 22:35:38 it charge to cells directly Jul 30 22:35:51 passing by even cells safety things Jul 30 22:35:56 aaaand you'll confuse hell outa bq27200 Jul 30 22:36:11 BTw, people who want to use N900's testpads for charging, or direct battery charge, won't have problems to kick out thin aluminum in "fake" slot Jul 30 22:36:15 and insert whatever they want there Jul 30 22:36:19 he said he uses it to draw power and/or to charge it with an external charger Jul 30 22:36:20 DocScrutinizer, exactly Jul 30 22:36:33 kerio, external = one as big as smal netbook Jul 30 22:36:46 or, alternatively, one small, but not very intelligent Jul 30 22:36:47 DocScrutinizer05: battery indicators are guesswork anyway Jul 30 22:36:53 kerio, no, they're not Jul 30 22:36:57 bme is guesswork Jul 30 22:37:10 bq27200 is very accurate in measurements Jul 30 22:37:16 http://talk.maemo.org/showpost.php?p=981077&postcount=448 omg i never saw this one Jul 30 22:37:21 can't it be reset somehow? Jul 30 22:37:29 I did zillions of comparissions between external charger-measurement one, and bq27200, and th latter is excellent Jul 30 22:37:30 if you know you're doing weird stuff with the battery Jul 30 22:37:48 will it be confused even if you *discharge* your battery? Jul 30 22:37:54 kerio, sure, sure, but all direct charge solutions confuse bq27200 Jul 30 22:38:04 no, unless You power something else with it Jul 30 22:38:08 but, when You charge battery Jul 30 22:38:12 it won't know you've charged it Jul 30 22:38:15 it will see increased voltage Jul 30 22:38:17 i meant powering something else in addition to the phone Jul 30 22:38:21 and wil lthink that Your battery have 32573645634753 capacity Jul 30 22:38:26 haha Jul 30 22:38:33 but it will refuse such data Jul 30 22:38:45 so You will stay with de-calibrated capacity forever ;) Jul 30 22:39:02 bq27200 doesn't accept too much of a change in % from one cycle to another Jul 30 22:39:08 forever = until you remove the battery? Jul 30 22:39:20 forever - until You calibrate it again and stop charging via direct-to-cell method Jul 30 22:39:33 in my signature, there is a guide for calibrating it Jul 30 22:39:43 and even for measuring capacity of battery using device only. Jul 30 22:40:03 accurately measuring (no less than using external dedicated LiIon charger with measuring capabilities) Jul 30 22:40:48 generally speaking, bq27200 is even better than them, cause it doesn't have problems with real-time changing current draw Jul 30 22:40:52 it just keep track about it Jul 30 22:41:00 external chargers just discharge battery at some constant rate Jul 30 22:41:05 like 500 mA or whatever You set Jul 30 22:41:14 and divide discharging time by this rate Jul 30 22:41:30 actually, multiply it, not divide Jul 30 22:41:43 this is why You get mA *h Jul 30 22:41:46 = mAh Jul 30 22:42:31 500mA rate and 3h discharging until full = 500 mA * 3H = 1500 mAh, basic math/physic ;) Jul 30 22:43:12 in fact it is 500 * mA *3*h, while multiplying, You can change order of elements freely, so You reduce numbers to one (1500) and put mA and h close to each other = mAh Jul 30 22:43:59 bq27200 is more uber, as it calculate it constantly, based on momnentary currents too Jul 30 22:44:24 so if it have 500 mA in half of second, and 50 for next 5 seconds, he know how to calculate it properly ;) Jul 30 22:45:59 normally, this kijnd of chip is bundled with batteries - making them more expensive, and allowing sucker manufacturers to impose limits, like "stop to work after xxx cycles" Jul 30 22:46:46 in our case, it is inside device, so no such tricks and we have read-only access to it - good side is that batteries are easidly moddable, bad side is that it get's confused if You hotswpa batteries, or generally, change them ;) Jul 30 22:46:55 honestly, I think it's for +, after all Jul 30 22:47:09 dealing with chips inside batteries is PITA Jul 30 22:47:19 Estel_: the chip resets itself if you remove the battery and power it down, right? Jul 30 22:47:38 no, unless You have internal battery (this one that keeps clock) fubar Jul 30 22:47:42 like, mostly everyone now Jul 30 22:47:44 ;) Jul 30 22:47:45 oh, i see Jul 30 22:47:48 hm Jul 30 22:47:52 it actually works for me i think Jul 30 22:48:00 try rmeoving battery for hour Jul 30 22:48:01 ;0 Jul 30 22:48:13 for me even few minutes/or even seconds is enough Jul 30 22:48:26 in fact, this green thing from inside "battery" is also outside Jul 30 22:48:56 Estel_: WUT? Jul 30 22:48:57 quite easy to desolder and solder new, though Jul 30 22:49:07 dafuq Jul 30 22:49:10 green thing? Jul 30 22:49:13 aaah bupbat Jul 30 22:49:16 DocScrutinizer, looking at it closely (this small coin thing) it's not Jul 30 22:49:17 yes Jul 30 22:49:33 in fact, every device I've disassembled have little ring of green thing outside Jul 30 22:49:36 ew Jul 30 22:49:39 :s Jul 30 22:49:44 what it is, strictly? capacitor battery? Jul 30 22:49:47 is there a way to get a good one that will last? or are they all bad? Jul 30 22:49:53 yes, there is, it's cheap Jul 30 22:49:56 DocScrutinizer, have links Jul 30 22:50:04 ~bupbat Jul 30 22:50:04 somebody said bupbat was http://www.mouser.com/catalog/632/2009.pdf http://il.farnell.com/taiyo-yuden/pas414hr-va5r/cap-0-06f-3-3v-80ohm-4-8mm-coin/dp/1853000?Ntt=PAS414HR-VA5R, or use the capacitive type, LiIon are breaking during 12 months Jul 30 22:50:08 shipping is more expensive than battery, though Jul 30 22:50:09 k Jul 30 22:50:15 hehe, it's always like that Jul 30 22:50:19 best to buy 50 at once :D Jul 30 22:50:35 DocScrutinizer, any sense in buying higher cap one? Jul 30 22:50:42 nope Jul 30 22:50:48 what it is, from technical point of view? hybride of capacitor and battery? Jul 30 22:50:48 nonsense Jul 30 22:51:03 you dont need 12 monts of standby Jul 30 22:51:11 shouldn't capacitor discharge itself pretty quickly anyway? Jul 30 22:51:21 true Jul 30 22:51:24 lifetime Jul 30 22:51:34 maybe should use normal silver coin from timer?:P Jul 30 22:51:36 haha Jul 30 22:51:37 no, an ideal capacitor holds charge infinitely Jul 30 22:51:49 DocScrutinizer, but in real life we don't have ideal ones Jul 30 22:52:07 but still they hold charge over months and years Jul 30 22:52:25 sure, but I wonder what is different in this one, from normal electrolyte one Jul 30 22:52:27 or ceramic one Jul 30 22:52:32 that loses charge in seconds Jul 30 22:52:45 I got shocked from a old aplliance that been in the attic for years Jul 30 22:52:54 yea, and shocked You in exchange Jul 30 22:52:58 but it's rare case Jul 30 22:53:02 capacitor in filter still had charge Jul 30 22:53:22 good one, for sure Jul 30 22:53:50 ceramic C hold charge for years Jul 30 22:53:57 but, we can't jsut put electrolyte cyllinder there, instead of this 3.3 bupbat, that's why I'm asking what are the differencies Jul 30 22:54:12 use a goldcap Jul 30 22:54:14 cause it's for sure something slightly different Jul 30 22:54:20 1F 3.3V Jul 30 22:54:31 works, tested Jul 30 22:54:44 :) thanks Jul 30 22:54:50 normal one is 0.06f? Jul 30 22:55:01 so this 1F 3.3V have standby time... 20 years? Jul 30 22:55:18 0.1F is OK as well, still sufficient for days of standby Jul 30 22:55:45 I'm asking, as it's more cheap for me to get other goldcap locally, than this special type of bupbat from abroad Jul 30 22:56:01 heck, even ten of them is still less than shipping. Jul 30 22:58:34 use any huuuuuge capacitor that fits into case. 3V3 min Jul 30 22:58:45 lol, first result on local wanna-be-ebay site - 2.2F 5,3V goldcap Jul 30 22:58:52 still less expensive than bupbat Jul 30 22:59:05 sure but will it fit into case? Jul 30 22:59:48 of course not, but surprisingly, 3F is even cheaper and is same form size as our bupbat Jul 30 23:00:01 it msut be 3.3v, or can be higher voltage range? I think the latter Jul 30 23:00:07 s/msut/must/ Jul 30 23:00:09 Estel_ meant: it must be 3.3v, or can be higher voltage range? I think the latter Jul 30 23:00:12 higher of course ok Jul 30 23:00:16 i.e. it's "no less than 3.3", yep? Jul 30 23:00:22 yea, as with capacitors, normally Jul 30 23:00:51 3V3 min was what I wrote Jul 30 23:01:09 ? Jul 30 23:01:21 minimum 3.3 Volt Jul 30 23:01:24 3.3v, I've said it too Jul 30 23:01:28 yea Jul 30 23:01:46 ~dict minimum Jul 30 23:01:48 Dictionary 'minimum' (3 of 5): the smallest possible quantity ;; the least possible; "needed to enforce minimal standards"; "her grades were minimal"; "minimum wage"; "a minimal charge for the service" ;; the point on a curve where the tangent changes from negative on the left to positive on the right . Jul 30 23:02:18 http://allegro.pl/kondensator-gold-cap-3-f-5-5v-poziomy-8245-i2479426361.html 3F one, lol, 3USD Jul 30 23:02:21 even less Jul 30 23:02:36 alnegro? Jul 30 23:03:08 RiD: ????? Jul 30 23:03:11 oh, it's allegro :S Jul 30 23:03:26 * RiD blames the n900 screen Jul 30 23:03:31 suuure Jul 30 23:03:36 bhungry one thinks about bread Jul 30 23:03:42 i swear that looked like alnegro Jul 30 23:04:11 maybe this one, it's like another N900 in size: Jul 30 23:04:12 http://allegro.pl/jantzen-audio-silver-gold-z-cap-0-22-uf-dunskie-i2511777804.html Jul 30 23:04:18 4,7f, 250V Jul 30 23:04:41 at least You never need to type in time again Jul 30 23:04:58 uwaga chwilowo lol Jul 30 23:05:16 only 30 USD Jul 30 23:05:41 rid, don't do that. It's gibberish ;) Jul 30 23:08:07 yeah those HVHC critters are nice'n'funny. Usually from USP Jul 30 23:08:23 or power grid Jul 30 23:08:42 or electric trains Jul 30 23:09:41 MEH, that clearly ready "µF" Jul 30 23:10:10 reads* Jul 30 23:10:54 Jantzen Audio Jul 30 23:11:05 clearly a completely different domain Jul 30 23:13:20 * DocScrutinizer05 offers some guarnateed 100% oxigen free copper with passivation coating power cord, Schuko plug to open wire, for DIY. Length 90cm. Only 699EUR Jul 30 23:13:45 free, but only 700? Jul 30 23:13:57 i'm in Jul 30 23:15:03 ;) Jul 30 23:15:04 RiD: you're the newcomer troll practicant? Jul 30 23:15:30 sure, he learns from old masters ;) Jul 30 23:15:40 or are you pupnik actually? Jul 30 23:15:48 I'm here to replace whoever got banned Jul 30 23:16:30 http://monkeyiq.blogspot.com/2012/07/off-record-messaging-tutorial.html <- absolutely great article. I would *love* to see communicator on N900 using this Jul 30 23:16:37 duh, Portugal Jul 30 23:16:46 so prolly not pupnik Jul 30 23:16:58 oh i see the ban list is actually quite clean, other than bots Jul 30 23:17:15 and no, I'm not pupnik at all (i don't even know who is he/she) Jul 30 23:28:20 no worries, DocScrutinizer remind him ocne for a while, so for sure this one have eternal place in DocScrutinizer's heart ;) Jul 30 23:28:27 although, he always refuse to tell who he was :( Jul 30 23:28:41 sounds like pumpkin to me, anyway Jul 30 23:29:41 pumpkin? weren't you talking about pupnik Jul 30 23:30:16 it still sounds like pumpkin Jul 30 23:30:52 what happened to pupnik Jul 30 23:32:45 ~seen pupnik Jul 30 23:32:53 pupnik was last seen on IRC in channel #meego, 304d 23h 42m 7s ago, saying: 'tegra chipset?'. Jul 30 23:33:18 he died? Jul 30 23:33:27 i mean, he left #maemo? Jul 30 23:33:50 d'oh. ignore what my last 2 messages Jul 30 23:42:20 I'm both disappointed that the channel rarely exceeds 300 users lately, and happy that it stays stable at 271-274 Jul 30 23:43:23 and if you wanna know who's pupnik: there's chanlog in almost every relevant chan Jul 30 23:46:38 lol Jul 30 23:48:14 http://mg.pov.lt/maemo-irclog/search?q=pupnik Jul 30 23:53:43 pupnika and every meego conference visitor accepted TSA scaning their penis to see if they were circumcised Jul 30 23:53:49 I see :/ Jul 30 23:59:05 heh, pupnik Jul 31 00:24:06 ~hen Jul 31 00:24:07 do ~hostmode and ~factinfo hostmode, this is your maemo council :-/ Jul 31 00:24:24 Hey, DocScrutinizer05... What's up with this ^^^ Jul 31 00:24:39 ~hostmode Jul 31 00:24:39 it has been said that hostmode is piece of crap written by morons to fry other people's devices! beware! Jul 31 00:24:48 ....o_o? Jul 31 00:25:11 ~factinfo hostmode Jul 31 00:25:12 error: you do not have enough flags for that. (o required) Jul 31 00:25:12 hostmode -- created by MohammadAG <~MohammadA@62.219.120.20> at Mon May 10 21:17:16 2010 (812 days); last modified at Fri Jun 1 23:21:46 2012 by Estel_!~Estel@lagrange.cloud-7.de; it has been requested 68 times, last by RiD, 33s ago; it has been locked by DocScrutinizer. Jul 31 00:26:38 Woody14619: what's wrong with it? The link between estel and council? Jul 31 00:26:46 DocScrutinizer05, If you'd be so kind... nuke that ~hostmode "factinfo" Jul 31 00:26:51 Because Estel is not Council. Jul 31 00:26:55 ok Jul 31 00:27:02 ~unlock hen Jul 31 00:27:02 unlocking factoid hen, DocScrutinizer05 Jul 31 00:27:05 Hasn't been for a while. Jul 31 00:27:07 go ahead Jul 31 00:27:37 lol... I've no clue how to change things in infobot. :P Jul 31 00:27:45 ~forget hen Jul 31 00:27:46 DocScrutinizer05: i forgot hen Jul 31 00:27:53 ~hen is a chick Jul 31 00:27:53 okay, DocScrutinizer05 Jul 31 00:28:06 ~hostmode Jul 31 00:28:06 i heard hostmode is piece of crap written by morons to fry other people's devices! beware! Jul 31 00:28:11 lol. Jul 31 00:28:34 ~unlock hostmode Jul 31 00:28:34 unlocking factoid hostmode, DocScrutinizer05 Jul 31 00:28:39 As long as it's attributed to Estel... :P Jul 31 00:29:06 I'd be curious when/where he said that... Doesn't supprise me frankly, but... :P Jul 31 00:29:30 he defined it for infobot Jul 31 00:29:36 like I do now Jul 31 00:29:51 infobot: no, hostmode is something estel abused Jul 31 00:29:51 okay, DocScrutinizer05 Jul 31 00:31:15 June 1st? Really? Crack smoking.... :P Jul 31 00:31:56 unforget hostmode Jul 31 00:32:00 But yes, FYI: Estel is no longer Council. He announced his resignation on TMO and in meeting. http://maemo.org/community/council/community_council_meeting-july_13-2012/ Jul 31 00:32:02 ~unforget hostmode Jul 31 00:32:02 DocScrutinizer05: cannot undeleted 'hostmode' because it already exists! Jul 31 00:32:10 ~hostmode Jul 31 00:32:11 i heard hostmode is something estel abused Jul 31 00:32:11 ~hen Jul 31 00:32:12 from memory, hen is a chick Jul 31 00:32:24 ~forget hostmode Jul 31 00:32:24 DocScrutinizer05: i forgot hostmode Jul 31 00:32:34 ~unforget hostmode Jul 31 00:32:35 DocScrutinizer05: Successfully recovered 'hostmode'. Have fun now. Jul 31 00:32:39 ~hostmode Jul 31 00:32:39 hostmode is, like, http://talk.maemo.org/showthread.php?p=652330 Jul 31 00:32:45 HAH! Jul 31 00:32:50 ~lock hostmode Jul 31 00:32:50 locking factoid hostmode, DocScrutinizer05 Jul 31 00:33:18 ~hen Jul 31 00:33:18 i guess hen is a chick Jul 31 00:33:25 ~forget hen Jul 31 00:33:25 DocScrutinizer05: i forgot hen Jul 31 00:33:30 ~unforget hen Jul 31 00:33:31 DocScrutinizer05: Successfully recovered 'hen'. Have fun now. Jul 31 00:33:35 ~hen Jul 31 00:33:36 hmm... hen is hostmode-easy-now, or http://talk.maemo.org/showthread.php?t=65232 Jul 31 00:33:40 Thanks for the assist on that. :) Jul 31 00:33:41 ~lock hen Jul 31 00:33:42 DocScrutinizer05: locking factoid hen Jul 31 00:33:57 i was about to define hen Jul 31 00:34:02 but that one is just fine Jul 31 00:34:10 i was doing a huge paragraph Jul 31 00:35:26 Can't believe I missed that for this long. Only reason I saw it was because someone highlighted me just above someone using it... Jul 31 00:38:54 K, off to get dinner. Thanks again Doc & RiD. :) Jul 31 00:39:08 why thank me? Jul 31 00:39:12 o_O Jul 31 00:39:33 you were willing to help (even if it was locked. ;) Jul 31 00:40:04 k, off to eat. Have a good night. Jul 31 01:01:14 lol, DocScrutinizer, why pinging me to death via factoid You've locked Yourself? :) Jul 31 01:01:46 Woody14619, if You would actually read council mailist on times when I was council - there was actually a discussion about it, with ivgalvez and SD69 involved, IIRC Jul 31 01:02:20 I those stormy times I highlighted that DocScrutinizer locked this factoid in such state. Jul 31 01:02:34 of course it doesn't surprise me, that You have no clue what hppened on council mailing list :) Jul 31 01:03:06 ~hen Jul 31 01:03:07 hen is, like, hostmode-easy-now, or http://talk.maemo.org/showthread.php?t=65232 Jul 31 01:03:11 ~h-e-n Jul 31 01:03:12 do ~hostmode and ~factinfo hostmode, this is your maemo council :-/ Jul 31 01:03:17 ~hostmode Jul 31 01:03:18 hostmode is, like, http://talk.maemo.org/showthread.php?p=652330 Jul 31 01:03:50 DocScrutinizer, might wan't to nuke ~h-e-n too, as now it suggets that Council created hostmode, which is covered in thread on TMO Jul 31 01:04:13 just a little suggestion ;) Jul 31 01:12:29 weren't you sleeping? **** ENDING LOGGING AT Tue Jul 31 02:59:58 2012