April 9, 2007 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
April 9, 2007 [00:00:28] <fyysik> in BeOS? Terminal. Menu->Settings. Save as default then[00:00:43] <Jin> oh[00:00:47] <Jin> you just meant for the terminal[00:00:58] <fyysik> yup.[00:01:01] <CIA-17> hugosantos * r20621 /haiku/trunk/src/add-ons/kernel/network/stack/ (datalink.cpp domains.cpp interfaces.cpp interfaces.h):[00:01:01] <CIA-17> whenever an interface is deleted, call put_domain_datalink_protocols() so all readers are unregistered.[00:01:01] <CIA-17> - Unfortunely this requires RX lock to become a recursive lock.[00:01:26] <Jin> :([00:02:10] <fyysik> same did in BeIDE, with more settings, as i need syntax styling[00:05:19] *** dr_evil has quit IRC[00:05:44] * Jin gets an idea[00:06:25] <Jin> maybe I can code a program that'll repeatedly cap the screen, invert it, and overlay it over the desktop...too bad I Don't know anything about Be's APIs[00:06:55] <Ingenu> night[00:07:17] *** Ingenu has quit IRC[00:07:49] <fyysik> it reminds me idea of new CEO in company, Jin "It is total disorder! Halfs of workers is working, Half isn't. I definitely will invert that situation!"[00:08:02] <stargater> DaaT, please see the querry and big thx[00:10:40] *** mats has quit IRC[00:13:41] <Jin> bbl[00:17:35] *** jpd has joined #haiku[00:18:16] *** SunWuKung has joined #haiku[00:18:23] *** zl0minus has quit IRC[00:19:56] *** JamesB192 has joined #haiku[00:30:04] *** jpd has quit IRC[00:37:38] *** DaaT has quit IRC[00:55:14] *** SunWuKung has quit IRC[00:56:57] *** gotaku has joined #haiku[00:57:20] *** MYOB has quit IRC[01:07:46] *** shackan has joined #haiku[01:08:08] *** Leszek has quit IRC[01:09:44] *** shackan has quit IRC[01:10:32] <aldeck> bullet under haiku? clicky: http://haikubeat.free.fr/files/bullet/ :P[01:10:42] <aldeck> http://www.continuousphysics.com/Bullet/[01:11:46] <MrSunshine> 55min running haiku ? :)[01:11:58] <aldeck> yepa[01:12:07] <MrSunshine> and running opengl stuff during that time .. nice :)[01:12:19] <aldeck> and many at once :)[01:12:34] <gotaku> I think I'm going to work on the app_server... If I can.[01:12:39] <aldeck> i had to kill them to exit tough..[01:13:21] *** Jin has quit IRC[01:14:03] *** Jin has joined #Haiku[01:14:21] *** snipe714 has quit IRC[01:14:36] <aldeck> number crunching doesn't scare haiku apparently, it more flaky on the i/o[01:15:09] <aldeck> *it's[01:20:18] *** BuildFactory has quit IRC[01:24:38] *** shackan has joined #haiku[01:26:08] *** fyysik has quit IRC[01:26:23] *** wildur has joined #Haiku[01:29:31] *** pikapika has quit IRC[01:44:02] *** aldeck_ has joined #haiku[01:47:16] *** TLF has quit IRC[01:47:35] *** aldeck_ has quit IRC[01:47:48] *** aldeck_ has joined #haiku[01:50:32] <Jin> right when I was about to get mad that I couldn't find a winamp clone, wham, I see CL-Amp :P[02:00:17] *** aldeck has quit IRC[02:01:57] *** wildur has quit IRC[02:09:15] <CIA-17> hugosantos * r20622 /haiku/trunk/src/add-ons/kernel/network/stack/ (interfaces.cpp link.cpp routes.cpp utility.cpp utility.h):[02:09:15] <CIA-17> packet capture sockets are now properly notified when the device is going down.[02:09:15] <CIA-17> - modified LinkProtocol::ReadData() a bit so it can safely react to external changes such as the device being monitored going down.[02:09:15] <CIA-17> - fixed an issue in invalidate_routes() where in some cases the default route was being kept.[02:09:16] <CIA-17> - fixed another issue related with the fifo implementation, the notifier semaphore was being created with count of 1, instead of 0.[02:19:59] *** petterhj- has joined #haiku[02:22:11] *** petterhj has quit IRC[02:24:55] *** Mazon is now known as mazon[02:30:08] *** eNGIMa has quit IRC[02:31:07] *** petterhj has joined #haiku[02:31:28] *** xcasex has quit IRC[02:35:33] <Jin> /me dances around the channel[02:35:47] <Jin> minus the space[02:39:29] *** gr00ber has joined #haiku[02:40:19] *** petterhj- has quit IRC[02:54:09] <Jin> where the heck is everyone...?[02:55:05] *** doppeljot has joined #haiku[02:55:38] *** doppeljot has left #haiku[02:55:41] <gotaku> This is normal.[02:55:48] <_hugo> yawn[03:00:38] *** kokito has joined #haiku[03:05:02] *** aldeck_ has quit IRC[03:05:09] <geist> different times of the day you get a lot or not much at all[03:07:08] <gotaku> Wow, Haiku received 42 student applications for the Google SoC.[03:07:47] <gr00ber> of course they did[03:07:56] <shackan> wow[03:08:04] <gr00ber> no other number of applications would be possible[03:08:11] <shackan> gr00ber, lol[03:08:28] <shackan> gotaku, and how many slots are available?[03:08:47] <gr00ber> any sluts available?[03:08:53] * gr00ber gets ready to go[03:08:59] <shackan> gr00ber, not for you :P[03:09:06] <gr00ber> dang[03:09:09] <geist> gr00ber ftw[03:09:09] <gotaku> Looks like 7.[03:09:11] <shackan> it's only for students :P[03:09:28] <geist> students of the study of sluts[03:09:35] <shackan> eww[03:09:47] <shackan> man I'd love such a degree[03:09:50] <gr00ber> aka slutology[03:10:04] <gr00ber> field trips are prolly fun[03:10:13] <gotaku> I'm not going to enter this conversation.[03:10:20] <shackan> but expensive, I guess[03:10:47] <gr00ber> heh, depends on the slut dept. economy[03:10:56] <gr00ber> Slut Fuckulty[03:11:01] <shackan> ok ok, enough[03:11:14] <gr00ber> ok ok then[03:11:35] <gr00ber> it's 3:11 am... maybe i should sleep now[03:11:48] <shackan> same here[03:12:23] <geist> it's the dynamic dou of gr00ber and shackan[03:12:35] <geist> able to kill any conversation in the blink of an eye[03:12:49] <_hugo> *blink*[03:12:52] *** gr00ber has quit IRC[03:13:21] <_hugo> wish i was playing wii tennis[03:18:44] <MikeW> oh, it's the man of the moment![03:19:06] <_hugo> people get too excited over a couple commits[03:19:55] *** gotaku has quit IRC[03:20:10] <_hugo> yank the bug! plonk[03:20:20] <[Katisu]> yeaaahh..a couple...need to step it up a bit :)[03:21:17] <_hugo> we should add a virtual destructor to BStatable when not aiming at R1 ABI compatability (i.e. gcc4 build). The warning is so noisy[03:22:51] *** aldeck has joined #haiku[03:29:42] *** M_199 has joined #haiku[03:29:57] *** BuildFactory has joined #haiku[03:39:44] *** arikato has quit IRC[03:40:50] *** M199 has quit IRC[03:42:04] *** vision has joined #haiku[03:44:38] *** vision is now known as nik0[03:49:26] *** xcasex has joined #haiku[03:52:05] *** JamesB192 has quit IRC[03:59:35] *** MikeW has quit IRC[04:07:24] *** ablyss has joined #haiku[04:14:52] *** eNGIMa has joined #haiku[04:18:39] *** petterhj has quit IRC[04:21:55] *** absabs has joined #haiku[04:26:50] <aldeck> _hugo: you there?[04:29:35] <aldeck> _hugo: just wanted to show you a stack crawl that might interest you.. i'll create a ticket later.. http://haikubeat.free.fr/files/testing/2007-04-09_035306.png[04:32:24] <_hugo> thank you aldeck[04:32:46] <_hugo> if you create a ticket let me know of the number please[04:32:47] <aldeck> :)[04:32:53] <aldeck> ok[04:33:15] <_hugo> aldeck: had you tried beshare before?[04:34:15] <aldeck> _hugo: no, was browsing the tickets list and found an old bug about beshare, so i tested again with the new netstack[04:34:36] <_hugo> oh, good. i'll check it out myself, thanks for pointing it out[04:34:50] <aldeck> _hugo: http://dev.haiku-os.org/ticket/437[04:36:08] <aldeck> need sleep... night all[04:36:25] <_hugo> night[04:36:37] *** aldeck has quit IRC[06:09:50] *** stargater has quit IRC[06:22:34] *** eNGIMa has quit IRC[06:31:59] *** Sikosis has joined #haiku[06:35:49] *** rcjsuen has quit IRC[06:36:48] *** GreyGhost-Ubuntu has joined #haiku[06:38:02] <geist> hmm, need to respond to that scheduler question[06:38:13] <geist> he seems pretty intent on making this much more complicated than they need to be[06:38:20] <_hugo> :-)[06:38:25] <_hugo> it was a long mail, eheh[06:38:33] <geist> yeah, that's why I'm putting it off[06:38:52] <geist> not sure I really like job proposal anyway. 'create an O(1)' scheduler[06:38:59] <geist> well, technically speaking the existing one already is[06:39:10] <geist> it's major flaw is it doesn't SMP scale very well[06:39:18] <geist> and, frankly, it's just weird[06:39:25] <_hugo> its weird it doesnt?[06:39:28] <geist> but it matches the weirdness of beos's original scheduler[06:39:31] <_hugo> ah[06:39:44] *** Sloar has joined #haiku[06:39:47] <geist> so I'm a little concerned about folks screwing with it without really having a clear goal[06:39:57] <_hugo> your concerns make sense[06:40:12] <_hugo> any particular shortcoming besides the SMP issue with the current one?[06:40:22] <geist> I'd be much happier if it was just refactored to SMP scale a bit better, but still be relatively generic code-wise so that it can be tweaked later[06:40:36] <geist> so a lot of his uber optimizations he was talking about would make that tough[06:40:53] <geist> well, it's weird and non dynamic, etc[06:40:58] <geist> but that's exactly how the original beos one worked[06:41:21] <geist> so by changing it much it's entering new territory (not that that's a bad thing)[06:41:30] <_hugo> but maybe still early[06:41:39] <geist> but willy nilly screwing with schedulers is a great way to waste time[06:41:47] <_hugo> :-)[06:41:56] <geist> one of the reasons I never changed it on newos, though I was tempted countless times. if you have no real load, there's no point[06:41:59] *** shackan has quit IRC[06:42:17] <_hugo> im sure the guys wont put people working on useless stuff. at least thats my idea from what i've discussed with axel[06:42:35] <geist> yeah, though axel and i dont necessarilly agree on scheduler stuff, oddly enough[06:42:40] <_hugo> :-))[06:42:49] <geist> he had some ideas last I talked to him that were not how I'd go about it[06:43:19] <_hugo> yes, i remember Andre mentioning something about that in the mail[06:43:23] <geist> anyway, basically I'm inderested in getting the scheduler as a data structure fixed before the scheduler as an algorithm[06:43:30] <geist> the algorithm can be tweaked forever, and that's fine[06:43:41] <_hugo> he also mentions the data structure[06:44:01] <geist> yeah, but he doesn't mention the magic words: per-cpu run queues[06:44:13] <_hugo> ah :-)[06:44:21] <geist> which is how it needs to happen. however, when I looked at how the kernel was organized, it was gonna be a pretty tough change[06:44:46] <geist> specifically the overuse of grabbing the global THREAD_LOCK spinlock to kill rescheduling (for internals of semaphores and others)[06:44:49] <DeadYak> per cpu or per priority?[06:45:02] <geist> which wont really work on a SMP machine with multiple run queues (since there would be effectively one THREAD_LOCK per cpu)[06:45:07] <geist> per cpu[06:45:20] <DeadYak> I thought O(1) more or less required a queue per priority level?[06:45:29] <geist> it already has that[06:45:31] <DeadYak> ah.[06:45:52] <geist> but for smp scalaab ily the general design solution is to have a set of queues per cpu[06:46:05] <geist> it solves two problems[06:46:20] <geist> a) each cpu only has to deal with it's own queue, so the spinlock overhead isn't so bad[06:46:25] <DeadYak> so 1 run queue per priority per cpu?[06:46:30] <geist> and b) it tends to keep threads running the same cpu[06:46:45] <DeadYak> I was about to ask how it decides what tasks to put in which cpu's queues[06:46:59] <geist> DeadYak: the reason I tend to blur that distinction is most 'good' designs of a bunch of run queues put them as a single queue[06:47:04] <geist> sorted by priority[06:47:19] <geist> with pointers off into the middle of the queue so you can still insert at the correct spot O(1)[06:47:22] <DeadYak> with pointers...[06:47:23] <DeadYak> yeah.[06:47:29] <geist> so it's really just a single queue[06:47:44] <geist> or at least it optimizes as such. a find_thread() call then just grabs the head of the queue[06:47:57] <geist> that's precisely how beos did it[06:47:58] <DeadYak> *nod*[06:48:09] <DeadYak> that still leaves the question of how it decides what CPU to put a task on in the first place though[06:48:14] <DeadYak> which is what I'm wondering about[06:48:17] <geist> frankly, I dont know how one would design a non O(1) scheduler[06:48:26] <geist> i think that falls out of the old unixy designs that were IMO really stupid[06:48:57] <DeadYak> right[06:48:58] <geist> though to be more specific there are multiple parts of a scheduler[06:49:11] <geist> most of the designs I've written follow more or less the beos model (in structure, not in algorithm)[06:49:18] <geist> with some flourishes from NT[06:49:27] <geist> basically you have a queue of run queues (single list like I was saying before)[06:49:42] <DeadYak> right[06:49:48] <geist> N number of priorities, top of the queue is always picked first (so the actual find_thread is trivial)[06:50:10] <geist> when the thread blocks it's completely removed form the queue (something that guy on the mailing list seems to have some conceptual issues with)[06:50:23] <geist> and thus doesn't enter any scheduler consideration[06:50:30] <DeadYak> right[06:50:35] <geist> when the thread is added back to the queue is when some processing may happen[06:50:44] <DeadYak> did the old unix schedulers keep them all in the queue but mark them as not ready or something?[06:50:51] <geist> yeah[06:50:54] <DeadYak> because that's the only one I can see a scheduler being O(n)[06:51:10] <geist> well, they did some strange logic whereby all the runnable threads would be examined every so often[06:51:16] <geist> sorted, then run[06:51:27] <geist> when they were all run, it'd reexamine them, sort them, repeat[06:51:37] <geist> hence as you got more threads it wouldn't be O(1)[06:51:57] <DeadYak> *blink* what was the logic behind that at the time?[06:52:02] <geist> no idea[06:52:05] <DeadYak> that seems kind of brain dead[06:52:16] <geist> maybe it fell out of the older batch processing mentality[06:52:21] <DeadYak> ahh[06:52:24] <DeadYak> that would make sense[06:52:35] <DeadYak> I've never really looked into the kind of schedulers on the old mainframes before[06:52:39] <geist> maybe it lets you do things a little more sophisticated sinc eyou're basically looking at all of them at once[06:52:50] <DeadYak> don't really know much about batch schedulers as a consequence[06:53:07] <geist> linux nowadays has some sort of hybrid in that they have two sets of queues[06:53:12] <geist> once the thread is run it's added to the second queue[06:53:24] <DeadYak> Linux has a bunch of different schedulers to pick from I thought?[06:53:25] <geist> and once the first queue is empty, it switches to the second one and runs it[06:53:32] <geist> no, solaris has that model[06:53:43] <DeadYak> ah, I'm thinking of I/O schedulers[06:53:54] <geist> per thread you can pick the scheduling algorithm though it basically ends up in the same run queues[06:54:07] <geist> you're just picking the strategy for dynamically modifying your priority[06:54:16] <geist> beos is odd in that the scheduler is entirely static[06:54:22] <DeadYak> static how?[06:54:25] <geist> your priority never changes[06:54:30] <DeadYak> ah[06:54:33] <geist> unless you do it programmatically[06:54:45] <DeadYak> NT doesn't change priorities but gives you more time slices iirc[06:54:45] <geist> however, it has an odd behavior of actually randomly skipping the head of the queue[06:54:56] <DeadYak> yeah, jbq mentioned the randomizer[06:55:08] <DeadYak> I thought the intent behind that was to prevent an RT thread from completely taking over the proc?[06:55:11] <geist> DeadYak: it does, you move within a range of priorities, +- form your base priority[06:55:22] <geist> (the NT model)[06:55:24] <DeadYak> ah[06:55:35] <geist> sorry if I'm like 2 sentences behind you, you keep changing topics[06:55:39] <DeadYak> I thought NT's way of boosting fg processes was by giving them 3 time slices[06:55:54] <geist> fg is a different thing[06:56:00] <DeadYak> right, it's just your mentioning BeOS not changing priorities made me think of NT because I thought it did that too[06:56:03] <DeadYak> guess I was wrong[06:56:03] <geist> that tweaks the scheduler on another axis[06:56:06] <GreyGhost-Ubuntu> just installed vmware player :)[06:56:31] <geist> NT has a different concept of priority[06:56:42] <DeadYak> in what sense?[06:56:47] <geist> being a different priority effects what queue you're in *and* your quantum[06:56:58] <DeadYak> ah[06:57:00] <geist> tweaking quantum is not something I have a good feel for[06:57:10] <DeadYak> quantum being scheduler time slice?[06:57:17] <geist> and being foreground I think tweaks your quantum, independent of your priority[06:57:19] <geist> right[06:57:20] *** GreyGhost-Ubuntu has left #haiku[06:57:35] <geist> the time slice you're given before the scheduler takes it away from you[06:57:36] <DeadYak> if I remember correctly being fg gives you 3 quanta instead of 1[06:57:49] <geist> for dangeros I more or less copied the NT scheduler, and it worked pretty well[06:58:00] <DeadYak> that might've changed by now though, that was according to my NT book from like 3.51[06:58:06] <geist> yeah, 3x the usual quantum, I think[06:58:43] <geist> you're still just as likely to get trumped by a higher priority, but when you do have the cpu, you can have it longer before getting preempted by other threads in the same queue[06:58:52] <geist> but anyway, offtopic[06:58:54] <DeadYak> right[06:59:29] <geist> beoss scheduler is totally static, but being in a different priority queue really sets the *probability* that you'll get preempted by a higher thread[06:59:38] <geist> which is a AFAIK pretty unique strategy[06:59:46] <DeadYak> how so?[06:59:54] <geist> so you're basically picking the percentage of the total cpu you'll get, based off your priority[06:59:55] *** shackan has joined #haiku[06:59:59] <DeadYak> I remember jbq saying something funky about it reprogramming the scheduler every 3 msec or something along those lines[07:00:03] <DeadYak> but didn't really clarify much on that[07:00:24] <geist> because the way the beos/newos/haiku scheduler works is by rolling a random number every time it finds a thread in the queue[07:00:35] <geist> and a percentage of the time it'll skip that thread and pick the next one[07:00:46] <geist> so how far down you are on the queue you have a less and less chance of running[07:01:02] <DeadYak> newos does that too?[07:01:09] <geist> yeah, I more or less copied the logic[07:01:33] <DeadYak> gotcha[07:01:34] <geist> I dont really like it that much, but without real load there's no point tweaking it[07:01:41] <DeadYak> does that kind of algorithm scale?[07:02:12] <geist> I can't imagine it does. of course it depends on which scheduler case you're looking for[07:02:26] <geist> collision of higher priority to lower priority, multiple threads in the same queue, etc[07:02:45] <geist> and remember, schedulers are totally unused in the case of there only being one thread running at a time[07:02:58] <geist> which on a modern interactive system is a good chunk of the time[07:03:17] <DeadYak> right[07:04:39] <geist> anyway, seperate queues per cpu introduces some more fun stuff to solve[07:04:46] <DeadYak> like?[07:04:53] <geist> like when to migrate threads between queues, what you do with new threads, etc[07:05:22] <DeadYak> how was axel planning on doing it without per cpu queues?[07:05:26] <DeadYak> locks on one queue?[07:05:42] <geist> didnt;' think axel was planning on it much[07:05:59] <geist> it's probably not too terribly difficult. I looked at what it would take[07:06:19] <geist> the biggest problem is the stuff outside of the scheduler that sort of rely on the THREAD_LOCK being global[07:06:38] <geist> but even then it probably is just a case by case basis to figure out the best alternate solution for those[07:07:16] <geist> right now if you grab the THREAD_LOCk you've effectively disabled all context switching on all cpus, so it's easy to then go muck around with the scheduler at that point[07:07:37] <geist> or at least be guaranteed that nothing will preempt you, or another cpu move a thread out from underneath you, etc[07:07:46] <DeadYak> right[07:07:50] <geist> the sem code uses it i know when it's resuming threads[07:08:12] <geist> most of that will probably translate okay to just grabbing the spinlock that protects the local cpu's queue[07:08:30] <DeadYak> right[07:08:33] <geist> and always waking threads up to the local cpu[07:08:34] <DeadYak> hmm[07:09:04] <geist> on the other hand, depending on how hard you want to keep the cpu's balanced, you might want to wake new threads up to the lightest occupied cpu[07:09:18] <geist> on the other hand, every time you do that you begin to negate the benefits of having seperate queus[07:09:25] <geist> so there's definitely some tradeoffs there[07:09:28] <DeadYak> that and cache pollution[07:10:04] <geist> the single queue strategy should perfectly keep the cpus balanced, provided you send reschedule ICIs every time a new thread is woken up[07:10:11] <geist> but that's pretty expensive[07:10:21] <geist> but that's close to what it does now[07:10:40] <geist> it might even do that exactly. lemme check[07:11:38] <geist> yeah, it's actually kind of broken right now :)[07:11:42] <DeadYak> so how does something like what NT does where you can actually force affinity on a per-proc basis work?[07:11:53] <DeadYak> per process I mean[07:12:09] <geist> that's easy. you just set a bit that makes it not eligable for transfer[07:12:20] <geist> and that pins it to a particular run queue[07:12:25] <geist> if you have a run queue per cpu[07:12:27] <DeadYak> right, but NT lets you check a list of boxes for all the CPUs in the system[07:12:31] <geist> if not, it's a bit more complicated[07:12:33] <DeadYak> so it's finer than just a bit[07:12:50] <geist> well, that's true. probably a bitfield[07:12:56] <geist> betcha 32 cpus is the limit[07:13:00] <DeadYak> then again if they were using a bitfield, that'd explain the 32 CPU limit NT used to have....[07:13:09] <geist> yup, I think that's precisely why[07:13:12] <DeadYak> it was at one point, I dunno if it still is[07:13:19] <DeadYak> that might've changed for 2k3 or Vista[07:13:20] <geist> anyway, yeah I just looked at the source[07:13:33] <geist> when haiku releases a sem, it only reschedules the local cpu[07:13:52] <DeadYak> oops[07:13:55] <geist> so that actually means one or more other cpus may be idle at that moment[07:14:00] <geist> and it'll remain as such[07:14:12] <geist> however, checking to see what the other cpus are doing at the time is relatively expensie[07:14:15] <geist> so it's a tradeoff[07:15:13] <geist> the most expensive algorithm would be to force a global reschedule, and let the cpus fight it out[07:15:44] <geist> lesser expensive would be to iterate through each of the cpu's current threads, seeing whatpriority they're running and reschedule any of the ones that have a lower priority thread than the one you just woke up (still sucks)[07:16:05] <geist> and lesser expensive strategy would be to only reschedule the ones that are idle[07:16:25] <geist> and the lest expensive is to just ignore the other cpus and let it get sorted out in the next time slice (what it does now)[07:16:56] <geist> perhaps a middle ground would be that if you woke up a RT thread you force a global reschedule, otherwise, blah[07:17:22] <DeadYak> well, it seems reasonable when you consider that the goal here isn't exactly to run on 64-way servers :)[07:17:44] <geist> yeah, but the goal is to be low latency[07:17:52] <DeadYak> true[07:17:58] <geist> which would tend to favor expensive but fair (global reschedule)[07:18:30] <geist> to be 100% low latency, you probably can't afford to have per cpu queues, and the single queue is the best strategy[07:19:10] <geist> since at best you'd migrate threads around between the cpus a few times a second, so you're probably not fair all the time (have a bunch of higher priority threads on one cpu, lowers on another, etc)[07:19:17] <geist> basically maf is hard[07:19:27] * DeadYak nods[07:21:38] <geist> hmm, looks liek someone added a quick hack a while back that temporarily boosts the priority of a thread by 1 when getting woken from a sem, for one schedule period[07:21:52] <DeadYak> is that good or bad?[07:22:03] <geist> it's okay[07:22:13] <geist> i seriously doubt it's all that effective though[07:23:17] <geist> mmm, 2 buck chuck[07:28:21] <DeadYak> chuck?[07:29:23] <geist> 2 dollar wine, Charles Shaw[07:29:36] <geist> http://en.wikipedia.org/wiki/Charles_Shaw_wine[07:30:14] <DeadYak> ah[07:34:43] * JonathanThompson screams into the ether in a manner with no meaning[07:35:26] <DeadYak> ha[07:35:54] <JonathanThompson> Just ask geist about my latest bad cow puns :)[07:38:10] <geist> dont udder another word[07:38:21] <JonathanThompson> I promise, no moo! :)[07:53:49] *** shackan has quit IRC[07:59:01] *** kr1stof_ is now known as kr1stof[08:01:11] *** kokito has quit IRC[08:05:04] *** shackan has joined #haiku[08:21:27] *** GreyGhost has joined #haiku[08:26:24] *** Sikosis has quit IRC[08:29:46] *** chaotic has joined #haiku[08:30:07] *** Begasus has joined #haiku[08:30:20] * JonathanThompson welcomes Begasus with a bad cow pun[08:31:24] <chaotic> Hello, i'm new here.[08:31:43] * JonathanThompson welcomes chaotic with a bad cow pun[08:32:00] <JonathanThompson> If you stay here long enough, you'll think some or all of us are udder our minds :)[08:32:07] * arachnist welcomes JonathanThompson with a bad cow[08:32:35] <arachnist> look out, she's about to eat your hair[08:32:42] <chaotic> i hope so[08:32:49] <JonathanThompson> That'll be slim pickings :)[08:33:02] * JonathanThompson doesn't have much hair to display a cowlick with[08:35:10] *** Begasus is now known as Begasus_bbl[08:35:17] <chaotic> oh, sorry, my english is a little bit terrible[08:35:41] <JonathanThompson> If you get confused by mine, it may be due to me using a lot of word-play.[08:35:47] <arachnist> don't worry, neather is mine[08:36:07] <MrSunshine> haha :P[08:36:14] <MrSunshine> haiku welcomes all ... :)[08:36:47] <JonathanThompson> Pirates playing with harpoons is all fun and games until someone loses an aye-aye![08:37:04] <arachnist> :>[08:37:19] <arachnist> aye-aye, captain! ;)[08:38:39] * JonathanThompson wonders how mant Haiku community people have what's now a collector's item in the form of Zeta 1.51[08:39:36] *** GreyGhost has quit IRC[08:42:18] <chaotic> someone her that spaeks german ???[08:42:38] <JonathanThompson> Mine is nicht so gut :)[08:42:46] <JonathanThompson> (only 1 semester in 2001)[08:43:13] <JonathanThompson> Your english is better than my german :)[08:43:42] <JonathanThompson> dr_evil (when he's on) does.[08:44:00] <chaotic> oh cool, lets haer[08:44:25] <JonathanThompson> I can mangle german, what little I learned.[08:44:28] <JonathanThompson> I forgot a lot of it.[08:45:52] <arachnist> einz, zwei, drei...zein, that's all i know :>[08:46:11] <arachnist> and 'ich heise robert'[08:46:13] <arachnist> ;>[08:46:29] <JonathanThompson> "Ist kaput!"[08:46:30] <JonathanThompson> :)[08:46:37] <chaotic> you cannot have forgotten nevertheless everything[08:46:53][08:46:54] <JonathanThompson> Not for lack of trying! :) (just kidding)[08:47:15] * JonathanThompson isn't certain how to type esset in BeOS[08:47:52][08:47:55] <JonathanThompson> B[08:48:11] * JonathanThompson needs to run at a lower resolution to see what difference that made.[08:48:28] <chaotic> who in the HAIKU tem is responsible for the vmware image[08:48:39] <chaotic> tem = team[08:48:55] <arachnist> german dvorak is a little wierd, so i'll stick to us-english :>[08:49:45] <arachnist> (who in their right mind would swap "u" and "i", or "r" and "t"[08:50:34] <JonathanThompson> You mention "right mind" like computers belonged with that :)[08:50:48] <arachnist> :P[08:51:19] <JonathanThompson> The original QWERTY keyboard was designed to be inefficient, intentionally.[08:51:49] <arachnist> that's why i use dvorak keyboard layout[08:51:56] <JonathanThompson> This is because mechanical typewriters using that technology were too slow to allow people to type things very fast without getting jammed.[08:52:45] <JonathanThompson> Perhaps the first keyboard layout would have been completely different, had IBM created the Selectric using the daisywheel :P[08:53:45] <arachnist> well, i type at about the same speed on both dvorak and qwerty keyboards, but i don't get any wrist-ache[08:54:00] <arachnist> (on dvorak that is)[08:54:33] <JonathanThompson> The Apple 2c is the first computer I'm aware of that had native support for Dvorak keyboard layouts.[08:54:46] <JonathanThompson> Just press in a switch on the computer by the keyboard, and you're all set.[08:54:52] <arachnist> it had a hardware switch, didn't it? :>[08:54:57] <JonathanThompson> (as long as you didn't have to look at the keyboard!)[08:55:03] <JonathanThompson> Yes.[08:55:25] <arachnist> btw, rearranging keys on my keyboard was a pain in the arse[08:55:35] <JonathanThompson> I remember a commercial or ad where the world record holder for typing set a new speed record on an Apple 2c :P[08:55:56] <arachnist> it has a laptop-like keys[08:56:05] * JonathanThompson wonders how quickly she wore through that thing, typing around 200 wpm[08:57:07] <arachnist> that reminds me that i'll need to get myself a new keyboard[08:57:30] <arachnist> cleaning this keyboard is also a pain[08:58:05] <JonathanThompson> I used my first Microsoft keyboard for about 11 years before it died spontaneously in November.[08:58:16] <JonathanThompson> It was the Microsoft Natural Keyboard version 1.0.[08:58:19] <arachnist> blah[08:58:24] <arachnist> i hate "natural keyboards"[08:58:27] <geist> ß[08:58:30] <JonathanThompson> It was quite dirty by that time, too :)[08:58:36] <arachnist> they feel so awkward to me...[08:58:37] <geist> did that come through as an esset?[08:58:44] <arachnist> 085830 | @ geist | ß[08:58:49] <JonathanThompson> Your character there didn't :)[08:58:53] <arachnist> looks like one to me[08:58:59] <geist> weird, terminal actually translated it to 'ss'[08:59:11] <JonathanThompson> B[08:59:28] <geist> i had typed it into textedit here and copied it into terminal[08:59:35] <arachnist> geist: irssi with '/set recode_transliterate on' ?[08:59:53] <arachnist> ćłąóęńśźż[08:59:54] <geist> ß[09:00:04] <arachnist> and here are some polish funky chars[09:00:09] <geist> ah, it wason. I turned it off and I get \ublahblah[09:00:26] <arachnist> geist: seems like you have a non-utf terminal[09:00:39] <geist> oh well, unicode via a ssh through screen to irssi is asking for trouble[09:01:10] <arachnist> get yourself an utf8 terminal, screen -Ur and /set term_charset UTF-8[09:01:27] <geist> ß[09:01:42] <arachnist> :>[09:01:42] <arachnist> looks normal to me[09:01:51] <arachnist> geist: can you see those? |> ćłąóńśźżę[09:01:58] <geist> ßß[09:02:03] <geist> nope, it's trashed[09:02:09] <JonathanThompson> Perhaps geist's attempts haven't come across as I expected due to font issues.[09:02:10] <geist> but in a different way than before :)[09:02:28] <geist> ?[09:02:43] <JonathanThompson> ![09:02:48] <arachnist> oh, and it'd be nice if your system was utf-8 too[09:02:59] <geist> I'm pretty sure it is. it's terminal on OSX[09:03:06] <JonathanThompson> Better utf-8 than wtf :)[09:03:07] <geist> it even has UTF-8 as an option (which it's set to)[09:03:16] <arachnist> hmm[09:03:25] <arachnist> echo $LANG $LC_ALL ?[09:03:35] <geist> on which machine? :)[09:03:42] <arachnist> both[09:03:43] <geist> there are multiple computers involved here[09:04:04] <geist> both are empty[09:04:06] <arachnist> well, i ssh from one vserver on this linux box to another ;>[09:04:30] <geist> oh well, it's not like I actually care[09:04:43] <arachnist> and OS is the other box running?[09:04:48] <geist> linux[09:05:07] <chaotic> I would like not to interrupt your discussion, but give it here also somewhat over HAIKU?[09:05:12] <arachnist> (i used to have problems with utf stuff on fbsd)[09:05:33] <arachnist> ok, no idea then[09:05:51] <JonathanThompson> Ask geist, he's had some important contributions to Haiku :)[09:06:04] <geist> main screen turn on[09:06:21] <arachnist> chaotic: first we needhmm[09:06:48] <arachnist> first i need to move my cat[09:07:00] <arachnist> it sits in front of my screen and meows[09:07:09] <chaotic> who in the HAIKU team is responsible for the vmware image[09:07:23] <JonathanThompson> Not on top, where it's warm? Or is it an LCD monitor?[09:07:37] <arachnist> JonathanThompson: it's an lcd[09:07:42] <JonathanThompson> Ah.[09:08:09] <JonathanThompson> I haven't updated my main machine to LCD, because I still have perfectly workable 19" monitors that run at 2048*1536.[09:12:48] <arachnist> i used to have a crappy 17" crt, so i was happy upgrading to 19" widescreen lcd[09:12:58] <arachnist> 1440x900 ftw! ;>[09:13:20] <arachnist> need to get myself another one[09:16:09] <geist> oh weird, this song is in a weird ass tuning[09:16:12] <geist> CCCCFF[09:27:00] *** _hugo-haiku has joined #haiku[09:27:59] <MrSunshine> gah let it be weekend again .... :)[09:28:04] <_hugo-haiku> plonk.[09:28:09] *** _hugo-haiku has quit IRC[09:33:53] *** rhu has quit IRC[09:34:02] *** Euver has joined #haiku[09:35:00] *** PORDO has joined #haiku[09:35:11] <PORDO> can anyone help me obtain the data from the icon set election?[09:35:22] <PORDO> i'd like to analyze it as an election method researcher.[09:35:31] <PORDO> it was a rare case of a large range voting election that could have potentially been contentious.[09:36:17] <geist> interesting[09:37:34] *** chaotic has quit IRC[09:37:48] <PORDO> i'm trying to get range voting used at San Francisco State University.[09:38:13] <PORDO> and at lots of other organizations like the SF Labor Council. Range Voting is a HUGE improvement over plurality - as big of an improvement as plurality is over random selection.[09:39:01] <geist> hmm, thought SF did something like that for city council seats?[09:39:09] <geist> probably not range voting, but it was something else odd[09:39:31] <geist> like you pick the top n people you want[09:44:14] <geist> well, okay[09:44:18] * geist sleeps[09:47:32] *** eNGIMa has joined #haiku[09:48:05] <PORDO> they use IRV, which is terrible.[09:48:13] <PORDO> it's not even IRV - it's a bastardized form of it.[09:48:17] <PORDO> http://RangeVoting.org/IRV.html[09:48:34] <PORDO> look at IRV's utility efficiency here - MUCH worse than Range Voting: http://rangevoting.org/vsi.html[09:52:33] *** Begasus_bbl is now known as Begasus[10:04:04] <Begasus> hmmm ... keyboard isn't working in Haiku ...[10:07:28] <PORDO> you use haiku?[10:07:47] <PORDO> awesome. hey, do you know how i can get those election results from the icon set election?[10:07:55] <PORDO> i'd like to use them for scientific research for the Center for Range Voting.[10:08:00] <PORDO> http://RangeVoting.org/[10:08:44] *** SunWuKung has joined #haiku[10:09:02] <PORDO> Begasus ?[10:09:08] <PORDO> SunWuKung do you work on Haiku?[10:09:42] <SunWuKung> nah...[10:11:00] <Begasus> I test Haiku from time to time PORDO[10:11:24] <Begasus> and about the result ... can't help you out there sorry[10:11:48] <Begasus> result -- election *[10:12:03] <PORDO> :( thanks.[10:12:18] <PORDO> it's got to be the largest potentially competative Range Voting election with a record of ballots.[10:13:14] <PORDO> it would be nice to know, for instance, what fraction of voters were strategic vs. honest; that is, willing to strategically min/max all their scores to get the best results.[10:14:27] <SunWuKung> PORDO: same as Begasus, I use Haiku[10:14:51] <PORDO> wow. how stable is it?[10:14:55] <PORDO> i used BeOS _years_ ago.[10:15:00] <PORDO> i miss it, but ubuntu is getting so good.[10:15:33] <MrSunshine> 55 min is the best ive seen so far .. and thats with running opengl apps etc :)[10:15:44] <PORDO> with or without?[10:15:46] <MrSunshine> dunno how long the record so far is tho[10:15:55] <MrSunshine> huh ?[10:16:00] <PORDO> but while it's up, does it seem "fully functional"[10:16:08] <PORDO> do most beos apps work ?[10:16:27] <Begasus> SunWuKung running the latest rev?[10:17:14] <MrSunshine> PORDO, havent tried it in like half a year so i dunno now .. saw pictures of python working with bethon yesterday .. and ive seen BeIDE run and alot of other apps ... Quake 3 ... Vision .. firefox[10:17:15] <PORDO> it is so beyond me why they decided to make it binary compatible. i would have just gone for an OS strongly _based on_ beos, with a similar API, such that the BeOS programs could have find/replace done in their source code and be easily compilable.[10:17:46] <PORDO> the binary compatibility just seems insane to me.[10:17:57] <MrSunshine> hehe :)[10:18:03] <MrSunshine> well its what they are aiming for :)[10:18:09] <PORDO> haiku really needs to have multi-user and practical things i think, more than binary support with beos.[10:18:28] <MrSunshine> PORDO, i think its cause ALOT of apps are dead, the source code is gone etc ... and without apps the OS just wont hit ... :)[10:18:29] <PORDO> i don't see any point in it. you could just recompile the beos apps.[10:18:41] <PORDO> hrm. perhaps perhaps.[10:18:49] <MrSunshine> PORDO, alot of commercial apps also that its impossible to get to the source code of[10:18:56] <PORDO> yeah. good point.[10:19:21] <PORDO> i've still found nothing that compares to beos in terms of that sense of responsiveness and lightness.[10:19:26] <PORDO> and the beos icons are just SICK.[10:19:47] <PORDO> they are cooler than Gnome, cooler than OSX - just amazingly modern yet gentle and almost sort of cartoonish.[10:20:14] <PORDO> i hope they get somewhere with it. i really really do. for right now, i just need their election results though. :)[10:22:05] *** mazon is now known as Mazon[10:29:10] <Begasus> drop a note in the mailinglist PORDO :)[10:29:15] <PORDO> how?[10:29:28] <PORDO> do i have to register first? :([10:29:33] <Begasus> google for the haiku mailinglist[10:29:38] <PORDO> k. thx[10:29:49] <Begasus> not sure .. but it doesn't kill you ;)[10:30:31] <CIA-17> stippi * r20623 /haiku/trunk/src/apps/icon-o-matic/ (25 files in 8 dirs): (log message trimmed)[10:30:31] <CIA-17> culmulative update...[10:30:31] <CIA-17> * holding down space, or using the third mouse button, will force[10:30:31] <CIA-17> the "pan canvas" mode[10:30:31] <CIA-17> * using the mouse wheel zooms in and out[10:30:32] <CIA-17> * fixed issues with the undo commands when nudging something with[10:30:33] <CIA-17> the cursor keys[10:31:02] *** genkie has joined #haiku[10:31:15] <Begasus> moin genkie[10:31:19] <genkie> moin[10:39:31] *** Begasus has quit IRC[10:39:54] *** Begasus has joined #haiku[10:40:19] <PORDO> i sent an email to the list Begasus. Hopefully I get the election results soon. :) Say, do you know how many people voted in those icon set elections?[10:40:44] <Begasus> can't tell PORDO ... been a while[10:40:52] <PORDO> cool[10:41:23] <Begasus> best place to look for answers there ;)[10:41:47] <Begasus> so lets see if the keyboard prob is also on the pc ...[10:42:14] *** Ingenu has joined #Haiku[10:42:40] <Begasus> nope .. so probly a laptop/keyboard isseu ..[10:42:45] <Begasus> morning Ingenu[10:43:27] *** chaotic has joined #haiku[10:44:29] <chaotic> hello ???[10:44:55] <Begasus> hi[10:46:15] *** steffen_f has joined #haiku[10:46:37] <chaotic> somebody here to speak german ??[10:47:17] <arachnist> my grandma, she's in the next room[10:47:27] <arachnist> (really!)[10:47:50] *** SunWuKung has quit IRC[10:48:08] <Begasus> da sind ja einige chaotic[10:48:43][10:49:13] <Begasus> bin da aber selber kein dev ;)[10:49:31] <Begasus> die meisten dev's sind nicht immer im channel[10:49:59][10:51:07] <Begasus> was nutzt du jetz fuhr IRC chaotic?[10:51:23] <Begasus> kanst vieleicht umschalten nach 'unicode' im app da ...[10:51:32][10:51:55] <chaotic> bin mit opera unterwegs[10:52:13] <chaotic> bin mal schnell fruehstuecken[10:52:26] <Begasus> ah k kp ... nur die umlauten sehen nicht aus so ;)[10:52:35] <Begasus> guten apetit[10:54:40] *** Jorsa has joined #haiku[10:54:45] *** Jorsa has left #haiku[10:55:56] *** Euver has quit IRC[10:56:21] *** Euver has joined #haiku[11:02:43] *** Lelldorin1 has joined #haiku[11:08:24] *** shackan has quit IRC[11:20:10] *** kb7sqi has quit IRC[11:22:59] *** kr1stof has quit IRC[11:23:23] *** aldeck has joined #haiku[11:23:58] *** doppeljot has joined #haiku[11:25:19] *** Begasus is now known as Begasus_bbl[11:26:10] *** genkie has quit IRC[11:26:38] *** PORDO has quit IRC[11:30:53] *** kb7sqi has joined #haiku[11:35:04] <CIA-17> hugosantos * r20624 /haiku/trunk/src/add-ons/kernel/network/protocols/tcp/ (7 files): (log message trimmed)[11:35:04] <CIA-17> fixed multiple TCP issues.[11:35:04] <CIA-17> - Moved FIN handling to after Segment Text handling, it matches the RFC and fixes a possible issue with a user missing some final segment data when receiving a FIN.[11:35:04] <CIA-17> - Fixed the FIN handling to better terminate connections (no more senseless chatter in the end with RSTs etc).[11:35:06] <CIA-17> - When Bind()ing, set the socket address so a call to getsockname() gets the allocated port correctly.[11:35:08] <CIA-17> - After reaching the ESTABLISHED state, try to wake all pending threads (still needs work).[11:35:11] <CIA-17> - Corrected the maximum amount of time we wait in SendData() to respect socket->send.timeout.[11:37:30] *** Sloar has quit IRC[11:47:50] *** doppeljot has quit IRC[11:49:29] *** numbah has joined #haiku[11:51:11] *** Lelldorin1 has joined #haiku[11:55:50] *** Leszek has joined #haiku[11:56:09] <Leszek> hi[11:58:33] *** aldeck has quit IRC[12:04:30] *** Begasus_bbl has quit IRC[12:04:52] *** Begasus has joined #haiku[12:05:48] *** wildur has joined #Haiku[12:10:36] *** steffen_f has quit IRC[12:10:47] *** Karina`` has quit IRC[12:11:13] *** Karina`` has joined #haiku[12:14:21] *** Sil2100 has joined #haiku[12:16:56] *** aldeck has joined #haiku[12:17:29] *** arikato has joined #haiku[12:23:39] *** petterhj has joined #haiku[12:27:42] *** TheDave has joined #haiku[12:27:45] <TheDave> o/[12:28:12] <arachnist> \o[12:28:42] <TheDave> quick question, did running Haiku in QEMU break in the past few days or should I look into my qemu config?[12:29:11] <arachnist> i run haiku in qemu + kvm, works nice[12:29:21] <TheDave> atm it just hangs at a blue screen with a pointer for me[12:29:33] <TheDave> panics if i try to run it in safe mode[12:30:53] <arachnist> TheDave: qemu on windows?[12:30:58] <TheDave> linux[12:32:12] <arachnist> hmm[12:32:15] <arachnist> strange[12:32:35] <TheDave> serial console doesn't output anythikng strange besides the usual driver errors[12:32:49] *** Euver has quit IRC[12:34:18] <arachnist> "PANIC: remove page 0x928a2c80 from cache 0x90965840: page still has mappings!\n\nWelcome to Kernel Debugging Land...\nRunninng on CPU 0\nkdebug>"[12:35:08] <arachnist> hrm... not repeatable...[12:35:15] <TheDave> different addresses but that's the crash i have[12:36:25] <arachnist> hmm[12:36:29] <TheDave> tried rebuilding it 3 times and it runs on real hardware[12:36:30] <arachnist> seems to be related to cdrom[12:36:43] <arachnist> arachnist@newnest haiku % qemu-system-x86_64 -net user -net nic -m 768 -hda haiku.image -cdrom /pub/iso/kubuntu-7.04-beta-alternate-amd64.iso -boot c <| did crash[12:36:54] <arachnist> qemu-system-x86_64 -net user -net nic -m 768 -hda haiku.image -boot c <| didn't crash[12:37:13] <TheDave> i just ran it with "qemu -m 512 -hda haiku.image"[12:38:01] <arachnist> hmm[12:38:13] <arachnist> qemu-system-x86_64 -net user -net nic -m 768 -hda haiku.image -cdrom /pub/iso/ReactOS.iso -boot c <| didn't crash[12:38:17] <arachnist> strange[12:38:22] <TheDave> for me it enters KDL right after loading the dev/zero driver, don't know if that's related[12:38:44] <TheDave> *relevant[12:39:04] <arachnist> hmm[12:39:40] <TheDave> want me to send the info somewhere?[12:39:59] <arachnist> hmm[12:40:04] <arachnist> it doesn't crash anymore here[12:40:13] <arachnist> probably was related to kvm paging issues[12:40:45] <TheDave> i don't have either kvm nor kqemu loaded[12:42:20] <arachnist> well, i couldn't try without kvm here, because qmeu segfaults here every time i try to run it with -no-kvm :>[12:42:43] <TheDave> hehe, yikes[12:45:11] <TheDave> what info from kdl would be good to dump?[12:52:22] *** GreyGhost has joined #haiku[12:53:18] *** Ball000 has quit IRC[13:02:38] *** emitrax has joined #haiku[13:05:11] *** petterhj- has joined #haiku[13:05:28] *** petterhj has quit IRC[13:05:41] *** petterhj has joined #haiku[13:11:25] *** dr_evil has joined #haiku[13:11:34] *** ChanServ sets mode: +o dr_evil[13:13:51] *** emitrax has quit IRC[13:14:00] *** petterhj- has quit IRC[13:21:44] *** Ingenu has quit IRC[13:26:21] *** Ingenu has joined #Haiku[13:35:49] *** Teknomancer has joined #haiku[13:39:43] *** rcjsuen has joined #haiku[13:48:07] *** MYOB has joined #haiku[13:56:21] *** rhu has joined #haiku[13:57:54] *** Leszek has quit IRC[14:02:23] *** Sil2100 has quit IRC[14:02:47] *** GreyGhost has quit IRC[14:20:15] *** tombhad-AC has joined #haiku[14:22:29] *** Teknomancer has quit IRC[14:25:49] *** tombhadAC has quit IRC[14:27:10] *** tombhadAC has joined #haiku[14:29:35] *** Begasus has quit IRC[14:29:53] *** Begasus has joined #haiku[14:45:52] *** Euver has joined #haiku[14:47:43] *** Begasus has quit IRC[14:47:49] *** Begas_ has joined #haiku[14:49:45] *** nik0 has quit IRC[14:55:56] *** Ingenu has quit IRC[15:02:58] *** doppeljot has joined #haiku[15:05:33] *** doppeljot has left #haiku[15:18:17] *** shackan has joined #haiku[15:29:55] *** Begas_ is now known as Begasus_bbl[15:31:09] *** Ingenu has joined #Haiku[15:33:07] *** TLF has joined #haiku[15:40:18] *** Ingenu has quit IRC[15:41:16] *** lorglas has joined #haiku[15:43:36] *** Ingenu has joined #Haiku[15:43:43] *** lorglas has quit IRC[15:46:04] *** StylusEater_Work has joined #haiku[15:49:11] *** stargater has joined #haiku[15:49:33] <stargater> hi[15:53:16] <Jin> /me yawns[15:53:27] <Jin> with a space on top[15:53:30] <Jin> hi all[15:53:51] <Ingenu> 'lo[15:56:55] *** guildencrantz has joined #haiku[15:59:18] <absabs> hi[16:03:46] *** stargater has quit IRC[16:05:45] *** Stargater has joined #haiku[16:05:47] <Stargater> re[16:11:28] *** MYOB has quit IRC[16:16:04] <Stargater> Begasus_bbl around ?[16:18:10] <Stargater> anyone used a ipw2100 driver in haiku ?[16:19:57] *** shackan has quit IRC[16:23:43] *** MYOB has joined #haiku[16:25:11] *** cyrus82 has joined #haiku[16:35:31] *** Sloar has joined #haiku[16:36:59] *** Stargater has quit IRC[16:40:10] *** cyrus82 has quit IRC[16:52:11] *** wildur has quit IRC[16:55:12] <StylusEater_Work> hi[16:55:31] *** xcasex_ has joined #haiku[17:03:14] *** Teknomancer has joined #haiku[17:06:45] *** GreyGhost has joined #haiku[17:11:15] *** kokito has joined #haiku[17:11:54] *** xcasex has quit IRC[17:44:54] *** Witukind has joined #haiku[17:45:21] *** Witukind has quit IRC[17:58:15] *** absabs has quit IRC[18:04:33] *** guma has joined #haiku[18:10:37] *** b3w_ has joined #haiku[18:16:22] *** vision has joined #haiku[18:22:51] *** vision has quit IRC[18:25:26] *** b3w has quit IRC[18:27:12] *** mmu_man has joined #haiku[18:27:12] *** ChanServ sets mode: +o mmu_man[18:30:03] <dr_evil> plop[18:30:32] <mmu_man> plop![18:30:44] <mmu_man> catsup 3days[18:30:55] <mmu_man> :p[18:31:25] <Teknomancer> atrus catsup :([18:33:05] <mmu_man> eh[18:33:07] <mmu_man> http://www.osnews.com/story.php/17650/Debian-GNULinux-4.0-Released[18:33:08] <mmu_man> lol[18:33:13] <dr_evil> too much booze mmu?[18:33:14] <mmu_man> quite late for an april fools[18:33:24] <mmu_man> dr_evil was doing kung fu al week end[18:33:38] <mmu_man> well actually I did more tai chi and qi gong[18:33:52] <Teknomancer> yin and yang[18:33:56] <dr_evil> I've got a cold since thursday[18:33:57] <mmu_man> flying cisors don't get along very well wih a hernia :)[18:34:06] <Teknomancer> bruce lee, jackie chan[18:34:07] <mmu_man> :-([18:34:26] <mmu_man> Teknomancer no film, no wires :p[18:34:41] <Teknomancer> yeah :)[18:34:43] <mmu_man> hmm, yeah we did some filming though =)[18:34:54] *** GreyGhost has left #haiku[18:37:22] *** MYOB has quit IRC[18:37:42] *** MYOB has joined #haiku[18:38:34] <dr_evil> I still want to by a mobile phone that can film 640*480 at 25 fps or better, and capture 3 Mpixel or more photos[18:38:35] <mmu_man> not much news for a whole week end...[18:38:46] <mmu_man> let's see what you all did on svn :)[18:38:57] <MYOB> mmu_man a 4 day weekend (as it was in much of europe) generally means heavy drinking not coding ;)[18:39:08] <mmu_man> right[18:39:37] <mmu_man> I forget I'm only one of few who don't drink :)[18:40:07] <Teknomancer> i haven't touched alcholic drinks for 22 years now[18:40:17] <Teknomancer> :P[18:40:29] <MYOB> Teknomancer are you even "allowed" ?[18:40:51] <Teknomancer> MYOB well most of my friends do drink, but if their parents find out they'll be in trouble :)[18:41:10] <Teknomancer> in some "high class" parties little drinking is common here too[18:41:12] <MYOB> I remember one of the Indians I worked with in BT telling me he wasn't, While eating a burger after the pub. He wasn't 'allowed' eat that either :D[18:41:32] <Teknomancer> but in general u won't find people drinking casually as much as u do there[18:41:41] <Teknomancer> yeah i'm a vegetarian too :P[18:41:50] <mmu_man> appart champain or clairette I don't like much alcohol anyway[18:42:01] <mmu_man> bear I can't even stand the smell of it[18:42:07] <mmu_man> beer[18:42:10] <Teknomancer> but its not that much of an 'allowed' thing, even if i eat meat my mom/dad won't say anything but i just don't want to[18:42:34] <MYOB> that said if you really wanted to get him on the hypocritical, you got him going about the Mumbai based call centre we were always cleaning up after....[18:42:45] <mmu_man> you don't want to or they don't want to ? :)[18:42:53] <Teknomancer> mmu_man i don't want to[18:43:25] <Teknomancer> i'm perfectly happy with the vast variety in vegetarian food :P[18:43:53] *** stargater has joined #haiku[18:44:05] <MYOB> vegetarian food and no booze would drive me insane[18:44:50] <Teknomancer> i heard its extremely bitter :S pfft, i'd rather drink something sweet[18:45:34] <MYOB> like vodka + irn bru then ;)[18:48:48] <stargater> re[18:55:51] <_hugo> yawn[18:58:28] <mmu_man> 7 wow :)[18:58:37] <mmu_man> do we have enough mentors ?[18:58:47] <mmu_man> I guess I'll just register anyway[18:58:48] *** TLF has quit IRC[18:59:03] <_hugo> mmu_man: torture waldemar for that information[18:59:08] <Teknomancer> i just finished being a student[18:59:09] <Teknomancer> damn[18:59:28] <mmu_man> I'll register anyway, one more backup is always useful[18:59:39] <_hugo> :-)[18:59:40] <mmu_man> Teknomancer get a Ph.D :)[18:59:52] <Teknomancer> mmu_man easier said than done :P[18:59:59] <mmu_man> I know :)[19:00:16] <mmu_man> well, you can do a Ph.D, that doesn't mean you'll pass it :)[19:01:28] <MYOB> eww college[19:01:29] *** jevin has joined #haiku[19:01:29] <Teknomancer> no even to join would take time[19:02:50] <kokito> good morning folks[19:02:55] <mmu_man> plop[19:03:30] <MYOB> economy here is in such a way that going to college is entirely pointless[19:03:33] <kokito> mmu_man, may help to have one more mentor. would be a good idea to register.[19:03:44] <MYOB> so PhD just sounds SCARY[19:03:53] <kokito> how's that MYOB?[19:04:15] <MYOB> kokito total lack of supply for IT jobs, so they hire the people they think can do it[19:04:27] <mmu_man> kokito ok[19:04:28] <MYOB> having gone to college for 4 years is a serious disadvantage versus having worked for those 4 years[19:04:40] <mmu_man> just need to figure out the correct tishirt size :P[19:04:41] <MYOB> I managed graduates in my last job...[19:05:19] <MYOB> (lack of supply = lack of candidates, not lack of jobs)[19:05:33] <kokito> so, not having a degree is not a disadvantage then?[19:05:39] <MYOB> effectively anyone can get an IT job here but the better ones go to those who have experience not a bit of paper[19:05:49] <_hugo> MYOB: where's 'here'?[19:05:52] <MYOB> kokito its not, no[19:05:54] <MYOB> _hugo Ireland[19:06:04] <kokito> hmmm... interesting[19:06:10] <_hugo> interesting indeed[19:06:13] <_hugo> i wouldnt expect that[19:06:22] <mmu_man> in france the better job goes to noone[19:06:28] <Teknomancer> so anyone can become kernel engineers there??[19:06:34] <mmu_man> cause when you have diplomas they say you dont have experience[19:06:41] <Teknomancer> or take up jobs like that?[19:06:53] <mmu_man> and when you have experience they say you don't have diplomas[19:06:56] <mmu_man> :D[19:07:07] <MYOB> Teknomancer well, considering the only OS firm that is here is Sun, not 'anyone' as they don't employ many ;)[19:07:10] <kokito> LOL[19:07:22] <kokito> brb[19:07:25] *** kokito has quit IRC[19:07:25] <Teknomancer> MYOB yeah but what about stuff like embedded programming etc[19:07:40] <Teknomancer> not necessarily making a desktop/server os[19:07:57] <MYOB> Teknomancer once you have the experience and the language(s) (both code and spoken) nothing else matters[19:08:48] <Teknomancer> MYOB yes here too but they do expect a BASIC degree not necessarily a PHd. unless u're a total prodigy[19:08:50] <MYOB> that said, I didn't know VBA and I still got my current job...[19:08:57] *** kokito has joined #haiku[19:10:46] <dr_evil> w00t thats innovative: http://worsethanfailure.com/Articles/Some_Call_it_Concatenation.aspx[19:12:36] *** vision has joined #haiku[19:13:00] <StylusEater_Work> dr_evil: tehe[19:13:38] <MYOB> how the fuck could someone actually write that?[19:14:03] <dr_evil> I have no idea[19:14:28] <dr_evil> however, I recently used this at work:[19:14:29] <dr_evil> volatile int i=0; i /= i;[19:14:48] <geist> boom[19:14:55] <StylusEater_Work> div by zero...[19:15:45] * StylusEater_Work assumes[19:15:46] <dr_evil> yes, but using a hook into windows exception handling, and logging all exceptions to a file, it proved beneficial for debugging instead of a simple abort()[19:15:59] <StylusEater_Work> ah[19:16:06] <StylusEater_Work> make it die then catch it[19:16:08] <StylusEater_Work> that works[19:16:30] <StylusEater_Work> I'm still puzzled why someone would write what you pointed out earlier[19:16:33] <StylusEater_Work> simply awful[19:16:42] <MYOB> I need to submit our entire current product like to worsethanfailure as an example of "how NOT to write a 3,000 euro application"[19:17:06] <dr_evil> hmm, 3 kEUR, thats it?[19:17:29] * StylusEater_Work is in similar boat[19:17:42] * StylusEater_Work thinks OO perl is fun... GAHHH!!!! help me![19:17:51] <MYOB> dr_evil 3k/euro + 850 a CAL + about the same again in annual maintaince fees, etc[19:17:59] <kokito> mmu_man, let me know when you register[19:18:04] <MYOB> and all for an application that runs in an access mde![19:18:12] <mmu_man> k[19:18:28] <mmu_man> need to find a shirt that has US size mark[19:18:51] <mmu_man> or ask google about their definition :))[19:18:52] <dr_evil> US? that would be L or M for you[19:19:00] <mmu_man> probably yes[19:19:01] <dr_evil> more M[19:19:38] <mmu_man> as I don't want a pyjama :)[19:20:04] <MYOB> I'm worried you know what size clothes mmu_man takes![19:20:21] <mmu_man> we met several times at BeGeistert[19:20:33] <kokito> mmu_man, M size should work for you[19:20:34] <mmu_man> but always in public, no worries :P[19:20:50] <mmu_man> kokito ok[19:21:46] <dr_evil> MYOB dirty mind[19:22:27] <MYOB> dr_evil you're the one deciding that I was implying something diry...[19:22:30] *** mats has joined #haiku[19:24:34] <mmu_man> at least I never went into BGA's basement :))[19:24:47][19:24:57] <MYOB> I -could- have been implying mmu_man made you do his washing or something, not that you'd have seen the inside of his clothes for other reasons...[19:25:34] <dr_evil> yeah sure, as you always do[19:25:45] <mmu_man> it could possibly be eventually ... yeah ok :p[19:26:34] <MYOB> its usually MikeW who makes the dirty jokes in here, not me[19:27:17] <mmu_man> shrug, 300 mails[19:27:30] <Teknomancer> k going to try and sleep[19:27:31] <Teknomancer> nite all[19:27:56] *** Teknomancer has quit IRC[19:29:36] <dr_evil> hmm, I'm awake for 6 hours only[19:36:35] <stargater> 300 mails = 250 spam mails ?[19:38:38] <mmu_man> =250 from ffmpeg-devel :-([19:38:45] *** JamesB192 has joined #haiku[19:39:27] <stargater> hehe // i ask me is ffmpeg beos like and native :-)[19:46:37] *** vision has quit IRC[19:49:50] *** TLF has joined #haiku[19:53:52] *** Mazon is now known as mazon[19:55:28] *** mazon is now known as Mazon[19:57:14] <MYOB> mmu_man imagine having it + vlc-devel... = 95% of my mail :S[19:57:34] <dr_evil> uhhh that hurts, too: http://worsethanfailure.com/Comments/Some_Call_it_Concatenation.aspx?pg=2#130999[20:00:18] <dr_evil> hmm, tv homepage is overloaded[20:00:25] <mmu_man> pfff, so much ?[20:00:35] <mmu_man> while it's so simple in brainfuck :)[20:03:14] <mmu_man> hmm about tv... will my DSL TV work this week ?[20:03:15] <dr_evil> perhaps something like std::string operator+(const char *, const char *)[20:03:32] <mmu_man> hmm no :-([20:03:50] <mmu_man> had 3 days with all the big channels non working :-((([20:03:51] <dr_evil> EPG might work[20:03:54] <mmu_man> and it's still there[20:04:27] <dr_evil> CSI New York next on VOX[20:05:08] <stargater> 20.15 dr_evil ON VOX ?[20:05:59] <Jin> where can I find some sort of a tutorial on programming Be applications?[20:06:12] <Jin> is this what the Be Book is, or is that just a reference?[20:06:31] <dr_evil> stargater yep[20:08:16] <Jin> also, is there any way I can "alt+tab" between windows, so to speak? I know it may sound silly but I'm just so used to that, and using the mouse for everything isn't my style[20:08:33] <stargater> Jin, ->http://www.oreilly.com/catalog/beosprog/book/[20:08:40] <stargater> dr_evil, thx[20:09:53] <stargater> Jin, ->http://bebits.com/app/3019[20:10:39] <stargater> BeBook is the API Reference[20:10:51] <Jin> thanks and thanks[20:10:59] <Jin> that's what I thought[20:10:59] <stargater> np :-)[20:11:00] <Jin> wasn't 100% sure[20:11:48] *** chaotic has left #haiku[20:11:51] *** chaotic has joined #haiku[20:11:59] *** chaotic has left #haiku[20:12:34] *** Kernel86 has joined #Haiku[20:19:41] <StylusEater_Work> did that "one company" (the actual name escapes me at the moment) finally release the api reference online?[20:20:40] *** Lelldorin1 has joined #haiku[20:20:48] *** Lelldorin1 has quit IRC[20:26:30] <Jin> hmm[20:26:45] <StylusEater_Work> it was mentioned on haiku's website last week[20:26:50] <StylusEater_Work> I didn't see the docs online yet[20:27:10] <Jin> I just found this "Modem Driver" on bebits that claims to support AOpen modems, which is the one I currently have in this computer, however, both the download links are down, any advice on that? I'm reading the talkback now, in the chance someone might be mirroring it...[20:27:25] <mmu_man> try beshare[20:27:43] <mmu_man> or http://web.archive.org with the download link[20:29:20] *** Hoern has joined #Haiku[20:29:47] *** nielx has joined #haiku[20:40:42] <Jin> all I can find on beshare is an older version without proper support[20:41:46] <MYOB> its not a driver, its a set of init strings. If your modem is newer than 1995 or so you should be able to set the init string to "ATZ"[20:41:55] <MYOB> if that doesn't work, its extremely unlikely to work with BeOS at all[20:42:17] <MYOB> what actual modem does the machine have? AOpen don't make modems...[20:43:21] <MYOB> if its a modem in an aopen machine, its likely a winmodem, and with the exception of some Lucent ones, they are not supported by BeOS[20:44:09] <Jin> afaik it's a conexant hsf chipset[20:44:24] <MYOB> not supported then[20:44:45] <Jin> damn[20:45:06] <MYOB> its not a modem anyway, its a cheap soundcard with a phone port[20:45:23] *** mmu has joined #haiku[20:50:25] *** mmu_man has quit IRC[20:50:25] *** mmu is now known as mmu_man[20:50:42] *** shackan has joined #haiku[20:53:48] *** shackan has quit IRC[20:56:08] *** nielx_ has joined #haiku[20:56:55] *** tombhadAC has quit IRC[20:57:13] *** nielx__ has joined #haiku[20:59:20] *** tombhadAC has joined #haiku[20:59:48] *** xcasex has joined #haiku[21:03:14] *** MikeW has joined #haiku[21:05:20] *** nielx_ has quit IRC[21:05:35] <MYOB> if you're stuck in dialup land, treat yourself to an external serial Hayes/Zoom modem[21:05:40] <MYOB> it'll be (slightly) faster too[21:06:20] <MikeW> I want the vulcan appkiller for windows[21:06:34] <MikeW> msn messenger went titsup a few minutes ago and taskmanager isn't getting rid of it[21:06:39] <MYOB> MikeW I used to have an app called XKill for windows IIRC[21:07:46] <MYOB> did something similar[21:07:51] *** xcasex_ has quit IRC[21:07:53] *** Begasus_bbl is now known as Begasus[21:08:22] <Begasus> re[21:10:21] <MikeW> actually I can't kill any task[21:10:28] <MikeW> firefox has also died[21:10:29] <MikeW> reboot time[21:10:31] <mmu_man> MYOB Yes I had that one in 9x as well[21:10:47] <MikeW> my antivirus checker service hasn't started either :/[21:11:56] *** MikeW has quit IRC[21:12:33] *** Begasus is now known as Begasus_bbl[21:12:40] *** proctop has quit IRC[21:13:48] *** nielx has quit IRC[21:19:22] *** PORDO has joined #haiku[21:20:13] *** MikeW has joined #haiku[21:23:08] <mmu_man> Unknown suffix on <src!add-ons!accelerants!intel_extreme>accelerants - see UserObject rule in Jamfile(5).[21:23:33] <mmu_man> hmm guess I must rebuid jam[21:28:01] *** sarcas has joined #haiku[21:30:32] <Jin> la lalala lala la lalala[21:39:06] * JamesB192 always had his hardware modems bomb out at minute 481, then he discovered it was a policy on the other end.[21:42:14] *** DaaT has joined #haiku[21:46:28] *** mmu_man has quit IRC[21:54:03] *** shackan has joined #haiku[21:56:57] *** TLF has quit IRC[22:06:58] *** jpd has joined #haiku[22:08:31] *** wildur has joined #Haiku[22:25:44] *** karijes has joined #haiku[22:26:12] *** petterhj- has joined #haiku[22:26:13] *** petterhj has quit IRC[22:29:39] *** nielx__ has quit IRC[22:36:20] *** guma has quit IRC[22:37:31] *** jpd has quit IRC[22:39:47] *** pikapika has joined #haiku[22:41:28] <pikapika> hello[22:47:00] *** [Beta] has quit IRC[22:51:30] *** Euver has quit IRC[22:54:46] *** StylusEater_Work has left #haiku[22:54:46] *** [Beta] has joined #haiku[23:04:04] *** MikeW has quit IRC[23:04:09] *** oco has joined #haiku[23:16:42] *** Witukind has joined #haiku[23:28:01] *** emitrax has joined #haiku[23:30:38] *** mphipps has joined #haiku[23:33:40] <mphipps> Hey All![23:34:43] *** proctop has joined #haiku[23:34:44] <xcasex> hey mphipps[23:35:01] <xcasex> i heard you have a license to redistribute beos r4.5 xD[23:35:28] <proctop> good old r4.5.. *sigh*[23:35:40] <pikapika> the last ppc one no ?[23:35:46] <mphipps> Ummm - no. :-)[23:36:07] *** emitrax has left #haiku[23:36:16] <xcasex> pikapika: yeah and i was just making a cheap joke :p[23:36:17] <proctop> R5 was ppc too, IIRC... just not the personal edition.[23:36:19] <mphipps> We were lucky to get the rights for the BeBook and the newsletters.[23:36:35] <xcasex> proctop: was it generic ppc tho?[23:36:44] <proctop> dunno...[23:36:44] <xcasex> mphipps: indeed :D[23:37:03] <Ingenu> now that everyone runs x86/x64...[23:37:42] <xcasex> Ingenu: i have a few ppc machines :p[23:37:51] <Ingenu> well there are no new ones[23:38:21] <mphipps> While I liked the PPC architecture, I can't see why anyone would want to continue to work on it.[23:38:22] <Ingenu> anyway, bed time, night[23:38:43] *** Ingenu has quit IRC[23:38:57] <xcasex> just fun..[23:39:02] <xcasex> harmless.. fun.[23:39:10] <mphipps> I guess.[23:39:23] <mphipps> MY idea of fun is to have millions of adoring fans. :-)[23:39:51] <NeonLicht> mine is to have millions of euros XD[23:40:12] <mphipps> NeonLicht - OSS is NOT for you. :-D[23:40:14] <NeonLicht> with them you'll have the adoring fans XDDD[23:40:28] <mphipps> I have aquired a lot more fans than Euros or Dollars. :O[23:40:36] <MYOB> redistribute R4.5? what?[23:41:17] <xcasex> MYOB: zeta pun[23:41:23] <xcasex> mphipps: i agree :D[23:41:28] <MYOB> ah[23:41:37] <NeonLicht> of course OSS is for me, mphipps... I use OSS and save my money... so that at some point I'll have all those millions of euros XDDDDD[23:41:43] <MYOB> I'm tired, its too late to be making jokes :P[23:42:08] <NeonLicht> MYOB: just laugh 'bout jokes the other make XD[23:50:06] *** proctop has quit IRC[23:58:27] <CIA-17> stippi * r20625 /haiku/trunk/src/apps/icon-o-matic/shape/commands/ (FlipPointsCommand.cpp FlipPointsCommand.h): fix the build... sorry guys!