April 6, 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 6, 2007 [00:00:13] <_hugo> we should add NFSv4 support with beos attributes :>[00:00:39] <_hugo> lots of stuff to do first though, eheh[00:01:01] <DeadYak> whay are some characters bold looking in that screenshot?[00:01:05] <DeadYak> in the df output that is[00:01:19] <DeadYak> font rendering?[00:02:31] *** emitrax__ has quit IRC[00:02:43] <_hugo> i have no idea :-)[00:03:02] *** XXXGauronXXX has joined #haiku[00:03:10] <[Beta]> I gonna go for: rounding errors :^)[00:04:20] <mmu_man> _hugo yes it's on my TODO list :)[00:04:43] <DeadYak> it seems kind of random though, twice I see the 1 bolded, once the i in the middle of "media"[00:04:44] <mmu_man> DeadYak the 1 in media1 ?[00:04:45] <mmu_man> dunno[00:04:48] <mmu_man> maybe a bug[00:04:49] <_hugo> mmu_man: nice[00:04:52] *** XXXGauronXXX has quit IRC[00:05:14] <mmu_man> _hugo NFSV4 might even allow tunnelling queries, not sure though[00:05:23] <_hugo> that would be neat[00:05:38] <_hugo> but having a haiku nfs server and being able to use the attributes would be nice enough[00:06:57] *** Fanskapet has quit IRC[00:07:41] <DeadYak> bbl :)[00:07:51] <_hugo> see you DeadYak[00:08:00] <DeadYak> yep yep[00:08:05] *** Fanskapet has joined #haiku[00:08:05] <DeadYak> time to go get destroyed by friend in racquetball :P[00:08:10] *** MakCept has joined #haiku[00:08:11] <_hugo> ah![00:08:12] <_hugo> meep[00:08:25] <MakCept> meep?[00:08:42] <_hugo> MakCept: and you sprung into existance in this ephemeral place[00:08:54] <MakCept> yes[00:09:02] <MakCept> it's been a while since I've "really" been here[00:09:11] <MakCept> but was on once a week or so ago[00:11:02] *** emitrax_ has quit IRC[00:16:48] *** fyysik has quit IRC[00:19:32] *** Emilio has joined #haiku[00:21:59] *** MakCept has quit IRC[00:22:14] *** Emilio has left #haiku[00:22:59] <mmu_man> hmm zzz[00:22:59] *** mmu_man has quit IRC[00:23:29] *** JamesB192 has joined #haiku[00:27:57] *** fyysik has joined #haiku[00:37:59] *** TLF has quit IRC[00:38:07] *** fyysik has quit IRC[00:38:11] *** erika142123 has joined #haiku[00:38:22] <erika142123> hi all[00:39:16] <erika142123> anyone here[00:41:18] *** erika142123 has left #haiku[00:41:21] *** erika142123 has joined #haiku[00:41:38] <erika142123> good evening all[00:42:24] <erika142123> anyone here[00:43:04] <CIA-17> hugosantos * r20592 /haiku/trunk/src/add-ons/kernel/network/protocols/ (tcp/TCPEndpoint.cpp tcp/TCPEndpoint.h udp/udp.cpp):[00:43:04] <CIA-17> tiny TCP cleanups, move the read notifications to _NotifyReader().[00:43:04] <CIA-17> includes the size of the header we need to check that as well. IPv4 is correct[00:43:05] <CIA-17> since we check for the size after prepending the header (due to it being[00:43:07] <CIA-17> possibly already included).[00:43:39] <erika142123> hi is anyone here[00:44:09] <dr_evil> the audience is listening[00:44:23] <erika142123> hi dr evil[00:46:33] <erika142123> quite room[00:47:08] <dr_evil> many from europe already sleeping, I'll leave now, too[00:47:11] *** [Nies] has quit IRC[00:47:31] *** erika142123 has left #haiku[00:47:59] * kokito wonders if mail_daemon and BeMail should work in Haiku now[00:49:00] <_hugo> kokito: what was the problem earlier?[00:49:49] <kokito> _hugo: have never tried it before, so I can't say whether there is a problem or not :)[00:50:07] <_hugo> oh[00:50:25] <_hugo> if you test it let me know, i would like to hear about any network related issues[00:52:12] <kokito> ok. but last I checked (a couple of months ago?), it was not possible to create email accounts.[00:52:30] <_hugo> i see[00:54:49] *** pikapika has quit IRC[00:55:06] *** [Katisu] has quit IRC[00:57:36] *** SunWuKung has joined #haiku[00:58:04] *** guildencrantz has quit IRC[01:00:54] *** SunWuKung has quit IRC[01:01:44] *** SunWuKung has joined #haiku[01:01:50] *** Jin has joined #Haiku[01:05:40] *** ekdahl has left #haiku[01:07:54] *** [Katisu] has joined #haiku[01:19:14] *** dtravis7 has joined #haiku[01:26:51] <CIA-17> hugosantos * r20593 /haiku/trunk/src/add-ons/kernel/network/protocols/tcp/TCPEndpoint.cpp: The Open Group base specification mentions that EINTR should be returned if the recv() is interrupted before _any data_ is available. So we actually check if there is data, and if so, push it to the user.[01:32:05] *** doppeljot has joined #haiku[01:32:30] *** oco has quit IRC[01:34:54] *** stargater has joined #haiku[01:34:59] <stargater> hi all[01:35:05] *** kb7sqi has quit IRC[01:36:21] *** kb7sqi has joined #haiku[01:37:44] <kokito> _hugo: ping stops working after the 4th or 5th ping[01:38:18] *** dtravis7 has left #haiku[01:38:33] <_hugo> kokito: you mean generally?[01:38:44] *** wildur has quit IRC[01:38:53] <_hugo> generically even[01:39:17] <_hugo> kokito: what network card are you using[01:39:23] <kokito> _hugo: vmware[01:39:49] <_hugo> pings do not stop here. what command line flags are you using?[01:40:02] <kokito> if I ping, say, osnews.com, it will return the first 5-6 packets and that's it.[01:40:20] <kokito> no flags[01:40:27] <kokito> just ping domain.com[01:40:30] <_hugo> kokito: is your vmware using ipro1000 or vlance?[01:40:34] *** SunWuKung has quit IRC[01:40:46] <kokito> hmmm... let me see[01:40:53] <kokito> vlance[01:41:01] <_hugo> there have been problems reported with vlance[01:41:05] <_hugo> being unreliable[01:41:09] <kokito> ic[01:41:11] <_hugo> you should be ok with ipro1000[01:41:22] <_hugo> until we actually fix the issues with vlance[01:41:48] <kokito> ok. will see if I can find out how to change the vmdk config file[01:42:11] <_hugo> ethernet0.virtualDev = "e1000"[01:42:11] *** SunWuKung has joined #haiku[01:42:20] <kokito> that's it?[01:42:23] <_hugo> yep[01:42:26] <_hugo> in the .vmx[01:43:58] *** tombhadAC has quit IRC[01:45:27] <kokito> _hugo, strange: the vmx file has "e1000"[01:45:52] <_hugo> kokito: what vmware are you using?[01:46:11] <kokito> free player[01:46:23] <_hugo> 1.0?[01:46:50] <_hugo> if so i'm using the same here[01:46:53] <kokito> think so[01:47:04] <_hugo> and my ethernet device is a ipro1000[01:47:20] <_hugo> i can even add multiple ipro1000s without issues[01:49:11] <kokito> mystery... :)[01:49:30] <_hugo> indeed[01:51:34] <kokito> downloading latest version[01:56:29] * Jin downloads kokito[01:56:32] *** MikeW___ has joined #haiku[01:56:48] <MikeW___> Anyone know if ar18 from the mailing list comes here?[01:57:01] *** SunWuKung has quit IRC[02:00:57] * MikeW___ proceeds to kick him in the teeth[02:01:01] <MikeW___> fucking piece of shite AT&T Message Center Version 1 (Oct 4 2006)[02:03:18] <Jin> o_O[02:03:37] * JamesB192 narrates 'From AT&T the company that brought down the network for want of a semicolon.'[02:03:55] <MikeW___> Jin: I'm guessing you don't read the mailing list through gmane web interface, gmail, or with a NNTP reader that reads from gmane[02:04:24] *** Yez has joined #Haiku[02:05:20] <Jin> no, I do not[02:06:07] *** josip_ has quit IRC[02:07:08] <MikeW___> Yez: oh, you are a comcast user[02:07:18] <MikeW___> Yez: Are you ar18 on the mailing list?[02:07:34] <Jin> what[02:07:35] <Jin> no[02:07:38] <kokito> _hugo: installed latest version of vmware player and download fresh image and vmx file, and now everything is sweet. :)[02:07:46] <_hugo> kokito: good to hear[02:08:44] <stargater> hmm i have the new network connections deskbar icon a withe barground in the icon ? anyone too ( i hove no device found )[02:09:55] <_hugo> stargater: do you have any network cards?[02:10:59] <kokito> LOL[02:12:36] *** aldeck has joined #haiku[02:12:41] <kokito> _hugo: wget works nice now :)[02:13:20] <_hugo> kokito: yep, fixed a bug in tcp a couple days ago that made it work nicely[02:13:36] <_hugo> i've been doing some stress tests and its holding up[02:16:30] <kokito> _hugo: downloading FF now at about 150K/s[02:16:35] <kokito> not bad :)[02:16:46] <_hugo> not sure FF will work. what version are you downloading?[02:17:17] <_hugo> also the TCP stack is not very performant yet, stability is a priority right now[02:17:28] <kokito> _hugo: I know. just playing around. :)[02:17:42] <_hugo> good. if you bump into any bugs let me know[02:17:49] <kokito> okidoki[02:21:44] <MikeW___> kokito: kevin?[02:21:59] <kokito> nope[02:24:18] *** aldeck has quit IRC[02:25:37] *** kb7sqi has quit IRC[02:26:01] <stargater> _hugo: wifi ipro 2100[02:27:03] <_hugo> stargater: ipw2100 perhaps?[02:27:52] <_hugo> you have to configure it manually to work. check ipw2100.settings[02:31:43] *** shackan has quit IRC[02:33:29] *** jevin has quit IRC[02:33:39] <stargater> ok i check it[02:33:42] <stargater> reboot[02:33:46] *** stargater has quit IRC[02:34:21] *** Mazon is now known as mazon[02:37:13] *** MikeW___ has quit IRC[02:37:54] *** doppeljot has quit IRC[02:43:54] *** kb7sqi has joined #haiku[02:44:01] *** Stargater has joined #haiku[02:44:16] <Stargater> _hugo i cant found the setting file[02:44:57] <Stargater> _hugo pc info in zeta says =[02:45:00] <Stargater> Netzwerkkarte: Intel - 82801DB - PRO/100 VE Network Connection[02:45:01][02:45:02] <Stargater>[02:45:02] <Stargater> Netzwerkkarte: Intel - - Intel(R) PRO/Wireless 2100 LAN Card Driver[02:45:03][02:46:00] <_hugo> Stargater: ipw2100 is not included in the current haiku build, so you have to compile it yourself and change what is necessary to include it. the configuration file is next to the source[02:46:43] <Stargater> hmmm why not is this on the haiku tree ?[02:46:53] <_hugo> it is in the haiku tree, just not in the build[02:47:09] <Stargater> ok why is this not in the build ?[02:47:16] <_hugo> i don't know[02:47:39] <Stargater> _hugo can you add this in the build ?[02:48:03] <_hugo> i won't change the current build in the svn, i'm sure there are good reasons for it not being added by default[02:48:17] <_hugo> feel free to test it locally[02:52:42] <Stargater> ok[02:55:10] <Stargater> _hugo o k i have compile and who put it in my haiku parttions ?[02:55:38] <Stargater> /HAIKU/beos/system/add-ons/kernel/drivers/bin ?[02:55:59] <Stargater> reboot[02:56:04] *** Stargater has quit IRC[03:02:05] *** slaad has joined #haiku[03:06:06] *** Stargater has joined #haiku[03:06:51] <Stargater> _hugo in zeta with qemu works ipw2100 and on real hardware i cam in boot proces in KDL[03:07:08] <Stargater> when the driver will load[03:07:35] <_hugo> im not sure i understand what you are saying[03:07:42] <_hugo> i dont know which driver zeta includes either[03:08:27] <Stargater> ok , in mast sleep , n8[03:10:09] *** SunWuKung has joined #haiku[03:12:02] <Jin> I don't understand that n8[03:12:12] <Jin> that's like, neight, and uh, how does that sound similar to night?[03:12:19] <Jin> is it one of those accent things?[03:12:24] *** steffen_f_ has joined #haiku[03:12:51] <silverpower> well, Americans didn't create txtspk as far as I'm aware.[03:13:02] <silverpower> they just imported it, probably.[03:13:30] <Jin> what?[03:14:18] <silverpower> or, my thought processes could be messed up from the caffeine withdrawal.[03:14:25] <silverpower> >_<[03:15:00] <silverpower> n8 comes from txtspk, though, IIRC.[03:15:04] *** aldeck has joined #haiku[03:17:01] * silverpower gives in and wanders off to make tea.[03:21:35] <slaad> It's German for night.[03:25:21] *** M199 has joined #haiku[03:25:35] *** umccullough_work has joined #haiku[03:26:34] *** steffen_f has quit IRC[03:29:33] *** Jin has quit IRC[03:31:03] *** SunWuKung has quit IRC[03:32:21] *** Kernel86 has joined #Haiku[03:34:25] *** SunWuKung has joined #haiku[03:41:36] *** M_199 has quit IRC[03:43:07] *** dr_evil has quit IRC[03:44:18] *** SunWuKung has quit IRC[03:44:40] *** SunWuKung has joined #haiku[03:58:28] <silverpower> umccullough_work, I got the toolchain to compile on PowerPC.[04:00:06] *** SunWuKung has quit IRC[04:02:35] <CIA-17> hugosantos * r20594 /haiku/trunk/src/add-ons/kernel/network/protocols/tcp/TCPEndpoint.cpp: check if receive timeout is infinite before doing the usual preparation for ABSOLUTE_TIMEOUT[04:02:37] <CIA-17> hugosantos * r20595 /haiku/trunk/src/bin/strace/ (NetworkTypes.cpp TypeHandler.h): added message_args decoding support to strace (used by net_stack's RECEIVE/SEND).[04:14:42] *** absabs has joined #haiku[04:33:32] <umccullough_work> silverpower, coolies![04:34:27] <silverpower> now, if my girlfriend would quit distracting me over IMs, I can start the build.[04:37:14] <CIA-17> bonefish * r20596 /haiku/trunk/ (Jamrules build/jam/MathRules):[04:37:14] <CIA-17> Added rules for performing basic integer arithmetics (+, -, *). The rules[04:37:14] <CIA-17> operate manually on digit lists, so they are certainly not fast and shouldn't[04:37:14] <CIA-17> be used excessively, but at least it's possible to do calculations in Jam now,[04:37:14] <CIA-17> should the need arise.[04:39:14] <silverpower> Where would I point configure to as far as the compile tools?[04:39:28] <umccullough_work> silverpower, explain?[04:39:38] <CIA-17> bonefish * r20597 /haiku/trunk/build/jam/MiscRules:[04:39:38] <CIA-17> Added rules NextID, which returns an numerical ID incremented with each[04:39:38] <CIA-17> invocation, and NewUniqueTarget, which returns a unique target name.[04:39:40] <umccullough_work> when you run ./configure --build-cross-tools it automatically sets you up[04:39:55] <umccullough_work> you should find in the generated/build folder a BuildSetup that is already set! :D[04:40:01] <silverpower> I know, but it's saying I don't have a valid cross-compiler.[04:40:02] <umccullough_work> or BuildConfig[04:40:11] <umccullough_work> silverpower, you don't run ./configure again[04:40:20] <umccullough_work> once you've done it, you're ready to jam[04:40:47] <silverpower> oh, dear. will that hose things? I'd hate to repeat the cross-compiler build...[04:41:04] <umccullough_work> no, ingo added that check to make sure it would just fail like you saw[04:41:07] <umccullough_work> it doesn't change the config[04:41:18] <silverpower> whew![04:41:19] <umccullough_work> before that it would reconfigure to use your host compiler blindly[04:41:36] <umccullough_work> the you'd have to use ./configure --set-cross-tools or something to fix it[04:41:37] <silverpower> Yeah, that's probably what hosed my attempt to build it last time.[04:41:42] <umccullough_work> (i forget the exact fix)[04:42:02] <umccullough_work> you should be good to go though[04:42:13] <umccullough_work> make sure you update jam to the version in the buildtools/jam[04:42:19] <silverpower> Is there a specific jam target?[04:42:34] <umccullough_work> you should probably start with a simple: jam -q haiku-image[04:42:55] <umccullough_work> otherwise it will go through all the motions to create an image[04:43:23] <silverpower> and that's what we need to start debugging the process, right?[04:43:31] <umccullough_work> basically :)[04:43:42] <umccullough_work> if something fails to build, that's where you gotta start :D[04:44:04] <umccullough_work> building the image requires a whole bunch of host tools to be compiled (like bfs_shell, etc.)[04:44:15] <silverpower> I'm not sure I'll be able to test the resulting image, though, if I get to that stage with no problem.[04:44:18] <umccullough_work> those will actually use the host compiler and headers, not the cross-compiler[04:44:38] <silverpower> and while my host compiler is very similar, it's not close enough for the task?[04:44:52] <umccullough_work> is it gcc?[04:45:35] <umccullough_work> the host compiler is used to build all the tools that are required to build the haiku targets[04:45:50] <umccullough_work> such as rc, xres, setattr, makebootable, etc.[04:45:51] <silverpower> yeah, gcc-4.1.2, I think, with Gentoo's packages. I'm under the impression that Haiku requires patches to the GCC it uses, though.[04:46:05] <silverpower> s/packages/patches[04:46:10] <umccullough_work> for those tools, you don't have to worry about it[04:46:26] <umccullough_work> they're built to run in the host OS anyway[04:46:30] <umccullough_work> not haiku[04:47:05] <umccullough_work> in other words ./configure --build-cross-tools only builds GCC, et al. - it doesn't build the tools that are used to build the image[04:47:20] <umccullough_work> the first time you try to jam the image is when it does those other tools[04:47:52] <silverpower> OK. I've still got the jam install left over from last time - I should rebuild it, right? Shouldn't take too long even on *this* thing.[04:48:11] <umccullough_work> jam install?[04:48:20] <umccullough_work> you mean the generated/objects directory?[04:48:36] <silverpower> No, I mean the jam tool.[04:48:44] <umccullough_work> that takes like 5 seconds to build :P[04:48:52] <umccullough_work> probably should update it[04:49:10] <umccullough_work> just go to buildtools/jam and type jam install[04:49:23] <umccullough_work> (which will replace your gentoo jam)[04:49:30] <umccullough_work> you might have to sudo[04:49:47] <silverpower> I thought it installed to /usr/local/bin?[04:49:54] <umccullough_work> don't remember where[04:50:01] <umccullough_work> you can just jam it and then copy it wherever[04:50:08] <umccullough_work> it'll create a jam0 binary[04:50:20] <silverpower> it did last time, I think.[04:50:38] <umccullough_work> you don't even have to rename it - you can use jam0 instead of replacing your gentoo jam ;)[04:50:45] <umccullough_work> jam0 -q haiku-image :D[04:51:01] <umccullough_work> or as someone in the mailing list suggested: hjam -q haiku-image[04:51:25] <_hugo> <- :-)[04:51:36] <_hugo> im using jam.haiku now though[04:52:11] <umccullough_work> that works too ;)[04:52:25] <silverpower> I'll probably rm the /usr/local/bin jam and install it as hjam.[04:52:44] <umccullough_work> i build on R5, so haiku jam is as good as anything I had already :D[04:52:49] <silverpower> since that makes the most sense, and I never know when a package might want Gentoo jam. :)[04:52:50] <_hugo> :-)[04:56:05] <aldeck> screenie :) http://haikubeat.free.fr/files/python/python_haiku2.png[04:56:08] <aldeck> hi there[04:56:18] <_hugo> nice aldeck[04:57:42] <aldeck> apparently anything using BView still doesn't work, but some test works like BAlert :)[04:58:38] <aldeck> wanna try? http://haikubeat.free.fr/files/python/haiku_python_vmware.zip :)[04:59:12] <umccullough_work> i might add that to my image building script[04:59:40] <umccullough_work> ah, so that's BeThon?[04:59:56] <aldeck> yep, almost worky[05:00:11] <aldeck> straight from bebits[05:01:27] <umccullough_work> i suspect if you start digging into the BView problem you might be able to get a few more bugs fixed ;)[05:03:20] <aldeck> yep, strange symbols missing and all ;)[05:05:08] <aldeck> "SetDiskMode__5BViewPcl" :)[05:05:45] <silverpower> OK, the build's starting.[05:06:04] <silverpower> anything vaguely resembling an ETA for a *full* build?[05:07:34] <silverpower> (this being a G3/500 - not exactly the fastest thing ever, especially being that it's a nearly 7yr-old machine...)[05:09:56] <aldeck> umccullough_work: python regression tests took the kernel down after a few minutes of success. too much processes i think[05:12:16] *** SunWuKung has joined #haiku[05:14:14] <umccullough_work> hmm...[05:14:37] <umccullough_work> silverpower, on my R5 box running a PIII 600 it takes a couple hours[05:14:45] <umccullough_work> but R5 is slow as shit[05:14:51] <aldeck> wow[05:14:59] <CIA-17> bonefish * r20598 /haiku/trunk/build/jam/ (BuildSetup MiscRules): (log message trimmed)[05:14:59] <CIA-17> New handy build system feature: It is now possible to pass a command[05:14:59] <CIA-17> line to jam that contains build targets and that will be executed by[05:14:59] <CIA-17> the build system after building the targets and replacing their[05:15:00] <CIA-17> occurrences in the command line by their paths. The keyword indicating[05:15:02] <CIA-17> such a command line is "run", targets are marked by a leading ":".[05:15:04] <CIA-17> E.g.:[05:15:12] <umccullough_work> aldeck, fortunately that only happens the FIRST time you build ;)[05:15:19] <aldeck> it's like 15min on a64 3200 / linux[05:15:22] <umccullough_work> after that it only recompiles whatever is needed based on changes[05:15:31] <aldeck> yep[05:15:45] <umccullough_work> linux is a hell of a lot faster at compiling Haiku - and a 3200+ is damn fast by comparison ;)[05:16:00] <umccullough_work> if R5 ran properly on my 3800+ i'd seriously consider using it[05:16:23] <umccullough_work> i suppose i may break down and install linux on it :P[05:16:52] <umccullough_work> silverpower, dr_evil claims it's 6-7 minutes on his Core 2 Duo[05:16:56] <umccullough_work> w/linux[05:18:54] <aldeck> hmm, haiku runs on my core 1 duo laptop, what about core 2 ?[05:26:34] <silverpower> okay, first error - where should I paste it, and how much?[05:28:27] <silverpower> KernelLd generated/objects/haiku/ppc/release/system/kernel/linkhack.so[05:28:27] <silverpower> /haiku-test/haiku/generated/cross-tools/bin/powerpc-apple-haiku-ld -shared -Bdynamic -o "generated/objects/haiku/ppc/release/system/kernel/linkhack.so" "generated/objects/haiku/ppc/release/system/kernel/linkhack.o" /haiku-test/haiku/generated/cross-tools/lib/gcc/powerpc-apple-haiku/4.1.2/libgcc.a ;[05:28:27] <silverpower> ...failed KernelLd generated/objects/haiku/ppc/release/system/kernel/linkhack.so ...[05:30:14] <aldeck> silverpower: the gcc4 is broken sometimes, that may be the cause[05:30:22] <aldeck> gcc4 build[05:30:57] <silverpower> mm. Unfortunately, it's practically a requirement for PowerPC Haiku, last I checked...[05:31:18] <aldeck> yep, usually korli does the fixes[05:32:17] <silverpower> mmkay.[05:32:20] <aldeck> http://cia.vc/stats/author/korli[05:32:31] <bashusr> ewww[05:32:33] <bashusr> link hack![05:32:37] <_hugo> the gcc4 build was not broken earlier today[05:32:37] <bashusr> hackers![05:32:51] <_hugo> i haven't tested the last few revisions, but should be ok[05:33:06] <aldeck> yep, i don't know[05:33:09] <silverpower> How would I go about getting more info on the build?[05:33:17] <_hugo> silverpower: what kind of info?[05:33:31] <umccullough_work> silverpower, that's not much info :([05:34:09] <silverpower> right, is there a verbosity flag I can pass to retry the build and get something *useful*?[05:34:18] <umccullough_work> try jam --help[05:34:28] <_hugo> silverpower: it should report the errors as is if you arent running with any special flags[05:34:38] <_hugo> silverpower: isnt there any errors reported before that KernelLd line?[05:34:42] <_hugo> s/isnt/arent/[05:35:11] <silverpower> no.[05:35:45] <_hugo> please try running the command that was printed below the KernelLd line[05:35:46] <_hugo> by hand[05:35:54] <umccullough_work> now that you have an environment, it might be a good time to hit the mailing list and see if ingo or anyone knows[05:36:47] <silverpower> _hugo, I get a segmentation fault.[05:36:53] <umccullough_work> ouchie[05:37:14] <_hugo> that explains it then[05:37:15] <umccullough_work> you did the --build-cross-tools-gcc4 right?[05:37:24] <_hugo> why its failing that is[05:37:25] <umccullough_work> i assume the 2.95.3 won't even build on PPC[05:37:28] <silverpower> Yes.[05:37:39] <silverpower> I built the gcc4 toolchain explicitly.[05:38:00] <umccullough_work> so, can you even run /haiku-test/.../powerpc-apple-haiku-ld by itself?[05:39:21] <silverpower> shodan haiku # /haiku-test/haiku/generated/cross-tools/bin/powerpc-apple-haiku-ld[05:39:21] <silverpower> shodan haiku #[05:39:32] <umccullough_work> well, that's good i guess...[05:39:35] <_hugo> something is borky there[05:39:44] <umccullough_work> it should spit back a message though right?[05:39:49] <umccullough_work> "no files" or whatever?[05:39:55] <_hugo> yes, about not having input files[05:39:56] <silverpower> yep, pasting error. no input files.[05:40:10] <umccullough_work> ah[05:40:13] <_hugo> ah ok. so the linker actually crashes trying to handle whatever was built[05:40:22] <umccullough_work> well, that's something anyhow...[05:40:30] *** ravtux has joined #haiku[05:40:33] <ravtux> hi[05:40:48] <silverpower> Gentoo-side gdb is currently being built. Sorry, I should've done it before I started, but it slipped my mind.[05:42:31] <umccullough_work> seems unfortunate that ld is crashing :(... might be able to find some info on the internet in reference to -shared or -Bdynamic to see if it's a known issue[05:42:40] <umccullough_work> on PPC that is ;)[05:43:10] <aldeck> nite[05:43:16] *** aldeck has quit IRC[05:43:40] <silverpower> I'm not sure, but wouldn't ld have linked some stuff before this stage?[05:43:53] <_hugo> it could be those input files in particular[05:44:42] <silverpower> Likely - the port isn't exactly maintained all that much from what I can tell.[05:45:07] <umccullough_work> considering the target name -- i guess one shouldn't be surprised ;)[05:45:37] <silverpower> True.[05:45:40] <umccullough_work> i wonder if it's a gcc 2.95.3 thing[05:45:58] <_hugo> i personally have no idea, Ingo should have some insight into that[05:46:38] <umccullough_work> yeah, he would probably be the most knowledgable - or perhaps Oliver Tappe[05:47:50] <umccullough_work> looks like ingo's awake (since he just committed 30 mins ago) - you might be able to get a response from him on the list[05:48:22] <geist> meep?[05:48:25] <silverpower> what do you mean by it perhaps being a gcc 2.95.3 thing? I've already said (and checking the binary versions of gcc/binutils agrees) that I'm on GCC4. Or do you mean it's doing something gcc4 doesn't like?[05:48:31] <umccullough_work> maybe geist knows...[05:48:39] <geist> moople?[05:48:51] <umccullough_work> silverpower, i mean the "linkhack" might be for 2.95.3 compatibility or something[05:49:09] <umccullough_work> geist, what is kernel/linkhack for?[05:49:11] <silverpower> ahh. Yeah, that makes some sense.[05:49:20] <geist> ah now we're getting somewhere[05:49:25] <geist> I'm absolutely responsible for linkhack[05:49:34] <geist> you'll see there is something similar on newosa[05:50:05] <silverpower> Well, I'm getting a segfault when ld tries to do something with it.[05:50:09] <geist> basically it was a trick to get the linker to generate the kernel as a shared library with a fixed link address[05:50:20] <geist> the only way to do it was to link the kernel with a dummy library[05:50:35] *** JamesB192 has quit IRC[05:50:40] <geist> at least that was the cast 3 or 4 years ago. it may be unnecessary, and it may not even be the correct way to do it, I just screwed around until it worked[05:51:06] *** kb7sqi has quit IRC[05:51:11] <umccullough_work> any idea why it would segfault gcc 4.1.2 ld on PPC?[05:51:12] <geist> generally speaking if you link a binary with -dynamic it exports the symbols as a shared lib, but also makes the code relocatable, which i did't want[05:51:21] <geist> umccullough_work: ld is not part of gcc, it's part of binutils[05:51:30] <geist> so the question is what variant of binutils is it?[05:51:31] <geist> 2.17?[05:51:35] <silverpower> Yes.[05:51:44] <umccullough_work> is that what it was updated to now?[05:51:45] <geist> and no, I dont know. I've managed to come across a handful of ways to segfault ld[05:51:52] <geist> in my years of using it[05:53:14] <silverpower> well, in the naive belief that it might actually work this time, I forgot to build gdb Gentoo-side, which I'm doing now.[05:53:19] <geist> axeld was screwing with ppc support at some point, not sure if he was using beos or linux to do it[05:53:34] <umccullough_work> ugh, he must have been using linux[05:53:51] <umccullough_work> i think there's pretty much no way he could run gcc4 on R5 right?[05:53:54] <geist> are you building on a ppc linux box or an x86?[05:54:04] <silverpower> PowerPC.[05:54:29] <geist> i betcha he was on a x86, FWIW[05:54:45] <umccullough_work> interesting[05:54:47] <geist> my ppc linux box is kind of on the fritz right now, somehow the toolchain got corrupted so that gcc crashes[05:54:55] <umccullough_work> he had that Genesi board[05:55:02] <geist> and since it's gentoo, if the toolchain is busted, you're kind of fucked[05:55:04] <umccullough_work> i always assumed he was using it to compile/test[05:55:26] <silverpower> Last time I spoke with axeld, he implied that the image tools were messed up on PowerPC and that he'd had to use x86 to get a usable image.[05:55:29] <geist> it's generally a pain to build and test on the same machine[05:55:46] <umccullough_work> :)[05:55:54] <geist> that may be true. I tried building on my G5 linux once and I think it failed in the image tools[05:56:02] <geist> missing an endian swap somewhere[05:56:15] <silverpower> were you running 32ul?[05:56:18] <geist> silverpower: is the tip of tree buildable up to your failure?[05:56:20] <geist> silverpower: yeah[05:56:37] <geist> if so, I can fire up the G5 and try to reproduce. oh yeah, and I forgot about my PS3[05:56:43] <geist> it has linux on it, 32ul. can try to build it there[05:57:04] <geist> hmm, now that I think about it i have at least 3 linux ppc machines around here :)[05:57:11] <umccullough_work> heh[05:57:17] <silverpower> geist, I'm not sure - umccullough_work started me out with 'hjam -q haiku-image'. Is there another target I could/should try?[05:57:23] <_hugo> geist: how does linux get booted in the ps3? what kind of firmware does it have?[05:57:35] <geist> _hugo: it'll explain in a bit, it's quite nifty[05:57:42] <_hugo> cool[05:57:49] <geist> silverpower: you have no patches against the tree, is what I'm asking[05:57:59] <geist> if I built the current svn, it should in theory hit the same failure?[05:58:15] <geist> if it's that linkhack thing, I can look at it. there's a good chance the linkhack is unnecessary[05:58:20] <silverpower> No patches, this is a clean tree as far as I'm aware. Not sure on revision.[05:58:25] <geist> though since most folks are still building on 2.95.3, it'll probably have to stay[05:58:34] <geist> svn info would tell you that[05:58:39] <geist> but I'm sure nothing changed in a while[05:58:40] <umccullough_work> can be made compiler-specific ;)[05:58:54] <geist> yeah, but it's kind of a major hack[05:58:55] <umccullough_work> i've seen some jamfiles somewhere with compiler checking logic[05:59:04] <umccullough_work> oh you mean to remove it?[05:59:06] <geist> I have to page all that back in[05:59:07] <umccullough_work> :P[05:59:08] <silverpower> Last Changed Rev: 20591[05:59:11] <geist> yeah, that's what I mean[05:59:49] <silverpower> I know a few revisions have prolly gone in since then, but I sincerely doubt it would matter *too* much.[06:00:02] <_hugo> it doesnt for the case in point[06:00:06] <umccullough_work> mostly just hugo's furious network activity ;)[06:00:13] <geist> _hugo: anyway, the PS3 is sort of nifty[06:00:18] <silverpower> that's what I figured.[06:00:28] <geist> so the main cpu is the cell, which has a dual threaded PPC + the 8 vector units[06:00:40] <geist> the PS3 always run a hypervisor[06:00:43] <geist> kind of like Xen[06:00:47] <_hugo> ah[06:00:55] <_hugo> even when running a game?[06:00:59] <geist> even when you're in game, the game is running on top of a little OS running on top of the hypervisor[06:01:04] <_hugo> i see[06:01:24] <geist> so they (Sony) ported linux to the PS3, released the patches[06:01:35] <geist> and in the system console you ca select what to boot next[06:01:49] <geist> it even lets you repartition the drive, which you basically have to do up front[06:01:54] <geist> carve out a chunk for linux[06:02:12] <geist> when booted into linux, the hard drive is virtual, so linux only sees the slice[06:02:13] <_hugo> can you put linux in the background and switch to another hypervisor domain? i.e. switch to a game[06:02:14] <_hugo> while running linux[06:02:16] *** Stargater has quit IRC[06:02:22] <geist> no, you can't[06:02:30] <geist> primarily because there wouldn't be enough memory[06:02:37] <_hugo> right[06:02:42] <_hugo> but it never gives up the whole hardware then[06:02:45] <geist> since every game is writtent o more or less assume it has 256MB[06:02:56] <geist> yeah[06:03:20] <geist> when in game (or in linux) the hypervisor keeps one of the SPUs for it's own stuff (probably crypto on the hard drive)[06:03:20] <_hugo> nifty[06:03:30] <geist> and one is disabled at the factory, so the clients get 6 of em[06:03:56] <geist> the bummer is that the linux client doesn't appear to get full access to the GPU. it only gets a linear framebuffer mapped by the hypervisor[06:04:04] <silverpower> What about the GPU inefficiencies? I saw the Snes9x youtube video, and the framerate was pretty lousy.[06:04:11] <geist> presumably the game os runs with a bit that gives it slightly higher priviledges and it gets the GPU registers[06:04:40] <_hugo> maybe the gpu registers are available but not currently used?[06:04:41] <geist> sony has been quiet about the GPU thing. prbably nvidia says 'no frackin way'[06:04:45] <_hugo> i mean, no driver[06:04:51] <geist> no, they're not avialable[06:04:59] <_hugo> i see[06:05:00] <geist> lots of folks have complained as much[06:05:20] <geist> wouldn't be documented anyway. sony has been extremely quiet about any specs on the GPU. total black hole about that[06:05:57] <_hugo> most console hardware has no public documentation i guess. even nintendo's stuff which is pretty low key have no docs[06:06:09] <geist> yeah.[06:06:18] <geist> MS has been surprisingly forthcoming about the Xenon however[06:06:19] <umccullough_work> but in this case, even the game devs get nothing?[06:06:25] <geist> but a lot of that is in ATI[06:06:32] <geist> oh the game devs get a lot :)[06:06:35] <_hugo> umccullough_work: they get a lot under NDAs[06:06:36] <umccullough_work> ah..ok[06:07:03] <geist> the library i work with for the ps3 gpu basically is a thin shim around the card's FIFO[06:07:10] <geist> you're more or less punching commands directly into it[06:07:39] <geist> gonna be a bitch for PS3 emulation on the PS4[06:07:46] <_hugo> eheh[06:08:09] <_hugo> geist: btw, only signed stuff gets used as the other higher priority stuff? e.g. to access the GPU registers etc[06:08:14] <umccullough_work> sounds like the ps4 is guaranteed to be nvidia in that case ;)[06:08:19] <_hugo> s/gets used/gets executed/[06:08:28] <geist> right[06:08:38] <geist> ie, stuff made with the PS3 SDK[06:09:08] *** rcjsuen has quit IRC[06:09:14] <geist> presuambly when the hypervisor loads it there is a bit that says what type of client it is and gives it more access taht way[06:09:15] <_hugo> i guess that makes sense. "unofficial stuff" with whom they have no agreements get no good hardware access[06:09:35] <geist> yup. it's quite amazing they did as much[06:09:44] <_hugo> well, considering ps2 linux[06:09:45] <geist> of course there are all sorts of reasons they would want to do the linux thing[06:09:55] <_hugo> and the whole "multimedia pc" thingy[06:09:58] <geist> the whole taxing thing in europe, IBM prpbably wanting them to do it[06:10:23] <geist> I tell ya what, the cell SPUs are pretty much insane[06:10:26] <_hugo> geist: does the main ppc have SMT?[06:10:32] <geist> yes. it's not terribly fast[06:10:49] <_hugo> how many hardware threads?[06:10:55] <geist> at least running linux, my G5 blows it out of the water by a factor of 3 or 4[06:11:00] <_hugo> oh[06:11:03] <_hugo> sucks then. :-)[06:11:06] <geist> 2 virtual threads, one hardware core[06:11:21] <geist> it's well documenbted, the cores are simplified.[06:11:25] <geist> totally in order, etc[06:11:45] <geist> I think the entire POWER line is moving in that direction, I thought i heard[06:12:02] <geist> they're doing the sparc thing and going to start stuffing craploads of em in a single die, run them at extremely high speeds[06:12:06] <_hugo> xbox 360's processor seems more atractive in terms of generic use, but i don't know about its performance[06:12:22] <geist> it's more or less identical, just has 3 hardware cores (6 threads)[06:12:33] <geist> the VMX bits are pretty heavily modified though[06:12:41] <geist> it has a monster vector unit[06:12:43] <_hugo> yeah, makes sense[06:12:52] <geist> though still pretty stall happy[06:13:20] <geist> like, I was doin some hacking on it the other day and I read that a dependency stall between any vmx instructions is an automatic 12 cycle penalty[06:13:23] <geist> pipeline stall[06:13:33] <_hugo> oh[06:13:52] <geist> of course there are 128 VMX registers, so you can generalyl schedule around that[06:14:08] <_hugo> does the ps3's cpu also have a vmx unit?[06:14:16] <geist> yes, but iot's the stock altivec[06:14:19] <geist> 32 registers, etc[06:14:27] <_hugo> considering the SPUs i would think they would ditch it[06:14:34] <geist> the xenon one added a few more instructions, most notably dot product[06:15:36] <geist> well, it's also the only floating point arch it has so you'd be pretty screwed without it[06:15:54] <geist> the SPUs are insane crazy fast, if you can write the code to use it[06:16:11] <geist> but what I've found is that even if you dont and you run just some lameass code with no attempt to vectorize, it's still really damn fast[06:16:13] <_hugo> right. and i would assume most games do so? even if only using a couple now[06:16:33] <geist> the instruction set has all sorts of crazy stalls, odd/even pileines, etc, but it's also running out of zero wait state ram[06:16:39] <geist> so no cache effects[06:16:44] <_hugo> right[06:16:51] <_hugo> how is the latency in the SPU's mem?[06:16:57] <geist> zero cycles[06:17:05] <_hugo> oh[06:17:17] <geist> that's the point, it's crazy fast[06:17:18] <_hugo> but i guess they dont have much memory[06:17:23] <geist> 256K a piece[06:17:41] <_hugo> ah, well. :-)[06:18:00] <geist> anyway, a lot of the games are prety heavily spuified[06:18:11] <geist> especially engines written by sony that dont need to be portable[06:18:32] <geist> saw some talks at GDC recently. for example that crazy motorsport game they just released is running almost entirely on the SPUs[06:18:43] <geist> the main cpu is really just scheduling stuff for the SPUs to do[06:18:55] <geist> that seems to be the strategy that the hard core games do[06:19:12] <_hugo> yeah, and makes sense considering the platform. i wouldnt be doing much in the main cpu. at least FP[06:19:13] <geist> thousands of jobs per frame, distributed across the SPUs[06:19:41] <_hugo> was trying to get a figure in the memory bandwidth between SPUs and main memory[06:19:55] <geist> 25GB/sec I believe[06:20:15] <geist> also there is local GPU ram, which has a second pipe[06:20:25] <_hugo> yeah, can go up to 51Gbit/sec too[06:20:30] <geist> so if you're moving between all of them equally you get a nice blend[06:20:31] <_hugo> at least according to wikipedia[06:20:39] <geist> that's probably where the 50GB/sec comes from[06:21:09] <geist> also the SPUs can dma between themselves[06:21:16] <geist> i diont know the bandwidth of that, probably crazy fast[06:21:31] <geist> the DMA engines on the SPUs seem to be extremely sophisticated[06:21:46] <geist> the sony sdk generally hides most of that behind a job based architecture, which frankly works fine[06:22:01] <_hugo> does the sdk supply a bunch of such libraries/code?[06:22:06] <geist> but you can request a raw spu and more or less write your own kernel, I think some games do that for one or two oft he spus, for long running stuff[06:22:09] <geist> yeah[06:22:14] <_hugo> but mostly unoptimized or?[06:22:30] <geist> seems to be pretty good[06:22:37] <geist> a lot of it is just thing shims over the top of hardware[06:22:48] <geist> and the non hardware stuff you dont tend to rely on for speed anyway[06:22:58] <_hugo> i see. i have this, probably wrong, idea that companies have to really optimize that stuff to push all bits to the maximum[06:23:00] <geist> reading from the disc, etc takes an eternity[06:23:23] <_hugo> ehehe[06:23:25] <_hugo> yes[06:23:26] <geist> yes and no. I think most places would like to, but the really good places spend time where it's important[06:23:49] <_hugo> i see[06:23:54] <geist> I'm amazed at how inefficient a lot of our game code is, but if you look at it with the profiler, a lot of that stuff is just a few percentage of the time[06:23:58] <geist> so it's not worth optimizing[06:24:09] <geist> a whole lot of stuff like[06:24:25] <geist> CFoo()::Get().Callsomething();[06:24:32] <geist> kind of stuff, where it grabs a static[06:24:53] <_hugo> well, i guess with the PS3 and its numbers some kind of optimizations start to be irrelevant. for the N64 i know companies did their own microcode to the SGI graphics chip to push it all to the maximum[06:24:54] <geist> and yeah, it's terrible, but it doesn't really show up on the rader because you really end up spending most of your time punching crap into the GPU[06:25:17] <_hugo> considering how complex GPU stuff is nowadays[06:25:36] <_hugo> should be pretty hard to optimize[06:25:36] <geist> yeah, I think that is the case. the modern game consoles are getting to be so fast and games so complex it's like a lot of "big" code[06:25:47] <geist> you dont have time to make it super awesome everywhere os you worry about what's important[06:25:52] <_hugo> yeah[06:26:09] <geist> which is kind of a bummer, I like making stuff super fast, knowing that there's nothing that I hadn't already looked over[06:26:16] <_hugo> eheh[06:26:21] *** SunWuKung has quit IRC[06:26:31] <umccullough_work> time to go home[06:26:37] <_hugo> see you umccullough_work[06:26:41] <silverpower> geist, do you have enough info to start looking into the issue?[06:26:44] <geist> time to get to hackin[06:26:47] <umccullough_work> later[06:26:49] *** umccullough_work has left #haiku[06:26:53] <geist> silverpower: yeah, i'll give it a whirl[06:27:23] <silverpower> geist, thanks. I need to sleep, and there's little more I can do atm, so I'll leave it in your hands. :)[06:31:31] <geist> ah, fired up the old G5 in linux[06:31:50] <geist> kind of feel sorry for this guy, I haven't used him much since leaving apple[06:31:53] <geist> used to be my main machine[06:45:13] *** kb7sqi has joined #haiku[06:50:27] *** Sloar has joined #haiku[06:55:09] *** kr1stof has joined #haiku[06:55:53] *** kr1stof has left #haiku[06:56:21] *** kr1stof has joined #haiku[06:57:01] <kr1stof> moin[07:11:33] *** bashusr has left #haiku[07:16:06] <geist> ugh, the svn server is extremely slow tonight[07:16:18] <geist> I'm *still* syncing my ppc box after 30 minutes or so[07:16:24] <geist> and it's disconnected from the server at least once[07:19:35] <_hugo> berlios is always slowish here :-/[07:36:06] *** GreyGhost has joined #haiku[07:40:41] *** umccullough has joined #haiku[07:51:12] *** absabs has quit IRC[07:51:53] *** absabs has joined #haiku[07:54:11] *** mmu_man has joined #haiku[07:54:11] *** ChanServ sets mode: +o mmu_man[07:56:47] <umccullough> hi mmu_man :)[07:56:48] *** mmu_man has quit IRC[07:56:55] *** mmu_man has joined #haiku[07:56:55] *** ChanServ sets mode: +o mmu_man[08:00:54] <geist> hmm, sure a lot of complaints about BStatable[08:01:14] <_hugo> has virtual methods, and no virtual destructor[08:01:51] <geist> right[08:01:59] <geist> guess that's a gcc 4.x warning[08:02:11] <_hugo> yeah, and a good one usually. eheh[08:02:17] <geist> well, or 3.x. I forget we're jumping two whole main revisions[08:02:32] <_hugo> i haven't check but probably adding the virtual destructor would break ABI[08:02:35] <_hugo> with R5[08:02:40] <_hugo> s/check/checked/[08:03:31] <geist> yeah. that warning is almost always a problem[08:04:17] <geist> only cases where I've seen it sort of okay is if you have private destructors and only destroy via some special mechanism wher eyou control what type of object it's being destructed as[08:05:29] <_hugo> well, in general if you don't have destructors to run on fields of derived classes you are ok[08:06:20] <_hugo> but i guess its more probable to derive the base class and add non-POD fields than not[08:06:35] <geist> right, and since this is obviously a base class for lots of stuff[08:06:51] <_hugo> yep[08:07:13] <geist> heck, it's even pure virtual[08:07:59] <_hugo> well, as long as the compiler has full visibility on the proper destructor you are ok :-) if someone starts deleting BStatables, eheh, troubles arise[08:08:22] * _hugo kicks Be Inc[08:08:41] <kokito> _hugo: are you in Brazil?[08:08:50] <_hugo> kokito: why do you say that? :-)[08:09:06] <kokito> trying to figure out your time zone[08:09:14] <_hugo> my time zone is GMT[08:09:19] <umccullough> heh, he's up WAY late ;)[08:09:27] * _hugo ducks[08:09:36] <_hugo> its day already actually[08:09:41] <umccullough> damn[08:09:44] <_hugo> kokito: im in Portugal btw[08:09:45] <kokito> that's what I was going to ask, _hugo: do yu ever sleep? :P[08:10:02] <_hugo> well, these days i sleep when im tired from doing code :-)[08:11:14] <_hugo> i've been the last hour searching and fixing a race condition so im getting tired, eheh[08:12:52] <geist> hmm, reproduced his linker crash[08:13:05] *** SunWuKung has joined #haiku[08:13:15] <_hugo> geist: nice[08:13:38] <umccullough> geist, silverpower is a female ;)[08:14:00] <geist> her[08:14:12] <geist> the crash attributed to silverpower[08:14:13] <umccullough> at least, last i recall...been a while[08:14:17] <umccullough> lol[08:14:32] <geist> well, okay. that fixed it[08:14:36] <kokito> yeah, you can never be sure these days :P[08:14:38] * _hugo yays[08:14:43] <geist> the first thing I tried doing was sticking *something* in linkhack.o[08:14:44] <geist> int foo()[08:14:48] <geist> that was enough[08:14:56] <umccullough> interesting[08:15:03] * JonathanThompson wonders who wants to subject their data to a Pretty Death Machine :)[08:15:04] <umccullough> so it otherwise works on x86?[08:15:08] * _hugo hands geist some cookies[08:16:08] <JonathanThompson> Sooner or later Haiku won't be a PDM, but until the kernel is stable and very close to bug-free, well, you're just asking for it.[08:16:21] <geist> oh wait. no, it didn't fix it[08:16:47] * _hugo revokes geist's cookies[08:16:58] * umccullough considers listening to Pretty Hate Machine at work tomorrow[08:17:53] <umccullough> mm... so i've got this 500gb HD on my AMD box here...just aching for linux[08:18:07] <umccullough> might as well get off of XP soon :P[08:20:32] <mmu_man> hey! who gave you the cookie jar ??[08:20:36] <mmu_man> :p[08:20:46] * mmu_man takes it back[08:20:49] <mmu_man> breakfast![08:21:44] * _hugo derefs mmu_man's cookies and crashes him[08:21:51] *** Teknomancer has joined #haiku[08:22:19] <umccullough> ew, that sounds like a stomach ache ;)[08:22:30] <geist> gee, i wish the toolchain built with debug symbols...[08:23:09] *** GreyGhostNew has joined #haiku[08:23:36] <CIA-17> hugosantos * r20599 /haiku/trunk/src/add-ons/kernel/network/devices/ethernet/ethernet.cpp: check if the device going down is the head of the check list and remove it.[08:23:47] <CIA-17> hugosantos * r20600 /haiku/trunk/src/add-ons/kernel/network/stack/ (7 files):[08:23:47] <CIA-17> started some work to properly handle device_removed() and setting interfaces down in general.[08:23:47] <CIA-17> - remove all routes that use interfaces going down.[08:23:47] <CIA-17> - when a device is going down, remove associated domain interfaces.[08:23:48] <CIA-17> - interfaces weren't getting a properly referenced device interface, fixed.[08:23:49] <CIA-17> - down call down_device_interface when deleting a interface, instead set it down and let the upcounts do its job.[08:24:09] *** GreyGhost has quit IRC[08:24:16] *** GreyGhostNew is now known as GreyGhost[08:24:39] * mmu_man reminds _hugo that he does kung fu[08:24:43] <_hugo> eh, s/down/dont/[08:25:03] * _hugo does kung foo[08:25:06] <umccullough> _hugo, i don't think CIA-17 cares ;)[08:25:07] <mmu_man> I'm used to escaping the Matrix crashes :p[08:25:33] * _hugo throws some foo bars at mmu_man[08:25:46] <_hugo> umccullough: unfortunely, i keep leaving typos[08:25:59] * mmu_man grabs them and throws them back[08:26:22] * _hugo crashes and goes to KDL[08:27:33] * mmu_man reboots _hugo[08:29:07] <_hugo> :-) i'll grab some food, brb[08:29:17] <kokito> bonjour mmu_man ;)[08:30:44] <mmu_man> plop kokito[08:31:12] <JonathanThompson> As long as you aren't leaving something slippery from the posterior portal, _hugo :)[08:32:57] <kokito> JonathanThompson: got my 6.6MB email? :P[08:33:02] <JonathanThompson> Yes.[08:33:21] <JonathanThompson> I can't seem to find any documentation on where to put the WonderBrush license file, at least not in english :P[08:33:24] *** SunWuKung has quit IRC[08:33:50] <JonathanThompson> Perhaps in config\settings\WonderBrush...[08:34:08] <Teknomancer> config/settings/WonderBrush :P[08:35:04] <JonathanThompson> Teknomancer, I suffer from a case of too many OS's that conflict in how they do things :)[08:35:14] <Teknomancer> haha[08:35:16] <_hugo> JonathanThompson: there is only one separator -> /[08:35:33] <JonathanThompson> Rgiht: the other is an imposter :P[08:35:50] <kokito> JonathanThompson: let me see[08:36:02] <JonathanThompson> See what?[08:36:03] <_hugo> look it in the interweb. oh wait, google.com/ <-[08:36:31] <_hugo> \ is an escape char, / is a separator[08:36:39] <JonathanThompson> I suspect the dokumentation is good, but... I don't read dutch.[08:36:53] <JonathanThompson> (I think that's dutch)[08:37:02] <kokito> JonathanThompson: I am going by memory, but I think it has to be placed in the WB directory under /config/settings[08:37:34] <geist> okay, building binutils with -O0 and -g[08:37:35] <JonathanThompson> That's my best most logical guess, if not the executable directory. If something follows Be guidelines, it would most likely go there (settings)[08:37:37] <geist> see if I can trace the failure[08:38:15] * JonathanThompson thinks perhaps in the next 6 months if he determines his employment is stable he'll be able to buy a new machine (truly new!) for the first time since 1999.[08:38:36] <JonathanThompson> At which point, I'm likely to go all out and get 8 cores :)[08:38:56] *** SunWuKung has joined #haiku[08:39:00] <umccullough> night night[08:39:06] <JonathanThompson> And hopefully Haiku will be ready by then.[08:39:07] *** umccullough has left #haiku[08:39:19] <geist> actually dont know if the kernel will rn on 8 cores[08:39:27] <mmu_man> like http://www.josephpalmer.com/cgi-local/View_Permalink.cgi?entry=2007/4/01/08:54:28:00 ?[08:39:44] <JonathanThompson> If not, there'd be a little work to do, geist :)[08:39:52] <JonathanThompson> I know it runs on dual p3's.[08:40:10] <geist> yay, libiberty for the win![08:40:56] * JonathanThompson wonders if geist has had an overload on Tic-tac-dough[08:41:07] <JonathanThompson> Or Hollywood Squares[08:41:16] <geist> waitin on gcc to build again[08:41:34] <geist> this ol mac runs linux like a champ. should use it more often[08:41:37] <geist> but gee, I got enough computers[08:42:00] <JonathanThompson> Someone on BeShare earlier were stating they have a CD produced in 1980.[08:42:50] * JonathanThompson must be getting tired[08:43:08] <geist> according to teh intarweb, CDs were developed around 1979[08:43:23] <JonathanThompson> Prototypes.[08:43:26] <geist> 1980 was when the standard was proposed[08:44:36] <mmu_man> http://www.osnews.com/comment.php?news_id=17641[08:44:39] <mmu_man> we can do it[08:44:41] <mmu_man> :)[08:44:46] <mmu_man> ok, other stuff first[08:45:27] <_hugo> eheh[08:47:36] <tic> moo[08:47:38] <Teknomancer> ZFS!!![08:48:29] <JonathanThompson> That way we can all easily keep on adding on new hard drive capacity in a very organic way, Teknomancer :)[08:48:53] <Teknomancer> haha well zfs's main feature i like is the 'supposed' data safety across reboots[08:49:21] <Teknomancer> not sure how well it actually works but random reboots not corrupting the fs would be nice[08:49:22] <JonathanThompson> Lightning storms do wonders for the wonders of man to try to combat :)[08:49:40] <geist> well, found the bug, but it's not real obvious why it got that way[08:49:54] <geist> some internal data strcture in LD is half initialized, and it derefs a null pointer[08:50:12] <JonathanThompson> Even if the filesystem structure isn't corrupted, if something was in the middle of writing something, chances are you still have *something* to worry about being in a weird state, regardless of journaling to any degree.[08:50:30] <JonathanThompson> OOHHHHH CRUNCH :)[08:50:54] <JonathanThompson> Probabbly a nice subtle logic bug to explain all that, geist.[08:51:09] <Teknomancer> LD <-- ?[08:51:16] <geist> it seems to think it has a GOT table, and it tries to patch it, but the data pointer in the structure is null[08:51:20] <JonathanThompson> <-- labeled LD[08:51:25] <geist> so it has a GOT table, but it is empty[08:52:47] <JonathanThompson> And that's with no optimization turned on?[08:53:05] <geist> yeah, i rebuilt with -O0 -g so i could trace it[08:53:13] <geist> sinc eit was in the middle of a monster function previously[08:55:23] <JonathanThompson> Sounds probable it was an incorrectly refactored function that's the cause.[08:55:32] <GreyGhost> w00t best ever game of GLtron i've played :D[08:55:54] <geist> it's probably a questionble input that's effectively asserting the linker[08:56:01] <geist> would be nice if it dealt with it, of course[08:58:56] <geist> i did note that the linkhack.so binary is the only one in the build system that isn't using a custom linker script[08:59:07] <geist> so who knows what format the default linker script is attempting to put together[09:00:41] <JonathanThompson> You wouldn't want to be bored with everything being stock anyway, and you know it :)[09:03:10] <tic> *yawn*[09:03:55] <slaad> tic![09:03:58] * JonathanThompson sticks finger (wrapped in chain mail) into tic's mouth while he's yawning[09:04:02] <tic> Heya![09:04:05] <tic> oh noes![09:04:10] <tic> my mouf huffs[09:04:17] <tic> fop iff[09:04:21] <JonathanThompson> Rusty chain mail :)[09:04:27] <tic> :[09:04:28] <tic> :)[09:04:31] <tic> 'sup, people?[09:05:11] * JonathanThompson wonders how long before Bernd is in court[09:06:19] * mmu_man wonders how long ACCESS is in court for NDA breakage and wandering[09:06:31] <mmu_man> oh well, who knows :)[09:06:50] <JonathanThompson> Either way, I suspect Bernd will be in court soon.[09:08:16] <GreyGhost> what disk format does haiku? ext 2/ 3 ?[09:08:35] <JonathanThompson> BFS[09:08:53] <JonathanThompson> Its own implementation without using original BeOS code.[09:09:01] <GreyGhost> hmm .. and can it be natively read by Windows or Ubuntu ?[09:09:12] <GreyGhost> read/write rather[09:09:13] <JonathanThompson> I'm not sure if it has been declared final release for that filesystem driver or not.[09:09:35] <GreyGhost> ok[09:09:54] <mmu_man> JonathanThompson maybe not on the side you think though[09:09:56] <mmu_man> anyway[09:09:57] <JonathanThompson> Of course, using Haiku in its pre-alpha state is horribly premature.[09:09:59] <mmu_man> bbl[09:10:00] *** mmu_man has quit IRC[09:10:03] <GreyGhost> lol .. first "BFS wikipedia" google search etyry .. http://en.wikipedia.org/wiki/Bowling_for_Soup[09:10:34] <GreyGhost> JonathanThompson ,i wanted to transfer files onto the image :([09:10:50] <geist> hrm, looks like the linkhack stuff is totally unnecessary on ppc-gcc4 at least[09:10:52] <JonathanThompson> You probably want to do a build of Haiku to do that.[09:11:03] <geist> probably will find that x86-gcc4 is the same[09:11:41] <geist> but it does seem to still be needed (at least without finding a better way to do it) on gcc 2.95.3 (x86 only of course)[09:11:57] <GreyGhost> JonathanThompson ,yeah .. i'll do that .. can i make build system auto include a directory to the image?[09:12:17] <JonathanThompson> The build system creates a BFS image, GreyGhost.[09:12:20] <GreyGhost> geist ,why hasnt Be world moved on to atleast gcc 3.x ?[09:12:36] <JonathanThompson> That's got a fairly complicated answer, GreyGhost :)[09:12:42] <JonathanThompson> Has to do with ABI of C++ compilers.[09:13:10] <GreyGhost> JonathanThompson , no i mean .. when it creates the image can i make him copy over a directory with my files?[09:13:15] <GreyGhost> oh k .. :)[09:13:26] <JonathanThompson> How compiled C++ code works at a low-level is different between 2.9x and a later version, making them incompatible.[09:13:32] <Teknomancer> gcc 4 is breaks bin compat[09:13:42] <JonathanThompson> That can be done, with a modification of the build, GreyGhost.[09:13:43] <geist> gcc 3.x does as well[09:13:46] <Teknomancer> and we first need to port gcc 4 to haiku[09:13:50] <Teknomancer> is that done?[09:13:53] <geist> there were two or three abi changes in there[09:14:03] <JonathanThompson> And it only takes one to break things.[09:14:04] <GreyGhost> JonathanThompson ,ok thanks :)[09:14:13] <geist> Teknomancer: yes. the buildtools dir has gcc4 in it as well[09:14:25] <geist> for ppc builds you have to use gcc4[09:14:26] <GreyGhost> oh k..[09:14:37] <Teknomancer> geist k nice because i think eventually somewhere down the line maybe a year or two we'd be moving to gcc 4[09:14:43] <GreyGhost> gcc 4 broke compatibility with a whole lot of gcc 3.x stuff anyway..[09:14:56] <Teknomancer> gcc 4 compiles slower[09:15:01] <Teknomancer> dunno if its a problem with the port[09:15:04] <Teknomancer> or gcc 4 is just slower[09:15:09] <geist> it's just slower[09:15:12] <geist> more optimizations[09:15:22] <Teknomancer> smaller bins?[09:15:22] <geist> doesn't come for free[09:15:34] <JonathanThompson> Also a more complicated structure to support long-term growth, right?[09:15:48] <geist> though i generally thought that they sped up a bunch of the other parts of gcc 4 which sort of made up for it[09:15:54] <GreyGhost> O_o .. 4 is fater than 3 .. atleast under windows ;) dont know bout 2[09:15:57] <geist> like the front end lexer is much faster[09:16:09] <geist> but yeah, it's slow[09:16:14] <geist> you can really tell if you run on a really old machine[09:16:19] <geist> 40mhz sparc or something[09:16:24] <geist> it really grinds along[09:16:26] <JonathanThompson> Like this thing would show it big time.[09:16:35] <Teknomancer> mine is a 1.73 Ghz Vaio[09:16:38] <JonathanThompson> Maybe not as much as a 40 Mhz sparc :)[09:16:41] <Teknomancer> and its noticably slower on that[09:17:27] <GreyGhost> does SDL work under Haiku ?[09:17:41] <GreyGhost> it has a BeOS 5 port so maybe ..[09:18:26] * JonathanThompson goes off to bed, or at least off to the couch[09:18:33] <Teknomancer> bye JonathanThompson[09:18:39] <JonathanThompson> Seeya, Teknomancer.[09:18:51] <JonathanThompson> (And all udders)[09:18:58] <GreyGhost> JonathanThompson cya .. night[09:19:04] <geist> yeah, same[09:20:11] <GreyGhost> O_o .. according to wikipedia .. BFS lacks a whole lotqa stuff..[09:20:54] *** jiuda_D`arkness has joined #haiku[09:21:26] <GreyGhost>[09:21:29] <geist>[09:21:30] <geist> such as?[09:22:11] <GreyGhost> http://en.wikipedia.org/wiki/Comparison_of_file_systems[09:22:22] <GreyGhost> i dont understand any of it .. nor their importance ;)[09:22:43] <Teknomancer> GreyGhost probably some linux zealot wrote that to show their own FS as the best :P[09:23:00] <GreyGhost> lol ;)[09:23:26] <GreyGhost> how large is a clean haiku source checkout?[09:23:36] <Teknomancer> actually the "Features" table shows BFS as quite good[09:24:16] <GreyGhost> yeah .. but the Allocation thingy doesnt eem to be happy...[09:25:04] *** emitrax has joined #haiku[09:25:34] <geist> that's becuase some of those columns dont make any sense[09:25:51] *** Euver has joined #haiku[09:26:11] <GreyGhost> hehe ok ;)[09:26:14] <geist> nor do some of them matter much anymore[09:26:21] <geist> block suballocation isn't that big of a deal anymore[09:26:37] <geist> packing the tail is totally a reiser thing (or some other tree based fs)[09:26:37] <GreyGhost> ok[09:26:54] <geist> allocate-on-flush would be nice though, but the beos fs layer doesn't allow for that[09:26:56] <GreyGhost> btw.. why does the registry process check ur math?[09:27:12] <GreyGhost> geist ,ok .. so its just fancyty stuff not really needed thingys?[09:27:29] <geist> well, some of it is platform limitations, or simply not applicable due to the design[09:27:52] <geist> and anyway, who cares[09:28:14] <GreyGhost> ok[09:30:35] <GreyGhost> so anyone.. how large is the clean SVN checkout?[09:38:18] <slaad> No idea, sorry....[09:38:23] <geist> src or src + buildtools?[09:38:53] <GreyGhost> yeah ... src +build tools[09:39:19] <GreyGhost> btw.. how are "source" incompatibilities known ?[09:39:25] <geist> geist@four ~/haiku$ du -s buildtools trunk[09:39:25] <geist> 923772 buildtools[09:39:26] <geist> 873062 trunk[09:39:53] <GreyGhost> geist ,ok thanks :)[09:40:32] <geist> note that half of that is the overhead of svn[09:40:47] <geist> since svn stores the head revision of each file in it's .svn dir[09:41:00] <GreyGhost> oh k..[09:41:09] <GreyGhost> so around 800 mbs[09:41:16] <geist> also that was on linux, so on beos the overhead may or may not be as much[09:41:29] <geist> yeah, but that would only be from a source snapshot[09:41:36] <GreyGhost> i'm going to be on linux ;)[09:41:37] <geist> you probably do want to svn co it, so you can update[09:41:53] <GreyGhost> yeah .. i want a check out .. will it be much larger ?[09:42:06] <geist> the previous numbers were a checkout[09:42:19] <GreyGhost> ok cool :)[09:42:23] * geist sleeps[09:42:44] <GreyGhost> geist ,night[09:43:32] * GreyGhost will checkout later today..[09:44:48] *** GreyGhost has quit IRC[09:48:15] *** Ingenu has joined #Haiku[09:48:33] *** mazon is now known as Mazon[09:57:18] *** jiuda_D`arkness has quit IRC[09:59:50] *** kokito has quit IRC[10:08:38] *** Lelldorin1 has joined #haiku[10:34:27] <raph_ael> hello[10:36:53] <Ingenu> hi[10:40:05] <raph_ael> :)[10:40:15] *** ravtux has left #haiku[10:42:25] *** arikato has quit IRC[10:44:42] *** arikato has joined #haiku[11:09:33] *** mmu_man has joined #haiku[11:09:33] *** ChanServ sets mode: +o mmu_man[11:33:16] *** genkie has joined #haiku[11:37:47] *** emitrax_ has joined #haiku[11:38:42] <Teknomancer> status_t[11:38:42] <Teknomancer> _user_shutdown(bool reboot)[11:38:43] <Teknomancer> {[11:38:43] <Teknomancer> if (geteuid() != 0)[11:38:44] <Teknomancer> return B_NOT_ALLOWED;[11:38:44] <Teknomancer> return shutdown(reboot);[11:38:45] <Teknomancer> }[11:38:47] <Teknomancer> hmmmm :)[11:41:50] *** Sloar has quit IRC[11:45:24] *** GreyGhost has joined #haiku[11:48:17] <CIA-17> hugosantos * r20601 /haiku/trunk/src/add-ons/kernel/network/ (4 files in 2 dirs):[11:48:17] <CIA-17> a bit more work towards proper locking including a fix to a refcount bug[11:48:17] <CIA-17> - fixed a issue in add_interface_to_domain where the device interface's refcount was always incremented since that function was getting the device interface handle and not returning it unconditionlly[11:48:17] <CIA-17> - if the ethernet device goes down, and the fd is close()ed, return B_FILE_ERROR instead of calling into the driver again[11:54:34] *** emitrax has quit IRC[12:26:10] <mmu_man> Teknomancer ?[12:26:16] <Teknomancer> mmu_man[12:26:52] *** dr_evil has joined #haiku[12:28:31] <mmu_man> plop[12:28:34] *** mmu_man sets mode: +o dr_evil[12:28:39] <mmu_man> brb[12:28:49] <Teknomancer> ok[12:29:22] <Teknomancer> looks like some parts of haiku are ready for MU but not others (like registrar)[12:33:31] <Lelldorin1> can i make the haku-image file bigger as the foldersize of the source with jam?[12:35:30] <Lelldorin1> sorry my english is not so good[12:35:45] <Lelldorin1> i have build the haiku source and want to create the imagefile of haiku[12:36:19] <Lelldorin1> and i want to create a bigger imagefile as i can get with the raw imagefile of the website[12:37:09] <Teknomancer> using dd ?[12:37:12] <Teknomancer> brb[12:37:35] <Lelldorin1> ok no other way?[12:37:38] <Lelldorin1> thanx[12:37:47] <Teknomancer> dunno, maybe there is[12:37:54] *** genkie has quit IRC[12:41:46] *** Shuren has quit IRC[12:44:35] *** ekdahl has joined #haiku[12:45:53] <GreyGhost> time to boot into Ubuntu and setup everything for compiling aiku now :)[12:46:02] *** GreyGhost has quit IRC[12:48:24] <Fanskapet> http://users.sidvicious.se/fanskapet/images/macbook/DSC00128.JPG[12:48:26] <Fanskapet> nice bag :P[12:49:51] <Teknomancer> be back in half an hour[12:49:58] *** Teknomancer has quit IRC[12:53:37] *** aldeck has joined #haiku[12:55:33] *** firefoxman_ has joined #haiku[12:55:34] *** Lelldorin1 has quit IRC[12:55:38] *** firefoxman_ has left #haiku[12:56:31] *** stargater has joined #haiku[12:56:48] <stargater> hi[13:11:08] <mmu_man> hmm after teh iIron they should do the macIron in white plastic (that would melt down :)[13:12:57] *** miqlas has quit IRC[13:14:49] <mmu_man> hmm can someone explain me this ???[13:14:58] <mmu_man> http://www.osnews.com/user.php?uid=9493[13:14:59] <mmu_man> Joined: 2006-09-30[13:15:05] <mmu_man> Number of Comments: 73 (30 voted up, 0 voted down)[13:15:05] <mmu_man>[13:15:06] <mmu_man> Average Comment Score: 1.85[13:15:17] <Ingenu> you've been rated[13:15:17] <mmu_man> http://www.osnews.com/user.php?uid=12211[13:15:22] <mmu_man> Joined: 2007-04-05[13:15:31] <mmu_man> Number of Comments: 3 (1 voted up, 0 voted down)[13:15:31] <mmu_man>[13:15:32] <mmu_man> Average Comment Score: 2.33[13:15:35] <mmu_man> *2.33* ??[13:15:48] <Ingenu> bah[13:15:56] <Ingenu> are you doing some epenis competition ?[13:16:00] <Ingenu> if not just discard that[13:16:01] <mmu_man> I have more upped comments in proportion...[13:16:21] <dr_evil> mmu_man yes thats strange, but perhaps Eugenia doesn't like you ;)[13:16:26] <mmu_man> no it just sounded weird seeing someone 1 day new getting a higher score[13:16:40] <mmu_man> dr_evil I wouldn't bet on that, indeed :))[13:16:51] <mmu_man> or she likes him omre[13:16:56] <mmu_man> or he paid more :D[13:17:23] <mmu_man> or maybe the way the y took old comments is odd[13:17:25] <mmu_man> anyway[13:17:37] *** wildur has joined #Haiku[13:18:14] <mmu_man> I find it funny someone who pops up from nowhere is granted such legitimacy, even if he really from where he says[13:18:23] <mmu_man> funny...[13:18:32] <mmu_man> oh well, not the place and I'm busy[13:25:12] *** aldeck has quit IRC[13:26:28] <dr_evil> mmu_man even my score is better than yours ;)[13:27:04] <mmu_man> arf[13:27:36] <mmu_man> <Calimero>That's unfair... That's sooo unfair...</Calimero>[13:27:51] <dr_evil> Number of Comments: 2 (2 voted up, 0 voted down)[13:27:51] <dr_evil> Average Comment Score: 2[13:27:58] <mmu_man> pff[13:28:24] <dr_evil> score must be calculated with a hash function ;)[13:28:27] <mmu_man> http://en.wikipedia.org/wiki/Calim%C3%A9ro[13:28:42] <mmu_man> Quote: It's unfair, they are big and I am small.[13:28:43] <mmu_man> :))[13:31:08] <Ingenu> :)[13:40:21] <dr_evil> mmu_man do you know the title of an anime where deamons are good and angels are evil?[13:40:21] *** [Beta] has quit IRC[13:40:40] <mmu_man> hmm no[13:40:49] <mmu_man> doesn't ring the bell[13:41:37] *** [Beta] has joined #haiku[13:43:25] *** Teknomancer has joined #haiku[13:46:11] *** kb7sqi has quit IRC[13:47:54] <mmu_man> bbl[13:47:55] *** kb7sqi has joined #haiku[13:47:56] *** mmu_man has quit IRC[13:53:42] *** stargater has quit IRC[13:56:44] *** tombhadAC has joined #haiku[14:05:39] *** rcjsuen has joined #haiku[14:06:13] *** mmu_man has joined #haiku[14:06:13] *** ChanServ sets mode: +o mmu_man[14:06:47] <dr_evil> probably was Angel Sanctuary[14:07:05] <mmu_man> ah[14:07:10] <mmu_man> dunno[14:07:10] *** kb7sqi has quit IRC[14:08:07] *** Euver has quit IRC[14:08:18] <dr_evil> I probably didn't pay enough attention when watching it[14:09:28] *** TLF has joined #haiku[14:12:12] *** gravy has joined #Haiku[14:21:56] *** DaaT has joined #haiku[14:30:31] *** slaine_ has joined #haiku[14:32:30] *** aldeck has joined #haiku[14:42:42] *** mmadia has joined #haiku[14:43:56] *** Mazon is now known as mazon[14:51:04] *** MrSunshine has quit IRC[14:51:20] *** MrSunshine has joined #haiku[15:01:44] *** mmadia has quit IRC[15:04:39] *** fyysik has joined #haiku[15:09:11] *** DaaT has quit IRC[15:10:07] *** jspiros_ has joined #haiku[15:24:27] *** jspiros has quit IRC[15:25:01] *** GreyGhost-Ubuntu has joined #haiku[15:26:16] <MrSunshine> _hugo, you a new dev btw? :)[15:30:55] <GreyGhost-Ubuntu> MrSunshine ,from the time i've been here.. two days.. i think someone said.. he was an old dev.. left and came back :) but i'm not sure. .[15:31:36] <MrSunshine> ahh :)[15:31:43] <MrSunshine> he seems to work good on the network stuff atm so :)[15:31:48] <MrSunshine> if its the hugo that is commiting :)[15:32:18] <[Beta]> it is :)[15:33:32] <MrSunshine> gah hangovers sucks :)[15:33:35] <MrSunshine> dammit[15:34:27] <aldeck> hehe, never heard of this :) http://bella.googlecode.com/svn/trunk/bella/headers/os/ , seems abandonned though[15:35:41] *** Sloar has joined #haiku[15:35:50] <aldeck> not much info either[15:36:37] <fyysik> MrSunshine - yup, he created FAT Tracker and GFX viewer for BeOS, for example, sold (with weird problems) Tracker to Zeta and then left:)[15:37:04] <MrSunshine> heh :)[15:37:15] <MrSunshine> great to see more devs here atlesat :)[15:41:17] <mmu_man> fyysik yeah sold, but yT had to rewrite it or so IIRC[15:41:23] <mmu_man> anyway I'm off[15:41:24] <mmu_man> +++[15:41:27] *** mmu_man has quit IRC[15:41:52] *** TTRanger has quit IRC[15:42:01] <kr1stof> I remember ShowDesktop by Hugo Santos[15:42:57] <kr1stof> ...and BeAndSee[15:48:30] <dr_evil> aldeck thats a good example of premature optimization http://bella.googlecode.com/svn/trunk/bella/src/kits/Interface/math.cpp[15:49:17] <aldeck> dr_evil: ohoh, yep you said it ![15:49:36] *** guildencrantz has joined #haiku[15:54:54] *** gravy_ has joined #Haiku[15:55:45] *** gravy has quit IRC[16:03:12] *** arikato has quit IRC[16:04:36] *** arikato has joined #haiku[16:08:02] *** Sloar has quit IRC[16:08:02] *** jspiros_ has quit IRC[16:08:03] *** ekdahl has quit IRC[16:08:04] *** M199 has quit IRC[16:08:04] *** Fanskapet has quit IRC[16:08:04] *** mats has quit IRC[16:08:07] *** eNGIMa|thiele has quit IRC[16:08:09] *** DeadYak has quit IRC[16:08:36] *** Sloar has joined #haiku[16:08:36] *** jspiros_ has joined #haiku[16:08:36] *** ekdahl has joined #haiku[16:08:36] *** M199 has joined #haiku[16:08:36] *** Fanskapet has joined #haiku[16:08:36] *** mats has joined #haiku[16:08:36] *** eNGIMa|thiele has joined #haiku[16:08:36] *** DeadYak has joined #haiku[16:15:49] *** umccullough has joined #haiku[16:19:16] *** slaad has quit IRC[16:25:28] <GreyGhost-Ubuntu> hmm.. all the buildtools are to be used from the haiku branch? and none like jam to be apt-get ed ?[16:25:57] <umccullough> GreyGhost-Ubuntu, saw some of your questions from last night...[16:26:11] <umccullough> you do need jam from haiku buildtools[16:26:16] <GreyGhost-Ubuntu> ok .. got my answer.. should have read the doc a bit further :([16:26:32] <umccullough> GreyGhost-Ubuntu, did you ever get that answer about adding custom files/dirs to the image?[16:26:35] <umccullough> it's actually easy[16:26:36] <GreyGhost-Ubuntu> umccullough ,ok .. and hello .. i'm really a n00b as u might have observed :([16:26:42] <umccullough> :)[16:26:55] <GreyGhost-Ubuntu> umccullough ,nope.. not the whole process atleast.. just that it caan be done..[16:26:57] *** kokito has joined #haiku[16:28:03] <umccullough> GreyGhost-Ubuntu, once you have things building, i can point you to a couple different ways - but the easiest I've found is to use the UserBuildConfig to run a late-image-script which can be used to (recursively) copy directories from the host system to the image using bfs_shell before it's unmounted[16:28:06] <GreyGhost-Ubuntu> umccullough ,if u could tell me the steps in short it would be nice :) .. and i'll tinker around once i'm building[16:28:18] <GreyGhost-Ubuntu> ok ..[16:28:38] <umccullough> you'll find a UserBuildConfig.sample in the build/jam directory I believe[16:29:03] <GreyGhost-Ubuntu> umccullough ok . i'll look at it :)[16:30:20] *** BGA has joined #haiku[16:30:20] *** ChanServ sets mode: +o BGA[16:32:30] <umccullough> GreyGhost-Ubuntu, here's an example of a script I use: ftp://ftp.binarychicken.com/haiku/late-image-script.sh[16:32:42] *** TTRanger has joined #Haiku[16:32:55] <umccullough> it recursively copies a whole bunch of structures that i set up however i wish to the root of the image[16:33:08] <umccullough> allows me to add basically whatever, apps, settings, drivers, etc.[16:33:19] <GreyGhost-Ubuntu> ahh.. ok .. cool .. thanks :)[16:33:33] <umccullough> granted, i build from BeOS R5 - so... my host is BFS as well[16:33:46] <GreyGhost-Ubuntu> ahh.. ok ..[16:34:05] <umccullough> some report that symlinks don't copy well with that method, btw[16:34:11] <GreyGhost-Ubuntu> umccullough: but i belive it should work under linux too ?[16:34:20] <GreyGhost-Ubuntu> ok..[16:34:21] <umccullough> GreyGhost-Ubuntu, it should[16:34:31] <GreyGhost-Ubuntu> ok .. i'll try that :)[16:34:36] <GreyGhost-Ubuntu> thanks for the script[16:34:40] * GreyGhost-Ubuntu copies it over[16:34:47] <umccullough> it's the same basic method used by the image build script to copy all the targets to the image[16:35:16] <umccullough> you just gotta have it execute the script at the right time in the build process - you'll see that in the UserBuildConfig.sample[16:36:19] <GreyGhost-Ubuntu> ok .. thanks[16:36:19] <umccullough> you can also add targets from the build system into the image using other features in the UserBuildConfig - but if you have a bunch of non-Haiku related things to copy, the late-image script is easier[16:36:43] <GreyGhost-Ubuntu> ahh.. ok ..[16:37:02] <umccullough> and BTW, once you get the cross-tools compiled, you do NOT need to run ./configure again -- for some reason everyone thinks you do :P[16:37:44] <umccullough> newbs that is[16:38:39] *** LeandroLuiz has joined #haiku[16:38:45] <GreyGhost-Ubuntu> ok ..thanks for all the info :D[16:39:16] <TTRanger> Can someone please help me with a scripting question?[16:42:50] *** emitrax__ has joined #haiku[16:44:19] <GreyGhost-Ubuntu> the build tools are pretty large ;)[16:46:11] *** LeandroLuiz has quit IRC[16:46:39] <GreyGhost-Ubuntu> btw.. a wiki would be nice for the docs.. it makes it much easier for ppl to contribute..[16:47:04] <umccullough> GreyGhost-Ubuntu, there was a wiki... but it's gone atm[16:47:31] <umccullough> ok, bye - going to work[16:47:38] *** umccullough has left #haiku[16:48:46] <GreyGhost-Ubuntu> v umccullough: ,cya[16:48:50] <GreyGhost-Ubuntu> wrr..too late[16:51:07] *** LeandroLuiz has joined #haiku[17:00:58] *** emitrax_ has quit IRC[17:06:18] *** fyysik has quit IRC[17:08:11] *** LeandroLuiz is now known as LeandroLuiz[OFF][17:09:29] *** fyysik has joined #haiku[17:09:46] <fyysik> heh, GPL danger:) http://www.osnews.com/story.php/17644/OpenBSD-bcw-Possible-GPL-Violation-Issues[17:13:09] *** Faust-C has joined #haiku[17:13:36] *** Teknomancer is now known as _tek[17:13:45] <Faust-C> can anyone point me in the direction of the available apps for haiku?[17:14:33] <kokito> bebits.com[17:14:42] <Faust-C> ty[17:15:13] <kokito> don't expect many to run though Faust-C; Haiku is still alpha[17:15:27] <Faust-C> im just looking atm[17:16:53] *** MikeW has joined #haiku[17:20:24] *** Faust-C has left #haiku[17:22:11] <GreyGhost-Ubuntu> hmm.. whats the /buildtools/legacy branch for.. is that 2.x gcc ?[17:22:28] <_tek> i sent a mail to the haiku mailing list[17:22:36] <_tek> but i don't seem to get it in my inbox :P[17:22:46] <dr_evil> GreyGhost-Ubuntu yes[17:23:03] <GreyGhost-Ubuntu> ok[17:26:50] <_tek> sent regarding the toolbar class[17:27:00] <_tek> not sure if it got sent even tho Beam says so :)[17:27:01] <ekdahl> _tek: I got it[17:27:06] <_tek> ekdahl ah cool :)[17:27:15] <_tek> guess it takes a while to get it for me[17:27:41] <kokito> _tek: your email did get to my inbox :)[17:27:52] <_tek> kokito but not mine ...[17:28:03] <ekdahl> an i also submitted the very same task to trac a couple of seconds before I got your mail :)[17:28:14] <_tek> ah :)[17:28:24] <kokito> _tek: I guess it can take a little time for the messages to be relayed to everyone on the list[17:28:44] <_tek> kokito hm, yeah could be how many ppl on it? 200 or so?[17:29:12] <kokito> _tek: don't know for sure, but I think mphipps said we were close to 1K subscribers[17:29:13] *** kb7sqi has joined #haiku[17:29:18] <_tek> oh wow[17:29:49] <kokito> even the website has around 1K registered users :)[17:29:54] <ekdahl> _tek: I don't think you get the messages you send yourself when using gmail, at least I don't[17:30:13] <_tek> ekdahl the previous message i sent i got it...[17:30:38] <ekdahl> the are on the web interface, but I never get them in Thunderbird[17:31:08] <ekdahl> are you using the webmail?[17:35:20] <ekdahl> anyway, I think that class is really needed[17:36:02] <ekdahl> _tek: do you think the toolbar buttons should be regular BButtons or some special class?[17:36:25] <_tek> well there should be BToolbarButton or so, or even something like BImageButton[17:36:29] <_tek> that can be added to the toolbar[17:36:46] <_tek> because of the "drop down menu" functionality it would be better if it were a special class[17:37:17] <_tek> i read mmu_man wanted to do something by inheriting the existing class, i'm not entirely sure if that might work out but then again i didn't read it very finely.[17:38:04] *** kb7sqi has quit IRC[17:38:41] <ekdahl> mm yeah maybe a special class is the way to go, wit a couple of different "modes": eg. big icon and label, big icon, small icon with label and small icon[17:38:41] *** slaine_ has quit IRC[17:38:59] <_tek> ekdahl yeah its already there to an extent in Beezer[17:39:09] <ekdahl> oh, nice[17:39:10] <_tek> i could probably rewrite about 40% of it and get it ready for haiku...[17:39:28] <_tek> currently the toolbar supports buttons, but making it support combo-boxes (BMenuFields etc.) should be possible as well[17:39:37] <_tek> maybe even make it accept generic BViews itself[17:39:55] <dr_evil> fyysik now you made me waste my time reading that discussion[17:41:20] <ekdahl> _tek: sounds like a plan then. lets wait and see what the "bosses" have to say about it then ;)[17:41:24] <_tek> hm, a few of the colors need to be changed from hardcoded to tint_color ...[17:41:59] <_tek> yeah[17:43:40] <ekdahl> btw, HVIF support would be a nice addition also[17:46:14] <fyysik> dr_evil - why did you read discussion? I read only newslink itself:)[17:47:16] <dr_evil> http://marc.info/?l=linux-wireless&m=117579116031296&w=2[17:47:34] <dr_evil> fyysik I can understand this reaction[17:48:43] <fyysik> public source, public statements, heh[17:49:33] <dr_evil> well, obviously the bsd developer made a mistake by simply copying parts of GPL code, but I think his response to the public attack was appropriate[17:50:44] <dr_evil> fyysik I recently compared the zeta sk98 driver with some source I wrote, and I wasn't happy either[17:51:00] <fyysik> yeah, it always reminds SCO case with discussions if copying if(i<1) is copyright violation[17:51:46] <fyysik> are licenses of Zeta's one and yours different?[17:52:41] <dr_evil> uhh well, they used my code, but removed the license and violate the conditions of use.[17:53:01] <dr_evil> but as they never published it, I'm not going to persue this further[17:53:07] <fyysik> same happened it seem with Siarzhuk's drivers IIRC[17:53:25] <dr_evil> i don'[17:53:35] <fyysik> which were part of distributed CD[17:53:36] <dr_evil> I don't know Siarzhuk, what kind of driver is it[17:53:49] <fyysik> sound driver + usb_mass_storage[17:54:20] <fyysik> don't remember though for which they dropped credits etc[17:54:45] <dr_evil> I see[17:54:53] <dr_evil> luckily, it seems to be all over now[17:55:28] <fyysik> is Ithamar still in Haiku development?[17:56:00] <dr_evil> i doubt that, at least he hasn't provided anything for a long time. but he's still welcome :)[17:56:56] <_tek> fyysik dont think so :([17:57:18] * fyysik wonders if "iVirus" term is trademarked already...[17:59:09] <dr_evil> fyysik http://www.theinquirer.net/default.aspx?article=38746 Open Source coders caught stealing Open Source code[18:00:33] <fyysik> dark/clean room approach?[18:02:40] <DeadYak> fyysik: Ithamar works at Tom-tom nowadays, he spends all his day doing linux kernel work[18:03:45] <_tek> he was working on the HDA driver :)[18:03:48] <_tek> guess he's too busy[18:03:56] <fyysik> btw, that approach reminds me experimental neurolinguistics. There was experiment with two schisofrenics sitting on sides of curtain. One has ready Lego-onjects, second - details. Goal of communications was second to assemble same object:)[18:05:40] <dr_evil> fyysik wireless reminds me that I wanted to buy WPA hardware and remove the long cable from this notebook[18:06:13] <dr_evil> it also reminds me that one of my neighbours sucessfully broke my WEP-128 bit key[18:07:30] <fyysik> sure, just read article about method to break it in 1 minute on 1.7 MHz Pentium[18:07:52] <fyysik> instead 20 minutes before[18:08:41] <Ingenu> GHz[18:08:44] <Ingenu> I'd guess[18:09:55] <DeadYak> nah, just weaknesses in WEP[18:10:04] <DeadYak> oh[18:10:05] <DeadYak> misread that[18:10:06] <DeadYak> but yeah[18:10:14] <DeadYak> that was for a 104bit key, dunno about for 128[18:10:23] <DeadYak> either way WEP is crap :/[18:11:05] <Ingenu> that's what my DS use[18:11:22] <fyysik> sure, GHz:)[18:12:28] <dr_evil> 104 is 128[18:12:47] <dr_evil> and 40 is 64[18:13:00] <DeadYak> are 24 bits of it reserved or something?[18:13:28] <dr_evil> there are only two WEP key length, i don't know whats in the other bits[18:13:35] <DeadYak> ah.[18:20:59] *** SunWuKung has quit IRC[18:31:05] *** Sil2100 has joined #haiku[18:31:30] <kokito> morning fellows[18:31:36] <Sil2100> Hi[18:32:43] *** XXXGauronXXX has joined #haiku[18:44:33] *** GreyGhost-Ubuntu has left #haiku[18:49:47] *** GreyGhost has joined #haiku[18:53:24] *** Ingenu has quit IRC[18:57:15] *** MikeW has quit IRC[19:02:07] *** fyysik has quit IRC[19:15:58] *** stargater has joined #haiku[19:16:17] <stargater> re[19:16:46] *** ekdahl has quit IRC[19:17:28] *** ekdahl has joined #haiku[19:17:56] *** ekdahl has quit IRC[19:18:09] *** absabs has left #haiku[19:18:43] *** ekdahl has joined #haiku[19:29:27] *** _tek has quit IRC[19:30:15] *** emitrax_ has joined #haiku[19:30:17] *** PascalR has joined #haiku[19:32:36] *** MikeW has joined #haiku[19:46:18] *** emitrax__ has quit IRC[19:47:02] *** JamesB192 has joined #haiku[19:50:44] *** Ingenu has joined #Haiku[19:51:43] *** GreyGhostNew has joined #haiku[19:58:13] *** Jin has joined #Haiku[20:08:41] *** doublegg has joined #haiku[20:11:28] *** GreyGhost has quit IRC[20:12:01] *** GreyGhostNew is now known as GreyGhost[20:15:05] *** TLF has quit IRC[20:18:28] *** _hugo has left #haiku[20:19:07] *** GreyGhost has left #haiku[20:31:24] *** JamesB192 has quit IRC[20:37:09] *** TLF has joined #haiku[20:37:42] *** doublegg has quit IRC[20:43:47] <MikeW> oh excellent, python[20:43:56] <MikeW> Anyone here from ITBug?[20:44:28] <MikeW> emitrax_ :)[20:56:30] *** arikato has quit IRC[20:56:42] *** _hugo has joined #haiku[21:00:52] *** sarcas has joined #haiku[21:04:38] *** Jin has quit IRC[21:17:36] <MikeW> Theo de raadt comes across as a major butthead[21:17:52] <MikeW> I was just reading his responses in this thread: http://marc.info/?t=117571775000002&r=4&w=2[21:18:48] <emitrax_> hi MikeW[21:19:21] <MikeW> emitrax_: Hey there. Was it you who was working on the python port for haiku?[21:19:39] <emitrax_> nope[21:19:43] <emitrax_> I'm totally new with Haiku[21:21:11] <MikeW> oh, sorry. I saw your nick mentioned in a link to a mailing list post in the same news article that showed python under haiku, and assumed you ported it[21:22:26] <aldeck> was me :P[21:23:35] *** emitrax__ has joined #haiku[21:24:54] *** kr1stof_ has joined #haiku[21:25:53] *** jpd has joined #haiku[21:25:54] <aldeck> i'm not doing a python port, just testing to see what's missing to make it run :) all the hard work has been done by Donn Cave http://www.bebits.com/app/1564[21:26:10] *** mazon is now known as Mazon[21:29:45] *** b3w_ has joined #haiku[21:33:34] *** ul1984_ has joined #haiku[21:39:17] *** kr1stof has quit IRC[21:39:25] *** emitrax_ has quit IRC[21:41:25] *** ul1984_ has quit IRC[21:43:54] *** sarcas has quit IRC[21:45:07] *** ul1984_ has joined #haiku[21:47:48] *** ul1984 has quit IRC[21:49:19] *** b3w has quit IRC[21:54:29] *** snipe714 has joined #haiku[21:56:34] *** dos4gw has joined #haiku[21:56:34] *** dr_evil has quit IRC[21:59:56] *** dos4gw is now known as dr_evil[21:59:58] *** ChanServ sets mode: +o dr_evil[22:00:45] *** Lelldorin1 has joined #haiku[22:00:53] *** Lelldorin1 has quit IRC[22:03:51] *** ul1984 has joined #haiku[22:04:43] *** guma has joined #haiku[22:12:16] *** ul1984_ has quit IRC[22:23:15] *** pikapika has joined #haiku[22:24:02] <pikapika> yo[22:25:07] <kokito> yo soy /me[22:25:42] <pikapika> :)[22:28:32] <kokito> ser o no ser...[22:31:35] * DeadYak pets spanish Hamlet kokito[22:31:48] <kokito> yo?[22:32:17] <DeadYak> at least I'm assuming that was spanish paraphrasing of "To be or not to be" :)[22:34:15] <kokito> yes, that's right DeadYak :)[22:34:51] *** eNGIMa has quit IRC[22:41:12] *** Sil2100 has quit IRC[22:43:17] *** JamesB192 has joined #haiku[22:48:51] *** ekdahl has left #haiku[22:53:13] *** snipe714 has quit IRC[22:57:53] <arachnist> http://img486.imageshack.us/img486/2417/zajacyc2.jpg <| no easter bunny this year[23:05:03] *** arikato has joined #haiku[23:08:52] *** shackan has joined #haiku[23:09:49] *** oco has joined #haiku[23:13:36] <CIA-17> bonefish * r20602 /haiku/trunk/build/ (5 files in 2 dirs): (log message trimmed)[23:13:36] <CIA-17> New build system feature to shorten the turn-around times when working[23:13:36] <CIA-17> with image files. E.g.[23:13:36] <CIA-17> jam -q update-image libbe.so kernel_x86[23:13:36] <CIA-17> will only build libbe.so and the kernel (if necessary) and copy them[23:13:37] <CIA-17> onto the already existing Haiku image. The MIME DB will not be[23:13:39] <CIA-17> reinstalled, and only those source directories will be copied for which[23:30:46] *** sarcas has joined #haiku[23:34:19] *** Sloar has quit IRC[23:43:56] *** Euver has joined #haiku[23:46:50] *** wildur has quit IRC[23:59:08] *** gotaku has joined #haiku