April 21, 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 21, 2007 [00:00:22] <mmu_man> the first one should be ok[00:04:41] <TTRanger> k thanks[00:05:15] *** TTRanger has quit IRC[00:10:22] *** stargater has joined #haiku[00:10:57] <stargater> hi[00:14:09] *** SiCuTDeUx has joined #haiku[00:20:32] *** jevin_ has joined #haiku[00:24:43] *** Deej575 has joined #haiku[00:33:48] <Pulko_Mandy> +++[00:33:50] *** Pulko_Mandy has quit IRC[00:35:39] *** Lelldorin1 has joined #haiku[00:35:43] *** Lelldorin1 has quit IRC[00:42:47] *** BryanV has joined #haiku[00:45:26] *** mmu_man has quit IRC[00:52:13] *** umccullough_work has joined #haiku[00:55:21] *** TLF has quit IRC[00:56:42] <_hugo> umccullough_work: will you be able to update your http server today? i have some changes in i would like to test (the image test)[00:58:28] *** MrRagga has quit IRC[00:58:37] *** MrRagga has joined #haiku[00:59:09] *** TTRanger has joined #haiku[01:00:47] *** guildencrantz has quit IRC[01:01:51] * TTRanger is broadcasting from BeOS Max...seems to be ok so far...[01:02:05] *** SiCuTDeUx has quit IRC[01:04:58] *** sarcas has joined #haiku[01:05:27] *** bobwhoops has quit IRC[01:05:38] *** bobwhoops has joined #haiku[01:08:00] *** sarcas has quit IRC[01:09:14] <umccullough_work> _hugo, i'll update it when i get home - probably in ~4 hours[01:09:49] <umccullough_work> _hugo, i must admit, the server is proving to be pretty stable lately[01:10:01] <umccullough_work> much better than when i first set it up on 20682 :)[01:10:17] *** rgb has joined #haiku[01:15:02] <_hugo> umccullough_work: cool[01:15:15] <_hugo> umccullough_work: hopefully we should be able to download bigger files with this latest revision as well[01:16:29] <umccullough_work> neat[01:16:34] <JamesB192> When I was you and went to school, I ask my teachers 'how can I think, with indoctrination day after day that causes my mind to shrink.' Yadda Yadda Ya. Whatever you think is wrong. You're listening to this song. Yadda Yadda Ya.[01:18:14] *** SiCuTDeUx has joined #haiku[01:18:14] <[Beta]> http://popey.com/Dell_XPS_Vanity_Lights would make a seriously nice Haiku laptop - blinken leds[01:18:23] <CIA-17> hugosantos * r20776 /haiku/trunk/src/add-ons/kernel/network/protocols/tcp/TCPEndpoint.cpp: fixed an issue with TCP where it possibly could not recover from a retransmission as the RCV.NXT was wrongly updated. When appending data we now set it to the sequence in the end of the receive queue.[01:20:46] *** SiCuTDeUx has quit IRC[01:21:08] *** SiCuTDeUx has joined #haiku[01:24:39] *** dr_evil has joined #haiku[01:37:16] *** Minbari has joined #haiku[01:38:05] *** rgb has quit IRC[01:43:18] *** tigerdog has quit IRC[01:45:24] *** jevin_ has quit IRC[01:54:16] *** dr_evil has quit IRC[01:55:01] *** tigerdog has joined #haiku[02:05:01] *** santagada has joined #haiku[02:05:21] *** kb7sqi has quit IRC[02:05:31] *** Mazon is now known as mazon[02:09:11] *** santagada has quit IRC[02:09:59] *** llama233 has joined #haiku[02:14:40] *** MikeW has quit IRC[02:16:44] *** ChrisHJW has quit IRC[02:22:35] *** pikapika has quit IRC[02:39:46] *** umccullough_work has quit IRC[02:48:29] *** llama233 has quit IRC[02:51:02] <BryanV> _hugo - you are my personal hero. :-)[02:51:43] <SiCuTDeUx> lol @ BryanV[02:51:48] <SiCuTDeUx> BryanV, why?[02:53:28] <BryanV> The amount of work on the network stuff he's done... it's almost like axel spawned or something[02:53:45] <_hugo> hey BryanV[02:53:53] *** SiCuTDeUx has quit IRC[02:53:54] <_hugo> thanks, but just improving what is already done[02:54:03] <BryanV> The work is really appreciated, _hugo[02:54:18] <_hugo> thanks[02:54:47] *** rcjsuen has quit IRC[02:55:10] <_hugo> it has been a fun ride so far :-)[02:55:21] <BryanV> :-D[02:59:20] <pyCube> heh[02:59:44] *** nonesuch has joined #haiku[03:08:57] *** FiddleStix has joined #haiku[03:21:34] <kokito> hey BryanV, just sent you an email ;)[03:36:56] *** M199 has joined #haiku[03:40:51] *** Master199 has quit IRC[03:40:51] *** M199 is now known as Master199[03:41:16] *** Minbari has quit IRC[03:49:38] *** JamesB192 has quit IRC[03:57:40] *** Crumbs has joined #haiku[04:02:51] <kokito> _hugo, in r20776, wget slows down and stops after 5 - 10 seconds from download start.[04:03:16] <_hugo> hm[04:03:32] <_hugo> kokito: which url?[04:03:33] <kokito> tested in vmware (both win and linux)[04:10:44] *** nonesuch has quit IRC[04:13:45] *** FiddleStix has quit IRC[04:23:51] <_hugo> ok, got the problem[04:23:59] <_hugo> as the TCP performance goes up, other problems surface[04:37:26] <CIA-17> hugosantos * r20777 /haiku/trunk/src/add-ons/kernel/network/protocols/tcp/TCPEndpoint.cpp: allow window updates to be sent after consuming data from the receive buffer. This fixes an issue where people were experiencing hangs in TCP downloads.[04:37:39] <_hugo> i can now max my internet link, neat[04:37:51] <_hugo> within haiku that is[04:43:15] <geist> woot[04:43:24] <geist> do we have a gigabit card driver?[04:44:01] <_hugo> yeah, intel pro 1000[04:44:09] <_hugo> but our allocator doesn't allows us to get that far[04:44:30] <_hugo> its pretty slow and its easy to get to out of memory situations (been there several times, eheh)[04:45:40] <geist> what allocator is getting hit?[04:46:05] <geist> for network stuff there's generally a single pool that stays active for that stuff[04:46:10] <_hugo> our net_buffer implementation (which isn't very optimized yet either) just uses malloc()[04:46:14] <_hugo> yeah[04:46:15] <geist> oh lame![04:46:20] <_hugo> but we don't have a slab allocator yet[04:46:33] <geist> even so, you should have it carve out a chunk[04:46:41] <geist> and fall back to malloc only if necessary[04:46:59] <_hugo> we could do that, but i think that effort would be better spent implementing the slab allocator[04:47:01] <geist> no wonder the port implementation is slow too, it probably mallocs the chunks as it moves data around[04:47:16] <geist> well, I'm thinkin the reason to do the carve is because it's trivial[04:47:19] <geist> until the slab allocator[04:47:36] <geist> but even then, i dont think the net stuff should hit the slab allocator that heavy either[04:47:45] <geist> you always need a pool that can be grabbed easily[04:48:05] <_hugo> well, for larger chunks it shouldnt, but for some inner parts of net_buffer it should[04:48:13] <_hugo> we have lots of tiny pieces[04:48:28] <_hugo> for net_buffer management[04:48:39] <geist> even for that[04:48:48] <geist> it needs to just be able to almost instantly grab a buffer[04:49:08] <_hugo> so what kind of allocator would you suggest?[04:49:15] <geist> a free list of buffers[04:49:31] <geist> pull form that, protected with just a spinlock[04:49:46] <geist> ie, optimize for the common case, which is something must needs a buffer and will probably give it back[04:50:01] <geist> and treat the edge case (there aren't enough in the pool) as the edge case[04:50:43] <geist> sometimes you gotta just grab a pool of ram and go with it, getting too fancy just adds overhead[04:50:44] <_hugo> that seams like something the slab could do[04:51:03] <geist> well sure, but I dont see anyone writing a slab allocator any time soon[04:51:08] <_hugo> we'll have to manage the list of buffers ourselves, and handle the fragmentation[04:51:09] <geist> folks have been talking about that for like 3 years now[04:51:35] <geist> no, i'm thinking of something far simpler[04:51:38] <geist> like what i had for newos[04:51:57] <geist> it just grabs a meg or 2 of ram at the start, carves it into 2k net buffers (header + some data)[04:52:03] <geist> and stuffs them into a free list[04:52:08] <_hugo> right[04:52:11] <geist> no allocator[04:52:17] <geist> you just grab one, stuff them back when you're done[04:52:22] <_hugo> but what if at some point we need more buffers than that?[04:52:28] <_hugo> do you grab more memory or?[04:52:29] <DeadYak> that's the edge case[04:52:33] <geist> right[04:52:34] <DeadYak> if you hit that point, then you malloc[04:52:35] <geist> that's the edge case[04:52:45] <_hugo> i don't think it is really[04:52:46] <geist> or create another pool (or go to the slab allocator)[04:52:52] <_hugo> it's not predictable how much memory you need[04:53:07] <_hugo> and that memory needs to be reclaimed by the system[04:53:08] <geist> absolutely, but for most cases it's still within reasonable bounds[04:53:29] <_hugo> sure, but when i start thinking about getting more buffers and returning them, i start thinking slab :-)[04:53:39] <geist> yes, but I dont see a slab allocator[04:53:44] <_hugo> but i agree that a simple solution[04:53:47] <_hugo> could be easily implemented[04:53:50] <geist> exactly[04:54:10] <DeadYak> I'm rusty on the specifics of what distinguishes a slab allocator :/[04:54:13] <geist> but I'd still do a simple free list, and only when it runs out or some tick happens you return to the slab[04:54:26] *** Crumbs has quit IRC[04:54:32] <geist> because net buffer allocation and freeing happens *all* the time[04:54:39] <_hugo> geist: yeah, i think i'll do something like that. thanks for the tip[04:54:41] <geist> and sometimes you might want to do it at interrupt context[04:54:50] <geist> which you can't do with a slab[04:55:07] <geist> if the isr wants to refill it's rx queue, for example[04:55:22] <_hugo> right, but the drivers don't handle net_buffers[04:55:34] <_hugo> and we don't have shared memory between drivers and the stack yet (sic)[04:55:43] <geist> or after a tx, it can just return the buffer to the free list[04:55:55] <geist> yes, but they will have to be rewritten[04:55:59] <_hugo> yep[04:56:04] <_hugo> but not for R1 i think[04:56:14] <geist> really? that's silly[04:56:36] <geist> otherwise you more or less 'fix' the driver api[04:56:48] <_hugo> i think the reasoning for that is that most the old binary drivers we'll use the old framework anyway[04:57:10] *** Deej575 has left #haiku[04:57:11] <_hugo> the feeling i got from discussing with axel is that might be moved for post R1[04:57:15] <_hugo> but maybe we'll do it before, dunno[04:57:42] <geist> in case you hadn't figured it out, I always design for the final goal[04:57:55] <_hugo> geist: and it makes sense to me :-)[04:58:06] <_hugo> its just that in this particular case there is a requirement[04:58:12] <geist> i dont like painting myself into corners, so where you can plan ahead (like use the net_buffers in particular ways) it generally pays off[04:58:17] <_hugo> to be compatible with older drivers[04:58:34] <geist> what older drivers? haiku is actually trying for binary compatible with beos nic drivers?[04:58:38] <_hugo> yep[04:58:41] <DeadYak> random curiosity, is there any hardware for which an older driver exists that Haiku doesn't have an src rev?[04:58:42] <geist> no shit[04:58:52] <_hugo> DeadYak: i think there are[04:58:58] <DeadYak> like what?[04:59:00] <geist> i thought driver binary compat was't an option[04:59:05] <DeadYak> ec9xx?[04:59:21] <_hugo> 3com[04:59:24] <_hugo> and stuff like that[04:59:37] <_hugo> we have no drivers for those yet[04:59:41] <geist> that's the part of this project that really annoys me[04:59:50] <geist> i have zero interest whatsoever in beos anymore[05:00:06] <geist> if i have to hack on newos via some beos clone project, well I guess that's my fate[05:00:20] <_hugo> geist: i agree. that was my idea in 2001/2002 as well. but some people feel diferently :-)[05:00:35] <geist> so it goes.[05:01:01] <Yez> for what it is worth, that is only for R1[05:01:18] <Yez> once R1 is done, R2 and beyond take us well into the future![05:01:23] <geist> right..[05:02:28] <DeadYak> R2 and beyond is what worries me more[05:02:53] <MrSun> gah .... now ive driven half drunk for the first time just to save a person from herself .. yey me :([05:03:17] <_hugo> geist: do you live in SF?[05:03:28] <MrSun> wtf does people do awake now ?[05:03:29] <MrSun> :)=[05:03:30] <geist> well, I live in sunnyvale, which is about 40 miles south[05:03:31] <_hugo> if you dont mind me asking :-)[05:03:34] <geist> i work in SF though[05:03:35] <_hugo> ah[05:03:41] <geist> and I keep meaning to move back to SF[05:04:20] *** FiddleStix has joined #haiku[05:04:39] <_hugo> commuting 40 miles doesn't sound very nice[05:04:48] <DeadYak> isn't Apple in Sunnyvale?[05:04:53] <DeadYak> er wait[05:04:54] <DeadYak> Cupertino[05:04:55] <DeadYak> never mind[05:04:56] <geist> i ride the train, which is convienient, but yeah, it's kind of a chore[05:05:04] <geist> yup, and cupertino is next to sunnyvale[05:05:13] <geist> I'm about 5 miles from main campus, which is why i moved down here[05:06:54] <_hugo> so the ipod devel shop is in SF?[05:07:01] <geist> i dont work for apple anymore[05:07:05] <geist> as of about 3 months ago[05:07:06] <DeadYak> oh?[05:07:08] <_hugo> oh[05:07:09] <MrSun> geist, heh .. that sounds like you are after campus girls ;)[05:07:10] <DeadYak> didn't know that[05:07:46] <geist> yup, I work at a game company up in the city[05:07:49] <DeadYak> ooh, which?[05:07:50] <_hugo> ah![05:07:55] <_hugo> right, i completely forgot :-)[05:07:56] <geist> called Secret Level, owned by Sega[05:07:59] <_hugo> nvm all the PS3 chatter :-P[05:08:00] <DeadYak> ah[05:08:05] <DeadYak> Sega do some neat stuff[05:08:08] <geist> exactly, I do mostly PS3 stuffs[05:08:33] <geist> bout to start a big rewrite of some gpu stuff next week to move more of the graphics bits to the SPUs[05:08:36] <geist> which should be fun[05:08:38] <DeadYak> didn't know Sega were doing any PS3 dev[05:08:43] <_hugo> sounds fun indeed[05:08:54] *** phrizek has joined #haiku[05:09:12] <geist> yeah, it's been extremel educational. I didn't know anything about modern GPUs and game development in general[05:10:02] <MrSun> humm, yT people has started working on stuff for haiku insted? :)[05:10:10] <MrSun> are they not allowed to work on haiku itself?[05:11:19] <CIA-17> hugosantos * r20778 /haiku/trunk/src/add-ons/kernel/network/protocols/tcp/ (TCPEndpoint.cpp TCPEndpoint.h): when sending pure ACKs, use an artificial window of 0.[05:12:16] <_hugo> i think axel will bug me if i introduce custom allocation for net_buffers, but what the hell, eheh[05:18:05] <geist> dont think of it as an allocator[05:18:19] <_hugo> well, it is a very simple one[05:18:20] <geist> think of it as a cache[05:18:28] <_hugo> sure, we allocate from the cache[05:18:31] <_hugo> which is pre-allocated[05:18:47] <geist> you dont even need that if you dont want[05:19:01] <geist> just have a lazy cache that doesn't immediately free old net_buffers[05:19:14] <geist> you can even create a timer that periodically frees some of them if it gets over some threshold[05:19:15] <_hugo> thats what i'll do[05:19:18] <_hugo> more or less[05:19:49] <geist> in newos i even had the ports stuff use the net_buffers for transport[05:19:52] <geist> worked pretty well[05:19:59] <MrSun> humm net_buffers contain what? :)[05:20:10] <_hugo> MrSun: network data, packets[05:20:26] <MrSun> k ... but if the trafic is very high and the timer is to slow then ? :)[05:20:46] <_hugo> don't worry about it, we'll put a clone of geist inside haiku releasing the buffers[05:20:55] <geist> it's nt the rate of traffic, it's the total size of all the sockets[05:20:57] <MrSun> ahh ;)[05:21:00] *** MauriceK has quit IRC[05:21:14] <geist> the worst case scenario is you open a ton of sockets, receive a ton of traffic on all of them, and never read[05:21:23] <geist> each one probably has 32 or 64k of data queued up[05:21:30] <_hugo> 64k by default[05:21:36] <geist> and frankly I dont know how most systems deal with it[05:21:37] <MrSun> ohh :)[05:21:47] <geist> except maybe reserve 64k of buffers when you open the socket[05:21:54] <_hugo> i don't see a problem with that[05:22:02] <_hugo> with getting a lot of traffic that is[05:22:07] <geist> or at least guarantee the total pool has enough to hold all of them[05:22:25] <_hugo> if you run out of buffers, you stop receiving[05:22:33] <_hugo> as in generally, not from the cache[05:22:38] <geist> later on when the nic api gets to be sane the nic may hold on to half a meg of buffers or so at a time as well[05:22:52] <geist> say the nic has 256 rx and tx queue entries, etc[05:23:07] <_hugo> right[05:23:18] <geist> and with gigabit it's probably even more[05:23:24] <geist> but that's also easy to solve[05:23:29] <_hugo> i think so, yeah[05:23:37] <geist> when the nic gets upped you can just increase the size of the net pool[05:24:05] <geist> I personally dont think net_buffers should come out of the general allocator directly[05:24:21] <geist> they should be managed as areas that are carved up by the net stack[05:24:26] <_hugo> well, net_buffers allocate two pieces[05:24:36] <_hugo> the data buffers themselves (2K), and the net_buffer control structure[05:24:39] <geist> since it can scale the size of the pool based on the activity, number of open sockets, number of nics, etc[05:24:46] <_hugo> the structure should come from the slab[05:24:47] <geist> same thing, doesn't matter[05:25:01] <_hugo> right now i'll implement the same kind of cache for both[05:25:17] <geist> you can also chop op an area for that as well, but yeah, putting the struct part in the general heap is i guess okay[05:25:28] <geist> except if the net buffers tend to stick around cached, they'll just chop up the heap[05:25:43] <geist> of course a slab allocator is better at that, since you would have the structs off in their own pool[05:25:45] <_hugo> there is also another structure which has the head, tail pointers etc, which is shoved inside the data node (2K)[05:26:02] <geist> yeah, iirc you guys didn't reuse the newos implementation[05:26:10] <_hugo> i have no idea :-)[05:26:29] <geist> i had a somewhat simpler implementation, since I didn't think most of the complexity of the classic mbuf from bsd was all that necessary anymore[05:26:40] <geist> at the expense of perhaps being more wasteful of mem[05:26:58] <_hugo> your buffers were of fixed size?[05:27:02] <geist> yep[05:27:05] <_hugo> 2K?[05:27:06] <geist> 2K iirc[05:27:16] <_hugo> well, you wouldnt be able to support jumbo frames then[05:27:22] <_hugo> without a lot of mem waste[05:27:23] <geist> sure, why not?[05:27:37] <geist> no, in fact it'd be less wasteful than allocating a big buffer[05:27:43] <_hugo> how would you put a single jumbo frame, lets say of 6K, within your structure?[05:27:50] <geist> you dont, you spread it across 3 of them[05:27:57] <_hugo> ah, so its the same[05:28:07] <geist> any gigabit card can do scatter gather[05:28:09] <_hugo> i thought you meant your buffers weren't linkable[05:28:14] <geist> they're *designed* for that[05:28:24] <geist> oh no, it's does all the usual stuff[05:28:37] <geist> what it can't do is have a naked structure pointing to some other storage spot[05:28:38] *** rcjsuen has joined #haiku[05:28:43] <geist> which the BSD mbufs let you do as well[05:28:49] <_hugo> it wasnt the problem of retrieving the frame from the nic but more if you had a single representation[05:29:05] <_hugo> you mean cloning?[05:29:23] <geist> yeah, something like that[05:29:29] <_hugo> right[05:29:31] <geist> i can clone in newos, but it actually copies the data[05:29:51] <_hugo> for the general case that isn't too problematic[05:29:58] <geist> yup[05:30:05] <_hugo> but if you introduce multicast, packet capturing etc, it starts to be[05:30:08] <geist> for the most part it works like a champ[05:30:24] <geist> yeah, that's probably the case[05:30:33] <_hugo> any multiple reader situation for that matter[05:31:18] <_hugo> our stack is capable of cloning but it isn't too smart about it[05:31:20] *** TTRanger has quit IRC[05:31:37] <_hugo> since eventually a protocol handler will do some action which involves doing a deep copy[05:31:42] <geist> yeah that seems like it could get really complicated really fast[05:32:33] <geist> also[05:32:41] <geist> bawlsy bawlsy bop[05:32:44] <geist> tiddy bop[05:32:49] *** Teknomancer has joined #haiku[05:32:56] <_hugo> woohoo[05:33:35] <_hugo> wish i was in paris right now, having a coffee in an obscure street[05:34:07] <_hugo> and get mugged[05:34:17] <geist> yeah, sittin here on my ass watching VH-1 is pretty lame[05:35:15] <_hugo> im hearing the birds singing, but doesnt make much sense since the sun is still a couple hours away[05:47:10] <MrSun> cant tehy just be happy anyways? :P[05:47:34] <_hugo> it isn't raining fish, no reason to be happy[05:47:39] <MrSun> haha[05:47:58] <geist> it's raining men![05:48:04] <_hugo> oh noes![05:48:08] <MrSun> gah[05:48:15] <MrSun> i want it to rain women for a change[05:48:18] <MrSun> but nooo[05:48:21] <MrSun> sigh ... :([05:48:46] <_hugo> that's some weird fetish[05:48:51] <MrSun> :P[05:49:01] <geist> http://www.fukung.net/v/354/103528028_l.jpg[05:49:38] <_hugo> i'm pursued by skeletons disguised with meat![05:49:47] <_hugo> sneaky skeletons[05:50:01] <MrSun> that cant be good[05:52:17] <geist> for you nerds out there[05:52:18] <geist> http://www.fukung.net/v/3754/Monitor%20Arch.jpg[05:52:37] <_hugo> geist: have you seen them all?[05:52:48] <geist> or http://www.fukung.net/v/3395/Fuck%20My%20Bluff.jpg[05:52:49] <_hugo> that person has walls made of monitors and such[05:53:39] <_hugo> nerds would prefer http://images.appleinsider.com/nab-apple-2007-1.jpg i think[05:53:51] <_hugo> or http://images.appleinsider.com/nab-apple-2007-35.jpg[05:58:55] *** umccullough has joined #haiku[05:59:14] <umccullough> geist, frys has lots of my money now ...[05:59:48] <umccullough> i almost bought a core 2 duo mobo/cpu combo...[05:59:57] <umccullough> but opted for the AMD64 X2 5600 combo instead[06:00:06] <_hugo> oh noes![06:00:09] <_hugo> core 2 is so nice[06:01:30] <geist> woot[06:09:12] *** avkig has quit IRC[06:11:19] *** avkig has joined #haiku[06:24:18] <Teknomancer> have a question guys :)[06:24:50] <Teknomancer> can someone tell me what are all the stuff in posix (and not Standard C library) and where its implemented in haiku ?[06:25:01] <Teknomancer> i might have an interview on posix :)[06:25:07] <Teknomancer> well unix c/C++ actually[06:25:19] <Teknomancer> signals, etc are in posix[06:25:44] <Teknomancer> hm, malloc.h is in posix directory in headers, does that mean malloc's posix ?[06:26:19] <pyCube> hi[06:26:29] <Teknomancer> hey pyCube[06:28:13] <JonathanThompson> Greetings earthlings and aliens.[06:28:16] <geist> malloc is std C[06:29:28] <JonathanThompson> Hey Teknomancer, here back from Google yet?[06:29:57] <JonathanThompson> Hear....[06:29:59] * JonathanThompson is tired[06:30:23] <pyCube> i hear google is opening something here soon[06:30:42] <JonathanThompson> Google is everywhere, and it seems Yahoo is similar.[06:31:12] <JonathanThompson> I had to laugh, though, someone sent out email saying goodbye to Yahoo, and it was a gmail account they listed as the way to contact them in the future :P[06:31:52] <pyCube> i wonder how long it'll be before i see a geek fight at lunch between guys with intel id badges and guys with google id badges[06:32:07] <JonathanThompson> Google/Yahoo geek fight :)[06:32:20] <JonathanThompson> That's within the realm of possibility here.[06:32:31] <pyCube> for some reason, the cuban restaurant i go to regularly gives me an "intel discount"[06:32:45] <JonathanThompson> You must look pentium-enabled :P[06:33:10] <Teknomancer> JonathanThompson no not yet, this one is for IBM, google position was not in my town so its getting delayed[06:33:21] * JonathanThompson wonders by comparison why it is several people have offered him a mall discount downtown Bellevue for mall employees...[06:33:44] <pyCube> hehe[06:34:00] <JonathanThompson> Good for 10% off pretzels :P[06:34:36] * JonathanThompson considers looking for a good computer chair this weekend...[06:35:02] <JonathanThompson> My current computer chair isn't exactly designed for the task: an antique wooden dining chair of little value.[06:35:21] <JonathanThompson> I must admit, though, it is structurally very solid for the task.[06:35:46] * _hugo sits in his couch[06:35:59] <JonathanThompson> I was slightly paranoid this afternoon when the site manager called me into talk to me :P[06:36:16] <JonathanThompson> At the end of my 3rd week, and knowing the previous contracted guy was terminated after 4.[06:36:28] <pyCube> um... how much structural integrity is required for your coding techniques?[06:36:29] *** tigerdog has quit IRC[06:36:32] <geist> TERMINATION[06:36:35] <JonathanThompson> :P[06:36:58] <JonathanThompson> The guy required too much hand-holding, and even after 4 weeks, he wasn't able to get the product to build.[06:37:11] <JonathanThompson> And yet he had previous unix experience, and this is linux we're dealing with.[06:37:26] <_hugo> oh no, corporate secrets[06:37:31] <pyCube> double half twist back flip if statement[06:37:34] <JonathanThompson> Not really, _hugo :)[06:38:41] <JonathanThompson> If a job description for Yahoo asks for linux experience and that's posted all over the world, that's hardly a trade secret :P[06:38:48] <JonathanThompson> Or corporate secret.[06:38:59] <_hugo> you have to click to accept the NDA before looking at the job page![06:39:09] <JonathanThompson> It's not that bad :P[06:39:18] * _hugo buys yahoo stock now that they are using linux[06:39:42] <JonathanThompson> Red Hat Enterprise, in fact.[06:39:53] * _hugo sells all his yahoo stock[06:39:55] <JonathanThompson> 4.something.[06:40:33] <JonathanThompson> My main linux box is an athlon 64 bit variant.[06:41:13] <_hugo> umccullough: around?[06:42:38] <umccullough> _hugo, yeah[06:42:53] <_hugo> umccullough: ignore the pressure, but were you able to update your server?[06:42:55] <umccullough> just gettin' the kids ready for bed[06:43:03] <_hugo> ok :-)[06:43:06] <umccullough> _hugo, not yet...i'll start it now :)[06:43:17] <_hugo> no problem, was just wondering[06:43:17] <JonathanThompson> I did learn today that they're expecting a very high level of independence out of me towards accomplishing whatever is assigned.[06:44:10] <umccullough> svn up running now[06:44:13] * JonathanThompson shrugs[06:44:16] <umccullough> bbib[06:47:49] <pyCube> thats good to hear[06:48:07] *** rcjsuen has quit IRC[06:48:16] <JonathanThompson> They're progressing me from simple to more complex over time.[06:48:21] <JonathanThompson> So it should remain interesting.[06:48:56] <JonathanThompson> The product I'm working on isn't as large as the CAD software I was working on for Kubotek, but it still isn't a small thing in terms of complexity.[06:48:57] <Teknomancer> [21:36] <JonathanThompson> At the end of my 3rd week, and knowing the previous contracted guy was terminated after 4. <-- JonathanThompson did u get it to build ???[06:49:16] <JonathanThompson> I diagnosed build errors and how to fix them, Teknomancer :)[06:49:25] <Teknomancer> but i guess u could easily dazzle them with ur TLB and VM knowledge :P[06:49:31] <Teknomancer> JonathanThompson ah cool[06:50:42] <JonathanThompson> As of yet, I've not used any SQL, but have instead written code directly controlling Berkeley DB stuff for part of things.[06:55:25] <pyCube> some sort of orm or something?[06:55:34] <JonathanThompson> ORM?[06:55:46] <pyCube> object relation mapper stuff[06:55:55] <JonathanThompson> Something like that :)[06:56:17] <JonathanThompson> I'm still learning.[06:56:23] <pyCube> berkeleydb eh?[06:56:39] <JonathanThompson> That's just part of things.[06:57:03] <JonathanThompson> They also use Postsgresql (eer, whatever)[06:57:20] <JonathanThompson> Tables may span several storage devices.[06:57:31] <JonathanThompson> Or quite a lot of them, actually.[06:58:09] <Teknomancer> do they use YFS :P[06:58:16] <JonathanThompson> ?[06:58:19] <JonathanThompson> YFS?[06:58:22] <Teknomancer> Yahoo File System ? :P[06:58:28] <Teknomancer> Google uses GFS for that sort of thing[06:58:30] <JonathanThompson> You could compare it to that :)[06:58:37] <Teknomancer> distributed FS with 64 MB block sizes :-P[06:59:00] <Teknomancer> k, must ready the damn documents for the interview , sigh...[06:59:24] <JonathanThompson> What, you don't know everything you'd expect to know off the top of your head, Teknomancer? :)[07:00:02] <Teknomancer> JonathanThompson well no time to prepare for anything now ... i guess i'll just attend it with what i know so far and hope for the best[07:00:13] <_hugo> sigh.. seems berlios is faulty. it was you umccullough![07:00:14] <JonathanThompson> When will you be there?[07:00:17] <JonathanThompson> Monday?[07:00:24] <Teknomancer> JonathanThompson today or tomorrow's the interview :([07:00:27] *** jspiros has quit IRC[07:00:41] <JonathanThompson> Hmmm... I guess they do things differently there :)[07:01:05] <Teknomancer> well i saw this opening in orkut's C++ community[07:01:14] <Teknomancer> when i saw the date it was 21st/22nd .. lol[07:01:31] <JonathanThompson> Of which month?[07:01:42] <Teknomancer> :) April[07:01:58] <JonathanThompson> Have they contacted you yet, or is this at a job fair?[07:02:18] <JonathanThompson> Or a "cattle call"?[07:02:31] <Teknomancer> JonathanThompson its an Employee referral program, and yeah i contact the guy and he sent me the details[07:02:36] * JonathanThompson takes memory allocation for $500, Alex[07:05:27] *** kad77|work has joined #haiku[07:05:37] *** kad77|work is now known as burg77[07:06:16] <burg77> whats hot today?[07:06:29] <JonathanThompson> Tamales, but it'll be cool tomorrow![07:06:44] <burg77> :D[07:07:07] * burg77 had some fresh chimis for supper[07:07:20] * JonathanThompson had 3 tacos for dinner, with hot sauce[07:07:34] <pyCube> enchilada and chile relleno here[07:07:47] <burg77> Anyone working with the haiku office project? or just that one guy?[07:08:01] <JonathanThompson> He's taking volunteers :)[07:08:07] * JonathanThompson nominates burg77[07:08:13] <burg77> sounds like a great idea[07:08:35] <JonathanThompson> It all hit me when I read about it as "Bit Soup" like the story "Stone Soup"[07:08:42] <burg77> yeah, I work about 65 hours a week ATM[07:08:42] <burg77> maybe later :D[07:08:57] * JonathanThompson wonders if burg77 works for Microsoft :P[07:09:02] <burg77> hahaha[07:09:07] <burg77> Medical IT[07:09:08] <JonathanThompson> I know a lot of people around here that work for Microsoft.[07:09:21] <burg77> private clinic group[07:09:23] <geist> isn't that because you live in seattle?[07:09:34] <JonathanThompson> Someone at church commented "Yes, Microsoft has flex time! You can work any 12 hours a day you wish!"[07:09:38] <umccullough> huh, what?[07:09:54] <JonathanThompson> Mercer Island, but I'm in running distance to Seattle over Lake Washington :P[07:09:55] <burg77> anyways, I hope its not OOffice cruft[07:10:06] <geist> well, bellvue[07:10:11] <geist> or redmond, I forget[07:10:12] <JonathanThompson> I moved from Bellevue.[07:10:25] <JonathanThompson> Redmond and Bellevue and Kirkland border each other.[07:10:35] <geist> btw, sister is loving baindbridge is.[07:10:39] <burg77> umc: I'm finally getting cable. $250 a month for 3mb down/ 768k up business class[07:10:40] <geist> bainbridge eve[07:10:43] <geist> even[07:10:46] <geist> damn, can't type today[07:10:59] * JonathanThompson coughs up a hairball aimed at burg77's statement[07:11:18] <pyCube> ACK! it s IT. http://www.best-horror-movies.com/images/it-pennywise-basement.jpg[07:11:21] <burg77> business class for the boonies. $3100 for the install, getting fee waived[07:11:32] <JonathanThompson> $3100????[07:11:39] <burg77> Yep[07:11:40] <pyCube> it's-it's f'ing rule.. btw[07:11:42] <umccullough> _hugo, if berlios is faulty, it's no surprise, geist looked at their storage mechanism quite a while ago and... well, they're lucky that thing is still running ;)[07:11:55] *** jspiros has joined #haiku[07:12:02] <geist> yeah, it was super close to running out of disk all the tim[07:12:11] <geist> looked like a linux box with a single reiserfs volume (/)[07:12:19] * JonathanThompson notes geist should go to bed since he's dropping characters[07:12:21] <umccullough> burg77, i assume you are kad77?[07:12:25] <burg77> Thye tried to tunnel down my drive ($11k), but I bitchslapped them and made them actually use the energy/telecom poles instead. ppl are so stupid[07:12:30] * burg77 is brian, yes[07:13:12] <_hugo> berlios is silly[07:13:21] <_hugo> can't update, can't commit now[07:13:22] <burg77> I have talked to SIX time warner ppl to get this arranged[07:13:53] <umccullough> _hugo, a year ago (or however long) it was so taxed that nobody could commit for days[07:14:00] <burg77> but, my primary employer is paying for my connection. :D :D :D[07:14:53] <geist> oh wow, can't log in right now[07:14:58] <_hugo> why aren't we hosting our own svn server?[07:15:07] <_hugo> seems silly to be using berlios[07:15:08] <burg77> good question[07:15:17] <geist> probably because it'd e pretty expensive[07:15:21] <umccullough> _hugo, waldemar has a trac ticket to at least mirror the repo on the haiku-os.org server[07:15:31] <burg77> lots of ppl have bandwidth[07:15:34] <geist> a single svn co of the entire thing uses up nearly a GB of bandwidth[07:15:37] <_hugo> geist: there are images to download[07:15:50] <umccullough> geist, not with compression though... right?[07:15:55] <burg77> I'm an admin at awkwartv.org, and mactijn had access to a 100Mbit line[07:16:06] <geist> how many bytes/mo?[07:16:15] <burg77> unlimited[07:16:17] *** knoedel has joined #haiku[07:16:30] <burg77> he works at an ISP :D[07:16:40] <geist> is that 'totally unlimited we dont care' or 'unlimited until you start moving a bunch of bytes and people notice'?[07:16:49] <umccullough> either way... there needs to at least be a mirror backup of the entire repo in case something goes horribly wrong ;)[07:16:54] <burg77> http://wiki.awkwardtv.org[07:16:56] <geist> i do think that someone does that[07:17:01] <_hugo> i have a mirror, in my git tree, eheh[07:17:11] <geist> i have a full svk sync of it too :)[07:17:12] <umccullough> _hugo, a mirror of the whole history?[07:17:21] <geist> with all of the history[07:17:22] <_hugo> umccullough: most of it, since r10000 or something[07:17:31] <geist> took me 2 or 3 days to sync it initially[07:17:32] <burg77> giest: he's in the .nl, so I think they don't care[07:17:40] <umccullough> berlios provides archives of the repo...[07:17:41] <_hugo> yeah, it took a long time to download it[07:17:46] <_hugo> i also have a svk one, from 20000 something[07:17:50] <geist> burg77: cause everyone is stoned and doesn't care?[07:17:50] <_hugo> but prefer git[07:18:05] <burg77> giest: if thats your take[07:18:25] <burg77> fiber in europe is quite prevalent[07:18:40] <umccullough> _hugo, i did get 20778 here and built..let me fire up the server[07:18:47] <_hugo> well, northern europe[07:18:53] <_hugo> not all of europe[07:18:53] <geist> again, it's not the speed, it's the bandwidth[07:18:56] <_hugo> umccullough: cool[07:19:00] <geist> but yeah, no doubt your buddy has a good hook up[07:19:16] <JonathanThompson> That is, data transfer in a month, geist?[07:19:17] * burg77 wishes he had it[07:19:24] <geist> right[07:19:33] * JonathanThompson thought bandwidth was an instantaneous measurement...[07:19:44] <geist> sigh. I thought someone would try to get me on that[07:19:47] <geist> you know what i mean![07:19:52] <JonathanThompson> :P[07:20:00] <burg77> 5Mbit down/2Mbit up for me is $550/mo, business QOS[07:20:01] <geist> should know better than to be wrong in front of a bunch of nerds[07:20:10] <geist> $550/mo?[07:20:13] <umccullough> _hugo, server running[07:20:14] <JonathanThompson> Damn straight, geist :)[07:20:14] <geist> holy shit that's expensive[07:20:17] <umccullough> still remember that image name?[07:20:22] <burg77> for hwere I live physically, yes[07:20:31] <_hugo> umccullough: let me see, one sec[07:20:47] <JonathanThompson> That cable service had better NEVER go down, burg77.[07:21:09] <burg77> yeah, too many ppl who don't belive in freedom, free markets, etc. taxes and special interests are key in my area[07:21:16] <_hugo> umccullough: the image name is lost in the ether, could you past it to me in private please[07:21:29] <geist> well, www.berlios.de doesn't even load right now[07:21:29] <umccullough> it's here /Haiku_RobinHood_r20682.png[07:21:34] *** stargater has quit IRC[07:21:46] <umccullough> _hugo, it loads up fast now :)[07:21:51] * burg77 governer is Jim Doyle if you follow politics[07:22:26] <_hugo> i completely lost the url[07:22:29] <_hugo> trying to find it[07:22:40] <geist> guess you're in idaho eh? written your manifesto yet?[07:22:41] <umccullough> hang on[07:23:12] <burg77> jonathon: that'll be three months credit if it does, my primary employer has a killer attorney. :D[07:23:17] <_hugo> umccullough: found it[07:23:22] <umccullough> oh ok[07:23:42] <burg77> geist: was that directed at me?[07:23:58] <JonathanThompson> Hwy burg77, were you aware you can use tab-completion?[07:24:02] <geist> sort of, but it was pretty half-assed[07:24:13] <burg77> jon: not in this client :D :D :D :D[07:24:14] <umccullough> yes, i was gonna say nick-completion too...[07:24:21] <_hugo> umccullough: its better, but not quite there yet[07:24:22] <umccullough> burg77, chatzilla does support it[07:24:29] * JonathanThompson feels sorry for over-charged, under-IRC-cliented burg77.[07:25:05] <geist> can't get this gosh darned linux box to fail anymore[07:25:08] <umccullough> _hugo, on a windows box, i recommend HydraIRC[07:25:10] <burg77> ok, then my my wine consumption is informing me I don't care. I assume you are all probably above 140 IQ[07:25:13] <_hugo> i need to set an upper limit on the retransmission timer, eheh, it is probably being kicked to seconds[07:25:20] <geist> burg77: mmm, wine[07:25:21] <_hugo> umccullough: no, the image download :-)[07:25:31] <umccullough> er[07:25:39] <umccullough> my bad, that message was for burg77[07:25:40] * JonathanThompson wonders if burg77 is drinking WINE or wine...[07:25:50] <burg77> thats vino[07:25:59] <umccullough> burg77, on a windows box I recommend HydraIRC :P[07:26:02] * JonathanThompson wonders if there are extra bits dribbling down burg77's face :)[07:26:33] <burg77> wonderful day working for the technically illiterate[07:26:40] <umccullough> speaking of alcohol...[07:26:44] * umccullough goes to get another beer[07:26:52] <burg77> Hydra, yeah I have taht..[07:26:55] <_hugo> umccullough: how much upload do you have?[07:27:08] <umccullough> 144kbps[07:27:11] <umccullough> 14kb/sec[07:27:21] <umccullough> my wife is using the internet...[07:27:28] <_hugo> just wondering.[07:27:30] <_hugo> thats even better[07:27:33] <_hugo> more congestion[07:28:08] <JonathanThompson> Better torture test :P[07:28:19] * JonathanThompson slings more bits at umccullough[07:28:27] *** knoedel has quit IRC[07:28:32] <umccullough> :)[07:28:33] <umccullough> :([07:28:38] <umccullough> my poor IDSL line[07:28:48] <JonathanThompson> At least it feels useful :)[07:28:53] * umccullough notes that JT still has credit on the haiku webpage[07:28:56] <_hugo> it is doable to download the image now, but since i don't have an upper limit in the retransmission timeout, due to all the congestion it takes a while[07:28:59] <umccullough> haiku webserver page that is[07:29:05] <JonathanThompson> I do?[07:29:16] <JonathanThompson> Oh.[07:29:23] <umccullough> http://binarychicken.com:575/[07:29:39] *** kad77 has joined #haiku[07:29:40] *** mmu_man has joined #haiku[07:29:41] *** ChanServ sets mode: +o mmu_man[07:29:46] <JonathanThompson> Ah, now you just need to spell my first name correctly :)[07:29:52] <kad77> yay. hydra[07:29:53] <geist> oh neat, the google homepage is nice[07:30:02] <geist> the new personalized thing that is[07:30:03] <umccullough> oh[07:30:10] <JonathanThompson> That sill being served by Haiku, umccullough?[07:30:12] * umccullough goes to fix it[07:30:17] <umccullough> yes[07:30:18] <kad77> ok, so how can I meet everyones standards more precisely now? :D :D[07:30:18] <JonathanThompson> (I guess it says that)[07:30:50] <JonathanThompson> Become a mind-reader, and (even better!) a mind-writer, so you can tell them what their standards are, kad77 :P[07:31:00] <kad77> dman[07:31:09] <umccullough> JonathanThompson, fixed[07:31:09] <kad77> oh, sorry "DAMN"[07:31:16] <JonathanThompson> Thanks, umccullough :)[07:31:23] <umccullough> :)[07:31:29] <JonathanThompson> I bet you weren't expecting such an answer, kad77 :P[07:31:36] <kad77> Its hard to spell when you've been drinking chardonney for 3 hours'[07:31:52] <JonathanThompson> I wouldn't know from anything remotely like that experience...[07:31:59] <umccullough> well, its hard to read what is spelled anyway[07:32:10] <kad77> words?[07:32:12] <umccullough> usually the person spelling at that point isn't having any problems ;)[07:32:14] * JonathanThompson wonders if he can torture umccullough's webserver into death by rapidly refreshing.[07:32:23] * kad77 tries[07:32:25] <umccullough> JonathanThompson, go for it[07:32:32] <kad77> jk[07:32:36] <umccullough> several have already done wget loops on it ;)[07:32:57] <umccullough> honestly, it's getting a lot more stable lately[07:33:04] <geist> what port is it? is the port 80 thing it?[07:33:07] <JonathanThompson> Then I suppose I'm not going to have much hopes of it unless there's some memory/resource leak that causes it to fail unexpectedly :P[07:33:20] <umccullough> not port 80[07:33:22] <umccullough> port 575[07:33:30] <JonathanThompson> It may help if a bunch of people from different locations hit it all at once.[07:33:32] <umccullough> looks dead now[07:33:36] <geist> alas, that does not work[07:33:37] <kad77> umc: are you still offended about how I expressed my option re: one gsoc project the other day?[07:34:10] *** kr1stof has joined #haiku[07:34:13] * JonathanThompson wonders if he killed umccullough's webserver :)[07:34:15] <umccullough> kad77, really i don't care much[07:34:22] <kad77> thank god.[07:34:43] <umccullough> JonathanThompson, let me restart the daemon[07:34:43] <kad77> that andre fella actually compared me to the VT killer[07:34:50] <JonathanThompson> Of course, maybe I just killed my instance of SeaMonkey :P[07:34:51] <umccullough> kad77, that's stupid[07:35:02] <kad77> I was about ready to leave this organization is thats the kind of ppl who frequent here[07:35:27] <kad77> well, he sounds intelligent, with all his scheduler talk to the list.[07:35:35] <kad77> but he is a child apparently[07:35:50] <umccullough> kad77, really - it's not worth getting upset over[07:35:54] <JonathanThompson> Some people are old on the outisde, but still soft and gooey on the inside :)[07:36:00] <kad77> i guess[07:36:06] <umccullough> it's just IRC...[07:36:26] <umccullough> JonathanThompson, server appears to be in a bad way[07:36:29] <JonathanThompson> Besides, remember this: developers need to have a certain amount of ego to keep from going nuts and getting all dejected.[07:36:37] <kad77> yeah, its been a hell of a week. I'm probably overreacting[07:36:38] <JonathanThompson> Yes, it does, umccullough :)[07:36:41] <_hugo> linux seems silly sometimes.[07:36:48] <_hugo> looking at capture from trying to get the picture[07:36:56] <JonathanThompson> Life seems silly sometimes, so I'm not going to get too concerned, _hugo :)[07:37:02] <umccullough> _hugo, have you been pulling that pic this whole time?[07:37:11] <_hugo> umccullough: nah, i tried a couple of times[07:37:12] <umccullough> heh... cuz the server is toast :)[07:37:21] <JonathanThompson> It must be me, umccullough :)[07:37:22] <_hugo> umccullough: hanged?[07:37:22] <umccullough> 1[07:37:22] <kad77> I had one of my bosses wifes try to get me fired be cause she couldn't log in due to our 90 day password policy[07:37:23] <umccullough> oops[07:37:35] <umccullough> _hugo, well, just not responding[07:37:36] <kad77> I guess I am just andgry at dumb fucking humans. :D[07:37:43] <umccullough> haiku doesn't show any signs, let me try to ping something[07:37:51] <umccullough> yeah, pings out[07:37:54] <geist> this is getting me all pumped about finishing the lwip port to my little embedded os[07:37:56] <umccullough> pings to google.com without issue[07:37:57] <_hugo> umccullough: so maybe RH is dead?[07:38:16] <umccullough> _hugo, yes, that's what i thought too, but after restarting the daemon it was still toast :/[07:38:18] <JonathanThompson> Interesting observation, umccullough.[07:38:20] <umccullough> let me restart it again[07:38:24] <_hugo> umccullough: try to wget something, wget google.com[07:38:29] <umccullough> k[07:38:38] <umccullough> yup[07:38:41] <umccullough> fine :)[07:38:53] <JonathanThompson> Perhaps the stack isn't flushing certain things out?[07:38:59] <_hugo> umccullough: stop RH and run netstat[07:39:12] <umccullough> oh i restarted RH and it's fine now[07:39:17] <_hugo> ah[07:39:20] <umccullough> maybe JT was hammering it when i restarted it before[07:39:28] <JonathanThompson> Could be :P[07:40:08] <_hugo> anywho, if linux actually behaved as it is expected, i should be able to download the picture[07:40:11] * JonathanThompson needs to create business cards that has his name on it, and states "Tortures computers for fun and profit"[07:41:21] <JonathanThompson> It does seem like I'm able to effectively kill the server, umccullough.[07:41:31] <_hugo> JonathanThompson: what are you doing?[07:41:39] <umccullough> oh dead again?[07:41:45] <JonathanThompson> I wonder if perhaps I'm closer to you ping-time=wise than the others?[07:41:49] <JonathanThompson> Seems like it.[07:41:55] <umccullough> JonathanThompson, my PoP is in seattle ;)[07:42:03] <JonathanThompson> :)[07:42:06] <umccullough> speakeasy[07:42:09] <JonathanThompson> Ah.[07:42:25] <JonathanThompson> Should be limited mostly by your IDSL line interface, then :P[07:42:30] <umccullough> my first hop is 50ms :P[07:42:38] <JonathanThompson> Perhaps that makes me best-suited to torment your server :P[07:42:44] <umccullough> probably[07:42:51] * JonathanThompson suspects umccullough's ISP server is <10 miles away[07:42:54] <umccullough> other than myself :D[07:42:55] <_hugo> i think ill add lighttpd to tests/, RH seems a bit cranky[07:43:04] <umccullough> _hugo, sounds good ;)[07:43:18] <geist> also, do we have netcat in da build?[07:43:27] <_hugo> geist: we do, but tty is weirdo[07:43:44] <umccullough> _hugo, maybe that should just be included in the image by default anyway - i mean an httpd is a good thing to have right?[07:43:54] <JonathanThompson> From here in BeOS 5.03 no BONE, I get 92-93 ms pings for binarychicken.com[07:43:55] <_hugo> im also sitting in a mtr port, which ill eventually add as well[07:44:15] <_hugo> umccullough: eheh, cant make those kind of political decisions :-P[07:44:32] <JonathanThompson> Is your Haiku server IP address 66.92.192.159, umccullough?[07:44:37] <umccullough> _hugo, well, you mean you can't make those kinds of decisions without getting crap on the commit list ;)[07:44:45] <umccullough> JonathanThompson, yes[07:44:51] <_hugo> umccullough: which i prefer :-)[07:44:54] <umccullough> JonathanThompson, that's my router[07:44:59] <JonathanThompson> I'm getting some irregular ping times :)[07:45:02] <umccullough> i use NAT to distribute the ports[07:45:02] <kad77> so, what are the top three things that need attention in the current haiku network stack?[07:45:04] <_hugo> umccullough: ill make it easy for you to add it to your build[07:45:13] <umccullough> JonathanThompson, probably cuz my wife is surfing the net[07:45:18] <JonathanThompson> Ah.[07:45:27] <_hugo> kad77: my lower back[07:45:29] <umccullough> and people are probably hammering my connection now ;)[07:45:40] <JonathanThompson> I'm in a ping loop to it.[07:45:49] <kad77> hugo: <unparseable>[07:45:55] <geist> pinging is probably hitting his router[07:46:08] <umccullough> yeah, ping is my router[07:46:13] <JonathanThompson> Fooey, then :P[07:46:20] <umccullough> port 80 is an apache server on a different box[07:46:29] <JonathanThompson> But it is still slow as molasses, or not getting data through from the webserver.[07:46:30] <geist> i'm all telnetting to 575[07:46:31] <geist> but alas[07:46:35] <umccullough> i have the ports spread across some 4 machines now[07:46:36] *** burg77 has quit IRC[07:47:00] <JonathanThompson> SeaMonkey/Firefox has enough bugs I'm not sure if it is one of those, or your system, umccullough.[07:47:09] <JonathanThompson> But when I do a refresh, I get nothing here now.[07:47:10] * umccullough thinks geist is trying to root my haiku box ;)[07:47:36] <kad77> axle ever hear back from coverity?[07:47:38] * JonathanThompson gets out NetPositive for a test of that theory[07:47:39] <geist> no, I'm trying hard to convince you that I haven't already rooted your box by pretending to still try[07:47:57] <umccullough> i'll restart rhdaemon[07:47:58] * _hugo adds a "debug" tcp option[07:48:11] <geist> i wrote the kernel, right? I should have a few backdoors[07:48:15] <umccullough> page loads again[07:48:39] <JonathanThompson> Hmmm... did you restart it, umccullough?[07:48:47] <umccullough> _hugo, there's probalby some bug in Haiku that causes RH to die :P[07:48:54] <umccullough> JonathanThompson, yes[07:49:07] <_hugo> probably there is[07:49:14] <JonathanThompson> Oh, darn, you changed the conditions on me before I got to verify if it was a SeaMonkey caching bug :P[07:49:26] <umccullough> i don't have the ambition or know-how to track it down...[07:49:48] <geist> hmm, maybe there's another FIN state machine problem here[07:49:55] <geist> try telnetting to it, hitting a key and enter[07:50:02] * kad77 feels umccullough has revealed why linux is where it is. :D[07:50:09] <JonathanThompson> What's funny is that Net+ seems to more correctly display the page than SeaMonkey for the counter.[07:50:12] <geist> note that the haiku box sends a FIN, but since the linux box still keeps it open[07:50:21] <geist> haiku keeps resending the FIN on retransmit[07:50:26] <geist> though linux seems to have acked it[07:50:33] <geist> it should just be a half open connection, right?[07:50:34] <_hugo> ah[07:50:45] <_hugo> it should not retransmit if the FIN was acked[07:50:57] <_hugo> let me check[07:51:04] * umccullough is glad to simply provide a testing ground for people who like to kill servers[07:51:20] <JonathanThompson> Everyone needs a hobby, umccullough :P[07:51:26] <kad77> run IIS6[07:51:26] * JonathanThompson gets out Server Gun[07:52:05] <JonathanThompson> I think either umccullough changed data, or I uncovered a bizarre Net+ bug :P[07:52:40] <geist> http://tkgeisel.com/stuff/RHtrace2.txt[07:52:43] <umccullough> changed data?[07:52:45] * kad77 had jerkwads from 3 countries try IIS6 exploits on a Exchange webmail addy today[07:52:58] <umccullough> looks toasted again[07:53:01] <JonathanThompson> I now have "[07:53:01] <JonathanThompson> HTTP/1.1 200 OK Date: Thu, 19 Apr 2007 20:51:15 GMT Last-Modified: Thu, 19 Apr 2007 20:30:46 GMT Server: RobinHood[07:53:02] <JonathanThompson> Accept-Ranges: bytes Content-Length: 1446 Content-Type: text/htm[07:53:08] <JonathanThompson> displayed at the top with Net+[07:53:31] <geist> hmm, might be false alarm[07:53:32] <umccullough> well, it died[07:53:36] <geist> it looks like linux isn't hacking the FIN[07:53:37] <geist> wonder why[07:53:39] *** kad77 is now known as kad77|log[07:53:45] <umccullough> oh, working again..[07:53:48] <_hugo> geist: yes, linux is weird[07:53:51] <kad77|log> well, have fun, happy hacking all.[07:53:54] <kad77|log> bye now[07:53:57] <umccullough> bye kad77|log[07:53:58] <_hugo> geist: this happens with data as well[07:54:08] <geist> does it not ack the FIN if the program hasn't read it all from the socket yet?[07:54:11] <_hugo> geist: i was downloading a big file from umccullough[07:54:24] <_hugo> and linux kept sending ACKs to segments that were already received[07:54:35] <JonathanThompson> Odd: now Net+ doesn't show the hits counter :P[07:54:36] <_hugo> i mean, to a sequence number already received[07:54:43] <_hugo> this is similiar[07:54:43] <geist> yeah, that's odd[07:54:44] <umccullough> JonathanThompson, the hit counter is javascript[07:54:50] <JonathanThompson> Ah...[07:54:57] <umccullough> 3rd party stats[07:55:04] <JonathanThompson> but it was updating (I think!) in Net+ and showing graphics.[07:55:18] <JonathanThompson> I thought it was perhaps php, having not bothered to look at the code.[07:55:22] <umccullough> hmm? no graphics on that page..[07:55:27] <_hugo> geist: it should ack the FIN when receiving it[07:55:41] <_hugo> geist: the app should just read() 0 bytes when the receive buffer runs out[07:55:44] <JonathanThompson> Funny thing is that the hit counter looked like a car odometer earlier under Net+.[07:55:51] <geist> yeah[07:55:59] <umccullough> JonathanThompson, odd...maybe the stats site sent the wrong stuff[07:56:00] <geist> there must be something you're doing wrong, but i dont know what it is[07:56:04] <JonathanThompson> :P[07:56:30] <_hugo> geist: this seems correct to me. in fact, i've seen haiku retransmit later and linux still acking the old sequence[07:56:41] <JonathanThompson> Nothing like having things displayed you can't account for existing to make bug reports entertainig :P[07:56:54] <geist> right, but there must be some reason they're rejecting it[07:57:06] <geist> i should note that i eventually get the data on the console[07:57:20] <umccullough> JonathanThompson, the default was to show a graphic, but i changed the default to plaintext for simplicity...maybe the default got used[07:57:36] <JonathanThompson> Apparently, but when did you change it?[07:57:46] <umccullough> JonathanThompson, when i set up the stat counter initially[07:57:49] <_hugo> geist: i dont see any reason to reject it (per RFCs)[07:57:50] <umccullough> days ago[07:57:56] <JonathanThompson> Very odd, umccullough.[07:58:11] <geist> nor do i, unless it's something not on the trace[07:58:44] <_hugo> geist: in your dump, its even neat that haiku first pushes 108 bytes with PSH, then sends FIN, then on retransmit pack them together, which is perfectly fine[07:58:57] * JonathanThompson wonders if umccullough's better half realizes her experience is being interrupted by someone purposely hitting "Refresh" for a test[07:59:00] <_hugo> linux keeps ACKing seq 1[07:59:22] <umccullough> JonathanThompson, fortunately she's using the downstream while you're mostly affecting the upstream ;)[07:59:23] <geist> it still does send the extraneous FP at the end, which linux RSTs[07:59:47] <_hugo> yes, indeed[08:00:03] <geist> why do you think linux's win is so small?[08:00:37] <geist> oh hey, wait a sec[08:00:38] <_hugo> in fact that last FP makes little sense, even the ack sequence is wrong[08:00:51] <geist> there are only 16 bits in the win field, right?[08:00:54] <_hugo> right[08:01:08] <_hugo> but linux is using window scaling[08:01:13] <_hugo> so thats 183 << 5[08:01:14] <geist> except we reject it[08:01:17] <pyCube> hehe.. stupid kids and their finding it fun to hit refresh as fast as they can on complex pages[08:01:18] <geist> right?[08:01:25] <_hugo> reject what?[08:01:33] <geist> but anyway, never mind. I was reading our 'win' as 65536, which doesn't make sense[08:01:52] <geist> i forget how the option works in the initial negotiation[08:01:53] <_hugo> our win is 65535, wscale 0[08:01:57] <_hugo> ah ok[08:02:02] <geist> you have to accept their escale?[08:02:05] <geist> wscale even[08:02:16] <_hugo> if i ack with wscale as well[08:02:18] <_hugo> i accept it[08:02:23] <_hugo> all values are acceptable though[08:02:30] <_hugo> as long as i understand window scaling[08:02:34] <geist> ah, okay. so if you dont ack with your own wscale, it's assume you don;t know about it[08:02:37] <geist> got it[08:02:39] <_hugo> yep[08:02:54] <geist> same with the other way around I guess. if the other side doesn't start with wscale, you can't use it?[08:03:01] <_hugo> right[08:03:34] <_hugo> sack permitted works like that as well, see that linux sends sackOK but we never send sackOK, so linux doesn't use SACK[08:04:01] <umccullough> pyCube, my page is hardly complex ;)[08:04:12] <geist> wait a sec. I'm not sure linux even considered this connected[08:04:21] <geist> notice that it duplicate acks 1 even before the first fin[08:04:30] <_hugo> ack 1 <- connected[08:04:35] <_hugo> acked our syn[08:04:36] <geist> swure, but it does it twice[08:04:36] <umccullough> but it does appear to be down again :P[08:04:56] <geist> they send three bytes, we ack it, they ack our SYN *again*[08:05:11] <_hugo> thats permitted[08:05:18] <geist> that seems weird[08:06:18] <_hugo> well, tcpdump doesn't show the sequence number[08:06:20] <_hugo> in pure acks[08:06:29] <geist> it shows the initial one on the SYNc[08:06:32] <_hugo> it should be 4 in the following ones[08:06:38] <geist> SYN, it's assumed they incremental from that[08:06:39] <_hugo> 4 as in .. initial + 4[08:07:22] <geist> they even retransmit 1:4[08:07:41] <_hugo> which is plain stupid considering our ack[08:07:55] <geist> right, which is what makes me believe something is seriously screwed[08:08:00] <_hugo> and it was 200ms later too, so plenty of time to process the ack[08:08:09] <geist> also, I'm gonna delete the bottom two lines, they're form a previous session[08:09:41] <_hugo> geist: could you retry this test but with window scaling disabled in linux?[08:09:47] *** pyCube has quit IRC[08:09:48] <_hugo> its net.ipv4.tcp_adv_win_scale i think[08:09:51] <geist> kay[08:09:53] <_hugo> the sysctl to change[08:11:14] *** ChrisHJW has joined #haiku[08:11:42] <_hugo> well, no, i changed that and it still uses window scaling[08:12:47] <_hugo> ah, net.ipv4.tcp_window_scaling[08:13:53] <geist> kay[08:14:08] <geist> though as a little more data I can reproduce it only after about the 2nd or 3rd try in a row[08:14:23] <_hugo> eheh.. this is so silly.. 9 dup acks from linux, concerning seq X[08:14:35] <_hugo> and im receiving the segments starting in seq X[08:14:37] <_hugo> a couple of times[08:14:38] <geist> you're seeing it too eh?[08:14:43] <_hugo> yeah[08:15:00] <mmu_man> already someone trying to root Haiku ? ;)[08:15:15] <JonathanThompson> Hey, it's a test :)[08:15:28] * umccullough doesn't mind if his Haiku box gets rooted[08:15:35] <umccullough> baron'd even[08:15:47] <JonathanThompson> Heck, it's still probably unstable enough it does it to itself sooner or later at random.[08:16:08] <geist> i'm having much better luck now[08:16:13] <umccullough> heh, i DARE you to root my box ... just so you can be frustrated as it KDLs[08:16:20] <JonathanThompson> lol[08:16:35] <geist> probablems again when i turn wscale back on[08:16:45] <JonathanThompson> "So secure, once you try something nasty, it disconnects instantly!"[08:16:59] <_hugo> geist: http://fivebits.net/~hugo/dump.txt[08:17:09] <_hugo> geist: linux keeps acking 4537[08:17:20] <_hugo> and haiku sends it a couple of times[08:17:39] * JonathanThompson thinks that text output reads a bit like the Martians script in "Mars Attacks!"[08:18:03] <geist> no fast retransmit I see?[08:18:04] <umccullough> heh, nop nop![08:18:31] <_hugo> geist: there is fast retransmit[08:18:50] <umccullough> that's a recent addition ;)[08:19:09] <geist> if it was you'd send the frame again the first tie you get that duplicate ack[08:19:24] <geist> though there is lag of course, but they were like 100ms away from each other[08:19:33] <geist> though i forget you're in brazil, so your link may be pretty bad :)[08:19:38] <umccullough> 100ms is probably less than my latency :P[08:19:46] *** pyCube has joined #haiku[08:19:47] <_hugo> im not in brazil :-P[08:19:55] <pyCube> me neither[08:19:56] <umccullough> Portugal![08:19:59] <JonathanThompson> I get about 91-93 ms pings for umccullough's router from here for good pings.[08:20:01] <_hugo> anyway, fast retransmit starts in the third received duplicate ack[08:20:17] <_hugo> a bunch of these acks might not actually reach the other side[08:20:40] <geist> yeah, anyway the only explanation for this is it's super lossy on his side[08:20:47] <geist> might be good to test against anpther haiku box.[08:20:52] * JonathanThompson starts singing a Jim Morrison mutation: " Ack on through to the other side!"[08:20:53] <geist> say, one right next to yours :)[08:20:57] <umccullough> heh[08:21:01] <umccullough> vmware[08:21:24] <JonathanThompson> If two Haiku boxes can't talk to each other, that's definitely a bug.[08:21:24] <umccullough> the rtl8139 driver might very well sucks[08:21:25] <umccullough> suck[08:21:26] <_hugo> geist: i'm getting a lot of bizzare behaviour like this with linux here[08:22:16] <geist> well, one *supposes* that linux's stack isn't screwy like that[08:22:41] <_hugo> i would expect so as well, but find it strange nonetheless[08:22:48] <JonathanThompson> Hah, and I had a linux box at Coinstar that wouldn't synch with DHCP for hours :)[08:23:27] <geist> okay, trying from OSX[08:23:36] <geist> totally different' stack[08:23:44] <geist> no way it would act the same[08:23:49] <_hugo> right[08:23:59] <_hugo> curious about the results[08:24:41] * JonathanThompson wonders if anyone has described a networked box as being "Totally stacked"[08:25:27] <geist> sam behavior[08:25:39] <_hugo> hm[08:25:42] <geist> getting the trace[08:26:31] <geist> http://tkgeisel.com/stuff/haiku-osx.txt[08:26:32] <_hugo> im only seeing something wrong, but which is acceptable, in the trace[08:26:42] <_hugo> is that haiku is sending a frame with 12 bytes[08:26:56] <_hugo> thats because SMSS is 1460, but 12 bytes are being consumed by the timestamp option[08:27:18] <_hugo> the 12 bytes segment should not be sent by us as to prevent SWS[08:27:40] <_hugo> maybe linux is rejecting the segments in fear of SWS, but doesnt make much sense[08:28:03] <_hugo> and this was in my trace, not yours[08:29:16] <_hugo> that ack 1 after receing the segment is killing me :-)[08:29:20] <geist> refresh it[08:29:33] <geist> i pasted a successful OSX -> linux http transfer of the same thing[08:30:29] <_hugo> seems the same to me (ignoring the FINs)[08:32:08] <_hugo> ok, there is a possibility that we dont see here[08:32:35] <_hugo> . ack 4 from haiku <- maybe this includes a seq number > initial + 1[08:32:44] <_hugo> that would explain the ack 1 from osx/linux[08:32:59] <_hugo> as in "you are using a newer sequence number, but we are missing data"[08:33:10] <geist> hmm[08:33:29] *** phrizek has quit IRC[08:34:44] <_hugo> i dont see on my traces though, the sequence numbers are fine[08:34:55] <_hugo> *dont see it[08:35:27] *** llama233 has joined #haiku[08:38:26] *** jsgotangco has joined #haiku[08:41:53] <geist> i assume the timestamps are merely there for mtt purposes?[08:42:03] <geist> it seems haiku's time stamps are all over the place[08:42:15] <geist> could that be messing up the other side's RTT calculations?[08:42:24] <_hugo> its not really necessary but our timestamps are in network order[08:42:27] <_hugo> no[08:42:37] <_hugo> our timestamp is opaque to them[08:42:40] <_hugo> and vice versa[08:44:05] <geist> it basically acts like on those traces that OSX/linux is flat out rejecting the full contents of the packets from haiku after the SYN, up until some point at which it starts paying attention[08:44:16] <geist> it doesn't even accept the ACK 4 from haiku, it eventuall retransmits 1:3[08:44:21] <geist> 1:4 that is[08:44:56] <_hugo> right[08:45:05] <_hugo> the checksums are correct as well[08:45:13] *** Lelldorin1 has joined #haiku[08:45:53] <geist> maybe double check the length fields and other such things are correct?[08:46:01] <geist> maybe they're wrong when it's below some certain size or whatnot[08:46:35] <geist> something that tcpdump isn't displaying[08:47:14] <_hugo> i use wireshark and i've checked all fields and they are correct[08:47:33] <geist> i'm gonna trace with netmon[08:49:37] <_hugo> between haikus it works much better[08:50:16] <geist> seems to work fine between windows[08:51:23] <_hugo> if this continues i'll eventually dig up linux to discover why it is rejecting those packets[08:51:24] <geist> windows isn't using timestamps, I should notice[08:51:38] <geist> know the sysctl to disable that offhand?[08:51:55] <_hugo> net.ipv4.tcp_timestamps[08:52:54] <_hugo> ugh[08:53:03] <_hugo> it actually works[08:53:13] <geist> no timestamps?[08:53:16] <_hugo> yes[08:53:21] <_hugo> downloaded the big image[08:53:22] <_hugo> with no problems[08:53:25] <geist> hmm, i betcha they're actually paying attention to it[08:53:27] <_hugo> 13.53KB/s[08:53:57] <geist> they see one of your timestamps go in some weird order or something and it rejects the packet[08:54:10] <_hugo> possibly[08:54:21] <_hugo> but how does it know the order?[08:54:33] <_hugo> it doesnt know the endianess of the system[08:54:42] <geist> and it's supposed to be opaque eh?[08:54:48] <geist> what are the meaning of the two different numbers?[08:55:00] <_hugo> i think it should be opaque, ill recheck the standard[08:55:01] <geist> are they up to you, or are you supposed to mirror one of theirs or something?[08:55:10] <_hugo> i mirror the received one[08:55:16] <_hugo> send (mine, theirs)[08:57:10] * geist double checks the mirroring is correct in the trace[08:58:48] <geist> weird, in my trace at least they mirror an old value on their first duplicate ack of 1[08:59:31] <_hugo> im rechecking the standard[08:59:50] *** emitrax has joined #haiku[08:59:59] <_hugo> its never mentioned that one should pay attention to a received TSval[09:00:03] <_hugo> only mirror it[09:03:57] <geist> looking at linux's source, they definitely seem to do a ntohl on both values[09:04:33] <geist> oh wait, no. never mind[09:04:44] <geist> oh wait, yeah, mind.[09:04:47] <_hugo> i just checked that i was doing ntohl/htonl twice on our value, so it was not actually in network order in the segment[09:05:03] <_hugo> i have a commit to fix that, but cant access berlio[09:05:05] <_hugo> berlios[09:05:31] <geist> yeah, it's definitely all screwed up n the wire, because we must still have it LE[09:05:37] <geist> though it should be opaque[09:06:00] *** BNickName has joined #haiku[09:06:03] <_hugo> umccullough: around?[09:06:22] * _hugo kicks berlios[09:06:37] <_hugo> geist: it might very well be this[09:06:52] <_hugo> maybe linux/osx look at timestamps for wrapped segments, or too old/too new[09:06:54] <_hugo> something like that[09:07:00] <geist> yup[09:07:25] <geist> I was just looking at their code, they definitely do a ntohl on *both* values and squirrel them away[09:07:36] <geist> which implies they might actually look at them[09:07:39] <_hugo> right[09:07:53] <geist> maybe it's asssumed that if your timestamp decrements then it must be out of order[09:07:58] <_hugo> maybe they are opaque in terms of clock resolution[09:08:18] <_hugo> "virtual timestamp clock" that is[09:08:21] <geist> though in our case they dont actually decrement[09:08:26] <_hugo> right[09:08:49] <_hugo> anywho, i will test with this fix in[09:08:50] <geist> in this particular case it's 222182144 -> 3108059904 -> 3141614336 -> 3158391552 -> 4098439936[09:09:35] <_hugo> running firefox makes me feel like i have 512mb of ram all over again[09:10:03] <geist> thought they (in this case) finally seem to accept our 913586944[09:10:17] <geist> oh duh, that's why[09:10:40] <geist> well, no, but it's much closer to our starting number (222182144)[09:10:43] <_hugo> connecting to a linux ftp feels faster in haiku now[09:10:59] <_hugo> after i changed the ntohl/htonl timestamps[09:11:14] <_hugo> let me try an upload[09:12:30] *** Ingenu has joined #Haiku[09:14:11] <geist> ah ha[09:14:24] <geist> read about PAWS[09:14:38] <geist> 'protect against wrapped sequence numbers'[09:15:02] <geist> uses the timestamps to decide if the sequence hasn't wrapped[09:15:03] <_hugo> yeah, i know PAWS, but i still havent read how it works[09:15:06] <_hugo> ah[09:15:10] <_hugo> nice..[09:15:19] <geist> http://www.freesoft.org/CIE/RFC/1323/13.htm[09:15:47] <_hugo> However, the receiver treats the timestamp as simply a monotone- increasing serial number, without any necessary connection to its clock. From the receiver's viewpoint, the timestamp is acting as a logical extension of the high-order bits of the sequence number.[09:16:04] <_hugo> the resolution is irrelevant, just needs to be monotone[09:16:10] <geist> yeah, it says it has to tick at least once per 4GB[09:16:12] <_hugo> *monotone increasing[09:16:19] <geist> "The timestamp clock must not be "too fast".[09:16:20] <geist> Its recycling time must be greater than MSL seconds. Since the clock (timestamp) is 32 bits and the worst-case MSL is 255 seconds, the maximum acceptable clock frequency is one tick every 59 ns."[09:16:40] <_hugo> iirc im using system_time() / 1024[09:16:59] <geist> sure but since it was the wrong endian, it appeared to jump ahead in time by large ticks[09:17:11] <_hugo> yeah, was just considering the frequency, which is ok[09:17:14] <geist> which is what caused it to be rejected until the timestamp got into a more acceptable[09:17:24] <_hugo> the problem was indeed the endianess[09:17:47] <geist> well, cool beans[09:18:00] <geist> i didn't know anything about PAWS[09:18:10] <_hugo> at least we figured this one out, eheh[09:18:57] <geist> system_time() / 1024 should be 1ms, which is pretty much what they recommend here[09:19:16] <geist> well ~1ms[09:19:19] <_hugo> yep, ~1ms[09:19:35] <_hugo> i considered it reasonable, but was thinking about decreasing it eventually[09:20:40] <_hugo> i thought 1ms was a good compromise considering the possible warp around[09:20:59] <_hugo> but didnt read much of rfc 1323 yet to be honest[09:21:02] <geist> now we just need to wait for berlios to fix their stuff[09:21:26] <_hugo> yeah[09:22:40] *** PulkoMandy has joined #haiku[09:25:43] *** emitrax_ has joined #haiku[09:26:53] *** Crumbs has joined #haiku[09:31:02] *** Teknomancer has quit IRC[09:31:05] *** Crumbs has quit IRC[09:39:51] *** Crumbs has joined #haiku[09:41:47] * geist sleep[09:41:52] *** Crumbs has quit IRC[09:42:15] *** emitrax has quit IRC[09:44:32] *** FiddleStix has quit IRC[09:44:33] *** mmu_man has quit IRC[09:49:45] *** Crumbs has joined #haiku[09:49:48] *** procton has joined #haiku[09:58:09] *** hUMUNGUs has joined #haiku[09:58:12] *** Crumbs has quit IRC[09:59:20] *** Crumbs has joined #haiku[10:01:02] *** hUMUNGUs has quit IRC[10:02:22] *** hUMUNGUs has joined #haiku[10:03:55] *** Pembros has joined #haiku[10:08:13] *** Crumbs has quit IRC[10:09:13] *** hUMUNGUs has quit IRC[10:09:16] *** Crumbs has joined #haiku[10:17:43] *** stargater has joined #haiku[10:17:55] <stargater> hi[10:24:04] *** Begasus has joined #haiku[10:24:44] <Begasus> g'morning peeps[10:25:20] <BNickName> yo Begasus[10:25:38] <Begasus> hoi BNickName[10:25:54] <Begasus> alles goed daar?[10:26:30] <BNickName> ja, al is het niet zo warm vandaag[10:27:19] <Begasus> voor mij hoeft het niet zo warm te zijn ;)[10:29:15] *** Crumbs has quit IRC[10:29:20] *** [Beta] has quit IRC[10:29:50] <BNickName> Begasus, we don't have an SVG to bitmap translator yet, do we? It would allow us to add some wallpapers to the image without increasing the size too much[10:30:12] *** Crumbs has joined #haiku[10:30:26] <BNickName> using SVG or something similar also alows you to change the color scheme before rendering[10:30:53] <Begasus> not to my knowlidge ... but IconOMatic should be able or not?[10:31:14] <BNickName> to an icon yes, not a whole desktop :)[10:31:53] <Begasus> a wallpaper is almost the same as an icon ... it's just larger ;)[10:33:09] <BNickName> I know, but the svg renderer should be extracted then and made into an svg to bitmap translator[10:34:18] <BNickName> I wrote svg to bpicture for jburton's picture demo, but it's not a translator, and not to bitmap[10:35:35] <Begasus> shouldn't be to hard to write one or?[10:36:00] * Begasus has no clue about how hard it is ;)[10:45:17] *** Teknomancer has joined #haiku[10:45:52] <Teknomancer> too late for today's interview :( will have to try tomorrow[10:46:19] <Begasus> moin Teknomancer ;)[10:46:23] <Teknomancer> hi Begasus[10:47:15] <BNickName> Begasus it's fairly easy as we have code to read the svg, just someone has to create a translator out of it :P[10:48:37] <Begasus> isn't there a cli in linux or so that could be ported over?[10:55:34] *** Jixt has joined #haiku[10:55:38] <Jixt> hi[10:57:10] <Teknomancer> hi Jixt[10:58:38] <Jixt> good news Teknomancer[10:58:57] <Jixt> PG will be hosted on osdrawer[11:00:06] <Jixt> Teknomancer: could you make an account on osdrawer? Then I can add you to the team.[11:01:43] <Teknomancer> Jixt ok[11:01:46] <Teknomancer> osdrawer.com ?[11:02:37] <Jixt> osdrawer.net[11:02:45] <Teknomancer> k seeing[11:03:40] <Teknomancer> Asia Chennai isn't there in timezones :P[11:03:40] <Teknomancer> grrrr[11:03:45] <Teknomancer> only calcutta :)[11:04:45] <Jixt> :s[11:04:52] <Jixt> send them a mail ;-)[11:05:01] <Teknomancer> You are now a registered user on OsDrawer.net, the online development environment for Open Source projects.[11:05:14] <Teknomancer> Jixt registeration done[11:05:29] <Jixt> don't forget to check your mail[11:06:20] <Teknomancer> Jixt done[11:06:34] <kokito> hello folks[11:06:40] <kokito> hey Teknomancer :)[11:07:19] <Teknomancer> hi kokito[11:07:24] <Jixt> ok Teknomancer, you have been added to the project[11:07:39] <Teknomancer> okies[11:07:40] <Jixt> today or tomorrow I will add the source code[11:08:01] <kokito> what project is this Jixt ?[11:08:24] <Teknomancer> Jixt ok and tell me how to checkout also, not used SVN much[11:08:34] <Teknomancer> Jixt and can i compile using zeta itself ?[11:09:37] <Jixt> you will be able to compile in in Zeta[11:09:42] <kokito> ah! PhotoGrabber. nice![11:09:50] <Jixt> I will send you a mail with details[11:10:24] <Teknomancer> Jixt hm, ok yeah need to know how much zeta-specifics are allowed in it, yeah will await ur mail[11:11:14] <Jixt> Teknomancer: what do you mean with 'how much zeta-specifics are allowed"?[11:11:37] <Teknomancer> Jixt like zetas CLV etc. that wont work on r5 etc.[11:12:01] *** siarzhuk has joined #haiku[11:12:33] <Jixt> ic, no problem. I am working on that[11:12:47] <Teknomancer> hm ok[11:14:03] <Teknomancer> btw anyone know how to compare 2 bitmaps in memory? Do i need to scan all the Bits() of the 2 bitmap ??[11:14:25] <Teknomancer> hmm, mayme memcmp could be useful here[11:14:59] <stargater> Jixt, Teknomancer Begasus hi :-) good news from GP :-)[11:15:10] <Teknomancer> hi stargater GP ?[11:15:14] <Teknomancer> or u mean PG ? :P[11:17:39] <stargater> ah PG i mean soory[11:17:53] *** Pembros has quit IRC[11:17:57] <stargater> Begasus, you are a developer ? ;-)[11:18:53] <Teknomancer> k gotta go eat something then prepare for tomorrow's interview :) must rehearse the "Tell me about yourself" answers :)[11:18:58] <Teknomancer> i'm bad at those HR stuff[11:19:19] <Teknomancer> l8r all[11:19:44] *** Teknomancer has quit IRC[11:20:33] <Jixt> hi stargater[11:22:10] <stargater> Jixt, i learning C/C++ a while :-) under linux with www.fltk.org toolkit , for yab port to win,linux,mac :-) i write e rebuild from bebits calendar control[11:22:13] *** llama233 has quit IRC[11:22:33] <stargater> my horro cpp code can you see here (but not LOL) http://yab-interpreter.svn.sourceforge.net/viewvc/yab-interpreter/flyab/CalendarControl/[11:23:13] <stargater> i hope i can next year code in haiku with the haiku API[11:23:28] <Jixt> looks nice stargater[11:23:56] <Jixt> nice to see some guys learning C++ to help Haiku ;)[11:26:11] <stargater> it is a long way to i help ore code for haiku[11:26:49] <Jixt> Rome hasn't been build in one day. One day you will get there ;)[11:27:09] <stargater> :-) thats true jixt , thats true :-)[11:27:14] <Jixt> stargater: about PG, please do not post anything yet[11:27:49] <stargater> i do not post , (i cant see, i cant liste, i can speek)[11:28:03] <Jixt> we are still working on it and when it is ready, a post will come on Bebug.[11:28:06] <Jixt> thx stargater[11:28:12] <MrSun> wtf ... so god wasted 2 days on making rome and only 4 days on the rest of the world!?[11:28:26] <Jixt> MrSun lol[11:28:39] <stargater> MrSun, :-)[11:29:49] <stargater> oh my littel boy wake up :-) so cu later .... <open the gate>--------------------#####<jump into the gate> (*)[11:29:50] *** stargater has quit IRC[11:34:37] *** jsgotangco has quit IRC[11:35:42] <Ingenu> JonathanThompson, ?[11:40:59] *** Crumbs has quit IRC[11:41:37] *** Crumbs has joined #haiku[11:51:03] *** Crumbs has quit IRC[11:52:12] *** Crumbs has joined #haiku[11:52:43] <Jixt> is there someone that can help me with a SVN question?[11:53:09] <Jixt> when I do an import SVN is complaining about SVN_EDITOR[11:53:24] <Jixt> what do need to is as SVN_EDITOR?[11:53:48] <Jixt> *What do I need to choose as SNV_EDITOR?[12:01:09] *** Crumbs has quit IRC[12:02:12] *** Crumbs has joined #haiku[12:09:54] *** wildur has joined #Haiku[12:12:11] *** Crumbs has quit IRC[12:13:15] *** Crumbs has joined #haiku[12:19:16] <siarzhuk> Jixt do you have basic vim knowledge? you can set it as default svn editor[12:21:32] <siarzhuk> just add export[12:21:46] <siarzhuk> export SVN_EDITOR=vim[12:21:57] <siarzhuk> into your ~/.profile[12:24:41] *** rcjsuen has joined #haiku[12:31:25] *** kokito has quit IRC[12:33:17] *** hUMUNGUs has joined #haiku[12:33:44] <Jixt> ok siarzhuk, thx[12:34:12] <Jixt> one little question about ssh[12:34:55] <siarzhuk> Jixt ?[12:35:02] <Jixt> I try to import a project in my repository, but ssh is searching for ssh-rand-helper[12:35:05] *** Pulko_Mandy has joined #haiku[12:35:07] <Jixt> which I don't have[12:35:49] <siarzhuk> rand heleper? never heard abpout it. how looks the whole error message?[12:35:51] <Jixt> http://yab-interpreter.svn.sourceforge.net/viewvc/yab-interpreter/flyab/CalendarControl/[12:35:54] <Jixt> whoops[12:36:00] <Jixt> "ssh-rand-helper produced insufficient data"[12:36:57] <Jixt> and I get it when I do "svn import . svn+ssh://......"[12:37:34] <siarzhuk> hm.... never seen, sorry. try to google this message.[12:37:47] <siarzhuk> btw, at the moment I canno update from haiku's svn[12:38:22] <siarzhuk> so, probably there some problems at the server now[12:38:32] <Jixt> ic[12:38:59] <Jixt> I'm trying to upload my project to osdrawer though[12:42:13] <Jixt> ok, I've downloaded the ssh package on the haiku-site[12:42:18] <Jixt> now it is working[12:42:33] <Jixt> the one of BeOs Max was not good :S[12:42:46] <siarzhuk> good luck[12:42:55] <Begasus> go for it ;)[12:43:21] *** Crumbs has quit IRC[12:43:24] <Jixt> thx[12:44:05] *** Crumbs has joined #haiku[12:47:14] *** _konrad has joined #haiku[12:49:17] *** _konrad has quit IRC[12:51:02] *** PulkoMandy has quit IRC[12:57:18] *** FiddleStix has joined #haiku[13:08:00] *** Begasus is now known as Begasus_bbl[13:08:23] *** Lelldorin1 has joined #haiku[13:14:19] *** Crumbs has quit IRC[13:14:43] *** FiddleStix has quit IRC[13:16:35] *** kr1stof has quit IRC[13:32:41] *** Euver has joined #haiku[13:38:40] *** stargater has joined #haiku[13:38:54] *** TLF has joined #haiku[13:42:50] *** FiddleStix has joined #haiku[13:43:19] <stargater> re[13:44:27] *** NeonLicht has quit IRC[13:59:42] *** NeonLicht has joined #haiku[14:03:22] *** Lelldorin1 has quit IRC[14:13:38] *** stargater has quit IRC[14:19:46] *** stargater has joined #haiku[14:19:59] <stargater> re[14:21:17] *** Lelldorin1 has joined #haiku[14:22:09] *** Euver has quit IRC[14:22:16] *** Euver has joined #haiku[14:26:31] <Jixt> argh[14:27:37] <stargater> argh ?[14:27:45] <Jixt> just removed my libstdc++ lib and I don't have my R5 cd with me[14:28:22] <Jixt> stargater: could you mail it to me?[14:28:40] <Jixt> libstdc++.r4.so in /boot/beos/system/lib[14:32:18] <Jixt> or anyone else may mail me :D[14:32:56] <stargater> i have no r5 here[14:33:01] <stargater> http://dl.hirschkaefer.de/gcc/[14:33:07][14:33:11][14:33:16] <siarzhuk> sorry[14:33:23] <Lelldorin1> Jixt. I CAN SEND YOU THE HAIKU ONE[14:33:32] <siarzhuk> it is dependent on version of gcc you have installed[14:33:41] <siarzhuk> be afraid of compatibility[14:33:42] <stargater> Jixt: see the link, you can download the gcc , thats include[14:33:58] <siarzhuk> yes. it is better to update complete gcc[14:35:00] <Jixt> I need the R5 one[14:35:20] <Lelldorin1> ok[14:35:26] <Lelldorin1> letme check my beos r5 cd[14:35:45] <Jixt> ok, thx[14:36:52] *** NeonLicht has quit IRC[14:37:02] <Lelldorin1> mail addy?[14:37:25] <Jixt> janvanhoye at pandora dot be[14:37:35] <stargater> Jixt: ore download the orginal r5 developer tool kit http://bebits.com/app/2680[14:37:57] <Lelldorin1> Jixt: out[14:38:11] <Jixt> yes stargater, but it does not contain that library[14:38:17] <Jixt> thx Lelldorin1![14:39:02] <Jixt> thx all btw ;-)[14:39:09] <Lelldorin1> :-)[14:39:12] <stargater> ok, but Lelldorin1 can help :-)[14:40:51] *** petterhj has joined #haiku[14:58:21] *** JamesB192 has joined #haiku[15:04:28] <Jixt> have to go[15:04:29] <Jixt> cu[15:06:17] <Lelldorin1> bye[15:06:40] *** Jixt has quit IRC[15:16:18] *** Euver has quit IRC[15:22:00] *** Sil2100 has joined #haiku[15:22:10] *** Lelldorin1 has quit IRC[15:22:12] <Sil2100> Hi[15:22:22] <stargater> hi sil[15:25:02] <Sil2100> Has anyone tried to build Haiku on a linux powered x64 machine using the linux32 wrapper, as Olivier suggested?[15:27:32] <stargater> no sorry[15:28:13] *** stargater has quit IRC[15:29:26] *** stargater has joined #haiku[15:29:29] <stargater> re[15:30:31] <Sil2100> re[15:30:42] <Sil2100> I mean, welcome back[15:31:19] <stargater> :-) linux without X is cool :-)[15:31:34] *** NeonLicht has joined #haiku[15:31:46] <stargater> wb NeonLicht[15:40:40] *** _Lucretia_ has quit IRC[15:44:43] *** _Lucretia_ has joined #haiku[15:46:34] *** dr_evil has joined #haiku[15:46:51] *** TLF has quit IRC[15:47:09] *** ChanServ sets mode: +o dr_evil[15:49:35] *** kad77|log is now known as kad77[15:54:55] *** JamesB192 has quit IRC[16:03:09] *** gravy has left #Haiku[16:03:15] *** gravy has joined #Haiku[16:07:23] *** NeonLicht has quit IRC[16:08:57] *** NeonLicht has joined #haiku[16:32:49] *** arikato has quit IRC[16:35:00] *** arikato has joined #haiku[16:37:57] *** Sloar has joined #haiku[16:44:27] *** Master199 has quit IRC[16:49:00] *** BGA has joined #haiku[16:49:01] *** ChanServ sets mode: +o BGA[16:52:34] *** sarcas has joined #haiku[16:56:15] *** MYOB has joined #haiku[16:56:18] <MYOB> umccullough you around?[16:57:59] *** NeonLicht has quit IRC[17:10:40] *** TTRanger has joined #haiku[17:11:09] <TTRanger> Anybody running BeOS Max here?[17:11:45] *** TTRanger has quit IRC[17:12:19] *** jspiros has quit IRC[17:12:32] <umccullough> MYOB, yeah?[17:12:58] <MYOB> umccullough you ever do any d.net stuff?[17:13:11] <MYOB> I -might- be able to use their CVS again now as I've lost my "scary" hostnam,e[17:13:12] <umccullough> MYOB, funny... i just got a reply from the admin there[17:13:44] <umccullough> couple days ago he finally emailed me back that he was gonna get my NDA from the person who owned the fax machine[17:13:51] <MYOB> heh[17:14:07] <umccullough> if you wanna grab the source, go for it[17:14:07] <MYOB> well I signed the NDA ages ago and have CVS except my hostname was blacklisted[17:14:21] <MYOB> need to get my network fully up and running again first though[17:14:22] <umccullough> i have both net_server and BONE, and Haiku here to test on ;)[17:14:32] <MYOB> only this laptop and the Mac are tested[17:14:40] <MYOB> I've only zeta w/r5 devkits...[17:15:05] <umccullough> well, if you build, i'll be happy to test the binaries[17:15:17] *** NeonLicht has joined #haiku[17:15:19] <umccullough> otherwise, I just need to create another PGP key and email it back to this guy before he gives me CVS access[17:15:32] <MYOB> other problem is I don't have a clue where my PGP keys are[17:15:38] <MYOB> I think they're on my old desktop[17:15:39] <umccullough> it probably wouldn't hurt to have 2 of us with CVS access actualy :D[17:15:54] <umccullough> heh, i created one for this 2 months ago...[17:16:00] <umccullough> and then i lost the passphrase![17:16:00] *** silverpower has quit IRC[17:16:06] <umccullough> so i can't unlock my secret key[17:16:59] <umccullough> MYOB, we can take this discussion to #teamhaiku if you want :)[17:19:14] <MYOB> I'm currently trying to figure out how to haul my printer upstairs so a bit busy ;)[17:20:44] <umccullough> yup :) - but thanks for reminding me, i'll go create a new PGP key[17:22:48] *** TTRanger has joined #Haiku[17:24:59] <umccullough> TTRanger, i use beos max v3.1[17:35:09] *** wildur_ has joined #Haiku[17:38:03] *** silverpower has joined #haiku[17:45:02] *** wildur has quit IRC[17:45:38] *** jspiros has joined #haiku[17:48:00] <_hugo> umccullough: around?[17:48:45] *** mazon is now known as Mazon[17:53:29] <umccullough> _hugo, yeah[17:55:10] <_hugo> umccullough: would you mind changing a couple lines in the source manually while berlios doesnt come back?[17:55:10] <_hugo> you may then 'svn revert' the file to update again when berlios is back[17:55:19] <umccullough> if you want, you can just send me a diff :)[17:55:33] <umccullough> umccullough at the gmail thingy[17:55:55] <umccullough> or post it on a paste site[17:56:03] <umccullough> rafb or whatever[17:57:33] <umccullough> _hugo, maybe you should create a separate vmware image with test files on it (like RH) and mount it as a second partition[17:58:12] <MYOB> do we have a working mkbfs?[17:58:23] <MYOB> trying to figure out how to get Haiku -on to- this laptop...[17:58:32] <umccullough> MYOB, no, but you can take another haiku.vmdk and clean it out once its mounted :)[17:58:34] <_hugo> i already have one with the R5 pe image, but using your smaller upload is better for some tests[17:58:45] <umccullough> _hugo, ok[17:58:49] <umccullough> i dont' mind :D[17:59:25] <MYOB> umccullough erm, no way to do that...[17:59:31] <umccullough> MYOB, seems like everyone has been struggling with that... and it seems none of the core devs have any good solutions yet[17:59:38] *** jsgotangco has joined #haiku[17:59:45] <MYOB> BeOS/Zeta don't see the IDe here, haiku does[17:59:50] <_hugo> umccullough: http://fivebits.net/~hugo/0002-ntohl.patch[18:00:33] <_hugo> umccullough: you may apply that with 'patch -p1 < 0002-ntohl.path' in trunk/[18:00:43] <_hugo> s/path/patch/[18:00:52] <_hugo> if you want to revert back to your svn revision just 'svn revert -R'[18:00:53] <umccullough> ok[18:01:07] <MYOB> can't be too far off being able to, well, feebly try to use Haiku normally ;)[18:01:15] <umccullough> yup, SVN isn't a problem for me ;)[18:01:16] <dr_evil> this patch looks reversed[18:01:26] <_hugo> dr_evil: it is correct[18:01:33] <_hugo> the ntohl/htonl is done elsewhere[18:01:56] <dr_evil> i see, then just the subject is *very* misleading[18:02:12] <_hugo> dr_evil: well, in the context of the source it is not[18:02:31] <_hugo> right now it is not being transmitted in network order, and it is after the patch :-)[18:02:36] <dr_evil> yes aggreed.[18:05:09] *** umccullough_r5 has joined #haiku[18:05:54] *** emitrax__ has joined #haiku[18:13:05] *** Sloar has quit IRC[18:13:20] <umccullough_r5> _hugo, patch didn't work for some reason, so i modified the files manually[18:13:32] <umccullough_r5> crap, i gotta close vision or jam will run out of memory[18:13:36] *** umccullough_r5 has quit IRC[18:13:45] <_hugo> umccullough: did you use patch -p1?[18:13:53] <MYOB> I don't want to know how low memory his PC is!![18:14:00] <umccullough> _hugo, i did[18:14:16] <umccullough> MYOB, you saw my webpage - it's only 256mb ;)[18:14:23] <umccullough> i also have FF open :P[18:14:27] <umccullough> *had[18:15:08] <umccullough> _hugo, no worries - it created .rej file but i didn't feel like looking at it[18:15:14] <_hugo> ok then[18:15:19] <umccullough> rather get it compiled to test ;)[18:15:48] <_hugo> umccullough: this patch will only affect us using linux and bsds, windows doesn't use tcp timestamps[18:16:06] <umccullough> ah, all that crap you and geist were complaining about last night?[18:16:07] <_hugo> umccullough: yesterday i was able to download your big image at 13kb/s[18:16:10] <_hugo> yep[18:16:15] <_hugo> we found the problem[18:16:16] <umccullough> 13kb/s is about max :D[18:16:36] <_hugo> :-)[18:16:49] <umccullough> while it's compiling, i gotta go do something, bbib[18:17:00] <_hugo> i also have to leave for some minutes, brb[18:17:06] *** Euver has joined #haiku[18:17:07] <umccullough> oh, its done...let me reboot[18:17:44] *** santagada has joined #haiku[18:19:20] <umccullough> _hugo, when you get back, it's running again - hope you remember the url ;)[18:19:33] <umccullough> let me know if it looks wrong and i'll double check to make sure the binary got rebuilt[18:20:18] <_hugo> let me test[18:21:06] <_hugo> yep[18:21:07] <_hugo> working[18:21:12] <umccullough> coolies :)[18:21:18] <umccullough> ok, afk now[18:21:25] <_hugo> thanks umccullough[18:22:30] *** emitrax_ has quit IRC[18:25:00] *** MikeW has joined #haiku[18:25:01] *** emitrax__ has quit IRC[18:27:41] *** emitrax__ has joined #haiku[18:31:03] *** hUMUNGUs has quit IRC[18:32:12] *** stargate1 has joined #haiku[18:35:39] *** [Beta] has joined #haiku[18:36:07] *** FiddleStix has quit IRC[18:39:35] *** stargater has quit IRC[18:39:43] *** rcjsuen_ has joined #haiku[18:53:28] *** Korli has joined #haiku[18:53:28] *** ChanServ sets mode: +o Korli[18:53:39] <Korli> hello[18:53:46] <_hugo> hi Korli[18:53:54] <Korli> is there a problem with berlios svn ?[18:54:04] <_hugo> im not able to commit since yesterday[18:54:17] <Korli> ok[18:54:25] <Korli> unable to update[18:55:17] <dr_evil> _hugo confirmed, doesn't work at all :([18:55:23] *** rcjsuen has quit IRC[18:55:45] <_hugo> its not the first time berlios fails like this since i have write access[18:57:08] *** Euver has quit IRC[18:57:13] <geist> darn, still down[18:57:43] <JonathanThompson> Did umccullough reboot his server with the timestamp changes added?[18:57:51] <_hugo> JonathanThompson: yep[18:58:04] <JonathanThompson> Seems I've overloaded it again :P[18:58:32] <_hugo> JonathanThompson: what are you doing in particular?[18:58:52] <JonathanThompson> Hitting refresh as quickly as possible from SeaMonkey or Net+.[18:59:05] <JonathanThompson> Before possibly everything CAN be completed for one page load.[19:00:43] <JonathanThompson> So, having read the discussion you and geist had related to timestamps, I was hoping that would solve even that problem, but apparently it still gets hosed.[19:00:55] *** rcjsuen_ is now known as rcjsuen[19:01:05] <JonathanThompson> umccullough was conjecturing that perhaps there's some kernel bug that trashing the memory of Robin Hood somehow.[19:01:12] <_hugo> that seems a different issue. im not even sure BeOS uses TCP timestamps[19:01:48] <JonathanThompson> Well, I don't think I'm directly on the net without a router betweenj.[19:02:00] <JonathanThompson> And it sounds like he has one on his end between his server, too.[19:02:38] <Korli> _hugo https://monitor.berlios.de/berlios-status/ for Berlios status[19:02:48] *** stargater has joined #haiku[19:02:50] <stargater> re[19:02:56] <_hugo> Korli: oh, cool, thanks[19:04:36] <dr_evil> load average: 88.50, 88.65, 88.18 whats going on there?[19:05:25] <Korli> dr_evil seems their NFS is misbehaving[19:06:48] <dr_evil> watching weeds avi and fixing dvb stuff... but no checkin of intermediate steps...[19:11:31] *** emitrax__ has quit IRC[19:11:46] *** santagada has quit IRC[19:12:18] *** stargater has quit IRC[19:12:29] *** stargate1 has quit IRC[19:14:02] <umccullough> JonathanThompson, its hung again apparently, but i'm going to disable it as I need my bandwidth now :P[19:14:11] *** santagada has joined #haiku[19:14:12] <umccullough> gotta RDP to work and fix some issues[19:14:24] <umccullough> so in other words, please stop hammering it :D[19:15:26] <_Lucretia_> just watched the google video, got a bone to pick[19:15:27] <_Lucretia_> "many oss projects encourage forks and alternate projects" er, no they dont[19:20:23] *** meianoite has joined #haiku[19:26:40] *** ablyss has quit IRC[19:40:19] *** rcjsuen_ has joined #haiku[19:45:30] <TTRanger> JonathanThompson[19:50:07] <pyCube> um.. what a lame statement, especially with the negative bent on it.. the 'open' part of oss sorta LETS forks and alt projects spin off..DUH[19:50:23] <pyCube> dont think its about this particular oss project or that one[19:51:17] <pyCube> "how dare you tmake use of the open nature of this project and spin off your own variant!! wtf is wrong with you, are you some linux/gpl nazi!!!"[19:51:45] <geist> well, that's a little extreme[19:51:52] <pyCube> yep[19:51:58] <geist> but yeah, I sort of disagree with the statements a bit[19:51:59] <pyCube> its called hyperbole[19:52:00] <BGA> pyCube: Well, whatever you may be reading onm it, this is not in any way what we mean.[19:52:22] <geist> OH NOES ITS THE BGA[19:52:33] <pyCube> BGA: just speaking in general[19:52:48][19:52:53] <BGA> geist: ' up?[19:53:47] <dr_evil> oh nooo, the BGA and the geist[19:53:52] *** JamesB192 has joined #haiku[19:54:30] *** JamesB192 has quit IRC[19:55:08] <dr_evil> yeah! success buidling dvb.media_node :)[19:55:09] <BGA> dr_evil: Hello.[19:55:54] <Korli> bye[19:55:57] *** Korli has quit IRC[19:56:09] *** JamesB192 has joined #haiku[19:56:24] *** rcjsuen has quit IRC[19:57:00] <pyCube> i am still a bit curious as to what is meant by "desktop os", and what makes other oses not desktop oses[19:57:35] *** TTRanger has quit IRC[19:57:51] *** JamesB192 has quit IRC[20:01:24] *** JamesB192 has joined #haiku[20:01:30] <Pulko_Mandy> pyCube: i think it's integrated GUI, as opposite to X server :)[20:01:49] *** rgb has joined #haiku[20:02:06] <petterhj> isn't X what makes Linux a desktop OS?[20:02:17] <pyCube> Pulko_Mandy: simply out of spite of *nix oses, or for some good reason?[20:02:33] <petterhj> pyCube: there's a difference between desktop and server OSes for an example[20:02:46] <pyCube> which is...?[20:03:08] <Pulko_Mandy> having a GUI or not :)[20:03:36] <pyCube> oses with an x server have guis[20:03:39] <Pulko_Mandy> but BeOS is tought from start to have a gui... so it's very well integrated[20:03:59] <Pulko_Mandy> yes but that doesn't make the os a desktop one[20:04:09] *** stargater has joined #haiku[20:04:10] <stargater> rr[20:04:20] <stargater> re (i mean)[20:04:21] <Pulko_Mandy> dos isn't a desktop os, even if there is windows 3.11 and gem and many others GUIs :)[20:04:55] <santagada> i don't think that guis is what make a desktop os[20:05:07] <pyCube> so what is it?[20:05:10] <santagada> is the focus on end user applications and not servers[20:05:25] <pyCube> ?[20:05:39] <santagada> like focusing on a coherent desktop and not on having the fastest web server[20:06:15] <santagada> on not having all the madness of X[20:06:30] <pyCube> how does having a solid netstack do anything but ADD to the end user experience?[20:07:17] <santagada> pyCube: services are ok[20:07:34] <santagada> the most performant web server is not[20:07:36] <Pulko_Mandy> pyCube: by moving developpers from some others tasks to work on it ?[20:07:58] <pyCube> Pulko_Mandy: thats totally assuming a small developer base[20:08:20] <[Beta]> Haiku will be very media & user friendly.[20:08:22] <santagada> pyCube: well linux is still not usable to the end user, as mac os x is[20:08:43] <santagada> and they have infinite amount of developers[20:09:17] <santagada> but they focus on making 30 diferente guis, and then making 300 themes to make the ugly guis a little less ugly[20:09:47] <santagada> and then make 3 or 4 abstraction layers on top of the sound sink[20:10:17] *** rcjsuen_ is now known as rcjsuen[20:10:19] <santagada> changing apis all the time also consumes a lot of developer man power[20:10:50] *** Yez has quit IRC[20:11:44] <Pulko_Mandy> well[20:12:04] <Pulko_Mandy> from the beginning linux is meant as a playground for developpers :)[20:12:11] <pyCube> i dunno.. i still dont get it.. i hear lots of principle and feel good stuff.. but not much of what it really means[20:12:44] <santagada> pyCube: it is an abstract concept...[20:13:01] <santagada> it needs feeling[20:13:03] <pyCube> but why is it better?[20:13:13] <santagada> pyCube: it is not better[20:13:16] <pyCube> better than, say, the the linux "model"[20:13:29] <santagada> it depends on what you want[20:13:40] <santagada> I want to use my computer to do my work[20:13:50] *** sarcas_ has joined #haiku[20:13:51] <santagada> that is why I'm on mac os x now[20:14:03] <santagada> I was using linux before mac os x[20:14:48] <santagada> it is great for end user application developers to be able to focus on their problems and be able to know the platform they are developing for[20:14:58] <santagada> that was always a problem with linux[20:15:16] <santagada> today they are using ESD and art[20:15:18] <santagada> artS[20:15:22] <siarzhuk> anybody have the problems connecting haiku svn today?[20:15:27] <santagada> tomorrow is everyone on gstreamer[20:15:34] <geist> yup. berlios is down[20:15:39] <pyCube> if its not better, then why always make the "linux is crap for reason X and Y" statements?[20:15:41] <santagada> the next day is liquid or something[20:15:44] *** stargater has quit IRC[20:15:51] <siarzhuk> geist thanks[20:16:09] <dr_evil> damn berlios. tv app compiles now, too. I want to check in[20:16:15] <kad77> what berlios alternatives have been tossed around?[20:16:20] <Pulko_Mandy> pyCube: that's when looking at it from our end-user approach[20:16:25] <santagada> pyCube: because it is, for motives X and Y[20:16:41] <santagada> pyCube: maybe you like it for Z[20:16:48] <Pulko_Mandy> but linux may have others uses where these Y and X may not be a problem[20:17:04] <[Beta]> i'm confused. why is everyone talking about linux ? :)[20:17:30] <Pulko_Mandy> because it's the one i use and know :)[20:17:31] <kad77> yeah, minix is way cooler[20:17:33] <santagada> I dunno[20:17:37] <pyCube> i just think haiku needs to be more than "the os that ISNT linux" in order to attract people[20:17:40] <geist> linux linux linux, linux[20:17:44] <geist> linux?[20:17:50] *** stargater has joined #haiku[20:17:51] <kad77> :D[20:17:53] <geist> linux linux linux...[20:17:55] <stargater> re[20:17:59] <santagada> i'm going to install BeOS R5 on vmware[20:18:04] <santagada> whish me good luck[20:18:41] <santagada> pyCube: I think of haiku as "the os that maybe is going to make me stop using mac os x someday"[20:18:57] <santagada> you can think of it as "A very nice toy"[20:19:34] <kad77> geist: has anyone discussed BerliOS alternatives? Another point: they don't have very good statistical reporting on the SVN[20:19:36] <santagada> pyCube: i think most of the core devs don't even think about linux that much[20:19:40] <pyCube> i think thats part of what doomed beos.. it was largely "the os that isnt windows".. to the point where being different seeemed more important than being functional[20:20:01] <kad77> uhm, no?[20:20:13] <geist> no idea. I personally dislike berlios[20:20:20] <geist> but I'm not a decision maker[20:20:27] *** sarcas has quit IRC[20:20:44] <kad77> is it the haiku foundation board members basically?[20:21:52] <santagada> pyCube: I never seem it as such... maybe you talked to the wrong people[20:21:53] <stargater> so updating ubuntu 6.10 to 7.04 (i hope my systems will run after update)[20:22:06] *** MikeW has quit IRC[20:22:09] <santagada> pyCube: but you should ask the haiku devs[20:24:05] <pyCube> santagada: i used beos exclusively for many years.. 1998-2002.. i was as avid a beos user/supporter as you could find[20:24:10] <santagada> actually I am going to download BeOS R5 PE[20:24:11] <pyCube> and i want haiku to succeed[20:24:30] <santagada> what do you guys think[20:25:15] <geist> kad77: I'm sure after this little incident the topic will come up on the mailing lists at least once[20:25:42] <[Beta]> stargater, make sure you have lots of disk space![20:26:29] <MYOB> "little incident" - what?[20:27:18] <[Beta]> haiku foundation ?[20:27:41] <santagada> you should try to go somewhere that has both subversion and trac[20:27:44] <geist> berlios being down for a few days[20:27:56] <kad77> giest: liekly, I'm just trying to get a handle on who is in what position in this organization. I'll review the website this weekend, and if that info is not clearly laid out there, I'll post about it to the openbeos list formally.[20:28:00] <geist> which it probably will by the time they get their stuff going[20:28:02] *** MikeW has joined #haiku[20:28:03] <stargater> [Beta]: to late, i have not check :( ahhhh[20:28:40] <[Beta]> stargater, just make sure its 2gb+, mine dled 1413 packages, yours is prolly less[20:29:08] *** sarcas has joined #haiku[20:30:06] <noph> someone who managed to get the latest revision to compile with ubuntu 7.04?[20:30:13] <noph> got build-essentials and flex installed[20:30:28] <noph> but buildning the buildtools fails[20:32:07] <noph> make[2]: *** [info-recursive] Error 1[20:32:23] <noph> leaving directory binutils/bfd/[20:32:30] <stargater> i understand whay the update not check the free spache[20:32:38] <geist> what is the spec if your machine?[20:32:39] <geist> is it x86-64?[20:32:48] <noph> x86 intel[20:32:55] <geist> plain 32bit?[20:33:14] <noph> yeah[20:34:27] <geist> unknown, if you could paste more of the error (on pastebin or something please) might be able to help[20:34:40] <noph> alright[20:34:48] <noph> w8 :D[20:34:49] *** sarcas_ has quit IRC[20:36:37] <noph> hmm, interresting, a mv fails in somewhere in the script.[20:37:07] <santagada> you lack the right access permission?[20:37:10] <santagada> :)[20:37:25] *** kokito has joined #haiku[20:40:06] <stargater> hi kokito[20:40:21] <kokito> guten morgen stargater :)[20:40:54] <noph> http://pastebin.ca/451652[20:41:04] <stargater> oh :-) the sun goes down here :-)[20:41:19] <noph> i think i took everything that could be of interrest.[20:41:27] <kokito> still going up here, stargater :)[20:42:05] <noph> geist: seems like a move don't work.[20:42:20] <geist> whacky[20:42:27] <pyCube> the wonders of a spherical planet.. :-p[20:42:29] <stargater> thats so cool with internet , we can talk and you and i living more then 1000 miles ,,,[20:42:41] <dr_evil> WARNING: `makeinfo' is missing on your system.[20:42:48] <dr_evil> apt-get install texinfo[20:42:51] <umccullough> heh, yep[20:42:52] <umccullough> #[20:42:52] <umccullough> You might want to install the `Texinfo' package or[20:42:52] <umccullough> #[20:42:52] <umccullough> the `GNU make' package.[20:43:19] <dr_evil> that one i s pretty easy noph[20:43:49] <dr_evil> is it too hard for you to read the warning message?[20:43:59] <noph> make is installed[20:44:18] <umccullough> type makeinfo[20:44:43] <dr_evil> noph read my message from 2 minutes ago[20:45:07] <noph> yeah, installing makeinfo now[20:45:10] <stargater> noph: sudo apt-get install texinfo[20:45:22] <noph> but i haven't really modified any texi files[20:45:25] <noph> :)[20:46:00] <noph> allright, new try[20:46:07] <geist> i'm a little surprised it isn't already instaled on this box[20:46:23] <geist> seems that'd be one of those things that just comes with any linux box with dev stuff on it[20:46:32] <umccullough> ubuntu doesn't come with anything :P[20:46:36] <noph> ;D true[20:46:51] <stargater> OS::depences_hell *linux;[20:47:13] <umccullough> i think somewhere on teh mailing list someone listed all the various packages they had to install to build on ubuntu - it was several[20:47:26] <noph> humnm, it's compiling stuff now :D[20:47:28] <noph> excellent[20:48:00] <noph> perhaps a refrase in the configure to make it a requirement.[20:51:17] <umccullough> noph, you're talking about gcc's configure script...[20:51:31] <umccullough> or binutils anyway[20:51:45] <noph> ah, yeah that might be true[20:51:53] <noph> anyway, autoheader missing :D[20:52:27] <noph> weird, thought those where basic things[20:52:42] <umccullough> noph, exactly - ubuntu doesn't even come with basic dev tools ;)[20:53:31] <noph> :)[20:54:15] <Sil2100> No wonder, Ubuntu is aimed to normal everyday users anyway[20:54:21] <Sil2100> But I also find that a bit strange[20:54:45] <noph> yeah, well i though the build-essentials package would conver most of these 'basic' things[20:55:16] <noph> well, I can't really say that it is hard to install the missing packages though :D[20:55:17] <umccullough> noph, you'll need bison also[20:55:34] <noph> i think bison is there.[20:55:42] <noph> *prays*[20:55:54] <umccullough> i read: build-essentials, flex, yacc, bison as the deps[20:56:05] <umccullough> might be others...[20:57:03] *** JamesB192 has quit IRC[20:58:35] <umccullough> ah, bryanv has a list on his blog apparently[20:58:44] <umccullough> http://bryan.varnernet.com/2007/04/10/setting-up-a-haiku-environment-on-ubuntu-edgy-eft-610/[20:58:46] <noph> ouh[20:59:39] <umccullough> would be nice to get that info onto the website as a sub-article for the "linux development page"...[20:59:46] *** meianoite has quit IRC[21:00:15] <umccullough> BryanV, there?[21:00:55] <noph> i'm compiling gcc now, so far it works.[21:01:08] <umccullough> noph, FYI, i think gawk is required for the haiku build - not the buildtools[21:01:17] <noph> ah[21:01:37] <umccullough> so, you'll eventually run into that one too ;)[21:03:03] <noph> haha, this was interresting! :D it says that gcc for cross compilation has been built successfully![21:03:09] <umccullough> yes[21:03:11] <noph> and then gives me like nine gawk not found :D[21:03:32] <umccullough> you can't say i didn't warn you ;)[21:03:48] <noph> ;D[21:04:14] <noph> allright, lets try again[21:04:24] <umccullough> wait[21:04:31] <umccullough> the build tools are done right?[21:04:38] <umccullough> so you just wanna 'jam haiku-image' now[21:04:50] <umccullough> or, did it fail before the build tools were complete?[21:05:03] <noph> well, not sure. it said after the compiled successfuly that it was missing gawk nine times or something[21:05:22] <noph> so i think ill run it again, just to make sure[21:05:30] <Sil2100> So it probably didn't build the buildtools correctly, IMO[21:06:10] <santagada> i still cannot build haiku in itself, is it right?[21:06:54] <stargater> ahh pizza is here from german "Hello Pizza"[21:06:55] <stargater> bbl[21:07:18] <noph> well, anyway. I should actually leave for some wine. :\[21:07:32] <noph> i hate when i want to do two things at the same time.[21:07:33] <Sil2100> Bon appetit, stargater[21:07:38] *** tgzuiop has joined #haiku[21:08:18] <santagada> umccullough: can I build haiku on itself?[21:09:04] <geist> nope, not stable enough[21:10:45] *** dr_evil has quit IRC[21:10:50] *** tgzuiop is now known as dr_evil[21:10:51] *** ChanServ sets mode: +o dr_evil[21:11:06] *** mmadia has joined #haiku[21:14:42] <santagada> geist: even if I like retry it 10 times ?[21:14:52] <geist> yes[21:14:53] <santagada> because I don't have a linux machine here[21:14:57] <noph> excellent, now it compiled just fine[21:15:05] <noph> thanks everyone for your feedback![21:15:17] <noph> ill leave for my wine now, bbl.[21:15:24] <geist> see ya[21:15:37] <umccullough> bye[21:15:48] <umccullough> santagada, it will be a glorious day when it is possible :)[21:15:57] <umccullough> but in the meantime, you can blame geist ;)[21:16:36] * geist blames himself[21:17:09] <geist> though to be fair, I have *generally* seen more haiku kernel bugs due to new stuff added for haiku than old design problems with newos[21:17:17] <umccullough> ;)[21:17:20] <geist> virtually all of the recent VM problems have been in new code[21:17:31] <geist> especialy early stuff, before axel got involved[21:18:21] <umccullough> either way, the amount of "blame" is absolutely miniscule compared to the amount of "thanks"[21:18:59] <umccullough> "looking a gift horse in the mouth" kinda thing ;)[21:19:45] *** MikeW has quit IRC[21:19:58] <umccullough> heh... i wonder how intelligent it is to go with my wife and show her how to shoot her bow with only a single arrow ;)[21:20:37] <umccullough> i imagine after the first shot...we may have to call it quits :P[21:20:50] <geist> might break the arrow I guess?[21:20:54] <MYOB> umccullough damn americans, do you always have to have weapons :P[21:20:55] <umccullough> or lose it[21:21:05] <umccullough> MYOB, seriously, a bow?[21:21:12] <umccullough> she intends to use it for hunting[21:21:14] <MYOB> umccullough sarcasm[21:21:17] <umccullough> ;)[21:21:18] <Sil2100> Or kill the neighbour[21:21:45] <umccullough> Sil2100, that's not a bad idea[21:21:50] <umccullough> ...not a good one either though ;)[21:21:55] <Sil2100> Well, that's probably a sacrafice everyone's ready for[21:21:57] <Sil2100> ^^[21:22:06] <umccullough> oh you mean we might accidentally kill them?[21:22:09] <Sil2100> As I though[21:22:20] <Sil2100> Yeah, *accidentally* yes[21:22:22] <umccullough> heh[21:22:39] <umccullough> actually will be using a large field... we're somehwat rural[21:23:39] * umccullough notes the large number of feral cats in teh back yard...[21:24:59] <Sil2100> I hope you don't intend to shoot them?[21:26:35] <umccullough> Sil2100, I personally won't[21:26:54] <umccullough> but really, our neighbor is pretty irresponsible about letting them breed and roam all over the place[21:27:11] <umccullough> best thing she could do would be spay/nueter them all[21:27:12] *** JamesB192 has joined #haiku[21:27:46] <Sil2100> Well, I think I understand you, since there's many stray cats around my place too[21:28:01] <stargater> re[21:28:03] *** tgzuiop has joined #haiku[21:28:08] <umccullough> Sil2100, i don't have a problem with cats in general - but I do have a problem with cat feces and urine all over my stuff[21:28:16] <Sil2100> But since they're not a bother to me (living in an appartment on the 4'th floor), I let them be[21:28:47] <Sil2100> Ah, that can be a bother indeed[21:28:59] <Sil2100> stargater: welcome back[21:29:19] *** dr_evil has quit IRC[21:29:21] *** tgzuiop is now known as dr_evil[21:29:22] *** ChanServ sets mode: +o dr_evil[21:30:04] <stargater> thx Sil2100[21:32:01] *** sarcas has quit IRC[21:35:18] *** MYOB2 has joined #haiku[21:35:24] *** MYOB has quit IRC[21:35:28] *** MYOB2 is now known as MYOB[21:35:52] *** tgzuiop has joined #haiku[21:36:28] *** stargater is now known as betka[21:36:42] *** dr_evil has quit IRC[21:36:45] *** tgzuiop is now known as dr_evil[21:36:46] *** ChanServ sets mode: +o dr_evil[21:36:56] <dr_evil> oh-so-stable-dsl[21:37:22] <siarzhuk> Alice? :-)[21:37:43] <dr_evil> yes[21:37:48] <dr_evil> Alice hates me[21:37:51] <siarzhuk> But you was warned - you have seen this quickly running girl ;-D[21:38:04] <santagada> geist: what is the priorities of the project haiku?[21:38:17] *** betka is now known as stargater[21:38:19] <santagada> geist: self-hosted is in the list?[21:40:54] <umccullough> santagada, i think "self-hosted" probably comes with "stable/working kernel"[21:41:18] <geist> right[21:41:18] <umccullough> since all the build tools are essentially there :)[21:41:27] <geist> exactly[21:41:42] <geist> and sinc eyoui can buid haiku on beos, and haiku is going to be beos binary compatible[21:41:46] <geist> it should be able to build itself[21:41:53] <umccullough> :)[21:42:08] <MYOB> ooh, my router does WPA2. never new that[21:42:41] <umccullough> of course, the other part that is missing is the ability to partition and mkbfs from within Haiku[21:42:42] <santagada> but are you working on making it stable enough to build itself or you are not doing that with this in mind[21:42:55] *** rcjsuen_ has joined #haiku[21:42:57] <santagada> like this would be a feature you would prioritize or not[21:43:04] <geist> there is no distinction[21:43:20] <geist> in fact, being able to build itself is probably be ultimate test of the kernel stability[21:43:37] <umccullough> i suspect that's one of the stability goals for R1, but I dont' know if anyone has explicitely stated that[21:43:50] <santagada> ahh ok[21:44:21] <JamesB192> I think that most of an open mkbfs is available from the image builder in the source tree.[21:44:30] <santagada> because I would like to test python on haiku, but for that probably I will get on the same problems as building the system[21:44:56] <santagada> if I get to that I will play on BeOS and then just try to run the partition on haiku[21:45:09] *** ablyss has joined #haiku[21:45:10] <santagada> having a shared partition between then[21:45:13] <JamesB192> There are also a couple of console disk partitioners around to inspire.[21:45:51] *** MikeW has joined #haiku[21:47:58] *** forQed has joined #haiku[21:50:34] <santagada> someone knows how to make a iso using a cue sheet?[21:50:49] <santagada> shit BeOS PE MAX instalation[21:51:04] <umccullough> santagada, haiku livecd is not iso9660[21:51:20] <umccullough> neither is beos max cd[21:51:32] <umccullough> you must burn it as a dual-track cd[21:51:45] <umccullough> where the second track is a BFS image[21:51:50] <santagada> shit I want to make an image and use it in vmware[21:52:04] <santagada> I don't have blank cds with me[21:52:32] <santagada> I wanto to combine it into a raw image file[21:52:35] <umccullough> santagada, windows?[21:52:41] <santagada> mac os x[21:52:56] <umccullough> mm... i'm not even sure what cd image formats vmware supports actually[21:53:08] <santagada> wouldn't it be easier to already distribute it as a image file[21:53:15] <umccullough> what format?[21:53:18] <santagada> iso basically[21:53:19] <umccullough> ccd? nrg?[21:53:29] <umccullough> it cannot be iso though[21:53:30] <santagada> iso only I think[21:53:47] <santagada> well beosmax come in a iso file though[21:53:47] <umccullough> heh, it's kind of hard to explain ;)[21:53:55] <umccullough> no, it doesn't actually[21:54:02] <santagada> yes it does[21:54:04] <umccullough> at least v3 never did[21:54:15] <santagada> and vmware knows how to read it[21:54:35] <santagada> it just don't know how to read a cue file[21:54:51] <umccullough> well, a .cue file is just a text file that specifies the track layout for the 2 images[21:54:53] <santagada> and then it cant boot the cd image[21:55:13] <santagada> I know that, go to vmware and try to explain that to then[21:55:36] <santagada> I really hate when things are this complicated for such an easy task[21:55:43] <umccullough> i've only booted vmware from a beosmax CD using a physical disc in a drive and raw access[21:56:12] <umccullough> santagada, if beos booted from iso9660, it would be easy...[21:56:22] <santagada> I will install cdrecord package and say to it to burn the cue sheet back to an iso[21:56:25] <umccullough> but iso9660 doesn't support attributed as needed by BeOS in order to boot[21:56:38] <umccullough> attributes*[21:56:40] <santagada> iso file is not iso9660 I think[21:56:50] <santagada> is just a raw file[21:56:53] <umccullough> really? then something is lying to you...that's exactly what .iso stands for[21:56:54] * JamesB192 wonders why santagada doesn't hit the build factory for vmware ready files.[21:57:09] <umccullough> santagada, usually a raw image is called .bin or some such[21:57:11] <geist> .iso is typically the logical sectors of a session[21:57:12] <santagada> JamesB192: I didn't know about them[21:57:26] <geist> and most of the time it has iso9660/udf/bfs/hfs/whatever on it[21:57:54] <santagada> well on my beosmax they have some .dmg boot files and one beosmax .iso file[21:57:55] <geist> since most cds (beos boot cds excepted) only have a single session, it's typically all you need[21:58:15] <santagada> geist: yes thats right[21:58:25] <santagada> well they could just shit a raw image[21:58:36] <umccullough> and beos boot CDs require dual tracks/sessions...[21:58:37] <geist> not really, beos boot cds have multiple sessions[21:58:38] <santagada> JamesB192: where is this build factory[21:58:39] <DeadYak> I don't think it's trivial to do a multitrack image file[21:58:48] <DeadYak> in fact I'm not sure that's even possible[21:58:53] <geist> one is iso9660 with the el torito stuff, and the other is a bfs image[21:58:58] <santagada> geist: well they could distribute a .raw or .bin[21:59:02] *** rcjsuen has quit IRC[21:59:05] <umccullough> DeadYak, the only way i know is to use a proprietary format like clonecd's ccd or nero's .nrg[21:59:10] <[Beta]> sounds like its time to make a multitrack image file spec.[21:59:11] <santagada> they can have more than one track[21:59:12] <umccullough> or .mds[21:59:15] <DeadYak> umccullough: right[21:59:18] <geist> or the ol cdrwin .bin+.cue file stuff[21:59:28] <DeadYak> what did the R5 Pro CD use? I remember it being a single track[21:59:31] <geist> or manually burn two sessions with cdrecord[21:59:32] <DeadYak> well[21:59:36] <DeadYak> one track for x86, one for PPC anyhow[21:59:37] <geist> no, it was multitrack[21:59:38] <umccullough> geist, yes :) - except apparently vmware doesn't support .cue :P[21:59:40] <DeadYak> oh[21:59:54] <geist> and before r5 the disc was big enough tol hold both x86 and ppc[22:00:05] <geist> so it was a iso9660 + x86 bfs + ppc bfs[22:00:16] <DeadYak> ah.[22:00:35] <geist> i think with r5 it got too big and since no one cared about ppc at that point they just made two discs[22:00:39] <santagada> well I know cdrecord can generate a raw image probably with the multiple tracks or it can't?[22:00:45] <DeadYak> geist: my R5 disc has both[22:00:50] <geist> you sure?[22:01:03] <DeadYak> geist: positive, I installed the same disc on my BeBox as on my Athlon[22:01:04] <umccullough> geist, my R5 pro 5.0.3 disc says it has PPC too[22:01:06] <geist> maybe I'm full of shit and I'm thinking aout the r6 image that never came[22:01:15] <umccullough> lol[22:01:17] <umccullough> maybe ;)[22:01:19] <geist> okay, i guess I'm thinking of r6[22:01:45] <santagada> JamesB192: where is the vmware factory that you talked about[22:04:47] <santagada> http://www.beosmaxfiles.org/ don't talk about vmware ready images[22:05:10] <DeadYak> haiku build factory he meant I believe[22:05:24] <umccullough> santagada, are you trying to install beos max or haiku?[22:05:43] <santagada> beos max[22:05:53] <santagada> I have haiku running ok on vmware[22:06:08] <santagada> but now I want beos so I can start to play with python in it[22:06:15] <umccullough> yeah, ok - you're gonna run into problems :)[22:06:23] <JamesB192> santagada: oops, meant haikuhost at http://haikuhost.com/housestrain/get.php?file=r20778_vmware.tar.bz2[22:06:44] <santagada> JamesB192: but that is for haiky[22:06:48] <santagada> haiku[22:06:53] <umccullough> best solution i've found is to burn the beos max disc to a physical CD-R and then use vmware to access it physically from your CDROM[22:06:58] <santagada> and I want beos max[22:07:15] <umccullough> if you find another way to pull it off, feel free to let me know :)[22:07:18] <santagada> umccullough: the problem is that I don't have blank cds here[22:07:23] <santagada> ahh ok[22:07:34] <umccullough> maybe vasper knows...[22:07:45] <santagada> I will try to make cdrecord burn a raw image and use that[22:07:57] <santagada> I don't know if it can do that[22:08:00] <JamesB192> Never mind then, I thought you were talikng about Haiku...[22:08:25] <umccullough> i would hope that if someone knew how to do it, they would stop distributing multiple images and a .cue...[22:08:38] <umccullough> or at least offer a single download option as well[22:08:55] *** MYOB has quit IRC[22:09:20] *** tgzuiop has joined #haiku[22:10:26] <santagada> cdrecord is on berlios also... shit this is not my day[22:10:28] *** dr_evil has quit IRC[22:10:29] *** tgzuiop is now known as dr_evil[22:10:30] *** ChanServ sets mode: +o dr_evil[22:10:54] <santagada> umccullough: I would like that also[22:11:47] <dr_evil> I just zipped the (not checked in) dvb sources on one machine, and put them into source tree of the other, (slowly) recompiling 1200 targets now, and I'm ready to test[22:12:40] <dr_evil> _hugo is it possible to start the network stack from within the kernel during boot, before mounting root device?[22:12:55] *** Begasus_bbl has quit IRC[22:14:40] <santagada> umccullough: I did it[22:14:44] <santagada> did it[22:14:49] <santagada> \o/[22:15:08] <santagada> just used the flopy image as a flopy and the iso as the cdrom image[22:15:13] <santagada> great !![22:15:51] <umccullough> santagada, oh yeah..i guess that works :)[22:15:56] <umccullough> didn't even consider that[22:16:14] * umccullough files it away in his brain[22:16:23] *** JamesB192 has quit IRC[22:17:43] <kokito> dr_evil, going to test dvb in Haiku?[22:18:38] <dr_evil> kokito yes, but I doubt that the codec interface does work good enough[22:18:54] <santagada> umccullough: I didn't also[22:19:00] <santagada> well network is not working[22:19:19] <dr_evil> kokito it builds now and is ported to used BMediaDecoder instead of external codec add-ons[22:19:35] <santagada> when I get back I will try to make it work[22:19:42] <dr_evil> but I think there is some tuning to haiku needed to make it work well enough[22:20:34] <kokito> cool dr_evil. it will be nice to see screenshots of dvb running Haiku. :)[22:21:53] <umccullough> santagada, ipro1000 tends to work better in vmware - check your .vmx file to see if it's set to emulate e1000[22:22:24] <dr_evil> http://svn.berlios.de/viewcvs/*checkout*/haiku/haiku/trunk/3rdparty/vmware/haiku.vmx?rev=19989[22:22:40] <dr_evil> err well, just use this file: trunk/3rdparty/vmware/haiku.vmx[22:22:50] <[Beta]> "Fred Fish compiled the famous Fish Disks of freely redistributable software for the Amiga, and was the very first person I met at Cygnus when I started working there in 1995. He was one of the resident GDB hackers at the time, until he left for Be in 1998. Fred was responsible for many important contributions to GDB and binutils over the years." - the Be Inc. ?[22:22:57] <umccullough> i suppose the Haiku vmx will work for beos max also...[22:23:15] *** rx_ has joined #haiku[22:23:20] <DeadYak> [Beta]: iirc he ported a good number of the *nix tools you find in BeOS[22:23:59] <umccullough> svn.berlios.de is dead from here[22:24:09] <[Beta]> It's sad to mention that he died yesterday.[22:24:12] <[Beta]> k DeadYak[22:24:17] *** Begasus has joined #haiku[22:24:27] <umccullough> well, outside i go...ttyl[22:24:30] *** umccullough has left #haiku[22:24:49] <DeadYak> he did? o.0[22:24:52] <DeadYak> of what?[22:24:54] <[Beta]> http://spindazzle.org/greenblog/index.php?/archives/60-guid.html[22:25:07] <[Beta]> I know little more, just saw it in google reader.[22:25:11] <dr_evil> yes dead here, too[22:25:44] <DeadYak> damn[22:32:43] *** TLF has joined #haiku[22:38:03] *** forQed has quit IRC[22:40:51] <Sil2100> Goodnight everyone[22:40:55] *** Sil2100 has quit IRC[22:40:56] <DeadYak> night[22:41:05] <dr_evil> kokito well, no video/audio, but at least the DVB media addon gets loaded and opens the driver[22:42:01] <kokito> dr_evil, it's a start :)[22:42:50] <dr_evil> kokito I managed to freeze haiku three times now[22:42:54] <dr_evil> in about 5 minutes[22:44:56] <kokito> dr_evil, that's an accomplishement! :P[22:48:51] <kokito> DeadYak!![22:55:01] <DeadYak> kokito?[22:55:42] <kokito> just saying hi DeadYak :P[22:55:48] <DeadYak> hi :)[22:55:53] *** tombhadAC has joined #haiku[22:56:05] <dr_evil> I think I just found a haiku bug[22:56:09] <dr_evil> when I do a[22:56:21] <dr_evil> ls /dev/dvb/cs23882/1[22:56:23] <kokito> squash it dr_evil!![22:56:30] <dr_evil> only the open hook is called[22:56:48] <dr_evil> but neither free or close hook[22:56:52] <DeadYak> could that be a devfs bug?[22:57:14] *** rx_ has quit IRC[22:57:15] <dr_evil> that would also explain why you can't get audio output after restarting media server[22:57:18] <dr_evil> DeadYak could be[22:59:53] *** rgb has quit IRC[23:02:06] *** MikeW has quit IRC[23:04:34] <dr_evil> need to inverstigate further, it's not ls, must happen on first use[23:09:18] <dr_evil> geist can you help me with that?[23:09:21] <dr_evil> ls /dev/dvb/cs23882/1[23:09:21] <dr_evil> cx23882: driver open[23:09:21] <dr_evil> cx23882: driver read[23:09:21] <dr_evil> cx23882: driver close[23:09:21] <dr_evil> cx23882: driver free[23:09:21] <dr_evil> .[23:09:23] <dr_evil> tail -f /dev/dvb/cs23882/1 (or opening media server and killing media_server)[23:09:25] <dr_evil> cx23882: driver open[23:09:27] <dr_evil> cx23882: driver read[23:09:30] <dr_evil> cx23882: driver close[23:09:32] <dr_evil> .[23:09:33] <dr_evil> ls /dev/dvb/cs23882/1[23:09:36] <dr_evil> cx23882: driver open (=> B_BUSY)[23:10:00] <dr_evil> so the free hook doesn't get called[23:10:44] *** umccullough has joined #haiku[23:11:24] <umccullough> question for the channel: since i'm using the stock fan/heatsink that came with my AMD64 X2 5600 - should I scrape off the thermal pad already on it and use AC5?[23:11:32] <umccullough> AS5 that is[23:16:02] <dr_evil> yes, reproduceable when killing the media server app, only close hook is called, not the free hook[23:18:45] <dr_evil> umccullough according to the vendor, http://www.arcticsilver.com/as5.htm it is cooler with as5[23:20:07] <dr_evil> I also scraped of a termal pad, cleaned the metal with resolvent, and used standard (no silver) thermal conductance paste[23:20:38] <umccullough> maybe I'll start with the stock thermal pad and monitor it for a couple months[23:20:47] <dr_evil> once, if you work accurate, and make the as5 very thin, it is better[23:21:07] *** meianoite has joined #haiku[23:21:13] <umccullough> i'm just worried because these new AMD procs have a giant heat spreader and i've read that AS5 doesn't work as well on larger areas[23:21:32] *** skoe has joined #haiku[23:21:35] <skoe> Hey[23:21:38] <meianoite> hi. anyone knows if the old site has been backed up somewhere?[23:21:54] <skoe> Could someone that's subscribed to the Mailing list tell me if the file i attached went through ok?[23:22:00] <umccullough> meianoite, try webarchive.org :)[23:22:14] <meianoite> does webarchive archives zip files?[23:22:23] <umccullough> meianoite, sometimes, yes[23:22:31] <umccullough> i've downloaded archives from webarchive before[23:22:44] <umccullough> meianoite, what are you looking for?[23:22:52] <meianoite> just a sec[23:24:28] <meianoite> here: http://haiku-os.org/documents/dev/experiments_in_scheduling[23:24:46] <meianoite> on the old site there used to be a link for Daniel's TVM code[23:24:56] <meianoite> *to[23:24:59] *** BNickName has quit IRC[23:25:12] <meianoite> but it's simply *nowhere* to be found[23:25:18] <meianoite> I have a somewhat modified copy of it[23:25:23] <meianoite> but I'm looking for the pristine version[23:27:08] *** stargater has quit IRC[23:28:10] *** tombhadAC has quit IRC[23:29:23] *** santagada has quit IRC[23:29:26] *** arikato has quit IRC[23:29:27] <Ingenu> night[23:29:35] *** Ingenu has quit IRC[23:29:44] *** stargater has joined #haiku[23:30:55] <stargater> re[23:31:10] <stargater> [Beta], update is finish and its runs[23:31:21] *** ablyss has quit IRC[23:31:21] *** rcjsuen_ has quit IRC[23:31:22] *** jsgotangco has quit IRC[23:31:22] *** silverpower has quit IRC[23:31:25] *** pyCube has quit IRC[23:31:26] *** MrRagga has quit IRC[23:31:28] *** MrSun has quit IRC[23:31:44] *** ablyss has joined #haiku[23:31:44] *** rcjsuen has joined #haiku[23:31:44] *** jsgotangco has joined #haiku[23:31:44] *** silverpower has joined #haiku[23:31:44] *** pyCube has joined #haiku[23:31:44] *** MrRagga has joined #haiku[23:31:44] *** MrSun has joined #haiku[23:32:20] <[Beta]> stargater, :)[23:32:57] <[Beta]> I hope Haiku updates will be as smooth (though I'd want them to complete in 1/10 of the time)[23:35:57] <dr_evil> geist it looks like vfs_free_io_context() calls close_fd() and put_fd(),[23:36:02] <umccullough> meianoite, so the file you're looking for is beatrice-t2-1.zip?[23:36:06] <umccullough> i can't find it anyway :P[23:36:27] <meianoite> umccullough I'm positive it wasn't named that ;)[23:36:40] <dr_evil> close_fd will call close if previous reference count was 1 and put_fd will call free if previous reference count was 1[23:36:49] *** Minbari has joined #haiku[23:37:15] <umccullough> meianoite, the old newsletter from 2002 references a file called that in /samples/[23:37:26] <dr_evil> somehow this isn't that correct?[23:37:27] <Pulko_Mandy> try http://www.filewatcher.com/ if you can find out the name[23:37:29] *** Master199 has joined #haiku[23:37:55] <meianoite> umccullough eh? I couldn't manage to find this newsletter article using webarchive.org[23:38:04] <meianoite> it loads haiku-os front page[23:38:08] *** siarzhuk has quit IRC[23:38:17] <meianoite> but if I try to navigate to any link it bounces me back to the new site[23:38:18] <umccullough> hang on, i'll get you a link[23:38:30] <meianoite> thanks[23:39:04] <Begasus> g'night peeps[23:39:17] <umccullough> oh, guess i can't get you a link...[23:39:21] *** Begasus has quit IRC[23:40:59] <umccullough> meianoite, if you go here: http://web.archive.org/web/20030128183524/http://openbeos.org/[23:41:11] <umccullough> click newsletter, then click newsletter archive and search for "Experimenting"[23:41:24] <umccullough> er "Experiements"[23:41:50] <umccullough> oh, i was reading the wrong article![23:42:05] <umccullough> Source Code:[23:42:06] <umccullough> http://open-beos.sourceforge.net/samples/tvm_scheduler.zip[23:42:11] <umccullough> that's what you're talking about?[23:42:14] * meianoite hugs umccullough[23:42:15] <meianoite> dude[23:42:25] <meianoite> high five! o/[23:42:38] <umccullough> :)[23:42:42] <umccullough> \o[23:43:43] *** stargater is now known as SprMa___[23:44:12] <[Beta]> yay, svn co[23:44:15] *** dr_evil is now known as SprMa____[23:44:21] <umccullough> :O[23:44:35] <umccullough> meianoite, i'm not sure why that link is missing from the new article - i can add it i think[23:45:16] *** MauriceK has joined #haiku[23:45:38] <meianoite> I'm talking to kokito right now regarding this[23:45:47] *** Minbari has quit IRC[23:45:47] <meianoite> there are probably gobs of missing files on the new site[23:46:02] <kokito> meianoite, the file seems to be here: http://open-beos.sourceforge.net/samples/tvm_scheduler.zip[23:46:05] *** skoe is now known as skoe|beos[23:46:21] <meianoite> kokito scroll up a bit :)[23:46:26] *** skoe has joined #haiku[23:46:37] <kokito> ah! ok, nice.[23:46:41] <meianoite> but I still love you both for that :)[23:46:48] <umccullough> kokito, i went ahead and posted the link into the article[23:46:50] <meianoite> in VERY a manly way, btw[23:47:00] *** skoe|beos has left #haiku[23:47:23] <umccullough> i could maybe upload the file to the site directly...[23:47:46] <umccullough> but i don't have much experience with that...so the link should suffice for now :P[23:48:00] <meianoite> kokito, umccullough: we might be missing content from the whole /samples dir[23:48:05] <meianoite> http://open-beos.sourceforge.net/samples/[23:48:29] <meianoite> it'd be cool to have them back[23:48:31] <umccullough> ah, the beatrice files are there too :)[23:48:47] <meianoite> I'm planning to post a blog entry that picks up where Daniel left[23:49:01] <meianoite> since I also used TVM as a testbed platform[23:49:41] <meianoite> so it'd be nice to have the original version available and reachable from his article[23:50:04] <umccullough> meianoite, i did add the link to his article[23:50:17] <umccullough> so at least unless SF loses it, it'll be there :)[23:50:18] * meianoite hugs umccullough once more[23:51:25] <meianoite> hm... I reloaded the article, and even flushed my cache, but there are still no links[23:51:29] <umccullough> really?[23:51:49] <umccullough> at the VERY end?[23:51:53] <umccullough> http://haiku-os.org/documents/dev/experiments_in_scheduling[23:52:51] <meianoite> eh... yeah. there. nothing yet.[23:52:52] <meianoite> any way that you can think of. It's fun.[23:52:52] <meianoite> Login or register to post comments :: email this page :: printer friendly version[23:52:52][23:53:35] *** SprMa____ is now known as dr_evil[23:53:59] <meianoite> umccullough I tried with IE as well, to no avail[23:54:04] <umccullough> weird[23:54:09] * meianoite goes on to throw up[23:54:11] <meianoite> IE. eew.[23:54:36] <umccullough> let me log out[23:54:45] <meianoite> btw, IE messes with the rendering quite a lot[23:54:46] <umccullough> ah[23:54:54] <umccullough> apparently it was only visible when i was loggedin...[23:55:00] <umccullough> i wonder if i have to publish a new rev or something[23:55:11] <meianoite> can't attached files be made public?[23:55:12] <meianoite> hmm.[23:55:18] <kokito> meianoite, about getting stuff into the website, you need to talk to Waldemar.[23:55:39] *** SprMa___ is now known as stargater[23:56:26] <meianoite> ok... and where can I find him, kokito? the stuff I'm trying to get into the site are not *mine*, but left-overs from the site migration[23:56:38] <meianoite> so in a sense it's actually his fault =/[23:56:43] <meianoite> and mine ain't but a bug report[23:57:27] <kokito> meianoite, just send an email to the Haiku list and Waldemar will respond (he is very responsive).[23:57:31] *** ajupiter has joined #haiku[23:59:10] <ajupiter> is the haiku disk image a disk image or a partition image?[23:59:38] <dr_evil> partition[23:59:50] <ajupiter> okay, should it be mountable from R5?