ScummVM IRC Log - 18 Jan 2012

[Back to Index]

[00:10] <-- Vaikungfu left irc: Ping timeout: 248 seconds
[00:13] <-- Mindkontrol left irc: Ping timeout: 276 seconds
[00:15] --> Mindkontrol joined #scummvm.
[00:20] <syke> doing pretty good
[00:20] <syke> we will hopefully be dads by the end of the year
[00:20] <syke> my music stuff is going okay; I'm going to do a kickstarter project to help fund the launch
[00:31] --> clone2727 joined #scummvm.
[00:31] #scummvm: mode change '+o clone2727' by ChanServ!ChanServ@services.
[00:31] <-- droid2727 left irc: Quit: Statement: Good-bye, meatbags!
[00:41] <-- Smartnow left irc: Quit: Smartnow
[00:47] syke (matt@nat-sv.bluecoat.com) left #scummvm.
[00:50] --> Smartnow joined #scummvm.
[00:58] <-- SylvainTV left irc: Ping timeout: 240 seconds
[01:08] <-- JimmyJ left irc: Read error: Connection reset by peer
[01:12] <CIA-26> Oystein Eftevaag master * r78baa27 / backends/midi/coreaudio.cpp : MACOSX: Moved an include a bit to make sure that some defines are there before they are actually used. - http://git.io/pFhZlA
[01:22] <-- Gentle left irc: Quit: No Ping reply in 180 seconds.
[01:22] --> Gentle joined #scummvm.
[01:55] <-- D0SFreak left irc: Ping timeout: 260 seconds
[02:05] <-- Kaidane_ left irc: Ping timeout: 240 seconds
[02:17] --> cyco_ joined #scummvm.
[02:18] <-- cyco left irc: Ping timeout: 240 seconds
[02:18] Nick change: cyco_ -> cyco
[02:19] --> Kaidane joined #scummvm.
[02:26] --> Tron__ joined #scummvm.
[02:28] <-- Tron left irc: Ping timeout: 252 seconds
[02:38] --> criezy_ joined #scummvm.
[02:38] #scummvm: mode change '+o criezy_' by ChanServ!ChanServ@services.
[02:38] <-- criezy left irc: Read error: Connection reset by peer
[02:39] Nick change: criezy_ -> criezy
[02:42] <CIA-26> Thierry Crozat master * race3bf6 / (engines/sword1/animation.cpp engines/sword1/animation.h): (log message trimmed)
[02:42] <CIA-26> SWORD1: Use color for subtitles during cutscene
[02:42] <CIA-26> During the game, different colors are used for subtitles depending
[02:42] <CIA-26> which character is speaking. This commit tries to use the same colors
[02:42] <CIA-26> for the cutscene subtitles. The color to use has to be specified in the
[02:42] <CIA-26> subtitle file between the frame end and the start of text using @1, @2,
[02:42] <CIA-26> @3 or @4 (for George, George as a narrator, Nicole and Maguire
[02:42] <CIA-26> Thierry Crozat master * r4546ca0 / engines/sword1/animation.cpp : SWORD1: Do not look for color indexes in cutscenes when there is no subtitles - http://git.io/U9hhCQ
[02:43] <-- Mindkontrol left irc: Ping timeout: 276 seconds
[02:45] --> Mindkontrol joined #scummvm.
[02:48] <clone2727> criezy: Bah, makin' my pull request not automatically mergable :P
[02:54] Action: Endy notes his IRC client loves to block CIA for flooding. :)
[02:58] <clone2727> Endy: But everyone loves commit messages!
[02:59] <Endy> Its just these commit messages are too full of awesome for my IRC client to handle. :)
[02:59] <clone2727> Ah, that makes more sense
[03:27] <-- Kaidane left irc: Ping timeout: 255 seconds
[03:31] <-- Javacat left irc: Quit: Please, try the fish
[04:23] Nick change: Adys_ -> Adys
[04:37] <-- clone2727 left irc: Quit: later
[04:39] <-- Mindkontrol left irc: Quit: Lost terminal
[05:12] <-- Gentle left irc: Quit: No Ping reply in 180 seconds.
[05:14] --> Gentle joined #scummvm.
[05:32] <-- Gentle left irc: Quit: No Ping reply in 180 seconds.
[05:33] --> Gentle joined #scummvm.
[05:57] <-- Gentle left irc: Quit: No Ping reply in 180 seconds.
[06:01] Nick change: Adys -> adys`nosopa
[06:02] --> Gentle joined #scummvm.
[06:07] <-- Gentle left irc: Client Quit
[06:07] --> Gentle joined #scummvm.
[06:11] [vEX] <-- (~vex@h-46-59-104-222.na.cust.bahnhof.se) left irc: Ping timeout: 276 seconds
[06:11] --> ajax16384 joined #scummvm.
[06:11] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[06:16] [vEX] --> (~vex@h-46-59-104-222.na.cust.bahnhof.se) joined #scummvm.
[06:16] #scummvm: mode change '+o [vEX]' by ChanServ!ChanServ@services.
[06:16] <-- Gentle left irc: Read error: Operation timed out
[06:17] --> Gentle joined #scummvm.
[06:21] <-- Strangerke left irc: Quit: Bbl
[06:46] Nick change: Tron__ -> Tron
[07:11] Tron (~Tron@srbk-4db7f229.pool.mediaWays.net) got netsplit.
[07:11] Harekiet (~freenode@dhcp-077-250-029-247.chello.nl) got netsplit.
[07:12] Tron (~Tron@srbk-4db7f229.pool.mediaWays.net) returned to #scummvm.
[07:12] Harekiet (~freenode@dhcp-077-250-029-247.chello.nl) returned to #scummvm.
[07:16] <-- _sev left irc: Quit: This computer has gone to sleep
[07:24] <-- _marc` left irc: Quit: _marc`
[07:31] --> Strangerke joined #scummvm.
[07:31] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[07:31] <Strangerke> hi guys
[07:35] --> bgKa joined #scummvm.
[07:41] --> _sev joined #scummvm.
[07:41] <-- _sev left irc: Changing host
[07:41] --> _sev joined #scummvm.
[07:41] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[07:45] <-- _sev left irc: Client Quit
[07:56] --> bushwakko joined #scummvm.
[08:06] <madmoose> Ahoy-hoy.
[08:17] --> Vaikungfu joined #scummvm.
[08:24] --> ny00123 joined #scummvm.
[08:26] --> _marc` joined #scummvm.
[08:33] --> dreammaster joined #scummvm.
[08:33] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[08:51] <DrMcCoy> moo
[08:56] <madmoose> g'morning, doc.
[09:02] <-- ajax16384 left irc: Quit: i am quit

[09:03] <ST> DrMoocCoy?
[09:04] <-- Vaikungfu left irc: Ping timeout: 252 seconds
[09:04] --> _sev joined #scummvm.
[09:04] <-- _sev left irc: Changing host
[09:04] --> _sev joined #scummvm.
[09:04] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[09:04] <DrMcCoy> Something like that
[09:06] --> ajax16384 joined #scummvm.
[09:06] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[09:18] <-- Gentle left irc: Quit: No Ping reply in 180 seconds.
[09:18] --> Gentle joined #scummvm.
[09:39] <-- ScummBot left irc: Ping timeout: 252 seconds
[09:43] <-- criezy left irc: Quit: criezy
[09:49] --> ScummBot joined #scummvm.
[10:02] --> Vaikungfu joined #scummvm.
[10:03] --> giucam joined #scummvm.
[10:06] <-- Vaikungfu left irc: Ping timeout: 252 seconds
[10:19] --> Vaikungfu joined #scummvm.
[10:38] --> D0SFreak joined #scummvm.
[10:53] <-- D0SFreak left irc: Ping timeout: 276 seconds
[11:02] --> D0SFreak joined #scummvm.
[11:12] --> Herrman joined #scummvm.
[11:35] <-- D0SFreak left irc: Ping timeout: 248 seconds
[11:35] <-- Kirben left irc:
[11:50] <-- dreammaster left irc: Quit: Leaving for the day
[11:51] --> D0SFreak joined #scummvm.
[12:00] Nick change: Endy -> Endy|AFK
[12:27] <-- cyco left irc: Ping timeout: 252 seconds
[12:27] --> cyco_ joined #scummvm.
[12:38] <-- Herrman left irc: Quit: Die letzten Worte des Zauberlehrlings: "Eigentlich bin ich ja noch nicht soweit, aber..."
[12:38] <-- dhewg left irc: Ping timeout: 240 seconds
[12:48] <-- D0SFreak left irc: Read error: Connection reset by peer
[12:49] --> D0SFreak joined #scummvm.
[13:01] --> aquadran joined #scummvm.
[13:01] <-- aquadran left irc: Changing host
[13:01] --> aquadran joined #scummvm.
[13:01] #scummvm: mode change '+o aquadran' by ChanServ!ChanServ@services.
[13:04] <-- aquadran left irc: Client Quit
[13:05] --> aquadran joined #scummvm.
[13:05] #scummvm: mode change '+o aquadran' by ChanServ!ChanServ@services.
[13:28] Nick change: cyco_ -> cyco
[13:34] --> dhewg joined #scummvm.
[13:34] #scummvm: mode change '+o dhewg' by ChanServ!ChanServ@services.
[13:54] --> DOSFreak joined #scummvm.
[13:58] <-- D0SFreak left irc: Ping timeout: 240 seconds
[14:19] --> LordHoto joined #scummvm.
[14:19] #scummvm: mode change '+o LordHoto' by ChanServ!ChanServ@services.
[14:26] --> clone2727 joined #scummvm.
[14:26] #scummvm: mode change '+o clone2727' by ChanServ!ChanServ@services.
[14:50] <-- dtcrshr left irc: Ping timeout: 240 seconds
[14:54] --> dtcrshr joined #scummvm.
[14:58] --> Smedles_ joined #scummvm.
[14:58] <-- Smedles left irc: Read error: Connection reset by peer
[14:59] <-- ajax16384 left irc: Quit: ajax16384
[15:00] <-- dtcrshr left irc: Ping timeout: 240 seconds
[15:03] --> bushwakko_ joined #scummvm.
[15:03] <-- bushwakko_ left irc: Client Quit
[15:03] --> dtcrshr joined #scummvm.
[15:07] <-- bushwakko left irc: Ping timeout: 272 seconds
[15:23] <-- clone2727 left irc: Quit: bbl
[15:29] <-- _marc` left irc: Quit: _marc`
[15:31] [md5] --> (~md5@ppp-94-65-50-230.home.otenet.gr) joined #scummvm.
[15:31] [md5] <-- (~md5@ppp-94-65-50-230.home.otenet.gr) left irc: Changing host
[15:31] [md5] --> (~md5@unaffiliated/md5/x-729473) joined #scummvm.
[15:31] #scummvm: mode change '+o [md5]' by ChanServ!ChanServ@services.
[15:48] --> kettuz joined #scummvm.
[15:49] --> giucam_ joined #scummvm.
[15:49] <-- giucam left irc: Disconnected by services
[15:49] Nick change: giucam_ -> giucam
[15:49] --> ajax16384 joined #scummvm.
[15:49] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[15:55] <-- Strangerke left irc: Quit: Page closed
[15:56] --> bgKa1 joined #scummvm.
[15:58] <-- bgKa left irc: Ping timeout: 252 seconds
[16:00] <DrMcCoy> Hmm
[16:01] --> <DrMcCoy> joined Anti-SOPA blackout?
[16:03] --> droid2727 joined #scummvm.
[16:03] #scummvm: mode change '+o droid2727' by ChanServ!ChanServ@services.
[16:20] --> risca joined #scummvm.
[16:21] <-- giucam left irc: Remote host closed the connection
[16:21] --> MetalSnake joined #scummvm.
[16:22] --> giucam joined #scummvm.
[16:23] --> hennymcc joined #scummvm.
[16:24] --> Hkz joined #scummvm.
[16:24] #scummvm: mode change '+o Hkz' by ChanServ!ChanServ@services.
[16:25] --> Mataniko_ joined #scummvm.
[16:25] <-- Mataniko left irc: Ping timeout: 240 seconds
[16:26] Nick change: Mataniko_ -> Mataniko
[16:31] --> _marc` joined #scummvm.
[16:32] <-- X-Scale left irc: Remote host closed the connection
[16:32] --> Herrman joined #scummvm.
[16:37] --> X-Scale joined #scummvm.
[16:38] Nick change: X-Scale -> Guest1078
[16:45] <LordHoto> DrMcCoy: yeah :-P
[16:48] <[md5]> I agree on this, tbh
[16:48] <[md5]> anti-SOPA = good
[16:48] <[md5]> http://en.wikipedia.org/wiki/Main_Page <--- blacked out for today
[16:49] --> <LordHoto> joined the protest too!!!!1111
[16:50] --> towerk_ joined #scummvm.
[16:52] <wjp> no objections from me if you want to do something
[16:53] <[md5]> we could make a black overlay over the site and forums
[16:54] <[md5]> I'll whip up something fast
[16:54] <-- hennymcc left irc: Ping timeout: 252 seconds

[17:29] --> risca joined #scummvm.
[17:29] <-- risca left irc: Read error: Connection reset by peer
[17:35] <-- KD0ETH_Wo left irc: Quit: Bye!
[17:44] --> hennymcc joined #scummvm.
[17:44] --> SylvainTV joined #scummvm.
[17:44] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[17:52] <-- Guest1078 left irc: Ping timeout: 272 seconds
[17:53] --> Guest1078 joined #scummvm.
[17:58] <[md5]> well, I got something done for SOPA, I guess I'll stick it a temporary copy
[18:04] <-- _sev left irc: Quit: This computer has gone to sleep
[18:06] --> mufflon joined #scummvm.
[18:16] --> Schnaks_ joined #scummvm.
[18:17] <-- Schnaks left irc: Ping timeout: 276 seconds
[18:17] Nick change: Schnaks_ -> Schnaks
[18:27] mufflon (holle@p5B17F9D7.dip.t-dialin.net) left #scummvm ("Ex-Chat").
[18:38] <-- James|GlideM left irc: Read error: Connection reset by peer
[18:43] --> criezy joined #scummvm.
[18:43] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[18:44] --> sirlemonhead joined #scummvm.
[18:46] <[md5]> here's a demo SOPA protest page: http://www.scummvm.org/sopa/index.html
[18:50] --> Krest joined #scummvm.
[18:50] <Krest> hi
[18:51] <LordHoto> hey Krest
[18:53] --> KD0ETH_Wo joined #scummvm.
[18:54] --> TAS-2012v joined #scummvm.
[18:56] <-- TAS_2012v left irc: Ping timeout: 252 seconds
[18:57] <-- Krest left irc: Quit: Krest
[19:15] --> Javacat joined #scummvm.
[19:23] <Mataniko> md5: looks good, but i'd use max! :)
[19:23] --> syke joined #scummvm.
[19:23] <droid2727> Yes, let's use Fingolfin
[19:24] <Mataniko> actually i'd use the forefathers from dott :)
[19:24] <droid2727> heh, that's not a bad idea :P
[19:24] <droid2727> "George found SOPA in the suggestion box"
[19:24] <syke> as a punching bag? ;)
[19:28] <Mataniko> and change the text to "Imagine a world without Tentacles"
[19:31] --> risca joined #scummvm.
[19:32] Nick change: Javacat -> NotJavacat
[19:33] <DrMcCoy> not Javacat? :(
[19:36] <-- hennymcc left irc: Quit: Verlassend
[19:37] <[md5]> Mataniko: feel free to write any nice anti-SOPA text, that actually makes sense :P
[19:37] <Mataniko> i would but i have to step out in a couple of mintues
[19:39] <Mataniko> don't get me wrong, i think the wikipedia text is fine
[19:39] <Mataniko> with the adjustments

[19:53] <-- giucam left irc: Read error: Operation timed out
[19:53] <-- LordHoto left irc: Quit: bbl
[19:55] <-- Tomaz^ left irc:
[20:00] --> giucam joined #scummvm.
[20:02] <dhewg> fuzzie: can you do the droid builds?
[20:02] <fuzzie> yep
[20:02] <dhewg> k nice
[20:02] <fuzzie> in fact was just discussing them :)
[20:02] <dhewg> ;)
[20:03] <fuzzie> dhewg: btw, was chatting to friends about their Wiis earlier, and several mentioned how awesome the ScummVM port was
[20:03] <fuzzie> as motivation to buy one
[20:03] <dhewg> heh
[20:03] <-- Hkz left irc: Remote host closed the connection
[20:03] <fuzzie> same for people on IRC, but, well, they're clearly nerds
[20:03] <fuzzie> but good work I guess :)
[20:04] <dhewg> well, i do think that i got to a pretty good that
[20:04] <dhewg> if only it wouldnt bitrot
[20:04] <dhewg> :D
[20:04] <dhewg> btw
[20:04] <dhewg> i just found a 11 month old patch for that
[20:04] <dhewg> upgrade to the latest toolchain
[20:04] <dhewg> but that requires a newer glibc
[20:05] <dhewg> which we do not have on builtbot
[20:05] <dhewg> did anything happen about that?
[20:05] <dhewg> and i meant "good state"
[20:05] <fuzzie> not other than a failed attempt to do some horrible preload thing which was doomed to failure
[20:05] --> Hkz joined #scummvm.
[20:05] #scummvm: mode change '+o Hkz' by ChanServ!ChanServ@services.
[20:06] <dhewg> whatever if was, attempting to hack around such a requirement is doomed to fail spectacular
[20:06] <dhewg> *it
[20:06] <-- ajax16384 left irc: Read error: Connection reset by peer
[20:08] Nick change: NotJavacat -> Javacat
[20:09] Nick change: Javacat -> NotJavacat
[20:10] Nick change: NotJavacat -> Javacat
[20:10] Nick change: Javacat -> NotJavacat
[20:12] <-- dtcrshr left irc: Ping timeout: 240 seconds
[20:13] <-- Smartnow left irc: Ping timeout: 252 seconds
[20:14] --> Smartnow joined #scummvm.
[20:20] --> _sev joined #scummvm.
[20:20] <-- _sev left irc: Changing host
[20:20] --> _sev joined #scummvm.
[20:20] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[20:21] Action: Hkz attempts to summon Strangerke
[20:26] --> dtcrshr joined #scummvm.
[20:28] <-- dtcrshr left irc: Max SendQ exceeded
[20:32] --> dtcrshr joined #scummvm.
[20:33] <-- dtcrshr left irc: Max SendQ exceeded
[20:41] Nick change: notajf|offline -> notajf
[20:41] <notajf> Hi! I'm a volunteer! Quick, make me do something before I vanish!
[20:45] Nick change: notajf -> vanished_volunte
[20:45] <fuzzie> well, I was going to point at the TODO list
[20:45] <fuzzie> but then I looked at the Android section and ended up removing all the entries because they were all done a long time ago
[20:46] <fuzzie> and other entries include things like kFeatureAutoComputeDirtyRects which has been gone for a year and a half
[20:47] <vanished_volunte> :/
[20:47] Nick change: vanished_volunte -> notajf
[20:47] --> LordHoto joined #scummvm.
[20:47] #scummvm: mode change '+o LordHoto' by ChanServ!ChanServ@services.
[20:47] <dhewg> add a "cleanup TODO" on top of TODO? :P
[20:47] <notajf> haha
[20:58] <-- DOSFreak left irc: Ping timeout: 240 seconds
[21:00] <-- Guest1078 left irc: Ping timeout: 272 seconds
[21:04] --> Guest1078 joined #scummvm.
[21:06] --> Tomaz^ joined #scummvm.
[21:15] Nick change: notajf -> notajf|offline
[21:19] --> tsoliman joined #scummvm.
[21:19] #scummvm: mode change '+o tsoliman' by ChanServ!ChanServ@services.
[21:22] <sirlemonhead> hmm if i'm looking at the IDA disassembly for a function which returns int, should I be seeing ax pushed onto the stack just before the function returns?
[21:22] <fuzzie> usually calling conventions just return the value in ax
[21:23] <sirlemonhead> so its assumed if its got a return type? that's good - this now makes sense
[21:23] <fuzzie> i mean pushing onto the stack before a return would be a pretty horrible calling convention on x86, since it has the return address on the stack and so you'd have to actually pop it as part of the return
[21:23] <fuzzie> it does happen but probably not in any code you're likely to see :)
[21:23] <-- tsoliman left irc: Client Quit
[21:23] <sirlemonhead> i saw a return type and them setting ax, but doing nothing else with it before returning. so that's all good :)
[21:24] <syke> the really early games had hand-coded assembly, which sometimes had really odd-looking patterns
[21:24] <fuzzie> there are compilers which would generate code that snaffles values out of the dead stack under some circumstances
[21:24] <syke> going from reversing stuff compiled post-1990 to pre-1990 can be a bit jarring
[21:24] <sirlemonhead> this one is from 1991 or so and was compiled with borland c++..so I assume they didn't use any/much hand coded assembly.......maybe :P
[21:25] <sirlemonhead> I have some functions where it seems after every line, there's a new "loc_" tag..is that normal? I see it a lot in functions that reside in overlays..
[21:33] <sirlemonhead> is ax pushed into the stack *before* a function call with a return type?
[21:33] --> DOSFreak joined #scummvm.
[21:33] <syke> sirlemonhead: sure, that's allowed.
[21:33] <syke> for instance, if the return value of one function is passed to another
[21:33] <syke> you might see
[21:34] <syke> call 0xd00d
[21:34] <syke> push ax // implicitly set by the function @ 0xd00d
[21:34] <sirlemonhead> but it wouldnt be normal if it was just a straightward function call? not nestted or anything?
[21:34] <syke> call 0xbabe // probably uses the last push as a parameter
[21:47] <-- bgKa1 left irc: Quit: Leaving.
[21:50] <-- MetalSnake left irc: Quit: MetalSnake
[21:52] Nick change: Endy|AFK -> Endy
[21:54] <droid2727> Endy!
[21:55] <Endy> clone-clone!
[21:56] <syke> .dcc send endy cornjob.jpg
[21:56] <syke> oopz
[21:56] <Endy> rofl
[21:57] <Endy> Family channel, family channel heh
[21:57] <syke> what? I was trying to send you a picture of my parent's produce farm
[21:57] <syke> you pervert
[21:59] <Endy> Can't deny that :P
[22:00] <syke> great to see that qfg3 appears to have really high compatibility
[22:05] --> digitall joined #scummvm.
[22:05] #scummvm: mode change '+o digitall' by ChanServ!ChanServ@services.
[22:09] <syke> what's the latest with qfg4?
[22:09] <[md5]> it's in progress, same as the rest of the SCI2 games
[22:09] <syke> any progress from 6 months ago or so?
[22:09] <[md5]> well, not much, I'm afraid
[22:10] <syke> I remember the opening cinematic played, but the menus were wonky and character selection were a bit wonky
[22:10] <[md5]> I'm hoping to get some good progress once I figure out the SCI3 executables
[22:10] Nick change: giucam -> LeMans
[22:11] <digitall> Evening... To be clear, the LD_PRELOAD stuff for WebOS didn't fail... I just can't quite work out a working solution...
[22:12] <digitall> It would be "easier" to just upgrade Buildbot's glibc... but I'm concerned that this would break things.
[22:13] Action: digitall is not a Debian expert...
[22:13] Nick change: LeMans -> giucam
[22:14] <fuzzie> digitall: it didn't give you a bunch of random errors whenever you tried running a binary?
[22:14] <digitall> fuzzie: This is all covered by the message I sent to scummvm-devel at the time...
[22:14] <digitall> But no-one commented then.
[22:15] <fuzzie> well, I'm sure several of us read it
[22:15] <digitall> *grumble*
[22:16] <fuzzie> but it's not particularly clear
[22:16] <fuzzie> I mean, it *looks* like it failed terribly and you got a bunch of random errors whenever you tried running a binary.
[22:16] <CIA-26> Strangerke master * r00a9d39 / (5 files in 2 dirs): TSAGE: R2R - Implement scene 1700 - http://git.io/T54ajQ
[22:16] <digitall> Well, I think part of the problem is that I really needed a way to disable ccache or similar
[22:17] <digitall> or force a full make clean...
[22:17] <fuzzie> i mean
[22:17] <fuzzie> you know that you're not actually running out of RAM there, right?
[22:17] <digitall> Note that stable-wince is currrently broken, but that is due to the compiler change and the cache not re-running configure
[22:18] <digitall> fuzzie: I don't really know anything, I'm no expert on doing this... just no-one else seems to care to try :-)
[22:18] <fuzzie> well, LD_PRELOADing glibc is sort of at the height of "unsupported crazy things" :p
[22:19] <digitall> Well, I think the issue is either a memory issue... or it may be due to a missing library dependency of debian's glibc build.
[22:19] <digitall> I was under the impression that it was static, but there were a few dependencies and I'm not sure if they are being met...
[22:20] <digitall> Since the compiler is run via the buildbot... not sure if STDERR messages are being passed up to stdio
[22:21] <fuzzie> i mean, the configure stage is failing way before it gets near a compiler
[22:21] <digitall> Might be ccache... Not sure if there is a way to force a clean build?
[22:21] <fuzzie> it's just failing to run anything at all
[22:22] <digitall> fuzzie: IIRC, I put back the old toolchain code after I failed... not sure why it is now failed altogether
[22:22] <fuzzie> the buildbot output was giving an empty value for LD_LIBRARY_PATH which doesn't make much sense but is slightly suspicious
[22:23] <-- Dominus left irc: Quit: Leaving.
[22:23] <digitall> I did have some hell there trying to work out what LD_LIBRARY_PATH should be in the default case.
[22:23] <digitall> Check my commits to buildbot SVN
[22:23] <digitall> I may have made a mistake.
[22:24] <DJWillis> digitall: if we do upgrade glibc on the build box I can start using the OpenPandora toolchain again as that needs a newer glibc and I could never get LD_PRELOAD to work for that ;). Infact considering both the WebOS GCC and OpenPandora GCC origionate from OpenEmbedded recipes I expect the same fix would work for both ;).
[22:25] <fuzzie> well, LD_PRELOAD rather than just LD_LIBRARY_PATH for old glibc seems a bit crazy and not likely to work anyhow
[22:25] <fuzzie> but not best time for me to think about it
[22:26] Nick change: giucam -> LeMans
[22:26] <DJWillis> fuzzie: yep, I think I just tinkered with lib path but I just gave up as it was on a big list of things I could make a failure of ;)
[22:28] <-- LeMans left irc: Quit: Ciao!
[22:28] <dhewg> i think our debian is at EOL anyway
[22:29] <dhewg> best thing is to just do an upgrade
[22:29] <digitall> dhewg: Yes... It needs a full backup, single user and then upgrade...
[22:29] <fuzzie> phft, single user :-p
[22:29] <[md5]> out of curiosity, why debian?
[22:29] <dhewg> yeah, single user :P
[22:29] <fuzzie> [md5]: why not? :p
[22:29] <dhewg> vm backup, life upgrade, remote reboot!
[22:30] <DJWillis> dhewg: agreed, and I expect the fallout will not be all that bad, infact with 1.4.1 done I guess now is a good time to do it (if we have issues we can fix without a massive project impact).
[22:30] <[md5]> well, we could use ubuntu?
[22:30] <dhewg> nooo
[22:30] <dhewg> debian!
[22:30] <fuzzie> what dhewg said
[22:30] <[md5]> or, for something more server-like, CentOS
[22:30] Action: [md5] loves CentOS
[22:30] <fuzzie> you do?
[22:30] --> giucam joined #scummvm.
[22:30] <[md5]> yeah
[22:30] <dhewg> debian!
[22:30] <DJWillis> [md5]: for what we use it for what would bloatbuntu give us?
[22:30] <dhewg> DJWillis: yeah hopefully, i mostly worry about /opt/toolchains
[22:31] <[md5]> DJWillis: well, more bleeding edge libraries, perhaps?
[22:31] <fuzzie> image it and try an upgrade locally?
[22:31] <digitall> For my two pins, I would say the choice is between Debian, Arch and Gentoo.
[22:31] <[md5]> they do have a long LTS plan
[22:31] <[md5]> Gentoo? O_o
[22:31] <digitall> But I would stick with Debian and try to upgrade first... Changing distros is probably not "worth" it.
[22:31] <syke> [md5]: take a quick poll to see what OS the contributing developers, then use something other than that. it'll help increase coverage, at the least.
[22:32] <dhewg> heh gentoo, like that poor vm isnt busy enough compiling all day long
[22:32] <-- giucam left irc: Client Quit
[22:32] <[md5]> that's what I was about to say
[22:32] <DJWillis> dhewg: yep, that said, I have done a lot of dist upgrades recently with 10-15 toolchains installed, ARM, MIPS, x86 etc. and all survived the upgrades, including some shitty 2.95 toolchains I exected to have to move to a VM ;) (don't ask WHY I have them)
[22:32] <[md5]> gentoo sounds like we're dooming that machine to eternal building... including the OS :p
[22:32] <fuzzie> [md5]: an LTS ubuntu having gone this long would be much worse library-wise
[22:32] <digitall> True, but Gentoo's support for cross compilers is very good: http://www.gentoo.org/proj/en/base/embedded/handbook/
[22:33] <fuzzie> but the VM is quite seriously limited in disk space, right?
[22:33] <digitall> Yes, I would recommend sticking with Debian... or Arch at the outside.
[22:33] <dhewg> digitall: yeah, that framework is nice, but wont really help for all our console toolchains
[22:33] <fuzzie> and rule #1 of crosscompiling is that trying to crosscompile without the exact binary toolchain which is known to work doesn't work
[22:33] <digitall> As you say Gentoo would require too much CPU.
[22:34] <[md5]> I haven't used Arch, I've heard good things about it though
[22:34] <digitall> Arch has the advantage that like Gentoo, it is a rolling distro...
[22:34] <syke> it sucks that GCC can't produce a profile-optimized cross-compiler. I was looking into this a few weeks ago as a way to further optimize a build at work.
[22:34] <DJWillis> digitall: I find Gentoo a pig for cross-compilers as they have ebuild for random snapshots that often don't work ;)
[22:34] <syke> native build time reduced by 20% with profiledbootstrap
[22:35] <-- ny00123 left irc: Quit: Leaving
[22:35] <digitall> DJWillis: Not sure when you tried it, but the point is that Crossdev allows you to dynamically make a toolchain by specifying the binutils, gcc etc.
[22:35] <DJWillis> I just vote we update the box and see what fails, is it worth moving to the mailing list to actually get this work planned in (in all seriousness), I am willing to help if I can.
[22:36] <digitall> http://en.gentoo-wiki.com/wiki/Crossdev
[22:36] <[md5]> great :)
[22:36] <digitall> DJWillis: I agree... but a full backup first please.
[22:37] --> Kaidane joined #scummvm.
[22:37] <fuzzie> well, if it's possible to back it up, it should be pretty trivial to just put it in someone's local VM and upgrade it and see
[22:37] <digitall> fuzzie: Agreed.
[22:38] <dhewg> backup should be no problem, it just has to be done on qvist's end
[22:38] <fuzzie> ah, well, see, despair.
[22:38] <[md5]> hm, if that is a VM, it should be relatively straightforward to set up a new one, and copy the stuff we want from the current one, right?
[22:39] <dhewg> considering we might have to restore the easy approach is just to power down the vm and backup the hdd images
[22:40] <[md5]> are there any other VMs running on that server?
[22:40] <dhewg> not sure
[22:40] <dhewg> but i know its vmware
[22:44] <[md5]> anyway, so the popular choices are Debian and Arch?
[22:44] <[md5]> and why do you hate CentOS? :P
[22:44] <digitall> Well, Debian since it is the "normal" upgrade path... Arch because it is similar i.e. build only what you need, well documented
[22:45] <dhewg> isnt centos using rpm?
[22:45] <digitall> non-commercial, rolling release.
[22:45] <digitall> The last 2-3 probably don't apply to Debian :-)
[22:45] <dhewg> rpm is a no-go, but maybe thats just me :P
[22:46] <digitall> No, I would tend to agree...
[22:46] <digitall> Though this is mainly a "taste" thing... CentOS is Red Hat Non-Commercial...
[22:46] <fuzzie> well, RHEL is quite nice if you have the support contracts and you need to run enterprise-ish stuff on it
[22:46] <digitall> It also comes down to support.
[22:46] <fuzzie> but there's kinda a lack of other reasons to do it, and then CentOS is like RHEL only without even being able to keep up with security releases, no?
[22:47] <digitall> *shrug*
[22:48] <fuzzie> my personal view on distros is "anything is fine as long as I don't have to administer it", so there is my bias
[22:49] <digitall> *ditto*
[22:49] Action: Endy has to administer CentOS VPS' for customers with cpanel.
[22:50] <fuzzie> i'm sorry. have a cookie.
[22:50] <digitall> I would prefer something with rolling support and we can generally just run the apt-get update regularly... or similar
[22:50] <Endy> I've since developed a even stonger Debian bias than I already had /oops :)
[22:50] <Endy> fuzzie: Ooooh Cookie!
[22:51] --> Strangerke joined #scummvm.
[22:51] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[22:52] <Endy> Strangerke: Good <Insert Local Timezone>!
[22:52] <fuzzie> Hkz: quick!
[22:52] <Strangerke> Hi Endy :) Thanks, good morning to you too :)
[22:53] <digitall> Strangerke: Evening... With your PR hat on, what do you think about the SOPA Blackout? Talk to [md5]
[22:53] <Hkz> fuzzie, yeah, i'm still not very good with summonings :P
[22:55] <fuzzie> i think at some point you will be a bit late on the SOPA blackout front, reddit will be back in 2 hours
[22:55] Action: Endy wonders what the webstats look like after the Ars article :P
[22:55] <fuzzie> Endy: there's a nice bump on the android market installs graph
[22:55] <Endy> fuzzie: Awesome!
[22:55] <-- Newa left irc: Read error: Operation timed out
[22:55] <[md5]> https://sourceforge.net/projects/scummvm/files/stats/timeline
[22:56] <fuzzie> we're still only at 24k currently-installed devices though. need more androiders!
[22:56] <droid2727> fuzzie: Including buildbot builds? :P
[22:56] <fuzzie> droid2727: no, but who'd install those?
[22:56] <fuzzie> silly.
[22:57] <[md5]> anyway, I mentioned CentOS cause it's basically a free enterprise system, hence less things to worry about... but yes, CentOS is a free RedHat, thus RPM-based
[22:57] <[md5]> but in the end of the day, yum does a similar job as aptitude does (even though they're based on different package systems)
[22:58] aquadran (aquadran@scummvm/undead/aquadran) left #scummvm.
[22:58] <dhewg> omg yum
[22:58] <dhewg> i hate that with a passion
[22:58] <fuzzie> people still use aptitude? :p
[22:58] <Endy> Some of the Redhat guys working on RHEL I've talked to this week don't have a very high (personal) opinion of CentOS either :) [md5]
[22:58] <[md5]> apt-get install foo/yum install foo
[22:58] <Endy> apt-get build-deps foo?
[22:59] <Endy> :)
[22:59] <[md5]> heh :p
[22:59] <fuzzie> yumyumdownload foo --please-don't-break-this-time
[22:59] <syke> the commandline interface usually isn't the problem -- it's the quality of the package meta-data, especially with regard to dependencies
[22:59] <[md5]> Endy: so what flavor would you recommend for a server OS?
[22:59] <[md5]> ok, sorry, that should be addressed to the channel, not just Endy
[22:59] <fuzzie> the eternal question is: what would you want to *do* with the server?
[23:00] <[md5]> heh :)
[23:00] <syke> what does server OS mean in this case? it sounds like you really want to optimize for ease of cross-compilation
[23:00] <fuzzie> if your answer is "run a whole heap of crosscompile toolchains which run nicely on Debian right now" then I think there's not much of a question
[23:00] <syke> not to necessarily have a "server"-optimized kernel, etc
[23:00] <Endy> [md5]: My bias is Debian - note this is largely due to the number of servers (physical and virtual) I have to maintain daily.
[23:00] <[md5]> I guess the usual LAMP stack can run on pretty much anything
[23:00] sirlemonhead (bduncan22@86-45-2-215-dynamic.b-ras2.prp.dublin.eircom.net) left #scummvm.
[23:00] <fuzzie> desks, tables, chests, etc?
[23:00] <[md5]> but for cross-compilation things are different
[23:01] <Endy> The only real thing thats shaken that bias was the _DUMB_ OpenSSL issue a while back.
[23:01] <Endy> Who needs to seed their random number generator? :P
[23:01] <[md5]> :p
[23:01] <[md5]> Sony!
[23:01] <syke> Endy: and LIBRARY_PATH
[23:01] <Endy> syke: Aye, that too.
[23:01] <syke> I want to skullrape the motherfuckers that had the idea, and the people who somehow allowed it to happen
[23:01] <digitall> I think Endy should volunteer to host a new VM instance and maintain it :-)
[23:02] Action: digitall runs
[23:02] <[md5]> he should indeed. Thank you for volunteering Endy :p
[23:02] Action: [md5] runs too
[23:02] <syke> WIRE HANGERS!!!$@!$@!$
[23:03] <Endy> For what, buildbot?
[23:03] Action: [md5] sticks marshmallows on syke's hangers and lights a bonfire :p
[23:03] <Endy> I've offered dual-xeon boxes for buildbot before. I have a strange suspicious people think I'm lazy and unresponsi................... ve
[23:03] <Endy> :)
[23:03] <digitall> Yes... The current VM is very slow... and the Debian host OS really needs an upgrade.
[23:04] <syke> what is the VM? bochs? qemu?
[23:04] <fuzzie> didn't someone mention vmware above?
[23:04] <[md5]> it was mentioned that it's VMWare
[23:04] <syke> sorry, missed that
[23:04] <Endy> Ugh.
[23:04] <fuzzie> Endy: i think you may possibly be far more responsive than the current vm owner
[23:04] <digitall> Aww... Well, a plurality of buildbots would not be bad... and there is no need to port the VM.
[23:04] <[md5]> so I would assume vsphere/esxi
[23:04] <Endy> fuzzie: Whos that?
[23:04] <[md5]> qvist
[23:05] <Endy> I can offer Xen VMs or a physical colo - I'd probably go physical for buildbot.
[23:05] <digitall> Just instance a new VM, add a new host OS and then copy the buildbot config, files and toolchains across....
[23:05] <Endy> LeChuck runs on one of the VMs - takes me about 15 minutes to instance a new one up :P
[23:05] <[md5]> and give everyone root access and security audit tools for added fun
[23:08] <Endy> Well, I'm back in the office next week - I might see what h/w I can throw together and make an offer.
[23:08] <digitall> OO.... Root Access to LeChuck!!!
[23:09] <digitall> Hey LeChuck... Want a ROOT BEER?
[23:09] <Endy> Given the number of platforms being built, I'm sure more grunt can't help :)
[23:09] <syke> Endy: I can pay for a RAM upgrade, if that helps. running compiles on a RAM disk cut my build times in half for GCC bootstrap.
[23:09] <droid2727> we can have a whole army of LeChucks!
[23:10] <Endy> syke: RAM likely to be the only issue - I've an abundance of dual-Xeon 2.3Ghz boxes lying around, but they're DDR2 ECC.
[23:10] <Endy> DDR2 ECC is expensive nowdays, and my hoarded pile is low :)
[23:11] <digitall> Anyway, Endy thanks and I hope you can sort out something.
[23:11] <syke> Endy: how many slots are free?
[23:12] <Endy> syke: I think most of them are 8-slot boards. I'm literally in the wrong state at the moment, so can't physically check :) (@ linux.conf.au for the week)
[23:12] <syke> ok
[23:12] <syke> let me know, I'm happy to paypal the funds
[23:12] <syke> looks like it's ~US$90 for a 4GB DDR2 ECC stick
[23:13] <Endy> Thats a lot less than we paid a few weeks ago
[23:13] <syke> I assume a pair (at least) will be needed to supplement whatever is existing?
[23:13] <syke> http://www.amazon.com/Kingston-ValueRAM-667MHz-Buffered-Desktop/dp/B00134A5PW/ref=pd_sim_sbs_e_1
[23:13] <Endy> Huh awesome. Why didn't we check Amazon? :P
[23:13] <syke> I did a google search and it was one of the links
[23:13] <Endy> syke: Aye, will need matching paired ram. Will be in touch anyway.
[23:13] <syke> ok
[23:13] <Endy> I'll run the offer past the Core Team first :)
[23:14] <syke> what's the size of the sources+output for the largest build done on the server?
[23:14] <digitall> syke: I'm not sure we have that kind of profiling information.
[23:14] <digitall> Anyone?
[23:15] <digitall> The current VM has 2G Physical and 1G Swap
[23:15] <syke> I'll check on this machine to get a ROM
[23:16] <digitall> So I would suspect we can at least do better than that.
[23:16] <-- TAS-2012v left irc: Ping timeout: 240 seconds
[23:17] <Endy> Heh, easily.
[23:17] <Endy> digitall: What does /proc/cpuinfo show, grunt-wise?
[23:17] Action: Endy back in a bit - switching rooms for a different talk
[23:17] <syke> is LTO being used for any builds? that can cause slightly higher RAM usage spikes during a build
[23:18] <fuzzie> i think you may be vastly overestimating the technical sophistication of our builds there
[23:19] <syke> LTO is just a flag
[23:20] <syke> looks like disk usage is ~650MB in the default build
[23:20] <syke> testing now with -O3 -flto -g
[23:20] <fuzzie> I mean, the native builds are almost irrelevant to our build times usually.
[23:21] <digitall> Endy: model name : Intel(R) Xeon(TM) CPU 2.80GHz
[23:21] <digitall> bogomips : 5585.00
[23:22] <-- echelog-1 left irc: Remote host closed the connection
[23:23] <qvist> fuzzie: perhaps.. the hardware is indeed not that new on the buildbot
[23:23] --> TAS_2012v joined #scummvm.
[23:24] <qvist> adn i think it's all about the cpu, and I can't upgrade that for now
[23:24] --> echelog-1 joined #scummvm.
[23:25] <fuzzie> well, it's more important having it be reliable than it be fast, I think
[23:25] <fuzzie> and it really has just kept going and going
[23:26] <fuzzie> does anyone know if it would be possible to just add buildslaves in some way where they could just handle builds if available?
[23:26] <syke> qvist: do you have swap statistics for a given build?
[23:27] <qvist> I have no statistics, I only maintain the hardware..
[23:27] <syke> I've had great luck just adding RAM and then building in a RAM disk, even on single-core machines
[23:27] <digitall> qvist: I assume we are responsible for the Host OS?
[23:27] <qvist> if ram is the issue, I think I can add some more
[23:27] <qvist> digitall: yes
[23:28] <syke> with -flto -O3 -g, the build is ~1.7GB
[23:28] <fuzzie> well, if we take a copy of the VM contents to test an upgrade, I imagine we can also test whether it would be helped by more RAM..
[23:28] <digitall> Hmm... The idea of having two replicated and synced buildbots could be a damn good idea...
[23:29] <digitall> Northern and Southern Hemisphere! :-)
[23:29] <qvist> I think ram can be added on the fly
[23:29] <digitall> Then can we get a Volcano Lair... and a crate of 20 assorted white cats :-)
[23:30] <-- Hkz left irc: Quit: sleep
[23:30] <fuzzie> digitall: well, I was just wondering whether we could simply offload some of the heavier builds
[23:30] <qvist> I can provide the vmdk file or do you want to copy only the files?
[23:30] <-- cyco left irc: Quit: cyco
[23:30] <digitall> qvist: Thanks.. The main issue at is maintenance of the Host OS and speed...
[23:31] --> Kirben joined #scummvm.
[23:31] #scummvm: mode change '+o Kirben' by ChanServ!ChanServ@services.
[23:31] <qvist> only scummvm on the jost, so you get all the cpu, I think memory can be added
[23:31] <fuzzie> I could just throw the vmdk file(s) into something, but I wouldn't have time until next week, so am maybe not the good choice here if there are people actually motivated.
[23:32] <digitall> The VMDK file would be the easiest way, but there is more on the Box than just buildbot and I think the Core Team should be consulted before we throw stuff around.
[23:32] <digitall> Is the VMDK everything including sda2 i.e. all 50G?
[23:32] <fuzzie> yes, private passwords/keys/etc on the disk
[23:32] <qvist> have to sllep now but tell me if I should do something
[23:33] <fuzzie> qvist: thanks!
[23:33] <digitall> Well, I think Endy will send a message to scummvm-devel...
[23:33] <digitall> qvist: Thanks, probably will be a minor discussion on scummvm-devel about the way forward here.
[23:34] <digitall> i.e. so WebOS, OpenPandora and future toolchains can be supported...
[23:34] <digitall> and we can add builders for more of the ports.
[23:34] <qvist> k, you have to mail me also since I don't "follow" scummvm-devel, night
[23:35] <digitall> ok
[23:50] --> sirlemonhead joined #scummvm.
[23:52] <Kirben> What is really so bad about SOPA? just been looking through notes on wikipedia, after protest suggestion of ScummVM web site. The idea of web sites having to police themselves is actually good, I'm tried of the endless links to illegal software/videos when searching these days, and web pages with only links to illegal stuff should be just as liable.
[23:53] <syke> Kirben: it's who decides what it infringing , and the speed at which ISPs would be forced to comply with takedowns, that is the problem
[23:54] <droid2727> and destroying the entire DNS system
[23:54] <syke> 14 years ago, when my old band started putting up MP3s of music we had written, produced, and published ourselves, we got dozens of automated takedown notices
[23:54] <syke> and so did the admin of the box/domain we were hosted on
[23:54] <syke> we were lucky -- the yak.net admins knew it was horseshit
[23:54] <syke> most people, especially nowadays, are not that lucky
[23:56] <-- SylvainTV left irc: Ping timeout: 240 seconds
[23:56] <digitall> It is similar to the patent issue... It will basically end up as a barrier to market entry to small players.
[23:56] <Kirben> So an immeditate take down is required, based on any complaint, rather than any verification at all? I din't know they were using automated take downs now.
[23:56] <syke> youtube has dozens of examples of legitimate uses of works being taken down
[23:57] <syke> due to secret agreements with UMG, Warner, Sony, etc
[23:57] <syke> and the remediation path is almost non-existant
[23:57] <digitall> Plus I don't want to be extraordinarily renditioned...
[23:57] <digitall> http://www.bbc.co.uk/news/uk-england-south-yorkshire-16544335
[23:58] <syke> the killer was a month or two ago when a bunch of rap stars got together to make a protest video
[23:58] <syke> which, without any legitimate reason, got taken off of youtube
[23:58] <Kirben> I have noticed several regular users reviews moving off YouTube, due to copyright reasons. But they were making use of scenes from movies/TVs, not sure on the legal side of that.
[23:59] <syke> recent fair use clarifications make it pretty clear, but it's not even about the effects on infringing content
[23:59] <syke> that is definitely a problem that needs to be solved
[23:59] <syke> it's about the non-infringing content
[00:00] --- Thu Jan 19 2012