[00:04] Kirben (kirben@c58-107-187-179.brasd3.vic.optusnet.com.au) joined #scummvm. [00:04] #scummvm: mode change '+o Kirben' by ChanServ!ChanServ@services. [00:06] ST (~ScottT@25.105.96.58.static.exetel.com.au) left irc: Ping timeout: 246 seconds [00:06] ST (~ScottT@25.105.96.58.static.exetel.com.au) joined #scummvm. [00:06] #scummvm: mode change '+o ST' by ChanServ!ChanServ@services. [00:30] Littleboy (~littleboy@c-76-24-3-227.hsd1.ma.comcast.net) joined #scummvm. [00:32] Action: DrMcCoy invites people to slap him [00:33] Action: ST slaps DrMcCoy for some reason [00:33] I got the wrong palette for the Saturn DS2 boxes :P [00:33] So they don't look as bad now? [00:34] There's two palettes in initial.idx/initial.glu, not just one as I thought [00:34] So no BGR555? [00:34] ST: Yup [00:34] digitall: They are BGR555 [00:34] Ah, but two different ones :-) [00:34] digitall: I just took MENU.PAL. The correct one for the box is PARTS.PAL (which I haven't seen when I looked for a palette) [00:35] :-~ [00:35] wee~ [00:35] free slapping [00:35] Lightkey: Slap me [00:35] http://drmccoy.de/zeugs/ds2saturnboxcorrect.png [00:36] I'll refrain from slapping since I've had a couple of senior moments today already... Grumble... data1.cab [00:40] DrMcCoy: Much nicer :) [00:44] Very nice :) [00:45] Strangerke_ (~a@cable-85.28.89.119.coditel.net) joined #scummvm. [00:46] Strangerke (~a@cable-85.28.89.119.coditel.net) left irc: Disconnected by services [00:46] Nick change: Strangerke_ -> Strangerke [00:46] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services. [00:59] Robin_Watts (~Javacat@82.153.193.216) left irc: Quit: Pop! [01:06] Nick change: Adys -> Shirvallah [01:09] DrMcCoy: haha [01:10] Nick change: Shirvallah -> Adys [01:18] Right.. Yep, I had basically missed the 4 OmniTV ads.. drat. Thanks _sev [01:18] I'll remember for the future I can just download the Amiga/Mac Cutscene pack, rather than having pain with installshield :-) [01:18] Lubomyr (~Miranda@194.44.255.186) left irc: Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org [01:24] LordHoto: So yet another new OPL emulator? :P [01:25] DrMcCoy: no the old one is gone [01:25] LordHoto: Ah :P [01:25] DrMcCoy: (i.e. the old DOSBox one) [01:29] what, what? [01:33] Lightkey: DO THEY SPEAK ENGLISH IN WHAT?!? [01:33] :P [01:35] Javacat (~ghsgh@unaffiliated/javacat) left irc: Ping timeout: 240 seconds [01:35] if there is a language called whitespace, why not? [01:37] Strangerke (~a@cable-85.28.89.119.coditel.net) left irc: Ping timeout: 256 seconds [01:37] Javacat (~ghsgh@unaffiliated/javacat) joined #scummvm. [01:38] DrMcCoy: the "new" DOSBox emulator is even less binary size (50KB with -O2 on amd64 for me instead of 88KB like the old) ;-P [01:38] \o/ [01:38] LordHoto: And it's even slower? :P [01:38] DrMcCoy: no idea, I don't have any audible difference ;-P [01:39] :P [01:39] hear even [01:40] Strangerke (~a@cable-85.28.89.119.coditel.net) joined #scummvm. [01:41] DrMcCoy: when there's some wince trunk build ready, I might give it a try... [01:41] DrMcCoy: the "old" one was a bit sluggish with Sam&Max on my mobile phone [02:00] raorn (~raorn@altlinux/developer/raorn) left irc: Read error: Connection reset by peer [02:00] Zaarin (~jalla@ti511110a080-0828.bb.online.no) left irc: Ping timeout: 256 seconds [02:03] StarG (~John@g224196010.adsl.alicedsl.de) got netsplit. [02:03] Gentle (~tier@193.22.254.209) got netsplit. [02:03] drantin (~drantin@pdpc/supporter/professional/drantin) got netsplit. [02:03] ebbex (erikberg@knuth.ping.uio.no) got netsplit. [02:05] StarG (~John@g224196010.adsl.alicedsl.de) returned to #scummvm. [02:05] Gentle (~tier@193.22.254.209) returned to #scummvm. [02:05] drantin (~drantin@pdpc/supporter/professional/drantin) returned to #scummvm. [02:05] ebbex (erikberg@knuth.ping.uio.no) returned to #scummvm. [02:05] Gentle (~tier@193.22.254.209) left irc: Max SendQ exceeded [02:06] Gentle (~tier@193.22.254.209) joined #scummvm. [02:08] aptitude show texlive-latex-extra|less [02:08] o_O'' [02:08] Javacat (~ghsgh@unaffiliated/javacat) left irc: Quit: zz [02:08] texlive-latex-extra-doc installed size 261 MB o_O''''' [02:34] LordHoto (~loom@p54B1E5CE.dip.t-dialin.net) left irc: Quit: night [02:43] Nevyn7 (jhgjfgl@81-236-19-16-no39.tbcn.telia.com) left irc: [02:46] lordhoto * r48179 /scummvm/trunk/sound/ (7 files in 2 dirs): Switch to the other DOSBox OPL emulator as suggested by the DOSBox developers. [02:46] lordhoto * r48180 /scummvm/trunk/sound/softsynth/opl/ (dosbox.cpp dosbox.h): Cleanup. [02:46] lordhoto * r48181 /scummvm/trunk/sound/softsynth/opl/dbopl.cpp: Use memset instead of a custom loop for zeroing the sample buffer. [02:46] lordhoto * r48182 /scummvm/trunk/sound/softsynth/opl/ (dbopl.cpp dbopl.h): Add a slight comment when the DBOPL sources where synched with DOSBox. [02:46] lordhoto * r48183 /scummvm/trunk/sound/ (fmopl.cpp fmopl.h): [02:46] Show an error message in case multiple OPL outputs are created instead of using an assert. [02:46] The new DBOPL emulator we are using should support multiple instances though. We *might* [02:46] consider allowing as many instances as the user wants. Of course since the original [02:46] games only had one OPL chip available, that should not be required. Also just in case [02:46] we might allow real hardware as playback device that would be out of the question again [02:46] too. [02:51] He leaves and then he commits? :P [02:56] CIA must be lagging. Either that, or I received the commits before they were committed ;) [03:09] GoGi (~gogi@pippo.9.6.gogi.tv) left irc: Remote host closed the connection [03:23] digitall (digitall@cpc1-hitc2-0-0-cust861.lutn.cable.ntl.com) left #scummvm. [03:49] Kerbox (~Kernox@ti0106a340-dhcp0478.bb.online.no) left irc: Read error: Connection reset by peer [03:50] Kerbox (~Kernox@ti0106a340-dhcp0478.bb.online.no) joined #scummvm. [04:19] ST can travel into the future?!? [04:56] "Safety not guaranteed" [04:57] Joke used twice in one day! [05:00] clone2727 (~clone2727@nat22-resnet-ext.rutgers.edu) left irc: Quit: later [05:00] :( [05:24] lol, I read your name as DrMcClone. [06:05] Damn, anyone got experience with JIS and Shift_JIS? I only find formulas to convert JIS to Shift_JIS, but not the other way round. And I'm getting crazy trying to reverse the thing myself [06:05] But it must be possible, because recode can do it :P [06:25] For example, 142 132 (0x8E84) could have come from both 59 100 (0x3B64) and 60 6 (0x3C06) [06:39] DrMcClone! Me likes :P [06:40] ... [07:00] Hmm, okay, apparently, not all 2-byte sequences are valid JIS codes. I hope that's enough :P [07:28] Strangerke (~a@cable-85.28.89.119.coditel.net) left irc: Quit: Bbl [08:33] johndoe123 (~john_doe@p5DE7B193.dip.t-dialin.net) joined #scummvm. [08:33] #scummvm: mode change '+o johndoe123' by ChanServ!ChanServ@services. [09:02] Tomaz^W (~tompsson@85.89.73.250) joined #scummvm. [09:04] ny00123 (~ny00123@bzq-79-182-58-1.red.bezeqint.net) joined #scummvm. [09:04] cmeme2 (~cmeme@boa.b9.com) left irc: Ping timeout: 265 seconds [09:11] Strangerke (~51f60aa1@gateway/web/freenode/x-lritgikpdzmscrwp) joined #scummvm. [09:11] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services. [09:11] hi guys [09:14] Hi [09:53] Sho_ (~EHS1@kde/hein) left irc: Remote host closed the connection [09:55] |Cable| (~cable@237.130.117.91.dynamic.mundo-r.com) left irc: Remote host closed the connection [10:01] raorn (~raorn@altlinux/developer/raorn) joined #scummvm. [10:12] oh great, so the missing jsMath font was just because it needs a cookie, what did I install this GB of software for, then?!? [10:24] Hkz (~noname@host8-143-dynamic.104-80-r.retail.telecomitalia.it) joined #scummvm. [10:24] #scummvm: mode change '+o Hkz' by ChanServ!ChanServ@services. [10:29] GouldFish (~gouldfish@87-194-1-237.bethere.co.uk) joined #scummvm. [10:51] Nevyn2 (jhgjfgl@81-236-19-16-no39.tbcn.telia.com) joined #scummvm. [10:54] fingolfin * r48184 /scummvm/trunk/ (6 files in 5 dirs): [10:54] Tweak makeVOCDiskStream [10:54] * now takes a SeekableReadStream *pointer* like (almost) all other [10:54] audiostream factories [10:54] * fix potential memory leak in it [10:54] * rename takeOwnershipOfStream to disposeAfterUse for consistency [10:55] fingolfin * r48185 /scummvm/trunk/ (4 files in 4 dirs): Remove unnecessary #includes [10:56] fingolfin * r48186 /scummvm/trunk/ (12 files in 6 dirs): Move some aspect ratio related stuff to graphics/scaler/aspect.h [10:56] fingolfin * r48187 /scummvm/trunk/ (6 files in 3 dirs): [10:56] SCALERS: Change 'bitformat' template params to 'colormasks'; [10:56] get rid of some #defines [10:57] fingolfin * r48188 /scummvm/trunk/graphics/scaler/thumbnail_intern.cpp: Code formatting [10:57] fingolfin * r48189 /scummvm/trunk/ (8 files in 3 dirs): SCALERS: Get rid of MAKE_WRAPPER; make RGBtoYUV internal [10:57] fingolfin * r48190 /scummvm/trunk/graphics/scaler.cpp: Move some var declarations into '#ifdef USE_NASM' [10:58] fingolfin * r48191 /scummvm/trunk/ (5 files in 2 dirs): [10:58] Rename PocketPCPortrait scaler to DownscaleHorizByThreeQuarters, [10:58] and move it to graphics/scaler. [10:58] fingolfin * r48192 /scummvm/trunk/ (7 files in 3 dirs): [10:58] Rename PocketPCLandscapeAspect to Normal1xAspect and move it to [10:58] graphics/scaler/aspect.cpp [10:59] fingolfin * r48193 /scummvm/trunk/graphics/scaler/downscaler.cpp: Remove redundant code [10:59] fingolfin * r48194 /scummvm/trunk/graphics/scaler/aspect.cpp: Rewrite Normal1xAspect (untested, may be broken) [10:59] fingolfin * r48195 /scummvm/trunk/graphics/scaler/aspect.cpp: Add new aspect ratio scaler variant, based on the Normal2xAspect ARM code [12:05] ST (~ScottT@25.105.96.58.static.exetel.com.au) left irc: Ping timeout: 240 seconds [12:05] ST (~ScottT@25.105.96.58.static.exetel.com.au) joined #scummvm. [12:05] #scummvm: mode change '+o ST' by ChanServ!ChanServ@services. [12:06] Robin_Watts (~Javacat@82.153.193.216) joined #scummvm. [12:06] #scummvm: mode change '+o Robin_Watts' by ChanServ!ChanServ@services. [12:28] D0SFreak (~D0SFreak@c-71-199-151-129.hsd1.ga.comcast.net) left irc: Ping timeout: 256 seconds [12:40] D0SFreak (~D0SFreak@72.1.88.158) joined #scummvm. [13:02] Nick change: michal[t] -> [michal]t [13:31] Kirben (kirben@c58-107-187-179.brasd3.vic.optusnet.com.au) left irc: [14:28] Schnaks (~Schnaks@p5486FA15.dip.t-dialin.net) joined #scummvm. [14:30] bushwakko (~bushwakko@fou162.telenor.ntnu.no) joined #scummvm. [14:44] MetalSnake (~snake@91-67-40-31-dynip.superkabel.de) joined #scummvm. [15:41] GoGi (~gogi@pippo.9.6.gogi.tv) joined #scummvm. [16:24] Anyone there who reads Japanese and can tell me whether the text I got out of my custom Shift_JIS -> JIS X 0208 conversion makes sense? http://drmccoy.de/zeugs/ds2saturnjap.png and http://drmccoy.de/zeugs/ds2saturnjap2.png [16:25] jvprat (~jvprat@unaffiliated/jvprat) left irc: Read error: Connection reset by peer [16:31] DrMcCoy: Why don't you send a mail to peres ? [16:32] Else, I can ask a friend, but you'll have to wait a bit [16:32] :P [16:32] Strangerke, or you could learn japanese! [16:33] Hkz : I tried. And failed. [16:33] Strangerke, you're doing it wrong! :P [16:34] Most likely, considering the number of languages I wasn't able to learn :P [16:35] :PPP [16:35] Kinda like me [16:35] I failed learning French [16:35] :P [16:35] DrMcCoy: Answer in less than 5 minutes. Please stay awake ! [16:35] DrMcCoy: That's only 1 language :P [16:36] :P [16:37] I failed (in order) : spanish, russian, japanese, klingon, arab and japanese again. [16:37] Strangerke, latin. [16:37] you forgot latin [16:37] Hkz : I don't call that a language. [16:37] ehi! [16:37] :P [16:37] DrMcCoy: "Yes, it's japanese" [16:38] DrMcCoy: "And it makes sense" [16:38] Strangerke: Yay :) [16:39] DrMcCoy: May I authorize my friend to disconnect ? :P [16:40] Strangerke: Yes. And say thanks from me :) [16:40] oki doki [16:43] jvprat (~jvprat@unaffiliated/jvprat) joined #scummvm. [16:43] #scummvm: mode change '+o jvprat' by ChanServ!ChanServ@services. [16:47] LordHoto (~Loom@p54B1F999.dip.t-dialin.net) joined #scummvm. [16:47] #scummvm: mode change '+o LordHoto' by ChanServ!ChanServ@services. [16:48] jvprat (~jvprat@unaffiliated/jvprat) left irc: Ping timeout: 256 seconds [16:50] Zaarin (~jalla@ti511110a080-0828.bb.online.no) joined #scummvm. [16:53] Tomaz^W (~tompsson@85.89.73.250) left irc: Ping timeout: 276 seconds [16:56] |Pixel| (pixel@nobis-crew.org) left irc: Remote host closed the connection [17:01] |Pixel| (pixel@nobis-crew.org) joined #scummvm. [17:01] bushwakko (~bushwakko@fou162.telenor.ntnu.no) left irc: Read error: Operation timed out [17:03] peres (~peres@unaffiliated/peres) joined #scummvm. [17:03] #scummvm: mode change '+o peres' by ChanServ!ChanServ@services. [17:04] jvprat (~jvprat@unaffiliated/jvprat) joined #scummvm. [17:04] #scummvm: mode change '+o jvprat' by ChanServ!ChanServ@services. [17:28] GoGi (~gogi@pippo.9.6.gogi.tv) left irc: Quit: Leaving [17:31] Vaikungfu (UPP@c-24-2-60-123.hsd1.ca.comcast.net) left irc: [17:38] Javacat (~ghsgh@unaffiliated/javacat) joined #scummvm. [17:40] Ceriand|work (~Ceriand@unaffiliated/ceriand) joined #scummvm. [17:44] a rumor! http://kotaku.com/5486393/rumor-lucasarts-releasing-monkey-island-2-special-edition [17:46] salty-horse (~ori@109.67.41.91) joined #scummvm. [17:49] lisppaste4 (~lisppaste@common-lisp.net) joined #scummvm. [17:55] graphics/scaler/hq2x.cpp:117: error: gBitFormat was not declared in this scope [17:57] missing extern int gBitFormat; ? [18:00] and in others too. maybe it should be defined in module scope? [18:03] peres (~peres@unaffiliated/peres) left irc: Quit: Leaving [18:04] LordHoto (~Loom@p54B1F999.dip.t-dialin.net) left irc: Quit: ... [18:06] bastardo (~bastardo@g226144042.adsl.alicedsl.de) joined #scummvm. [18:09] Strangerke (~51f60aa1@gateway/web/freenode/x-lritgikpdzmscrwp) left irc: Quit: Bbl [18:09] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) joined #scummvm. [18:27] johndoe123 (~john_doe@p5DE7B193.dip.t-dialin.net) left irc: [18:38] Zaarin (~jalla@ti511110a080-0828.bb.online.no) left irc: Quit: In the grand design, women were definitely drawn from a different set of blueprints. [18:40] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) left irc: Quit: bushwakko [18:49] StarG_ (~John@f053082186.adsl.alicedsl.de) joined #scummvm. [18:52] StarG (~John@g224196010.adsl.alicedsl.de) left irc: Ping timeout: 240 seconds [18:54] |Cable| (~cable@237.130.117.91.dynamic.mundo-r.com) joined #scummvm. [18:55] clone2727 (~clone2727@nat02-njh-ext.rutgers.edu) joined #scummvm. [18:56] #scummvm: mode change '+o clone2727' by ChanServ!ChanServ@services. [18:59] salty-horse (~ori@109.67.41.91) left irc: Ping timeout: 248 seconds [19:01] salty-horse (~ori@109.67.41.91) joined #scummvm. [19:13] WooShell (~Markus@woo.li) joined #scummvm. [19:24] nøbend [19:28] LordHoto (~loom@p54B1F999.dip.t-dialin.net) joined #scummvm. [19:28] #scummvm: mode change '+o LordHoto' by ChanServ!ChanServ@services. [19:30] Fingolfin (~Fingolfin@p548560AA.dip.t-dialin.net) joined #scummvm. [19:30] #scummvm: mode change '+o Fingolfin' by ChanServ!ChanServ@services. [19:31] Lubomyr (~Miranda@194.44.193.31) joined #scummvm. [19:34] lordhoto * r48196 /scummvm/trunk/graphics/scaler/ (hq2x.cpp hq3x.cpp): Fix compilation. [19:39] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) joined #scummvm. [19:40] waltervn (~walter@82-171-47-198.ip.telfort.nl) joined #scummvm. [19:40] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services. [19:43] lordhoto * r48197 /scummvm/trunk/graphics/scaler/ (hq2x.cpp hq2x.h hq3x.cpp hq3x.h): Use a template for hq2x and hq3x instead of using #include and defines. [19:54] salty-horse (~ori@109.67.41.91) left irc: Ping timeout: 268 seconds [19:56] salty-horse (~ori@bzq-109-67-41-91.red.bezeqint.net) joined #scummvm. [19:59] Zaarin (~jalla@ti511110a080-0828.bb.online.no) joined #scummvm. [20:17] Strangerke (~a@cable-85.28.89.119.coditel.net) joined #scummvm. [20:17] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services. [20:32] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) left irc: Quit: bushwakko [20:33] Fingolfin, is there somewhere public you can post the buildbot config file? (if the problem is there) -- also, someone with shell access could ask in #buildbot for advice [20:33] salty-horse: the config file is in our SVN [20:34] salty-horse: it's a python file [20:34] and that "someone with shell access" shoudl probably start by finding log files and looking at them, I [20:34] I think :) [20:34] that someone also should first turn up, we can then give him hints ;) [20:34] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) joined #scummvm. [20:35] clone2727 (~clone2727@nat02-njh-ext.rutgers.edu) left irc: Quit: Java user signed off [20:35] you don't *have* to refer to yourself in the third person :) [20:35] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) left irc: Client Quit [20:35] :) [20:38] salty-horse: yupp [20:38] did I mention that I am sick and tired of taking care of the buildbot? I'll just not touch it anymore, except for taking it offline if nobody is interested in it :) [20:38] fingolfin * r48198 /scummvm/trunk/backends/platform/wince/ (CEScaler.cpp CEScaler.h): WINCE: Rewrote SmartphoneLandscape scaler C version to match what the ARM assembler version does (untested) [20:39] fingolfin * r48199 /scummvm/trunk/engines/sci/engine/klists.cpp: SCI: Made some stuff static [20:42] Fingolfin, can I try and see what's going on on the server? I'm not familiar with buildbot, but I'm willing to try [20:44] Hkz (noname@host8-143-dynamic.104-80-r.retail.telecomitalia.it) left #scummvm ("SIGNOFF"). [20:44] Hkz (~noname@host8-143-dynamic.104-80-r.retail.telecomitalia.it) joined #scummvm. [20:44] #scummvm: mode change '+o Hkz' by ChanServ!ChanServ@services. [20:54] waltervn (~walter@82-171-47-198.ip.telfort.nl) left irc: Ping timeout: 245 seconds [20:54] Fingolfin: can you create shell accounts on that box, still in need of one to move the planet and I don't mind taking a look at the buildbot. [20:56] DJWillis, maybe upgrading to 0.7.12 will fix some problems [20:57] salty-horse: it crossed my mind, on one level it's tempting to upgrade and then just take the hit on the fallout and fix as we go. [20:58] it seems they changed some config stuff, so it will require editing of the config file :/ [20:59] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) joined #scummvm. [21:00] salty-horse: I can think of worse things, well some worse things anyway ;) [21:00] editing said file with notepad? [21:01] salty-horse: notepad, yep, nothing like spending hours fighting dos <> unix line endings ;-) [21:01] you can DO something about those in notepad? [21:01] wow [21:07] you can do it with a perl one liner [21:07] I WAS being facetious ;-) [21:08] waltervn (~walter@82-171-47-198.ip.telfort.nl) joined #scummvm. [21:08] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services. [21:10] D0SFreak (~D0SFreak@72.1.88.158) left irc: Ping timeout: 268 seconds [21:16] getisboy (~Family@71.174.111.44) joined #scummvm. [21:17] salty-horse, DJWillis yo, sorry, I was skyping and had IRC off screen [21:18] it's unlikely that a blind upgrade to buildbot 0.7.12 will fix anything, it's more likely to break things, because they tend to make incompatible config changes [21:18] in fact, this may well be the cause for the current issues we have -- they may be caused from my semi-blind update to 0.7.11p3 [21:20] http://github.com/djmitche/buildbot/blob/master/NEWS#L113 [21:21] wjp: here? [21:21] Fingolfin, do you remember which version did you upgrade *from*? [21:22] salty-horse: sure, the latest patch level rev of 0.7.10p [21:22] salty-horse: is that link to the things which are fixed in 0.7.11p3 supposed to tell us something? :) [21:23] us? [21:23] it's the place to look for api changes, and I don't spot anything. I may have to look at the "detailed" changelog [21:23] oh, you heiness [21:23] the royal us [21:24] pluralis majestatis [21:24] "us" = the channel, everybody who is participating in the discussion [21:24] e.g. DJWillis and me, at least [21:24] and ebbex [21:24] (I exclude salty-horse because he's probably not telling himself ;) [21:25] I don't usually keep notes in here [21:26] getisboy (Family@71.174.111.44) left #scummvm. [21:27] salty-horse: 0.7.12 introduces some backward incomaptible changes (as does almost every buildbot release), they even documented some in the NEWS file [21:27] e.g. "BuildmasterConfig no longer accepts 'bots' and 'sources' as keys" [21:27] and they have more, I am pretty sure. Which need not be an issue [21:27] yup, but not the .10 to .11 release [21:27] and even if it affects us, should be work aroundable [21:28] it's probably pointless to talk about it, if looking at the logs could tell you the problem right away :) [21:28] Fingolfin: yes [21:28] I am pretty sure they made some incompatible changes there, too, but we checked and none should hurt us. anyway, this is pointless, the logs should be analyzed instead of this purely theoretical discussion ;) [21:28] indeed [21:28] wjp: too late :) [21:29] ok :-) [21:38] syke (9194-ident@cesium.clock.org) joined #scummvm. [21:38] #scummvm: mode change '+o syke' by ChanServ!ChanServ@services. [21:39] OK, I am looking through some bug reports [21:40] one of them is on a well-known issue: We are currently not very good at handling the absence of a working digital audio mixer [21:40] hi [21:40] our Audio::Mixer has an isReady method, and in the distant past, we had some limited support for running without digital audio [21:40] (I'll just say "without audio" from now on ;) [21:41] problem is, quite some code relies on audio playback, e.g. scripts try to sync to it [21:41] anyway, right now, we end up in a semi-broken state if there is for some reason no audio support [21:41] (e.g. because you are on linux and no sound card driver is loaded) [21:41] so [21:42] either we just change ScummVM to error out in the case, right from the start [21:42] or, 2) we try to adapt all engines to work with it [21:42] or 3) we try to make Audio::Mixer handle the situation as good as possible, with [21:42] 3) can optionally be augmented by 2) ;) [21:43] we could do 3 as follows: If no audio is available, the mixer would simply immediately mark any data passed to it as "done" (deleting streams if necessary) [21:43] _sev|work_ (~sev@scummvm/undead/sev) left irc: Ping timeout: 276 seconds [21:43] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) left irc: Quit: bushwakko [21:44] this won't be great for syncing, but it'll be better than simply waiting for ever on all streams, and letting the stream list fill up endlessly [21:44] because without an audio thread, of course no audio data is ever consumed [21:44] we could also try to use a timer to fake real audio usage; but I wonder whether that's worth it, and whether it would work too well... hmm [21:45] anyway, this all is just meant to make the "default" behavior of the mixer as good as possible; of course ideally engines would also detect if audio is unavailable, and would take corresponding measures [21:45] such as refusing to start, showing an error alert that audio is absolutely necessary ; or by switching to an alternate syncing method; etc. [21:53] Fingolfin: I would guess that you'd have to wait the appropriate time, otherwise you'd run into speed and script hang problems [21:53] trying the simplest thing first makes sense, tho :) [21:54] well [21:54] for now, just about anything is better than the current situation [21:54] which is a guaranteed crash, whenever an engine tries to play a stream when no audio is available [21:55] failing faster, and more graciously, is a good start [21:55] what I propose is to just ignore playing the stream (and of course if the client code requested that we delete it after playback, do so, too) [21:55] but making games appear to be playable, only to hit a hang later, might be a worse experience [21:55] that makes sense [21:55] this will make many games work which don't work right now, and shouldn't result in crashes -- I'd expect any crash resulting from this to be a bug in the engine [21:56] so long as the 'done playing' callbacks don't happen too soon [21:56] there are no "done playing" callbacks :) [21:56] I would expect more script hangs/bugs than crashes [21:56] but an engien can query the stream resp. the sound handle it got [21:56] yeah, well, we used to get tons of script hangs (before we changed the code, making it crash in a weird place ;). Because we used to just queue play requests, but would never execute them [21:57] so now we'd queue them and then immediately unqueue them, roughly spoken [21:57] it's certainly not ideal, but then running without audio isn't ideal either [21:57] yea, it's worth a shot [21:57] of course a more sophisticated audio "emulation" might help, too [21:57] probably a quick playthru of one game in each engine will show if there's any immediately horrid issues [21:57] i.e. based on TimerManager, trying to fake "real" playback, time wise... [21:57] exactly [21:58] yea, that's what I was thinking would be necessary [21:58] before we try to implement a more elaborate solution, we should know what fails, where, and how ;) [21:58] yup [21:58] like text speeding by because it thinks the sample is done playing, etc [21:58] right now, I just know "If you start BASS w/o audio, it'll crash trying to create an audio rate converter for sample rate 0" ;) which I already fixed as described, locally [21:58] yup [21:59] or scripts somewhere doing something like subtracting the time difference between sound plays and using that delta for something [21:59] although too fast text IMO is better than hanging endlessly on the intro screen (as Monkey Island used to :) [21:59] anotherguest * r48200 /scummvm/branches/branch-1-1-0/backends/platform/symbian/src/portdefs.h: Re-enable the asm costume [21:59] heh [21:59] yes [21:59] (though the MI bug was due to MIDI, strictly speaking) [21:59] but both would be unplayable [22:00] which is where option 2) also comes in: Engine should refuse to start if they don't support playing w/o audio [22:00] we could just add such a check to all engines, now, and ask engine authors to revise them as they go [22:00] anotherguest * r48201 /scummvm/trunk/backends/platform/symbian/src/portdefs.h: Re-enable the arm asm costume drawer [22:00] that's the best situation, I think [22:00] my changes for now mainly aim to make it possible to engine authors to implement a mode that works w/o audio painlessly... I think no "automated" change will work for all engines/games anyway [22:01] agreed [22:02] hurm [22:02] wtf [22:03] I disabled audio in my copy of sdl.cpp -- I *removed* the call to SDL_OpenAudio(); forcing it to fail opening an audio device [22:03] yet BASS plays some audio before crashing into the debugger... [22:04] oh wait -- maybe this was some MIDI music :) [22:04] Action: Fingolfin is an idiot :) [22:04] puh, that freaked me out for a bit :) [22:06] heh [22:06] syke (9194-ident@cesium.clock.org) left irc: Quit: leaving [22:06] syke (2119-ident@cesium.clock.org) joined #scummvm. [22:06] #scummvm: mode change '+o syke' by ChanServ!ChanServ@services. [22:08] or BASS is demonic [22:12] syncing about it so more, I now think that we definitely shouldn't try to abuse TimerManager from the sound mixer directly [22:12] this could e.g. lead to locking errors [22:12] and it would complicate the sound mixer needlessly [22:12] D0SFreak (~D0SFreak@c-71-199-151-129.hsd1.ga.comcast.net) joined #scummvm. [22:12] instead it seems much more sane to expect backends to implement this emulation, if possible [22:12] e.g. in the SDL backend we could spawn a thread which calls the mixer callback [22:13] the mixer would also be told a fake sample rate, as configured [22:13] we'd still need to retain a way to signal client code that no actual audio consumption takes place [22:13] anyway, this is for the future to consider ;) [22:21] fingolfin * r48202 /scummvm/trunk/engines/sci/engine/ (klists.cpp kmovement.cpp kparse.cpp selector.h): SCI: Make EngineState parameter to INV_SEL explicit [22:21] fingolfin * r48203 /scummvm/trunk/sound/mixer.cpp: Fix bug #2872076 (MIXER: Division by 0 in rate conversion if w/o soundcd) [22:22] fingolfin * r48204 /scummvm/branches/branch-1-1-0/sound/mixer.cpp: Backport: Fix bug #2872076 (MIXER: Division by 0 in rate conversion if w/o soundcd) [22:22] _sev|work (~sev@scummvm/undead/sev) joined #scummvm. [22:22] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services. [22:28] Fingolfin: maybe a sleep(1)? [22:29] you mean g_system->delayMillis(1) ? what good would that do? [22:31] Fingolfin: I'll try and check the wince build later this week, but I can't promise - I'm off on business on saturday, so desperately trying to get stuff done. [22:31] Robin_Watts: OK.. well.. we still have this problem of not having a WinCE porter, in particular, somebody who builds release binaries [22:31] Robin_Watts: do you think you could do that, too ? (when the time is there) ? [22:32] I can certainly try. When is the release ? [22:32] we also should put up a news item looking for folks to help out... "we" ;) [22:32] Mar 21st: backend code freeze [22:32] Mar 24th: tagging [22:32] Mar 28th: release [22:32] Sure. I should be OK to help out then. [22:33] the royal I? [22:36] Robin_Watts: that 16bit patch seemed ok but not tested at length, time sort of vanished ;-) [22:39] Fingolfin: when we first starting doing the DirectX sound drivers in FreeSCI, sleep(1) was used as a placeholder for a couple of reasons [22:39] Something is up with gBitFormat. [22:40] fixed. [22:40] Robin_Watts: missing 'extern int gBitFormat;'? [22:41] yup. [22:41] just noticed it doing a quick local build. [22:42] robinwatts * r48205 /scummvm/trunk/backends/platform/wince/CEScaler.cpp: Move gBitFormat definition so it works in both sizes of an ifdef. [22:42] in the wince code? [22:42] I guess that's an yes ;-) [22:45] syke: fair enough, but that was FreeSCI and a direct X sound driver. This is ScummVM, and all sorts of different sound drivers (but none of them is directX, at least directly)... so I don't see why we should do a sleep(1). [22:46] clone2727 (~clone2727@nat22-resnet-ext.rutgers.edu) joined #scummvm. [22:46] #scummvm: mode change '+o clone2727' by ChanServ!ChanServ@services. [22:47] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) joined #scummvm. [22:47] bushwakko (~bushwakko@199.229.16.62.customer.cdi.no) left irc: Client Quit [22:49] ny00123 (~ny00123@bzq-79-182-58-1.red.bezeqint.net) left irc: Quit: Leaving [22:52] robinwatts * r48206 /scummvm/trunk/backends/platform/wince/wince-sdl.cpp: [22:52] Add missing include that caused build failures with Normal1xAspect and [22:52] Normal2xAspect being undefined. [22:53] Robin_Watts: thanks for the fixes, sorry for making them necessary. With buildbot being down, I had no chance of fixing these myself, else I would have, I swear :) [22:54] Lubomyr (~Miranda@194.44.193.31) left irc: Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org [22:55] Fingolfin: No worries. [22:55] I know precisely how easy such things are to do, and I'm in awe of the amount you get done. [22:56] It's chugging through the rest of the build now. [23:00] MetalSnake (~snake@91-67-40-31-dynip.superkabel.de) left irc: Quit: MetalSnake [23:01] ETOOTIRED. Telly calling. Will check results of build later. [23:05] GoGi (~gogi@pippo.9.6.gogi.tv) joined #scummvm. [23:10] Vaikungfu (UPP@c-24-2-60-123.hsd1.ca.comcast.net) joined #scummvm. [23:20] syke (2119-ident@cesium.clock.org) left irc: Quit: leaving [23:33] dreammaster (~paulfgilb@c114-77-95-66.lowrp3.vic.optusnet.com.au) joined #scummvm. [23:33] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services. [23:39] Fingolfin (~Fingolfin@p548560AA.dip.t-dialin.net) left irc: Quit: Fingolfin [23:45] WooShell (~Markus@woo.li) left irc: [23:55] Anyone eknow [23:55] Sorry, anyone knowledgable with the source about? [23:55] what source exactly? [23:55] dreammaster: Just ask, maybe someone can answer ;) [23:56] I'm familiar with engines/kom and engines/neverhood and engines/ags [23:56] Good point. :). I noticed that the MSVC projects won't build anymore - there seems to be some confusion about the #define of 'HAVE_NASM' versus 'USE_NASM'. I was curious as to whether there's a particular reason for the difference, and if not, if I should merge the two so it will recompile again properly. [00:00] --- Tue Mar 9 2010