[00:07] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) joined #scummvm. [00:16] Kirben (kirben@c220-239-194-17.brasd3.vic.optusnet.com.au) joined #scummvm. [00:16] #scummvm: mode change '+o Kirben' by ChanServ!ChanServ@services. [00:23] dtcrshr (~datacrush@unaffiliated/datacrusher) left irc: Quit: Saindo [00:29] chkr (~chkr@p579324A6.dip.t-dialin.net) left irc: Remote host closed the connection [00:43] hennymcc_ (~hennymcc@p4FC1608F.dip.t-dialin.net) left irc: Ping timeout: 272 seconds [00:48] Sylvain (~Sylvain@ALille-553-1-34-142.w90-34.abo.wanadoo.fr) left irc: Ping timeout: 264 seconds [00:58] debris (~deb@host52-157-dynamic.16-79-r.retail.telecomitalia.it) joined #scummvm. [01:02] debris (~deb@host52-157-dynamic.16-79-r.retail.telecomitalia.it) left irc: Client Quit [01:38] Vampire0_ (vampire@jEdit/Vampire) joined #scummvm. [01:42] Vampire0 (vampire@jEdit/Vampire) left irc: Ping timeout: 260 seconds [01:50] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) left irc: Quit: ERC Version 5.3 (IRC client for Emacs) [01:51] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) joined #scummvm. [01:53] Smartnow (~Smartnow@gateway/tor-sasl/smartnow) left irc: Ping timeout: 276 seconds [02:37] sirlemonhead (bduncan22@86-45-5-190-dynamic.b-ras2.prp.dublin.eircom.net) left #scummvm. [03:08] echelog-1 (~echelog-1@95.166.104.217) left irc: Disconnected by services [03:10] echelog-1 (~echelog-1@95.166.104.217) joined #scummvm. [03:11] _sev__ (~sev@109.87.142.103) joined #scummvm. [03:11] _sev__ (~sev@109.87.142.103) left irc: Changing host [03:11] _sev__ (~sev@scummvm/undead/sev) joined #scummvm. [03:11] #scummvm: mode change '+o _sev__' by ChanServ!ChanServ@services. [03:12] Upth (~ogmar@108-85-91-228.lightspeed.frokca.sbcglobal.net) joined #scummvm. [03:12] #scummvm: mode change '+v Upth' by ChanServ!ChanServ@services. [03:13] Upthorn (~ogmar@108-85-91-228.lightspeed.frokca.sbcglobal.net) left irc: Ping timeout: 265 seconds [03:13] _sev_ (~sev@scummvm/undead/sev) left irc: Read error: Connection reset by peer [03:20] umisguy (4cda425a@gateway/web/freenode/ip.76.218.66.90) left irc: Ping timeout: 245 seconds [03:35] fydo_ (fydo@209-20-66-28.slicehost.net) joined #scummvm. [03:39] fydo_ (fydo@209-20-66-28.slicehost.net) left irc: Client Quit [03:56] clone2727 (~clone2727@pool-71-251-38-73.nwrknj.east.verizon.net) left irc: Quit: later [04:01] LordHoto (~loom@unaffiliated/lordhoto) joined #scummvm. [04:01] #scummvm: mode change '+o LordHoto' by ChanServ!ChanServ@services. [04:40] Herrman (wolter@p50912E23.dip.t-dialin.net) left irc: Quit: Aufsteher aller Länder, bleibt doch mal gemeinsam liegen. - [04:58] kaltsasa (~KD0ETH_Wo@MA2982.grinnell.edu) joined #scummvm. [05:02] KD0ETH_Wo (~KD0ETH_Wo@132.161.219.52) left irc: Ping timeout: 265 seconds [05:03] LordHoto (~loom@unaffiliated/lordhoto) left irc: Quit: ... [05:17] _sev_ (~sev@scummvm/undead/sev) joined #scummvm. [05:17] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services. [05:17] _sev__ (~sev@scummvm/undead/sev) left irc: Read error: Connection reset by peer [05:22] CaelThunderwing (CaelThunde@d14-69-103-81.try.wideopenwest.com) joined #scummvm. [05:24] anyone here have had any luck burning a Sucessfull Dreamcast Disc? [05:29] everytime ScummVM Just makes the Console crash (Reboot) or freeze up (launched thru DCHakker) [05:33] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) left irc: Ping timeout: 250 seconds [05:46] rvalles (~rvalles@unaffiliated/rvalles) left irc: Quit: bbl [06:11] Kaidane (~Kaidane@adsl-065-005-218-152.sip.cae.bellsouth.net) left irc: Read error: Connection reset by peer [07:28] Nick change: _sev_ -> _sev [07:29] <_sev> We've reached 8,000,000 downloads [07:29] nice [07:29] <_sev> http://sourceforge.net/projects/scummvm/files/scummvm/stats/timeline?dates=2001-10-05+to+2012-04-14 [07:29] great :) [07:31] congratulations all, I guess :) [07:36] <_sev> http://www.youtube.com/watch?v=av8NTy5WkFc [07:36] digitall (~digitall@cpc2-hitc2-0-0-cust28.9-2.cable.virginmedia.com) joined #scummvm. [07:36] #scummvm: mode change '+o digitall' by ChanServ!ChanServ@services. [07:37] Congratulations to everyone on download milestone! [07:38] CaelThunderwing: No, but please note http://wiki.scummvm.org/index.php/Dreamcast [07:39] The DC has only 16M of RAM and thus will not run the larger engines.... [07:39] Does the v1.3.1 stable work for you? [07:40] well, rule #1 of trying Dreamcast builds is to first make sure your image runs in e.g. NullDC [07:44] Oh... and I think the debug builds from buildbot fail to run unless you have a debug serial cable connected... [07:45] i.e. See http://sourceforge.net/tracker/?func=detail&aid=3123788&group_id=37116&atid=418820 [07:55] Hi guys [08:03] Good Morning Strangerke. [08:07] ajax16384 (~User@ip196.net172.n37.ru) joined #scummvm. [08:07] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services. [08:28] James|GlideM (~James|Gli@cpc2-mapp11-2-0-cust447.12-4.cable.virginmedia.com) joined #scummvm. [08:28] #scummvm: mode change '+v James|GlideM' by ChanServ!ChanServ@services. [08:42] pok0j (~pok0j@abtg238.neoplus.adsl.tpnet.pl) joined #scummvm. [08:50] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) joined #scummvm. [09:09] mhilmi (~marwan@199-83-220-154.PUBLIC.monkeybrains.net) left irc: [09:20] kurtwr2 (kurtwr@c-71-193-60-169.hsd1.ca.comcast.net) joined #scummvm. [09:24] kurtwr (kurtwr@c-71-193-60-169.hsd1.ca.comcast.net) left irc: Ping timeout: 265 seconds [09:34] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) left irc: Quit: Ciao! [09:49] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) left irc: Quit: L0ngcat [09:53] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) joined #scummvm. [09:57] WooShell (woo@188-192-188-111-dynip.superkabel.de) joined #scummvm. [10:01] möin [10:05] vendu (~vendu@unaffiliated/vendu) joined #scummvm. [10:05] hola :) [10:25] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) left irc: Quit: L0ngcat [10:31] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) joined #scummvm. [10:33] SylvainTV (~Sylvain@ALille-553-1-34-142.w90-34.abo.wanadoo.fr) joined #scummvm. [10:33] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services. [10:35] James|GlideM (~James|Gli@cpc2-mapp11-2-0-cust447.12-4.cable.virginmedia.com) left irc: Read error: Connection reset by peer [10:38] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) joined #scummvm. [10:41] morgen [10:44] sirlemonhead (~bduncan22@86-45-5-190-dynamic.b-ras2.prp.dublin.eircom.net) joined #scummvm. [10:58] _sev_ (~sev@109.87.142.103) joined #scummvm. [10:58] _sev_ (~sev@109.87.142.103) left irc: Changing host [10:58] _sev_ (~sev@scummvm/undead/sev) joined #scummvm. [10:58] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services. [10:58] _sev (~sev@scummvm/undead/sev) left irc: Read error: Connection reset by peer [11:10] TAS_2012v (2012@c83-248-147-182.bredband.comhem.se) left irc: Ping timeout: 245 seconds [11:11] TAS_2012v (2012@c83-248-147-182.bredband.comhem.se) joined #scummvm. [11:18] pok0j (~pok0j@abtg238.neoplus.adsl.tpnet.pl) left irc: Ping timeout: 252 seconds [11:24] Sigh... Will probably need DrMcCoy and clone2727 to review https://github.com/scummvm/scummvm/pull/226 [11:25] should be uncontroversial... but my judgement of that is not apparently very good. [11:30] pok0j (~pok0j@abti147.neoplus.adsl.tpnet.pl) joined #scummvm. [12:06] |Cable| (~cable@130.133.165.83.dynamic.mundo-r.com) left irc: Read error: Connection reset by peer [12:12] Smartnow (~Smartnow@gateway/tor-sasl/smartnow) joined #scummvm. [12:21] Ori Avtalion master * r870ab35 / video/codecs/svq1.h : VIDEO: Change 'class' to 'struct' to match declaration - http://git.io/jaRX_g [12:23] digitall: thanks [12:23] fuzzie: If you mean that commit... that was salty-horse, not me! :-) [12:24] If you pull 226... that was me. [12:24] for the pull request :P [12:25] I was going to just merge... but I'm sure someone would complain... so will give a chance for that first... [12:25] well, it is a non-trivial change, so I would be nervous about it myself [12:25] *shrug* [12:26] It is actually fairly self contained... even though it is in common, it is only used by the dct, fft and rdft classes... [12:26] one problem we seem to have with pull requests is that people are unwilling to merge their own [12:26] and the only users of those are Bink, and the codecs I have looked at... [12:26] And I did check compile, performance, binary size and RAM profiling... [12:27] well, to seriously consider it on the basis of a profile, I'd want to see before/after massif reports [12:28] but I don't think that is particularly relevant [12:28] Yes... I think the problem here is social... Generally they open a PR when they are unsure if it is a "correct" change or otherwise controverisal. [12:28] massif? [12:28] massif is valgrind's heap profiler [12:28] I use gprof and memprog. [12:28] Ah right.. [12:28] Well, the point is that the change doesn't really affect the codec side code... [12:28] it gives you snapshot reports of memory allocations, with information about who is responsible (as call stacks) [12:29] but, yes, PRs can also be used as a "I am going to commit this, does anyone want to point out some obvious downside I missed before I do?" thing [12:29] There is a little overhead to build the tables at construct... but very low. [12:29] such as my dreamweb refactorings, my tinsel global renaming, etc [12:30] And the codec use the RAM tables in the same way as the static ROM ones... so shouldn't be any dynamic perfomance change in the codec. [12:31] True... But this is partly the issue with PRs ... lots of reasons for them, review/acceptance criteria/sign-off person is not clear.. [12:31] and it eventually comes down to one developer pushing merge... [12:32] yes, perhaps we need to start clearly marking as "I will merge this if there are no objections." vs "This needs a code review before it is merged, please."? [12:32] Yes... something like that.. Would be nice if Github had a decent set of fields for that... I see they have added a responsible reviewer and milestone field... [12:33] Nick change: masak_ -> masak [12:33] yes, it is getting better. [12:34] Incremental improvement... This is also another reason I tend to like presuming merging code contributions, unless there are major problems or serious design objections. [12:35] Well, as I said, I think it's difficult for any one developer to consider all the portability/etc implications when they're making non-minor changes to common code. [12:35] So it'd be nice to let people have a *relatively short* period of time to comment. [12:36] Yes... Would also be nice if you could set a review timer for merge on a PR... [12:36] But indeed, effective blocking of merges due to minor issues becomes spectacularly unproductive. [12:36] Yes. [12:42] L0ngcat_ (~Jonatan@ti0111a380-1086.bb.online.no) joined #scummvm. [12:42] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) left irc: Read error: Connection reset by peer [12:42] Nick change: L0ngcat_ -> L0ngcat [12:45] capiscuas (~capiscuas@ppp-58-8-165-2.revip2.asianet.co.th) joined #scummvm. [12:45] hi everybody [12:48] hey capiscuas :) [12:48] how's it going? [12:49] hi vendu, i have a question, while scummvm is running the games... does it decode the phrases strings and display them? in this case, will it be possible to apply some sort of fan-translations into the game? [12:50] capiscuas, you'd probably need a developer or someone to figure out how hard that would be :) [12:50] so in short, not qualified to answer :) [12:51] ah ok, thanks, i am reading this thread http://www.gamedev.net/topic/257181-translating-scumm-games/ [12:51] which mentions maybe the source code of the code is nneded [12:52] :) [12:52] Strangerke_ (~Strangerk@46.253.166.145) joined #scummvm. [12:52] It depends on the game. [12:53] I think capiscuas is a bit confused... ScummVM is a collection of engines... You are talking about the scumm engine [12:53] they mention about some scummtr that may do the trick [12:53] used for the Lucasarts games such as DOTT etc.? [12:54] capiscuas: That is not one of our tools... [12:54] Strangerke (~Strangerk@46.253.166.145) left irc: Ping timeout: 260 seconds [12:54] Nick change: Strangerke_ -> Strangerke [12:54] As fuzzie said, this would differ per engine... but we don't regularly support fan translation in any official manner. [12:55] but afaik for the LucasArts SCUMM games it is at least possible. [12:56] digitall, like always fan translations are not official [12:56] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services. [12:56] Yes... But it does incur some issues associated with fangames... and SCUMM has other issues. [12:56] http://wiki.scummvm.org/index.php/HOWTO-Fangames [12:58] It is not an official thing... it is a Lucasarts thing: [12:58] "Also, we highly discourage writing fangames for the SCUMM engine. As is indicated by the preceding fan-fiction link, LucasArts Legal have a history of pursuing any unauthorised use of their characters, stories etc. by fans. " [12:58] hmm, i was just refering to a translation [12:58] not creating another game based on lucasart characters [13:00] Yes... but even if you just distribute a patch file of strings... rather than a modified set of datafiles... [13:00] that is still a file of the game dialogue (albeit translated into a different language)... [13:02] Most fan friendly game companies would then probably welcome that or add the translation as an official patch on their site... [13:03] Unfortunately, if the current incarnation of LA notices this, they will probably send a C&D letter and a pack of attack lawyers :/ [13:03] Anyway, IANAL...but just making you aware of the issues here. [13:04] ok, that is not concern... i'm not hosting the translations... whoever wants to do go ahead with the risks... i was concerned about the open source tools to allow in-translation of scumm games [13:04] well, we don't really provide any scumm editing tools at all [13:04] For similar reasons.... and no one has written them. [13:04] but for the lucasarts scumm games, it is possible, and scummtr supports some stuff (but afaik is closed). [13:05] And nothing to do with us... [13:05] so how the strings are coded in the scumm game depen on each game? [13:05] Yes. [13:06] ok, thanks digitall [13:06] and scummvm supports hundreds of non-scumm games too. [13:14] ajax16384 (~User@ip196.net172.n37.ru) left irc: Ping timeout: 265 seconds [13:18] Javacat (~Javacat@unaffiliated/javacat) joined #scummvm. [13:22] vpelletier (~vincent@ALille-154-1-57-86.w86-198.abo.wanadoo.fr) joined #scummvm. [13:24] ajax16384 (kvirc@ip168.net174.n37.ru) joined #scummvm. [13:24] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services. [13:24] Kirben (kirben@c220-239-194-17.brasd3.vic.optusnet.com.au) left irc: [13:25] h00ligan (~User@ip70.net171.n37.ru) joined #scummvm. [13:25] #scummvm: mode change '+o h00ligan' by ChanServ!ChanServ@services. [13:26] Nick change: vpelletier -> Subdino [13:26] Nick change: Subdino -> vpelletier [13:27] ny00123 (~ny00123@bzq-79-181-214-190.red.bezeqint.net) joined #scummvm. [13:28] ajax16384 (kvirc@ip168.net174.n37.ru) left irc: Ping timeout: 276 seconds [13:31] Nick change: h00ligan -> ajax16384 [13:38] Drat... Do we have any wiki maintainer in the house? [13:38] LordHoto (~loom@unaffiliated/lordhoto) joined #scummvm. [13:38] #scummvm: mode change '+o LordHoto' by ChanServ!ChanServ@services. [13:39] Hmm... [13:39] Action: digitall checks the calibration on his summoning chicken-with-a-pulley-in-the-middle... [13:40] Kaidane (~Kaidane@adsl-065-005-218-152.sip.cae.bellsouth.net) joined #scummvm. [13:40] Anyway... just for note, I thought it was me... but the ScummVM wiki font size is apparently almost unreadably small for other people as well... [13:40] I just checked the CSS and the older ScummVM wiki skin uses percentage font sizing... [13:41] but the new ScummVMmodern wiki skin uses fixed point sizing and it is using 7.5pt for the font which is really too small. [13:41] It should be 12pt... or 10pt at a pinch to render readable for people with normal eyesight. [13:42] Or changed back to relative font-sizing... though I'm no CSS expert. [13:42] Hopefully criezy might be able to look at this when he is around. [13:42] Action: vpelletier thinks the wiki looks normal... or is not on the right page [13:42] anyway, I love small fonts [13:43] Javacat (~Javacat@unaffiliated/javacat) left irc: Quit: Please, try the fish [13:43] yes, but small fonts don't really make for a wide viewing audience :P [13:43] term-in-7pt kind of small... tilda-is-minus-one-extra-pixel-long kind of small [13:46] hm, MMDeluxe seems to have some dorky stuff in it, in German. [13:46] vpelletier: Is it with an orange gradient to white background? If so, then that is the Modern Theme... Rather than plain white [13:47] which is the classic ScummVM theme kin... [13:47] It is the modern theme skin for the wiki that is a bit too small to read the text comfortably... IMHO. [13:49] *shrug* It is borderline... SF.net uses a similarly small size for the bug text... but again that is a bit painfully small. [13:50] digitall: orange-yellow gradient on text's left, main area background white and orange gradient behind titles [13:50] but my screen DPI might be off, I usually don't check this [13:51] Yes, that is the Modern skin... It is the default for non-logged in users. [13:52] Hence the issue... i.e. I can change my skin if I'm logged in... but casual readers can't... and we should default to a more readable default [13:52] 10pt would be a good comprimise. [13:53] well, 8pt would be a big improvement. [13:53] checked dpi settings from xdpyinfo-announced screen physical sizes, they are off by 1cm (so close enough to correct seting, with 96dpi square) [13:53] LunaVorax_ (~LunaVorax@cau33-2-82-227-183-10.fbx.proxad.net) joined #scummvm. [13:54] ah, and firefox enforces a minimal font size [13:54] I did wonder... I can use the View->Zoom options to make it readable... but the point here is to have a more readable default. [13:55] so who controls the wiki? [13:55] fuzzie: http://wiki.scummvm.org/index.php/Wiki [13:55] It's one of the sites sitting on hosteurope, I saw. [13:55] Maintainers listed are criezy and jvprat. [13:56] Well, maintainer != has control though, I assume.. [13:56] I have an account with edit privs.. but to change skins etc. will probably need criezy. [13:56] criezy has full access AFAIK... He did the upgrades of mediawiki etc. [13:57] Probably a mail to him cc'ed to scummvm-devel is best idea here... [13:59] fuzzie: Shall I do this, or will you? [13:59] I would be grateful if you would. [13:59] I actually thought I'd done it before, but I dont' see it in the list archives. [14:02] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) joined #scummvm. [14:02] I am still sort of staring at MMDeluxe.. [14:02] The AGS compiler isn't quite helpful enough to include *all* the game script source in the game files. [14:04] And I guess the logic I'm looking for must be in the global game script. [14:06] Done. [14:07] :/ [14:07] Are you saying it bundles all the sources files in the binary?!? or just most?!? :-) [14:07] It helpfully includes all the room script source in the room files, in pre-3.0 versions. [14:08] Or that it compiles most of the files, but skips some to give the game author something to debug! :-) [14:08] Ah... [14:10] Thanks again, btw. [14:15] LeChuck joined #scummvm. [14:15] LunaVorax_ (~LunaVorax@cau33-2-82-227-183-10.fbx.proxad.net) left irc: Remote host closed the connection [14:22] Well, it's probably irrelevant overall, but it does make me despair every time I look at it. :) [14:22] A lot easier to decompile this way, since you can just pattern-match everything. [14:24] \o. [14:24] Action: digitall picks up his arm [14:27] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) left irc: Quit: ERC Version 5.3 (IRC client for Emacs) [14:28] kliph (~user@24-183-96-81.dhcp.mdsn.wi.charter.com) joined #scummvm. [14:40] Nick change: Vampire0_ -> Vampire0 [14:40] LordHoto: I think he gets to take a day off! ;-) [14:42] anyone have indy and the fate of atlantis play either mt-32 or general midi with an external synth? [14:43] vendu: I could... why? [14:43] digitall, for some reason or another, it seems to use soundcard audio for music too :| [14:44] the git version of scummvm is perhaps 1-2 weeks old [14:44] gonna try with the latest as i have built it :) [14:44] digitall, at the moment, i have a roland gm compatible synth hooked :) [14:44] soundblaster gameport midi [14:46] clone2727 (~clone2727@pool-71-251-38-73.nwrknj.east.verizon.net) joined #scummvm. [14:46] #scummvm: mode change '+o clone2727' by ChanServ!ChanServ@services. [14:47] clone2727: \o/ [14:48] digitall: Why did you assign that pull request to me? [14:49] Errm... because I thought you would be the "best" person to review it... i.e. the most interested party. [14:50] It isn't really clear what the "assigned to" field on a pull request is intended for... [14:51] i.e. is it meant to be the person to review... or only that person can do the merge.... [14:52] I think anyone can do the merge. [14:56] Yes, I know it is only an information field... but the point is the intended use is unclear. [14:56] vendu: Working fine here. [14:57] Probably that FFT one needs some work. Will review later if I remember. [14:57] digitall, real mt-32? [14:57] You need to set the MT-32 Tab to SEQ, and set the "Real MT-32 (disable GM)" box. [14:57] And ensure the Music driver is set to default. [14:58] Or vendu, did you mean the full sentence "Digitall, do you have a real MT-32?" :-) [14:58] ffmpeg does something smart for the FFT bit, I think creating a global table with the max needed bits. [14:59] digitall, yep :) [14:59] fuzzie: Yes, I did think I need to revisit the fft.* class, especially with respect to DECL_FFT macro. [14:59] digitall, i have it at SEQ now + general midi on... still plays through the soundcard :P [14:59] But incremental improvement. [14:59] digitall, it's the fm-towns version of indy [14:59] also [15:00] isn't the fft16() passing constant values around? [15:00] vendu: The FM-Towns version might not support MT-32 MIDI btw. [15:00] i.e. cosTable[1] and cosTable[3] for a cosTable with bits==4 ought to be predictable [15:00] vendu: Then I have no idea... and I would suggest that you find a not-obsucre variant... and test with that until you have your MIDI [15:00] system setup debugged. [15:01] LordHoto, digitall, just about to try the 'regular' cd dos version :) [15:01] fuzzie: Yes... as I said, I need to go back and rework the fft class. [15:01] The FM-Towns version of Indy4 is probably quite expensive. [15:02] I guess going back to ffmpeg and taking the *sane* half might be a good idea. [15:02] But the point here was to get rid of the static tables with performance regression! [15:02] fuzzie: Sane half... in FFMPEG.... LMAO. [15:02] LordHoto, digitall, the 'regular' cd version plays general midi :) [15:03] Indy4 shouldn't have any GM tracks, so it's auto-converted MT-32 MIDI. [15:24] Smartnow (~Smartnow@gateway/tor-sasl/smartnow) left irc: Remote host closed the connection [15:26] Schnaks (~Schnaks@p4FC4B3C3.dip.t-dialin.net) left irc: Ping timeout: 260 seconds [15:27] Schnaks (~Schnaks@p4FC4AE30.dip.t-dialin.net) joined #scummvm. [15:40] Smartnow (~Smartnow@gateway/tor-sasl/smartnow) joined #scummvm. [16:00] _marc` (~marc@2a00:1828:2000:76:6502:6561:6569:6581) left irc: Quit: _marc` [16:02] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) left irc: Quit: Ciao! [16:10] oh for .. argh. [16:10] The original engine dies in MMDeluxe at the same place as my engine dies, with the same error. [16:10] I just wasted, what, 3 hours trying to track that down? [16:18] fuzzie: Which bug? [16:19] the mmdeluxe one? [16:20] Ah, it doesn't happen if I use the engine version they were using. [16:20] yes [16:20] I guess they're expecting a queue which doesn't exist. [16:21] It sets Ed's room id to -1, and then when Ed is meant to step outside the mansion, it switched to Ed as the player character. [16:22] And modern AGS promptly schedules a switch to room -1, and explodes. [16:22] Schnaks (~Schnaks@p4FC4AE30.dip.t-dialin.net) left irc: [16:23] fuzzie: :( [16:24] So, no MMDeluxe for you! [16:24] AGS is sounding like SCI... but worse. [16:25] Well, the sensible approach is to pick some modern games to target. [16:28] But what I should do instead is implement keyboard input, so I can play platform games. [16:33] fuzzie: I guess that confirms that AGS isn't always backwards compatible. (I had seen some examples of that, but they were all very minor glitches.) [16:35] Well, of course the backwards compatibility is irrelevant for us anyway, since we can't recompile the games. [16:35] |Cable| (~cable@130.133.165.83.dynamic.mundo-r.com) joined #scummvm. [16:37] I take it there is no easy way to detect which version the game was made for/with. [16:37] There is. [16:38] Well, that's good at least. [16:39] So the 'sprinkling version checks all over the code' approach works fine. :) [16:40] There's a developer called JJS who has been doing a lot of backwards-compatibility work on the original codebase, but it seems to be mostly based on guesswork, which I guess isn't too surprising. [16:40] But is the source code for those earlier versions available, or would it have to be done on a case by case basis, as the glitches/bugs are noticed? [16:40] There is a distinct lack of any response about source for earlier versions. [16:41] So I imagine case-by-case REing is not a bad approach, if that continues to be the case. [16:43] I suspect that will be the best approach... It may be the case that the source code for earlier versions does not exist i.e. no RCS was used. [16:44] It's all in sourcesafe, apparently, and in any case there are porters with copies of the code for the more recent versions. [16:46] VSS isn't an RCS, it is a plot to destroy coder's souls! :) [17:13] I just discovered that room background palettes didn't work *at all* in 256-colour mode in my engine, thanks to eriktorbjorn. [17:14] That's a bit terrifying, it means that everything I've tested so far must have been running in >8bpp (since the conversion path from 8bpp->higher backgrounds was working). [17:21] I did? Er, I mean, of course I did! What did I do? :-) [17:24] Well, it seems that Trilby's Notes SE is, in fact, a 8bpp game. :) [17:24] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) joined #scummvm. [17:32] TAS_2012v (2012@c83-248-147-182.bredband.comhem.se) left irc: Ping timeout: 252 seconds [17:33] risca (~risca@wnpgmb0903w-ds01-249-233.dynamic.mtsallstream.net) joined #scummvm. [17:33] TAS_2012v (2012@c83-248-147-182.bredband.comhem.se) joined #scummvm. [17:35] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) joined #scummvm. [17:37] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) left irc: Client Quit [17:37] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) joined #scummvm. [17:38] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) left irc: Client Quit [17:43] fuzzie: are you around? [17:46] On advice, I'm changing my GSoC project application from the testing framework project to the saving anywhere project, and I'm not sure what to put for "deliverables" other than working savestates for some number of engines [17:47] hm, I am sort of cooking, bad timing [17:47] what I suggested to other students was: "I imagine the sensible thing to do would be to try and estimate the difficulty for some popular engines, and then base your proposed GSoC schedule on that guess." [17:49] But there are also other considerations; making it so that engines can autosave often, rather than truly being able to 'save anywhere' including in the middle of cutscenes, is probably sufficient for ScummVM's purposes. [18:01] yeah, but just adding a timed autosave is much less interesting [18:03] Mmhm. [18:03] Upth... You could implement some kind of process freezing? :-) [18:03] Another obvious thing to point out is that you can't do anything smart at the process level. [18:03] Drat! [18:03] Sorry. :P [18:04] I'm not quite sure what is meant by "anything smart" [18:04] So we can't override malloc and new and implement a ScummVM backend memory allocator... :-) [18:04] You can't start taking snapshots of RAM or anything. [18:04] oh, right. [18:05] digitall: not unless we want to do that for every backend, which would require fairly intimate hardware knowledge of every small device scummvm works on [18:06] And a lot of magical fairy dust for platforms like Android where the Java VM is in charge... :) [18:07] Not really... I meant that we could provide new, free, malloc etc. to the engine from the backend... and have a pool of memory chunks... [18:07] which are frozen and then restored... [18:07] Sure, but then you have a bunch of pointers in them. [18:07] But yes, I am only slightly serious... as yes, you would have to detect and swizzle the pointers... [18:07] oh, right they could just trampoline to the standard new malloc etc [18:08] If you can manage to do that, there's a lot of engineers at places like Google who would like to hear from you. :-) [18:08] after doing their special stuff [18:08] and then you have the state of the rest of the common/graphics etc. :-) [18:08] It would be interesting to see someone do a proof of concept for something like that... but I think it is way out of scope here. [18:08] actually a friend of mine (who is notoriously difficult to get in touch with) did something not altogether dissimilar [18:08] Not to forget detecting the size of variables, which you might need to swap bytes because of endian issues, right? [18:09] digitall, i have a high-speed multithreaded/scalable version of malloc et al for libc :) [18:09] LordHoto: Not sure.. I'm not talking about migrating that kind of process snapshot between platforms... just for one machine. [18:09] LordHoto: I think trying to make a far less ambitious savestate system won't work out if you try and make it so the states can be moved between devices. [18:09] uses bonwick-style (sun microsystems) magazine cache layer on top of a slab allocator [18:09] I'm not sure. [18:09] he made a program called hourglass, which fakes being a debugger to inject a DLL into windows programs that then hooks input, graphics, and systime calls [18:10] in order to provide a deterministic environment [18:10] Anyway, I think we need a hammer to crack this nute... That would be using a Tact Nuke! :P [18:11] Upth: So what happens when the OS decides to change the address space from underneath this mess? :) [18:11] As far as I can tell, it doesn't affect anything [18:12] It does have some issues dealing with threading, however, so it doesn't work terribly well on anything that uses threads for anything other than sound processing [18:13] Well, "sound processing" covers a lot of things. :) [18:13] fuzzie: Pff, I want my states moved between my devices! [18:14] actually detecting variable size wouldn't be terribly difficult if you hooked new [18:14] but would if you hooked malloc [18:14] Well, step one is to make sure all the engines are only using 565 for their 16bpp graphics. [18:15] And, well, bear in mind that any GSoC application actually has to be sort of realistic. :P [18:15] Upth: I'm pretty sure you will have some fun deciding what members you have in structures. [18:16] LordHoto: I always sort of imagined that a new struct would recursively make a new one of every member [18:16] Upth: I think you lost me there, what are you talking about? ;-) [18:16] Structs do not secretly contain pointers in them. :P [18:17] oh, I was thinking of classes [18:17] or c++ structs [18:17] struct and class is pretty much the same thing in C++ [18:18] but not the same as c structs [18:18] They are. [18:18] If you're only using C-compatible bits in them. [18:18] I do have another question to throw out here while people are around who might have answers [18:19] I'm not sure what to do for my patch submission [18:19] Hum, there is another student with that problem too. [18:19] I had been planning to make a UI for the event playback/recording framework [18:20] but that no longer seems like a good idea, as I won't be the one modifying the event playback/recording framework, and the UI could very possibly be completely incompatible with an engine-based method [18:20] Well, it's a bit too early to work out which method to use anyway. [18:23] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) left irc: Ping timeout: 272 seconds [18:24] To one student who was interested in GUI work, I suggested maybe looking into rollovers for the GUI elements which don't have one. [18:26] For saving .. well, AGOS has a quickLoadOrSave() for Simon the Sorceror 1/2 if you have one of them, and it seems like it should really be possible to try hooking that up to the GMM, but .. I think it's too much work for a patch. [18:26] GMM? [18:27] the Global Main Menu [18:27] as opposed to in-game onl [18:27] y [18:27] Ah. Unfortunately I don't have either of those games [18:29] There's also the other engines which don't have support for listing games or loading games from the launcher (e.g. Drascula). But again I'm not sure of the workload. [18:31] But I guess adding support for just listing games (via --list-saves) to Drascula should be trivial enough. [18:34] _sev_ (~sev@scummvm/undead/sev) left irc: Read error: Connection reset by peer [18:35] _sev_ (~sev@109.87.142.103) joined #scummvm. [18:35] _sev_ (~sev@109.87.142.103) left irc: Changing host [18:35] _sev_ (~sev@scummvm/undead/sev) joined #scummvm. [18:35] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services. [18:37] oh look [18:37] when I ctrl-f for list-saves, I found this: // TODO: Make the argument optional. If no argument is given, list all savegames for all configured targets [18:38] I am not sure whether we really want that though ;-). [18:38] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) joined #scummvm. [18:39] Well, also it would demonstrate less familiarity with the metaengine interface. [18:39] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) left irc: Client Quit [18:42] are there any free drascula engine games? [18:42] Yes, Drascula. :-) [18:42] Drascula is a free game. [18:42] It's why I suggested that one. [18:43] http://scummvm.org/downloads/#extras [18:43] The other freely-available games seem to have pretty good support for this stuff at a glance. [18:47] Upth: It would be nice to hear what engines you plan to look at and what exactly you mean by Savestate. [18:48] well, I plan to do SCUMM for certain, as SCUMM is the flagship engine [18:48] Yes, you mentioned that. [18:49] as for the other engines, I was going to ask here what the most popular engines are [18:49] But second, third, fourth etc. seems rather unprecise. [18:49] imprecise even. [18:49] It's probably better to pick engines you'd want to work on anyway. [18:50] alright, then SCI and AGI go in next, as I was a child of Sierra [18:50] I imagine it's probably overly optimistic to try and finish just those three. [18:52] as far as what I mean by savestates, I mean, well, have you ever used an emulator? [18:52] Yes, I use DOSBox regulary. [18:52] :-P [18:53] And I used qemu in the past too. [18:53] I know DOSBox rerecording has savestates, but I don't know if they're in mainline [18:53] At any rate, you should explain what savestates are. [18:55] The "Saving Everywhere" description seems to talk about plain old saved games, so it would be nice to hear about whether savestates are different and if so in which way they are different. [18:56] Well, the description carefully doesn't specify. :P [18:57] _sev__ (~sev@scummvm/undead/sev) joined #scummvm. [18:57] #scummvm: mode change '+o _sev__' by ChanServ!ChanServ@services. [19:01] digitall (digitall@cpc2-hitc2-0-0-cust28.9-2.cable.virginmedia.com) left #scummvm ("bbl"). [19:01] _sev_ (~sev@scummvm/undead/sev) left irc: Ping timeout: 265 seconds [19:03] L0ngcat (~Jonatan@ti0111a380-1086.bb.online.no) joined #scummvm. [19:04] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) left irc: Quit: Ciao! [19:06] risca (~risca@wnpgmb0903w-ds01-249-233.dynamic.mtsallstream.net) left irc: Quit: Lämnar [19:08] LordHoto, fuzzie, I have updated the task description to specify the definition of savestates, and be more somewhat more specific regarding engine modules I intend to work on. [19:09] capiscuas (~capiscuas@ppp-58-8-165-2.revip2.asianet.co.th) left irc: Quit: Leaving [19:11] Well in this case I think you can be happy when you manage to do this for two of the engines you mentioned. [19:16] Okay, I completed an actual game in my AGS engine. [19:17] hooray [19:17] A fairly short one, admittedly. [19:17] nice :-) [19:17] (also, hi) [19:17] (Specifically, Automation: http://i.imgur.com/rOE5p.png ) [19:21] That is a much clearer description of savestates. [19:30] Upth: Your Project Title line looks wrong btw. [19:31] fixed? [19:32] No, you still write "Project Title: Testing Framework (and perhaps TASing support)" [19:32] oh. dammit. [19:35] there. [19:51] risca (~risca@wi-secure-7602.cc.umanitoba.ca) joined #scummvm. [20:18] P2E (~tgz@74.60.157.211) left irc: Ping timeout: 245 seconds [20:20] P2E (~tgz@74.60.157.211) joined #scummvm. [20:33] vendu (~vendu@unaffiliated/vendu) left irc: Quit: Hey! Where'd my controlling terminal go? [21:00] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) joined #scummvm. [21:05] debris_ (~deb@host68-22-dynamic.25-79-r.retail.telecomitalia.it) left irc: Client Quit [21:08] Matthew Hoops master * r473a097 / (3 files in 2 dirs): [21:08] GRAPHICS: Make YUV410 conversion code use bilinear interpolation [21:08] SVQ1 no longer looks blocky and now looks a lot closer to what QuickTime outputs - http://git.io/zygufA [21:08] I note the conspicious lack of a footnote saying "PS it is about 50 times slower". [21:08] No exaggerations! [21:09] More like 4 :P [21:09] Well, for the purposes of that line, let us take a memset() to black as the comparator. [21:20] kurtwr (kurtwr@c-71-193-60-169.hsd1.ca.comcast.net) joined #scummvm. [21:24] kurtwr2 (kurtwr@c-71-193-60-169.hsd1.ca.comcast.net) left irc: Ping timeout: 260 seconds [21:45] debris (~effect@host68-22-dynamic.25-79-r.retail.telecomitalia.it) joined #scummvm. [21:46] debris (effect@host68-22-dynamic.25-79-r.retail.telecomitalia.it) left #scummvm. [21:47] ajax16384 (~User@ip70.net171.n37.ru) left irc: Read error: Connection reset by peer [21:47] Nick change: Upth -> Upthorn [22:12] _sev_ (~sev@scummvm/undead/sev) joined #scummvm. [22:13] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services. [22:13] _sev__ (~sev@scummvm/undead/sev) left irc: Read error: Connection reset by peer [22:21] droid2727 (~AndChat@12.sub-75-228-68.myvzw.com) joined #scummvm. [22:21] #scummvm: mode change '+o droid2727' by ChanServ!ChanServ@services. [22:24] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) joined #scummvm. [22:26] ny00123 (~ny00123@bzq-79-181-214-190.red.bezeqint.net) left irc: Quit: Leaving [22:34] vpelletier (~vincent@ALille-154-1-57-86.w86-198.abo.wanadoo.fr) left irc: Quit: Zzz [22:37] risca (~risca@wi-secure-7602.cc.umanitoba.ca) left irc: Quit: Lämnar [22:37] risca (~risca@wi-secure-7602.cc.umanitoba.ca) joined #scummvm. [22:37] risca (~risca@wi-secure-7602.cc.umanitoba.ca) left irc: Remote host closed the connection [22:55] giucam (~giulio@adsl-ull-111-254.49-151.net24.it) left irc: Ping timeout: 265 seconds [22:57] ok, so, I've got my patch working, how do I make a proper pull request? [22:57] push to your fork, create pull request? [22:59] it's the second part I'm unsure of [22:59] github has a pull request button [23:00] looks like I need to register as a github user [23:01] yes. [23:01] push to your *github* fork. [23:01] so, make github account, hit 'fork' on the web UI of the scummvm repo, then push a branch to that repo. [23:02] yeah, doing that now. [23:03] (then it'll give you a list of recently pushed branches at the top of the web UI for your own scummvm repo, with a 'pull request' button next to each, which is convenient.) [23:08] should I adorn the pull request in anyway to indicate it is gsoc related? [23:08] No. Do add it to your application (or add a comment), though. [23:22] done [23:26] WooShell (woo@188-192-188-111-dynip.superkabel.de) left irc: Quit: svc.startd: The system is coming down. syncing file systems... done. [23:27] droid2727 (~AndChat@12.sub-75-228-68.myvzw.com) left irc: Quit: Statement: Good-bye, meatbags! [23:38] MetalSnake (~MetalSnak@77-22-219-166-dynip.superkabel.de) left irc: Quit: Bye [23:41] MetalSnake (~MetalSnak@77-22-219-166-dynip.superkabel.de) joined #scummvm. [23:57] hennymcc_ (~hennymcc@p57A89F0C.dip.t-dialin.net) joined #scummvm. [00:00] --- Sun Apr 15 2012