[Back to Index]

[00:12] <-- girafe left irc: Read error: Connection reset by peer
[00:15] <snover> even in code, bellbot is kind of a jerk
[00:19] --> coltork joined #scummvm.
[00:21] <-- waltervn left irc: Quit: Leaving
[00:23] <coltork> Hi! I am a follower and Ive seen rests of previous developments of games Id like to ask if its planned to resume the current state
[00:23] <coltork> just out of curioisty
[00:23] <coltork> desktop adventures and igor uikokahonia
[00:24] <coltork> any info shared is appreciated!
[00:25] <snover> coltork: unlikely. http://wiki.scummvm.org/index.php/Igor
[00:26] <snover> if someone wants to do it then it will get done, otherwise not
[00:26] <coltork> thanls for the lin, snover
[00:27] <coltork> I take profit to congratulate the big job you have done with SCI. Its really awesome
[00:27] <snover> thanks!
[00:29] <coltork> and regarding dekadv, any plans to resume? https://github.com/digitall/scummvm-deskadv
[00:29] <Lightkey> salty-horse hasn't been here in weeks
[00:30] <snover> same thing with digitall. but i imagine they will pop up again before too long. they would be the best people to ask, coltork.
[00:32] <snover> oh, Space Quest 6. you have been so good at having the most common code that all the other interpreters have, until now, with this old version of the MIDI/sound code.
[00:32] <coltork> ok thanks1
[00:32] <-- coltork left irc: Quit: http://www.kiwiirc.com/ - A hand crafted IRC client
[00:42] --> omer_mor joined #scummvm.
[00:44] <-- omer_mor_ left irc: Ping timeout: 240 seconds
[00:52] <-- t0by left irc: Quit: t0by
[01:30] <-- Dominus left irc: Ping timeout: 260 seconds
[01:31] --> Dominus joined #scummvm.
[01:35] --> Joefish_ joined #scummvm.
[01:35] #scummvm: mode change '+v Joefish_' by ChanServ!ChanServ@services.
[01:39] <-- Joefish left irc: Ping timeout: 255 seconds
[01:52] <-- SylvainTV left irc: Read error: Connection reset by peer
[01:54] <-- m_kiewitz left irc: Read error: Connection reset by peer
[02:15] <-- borosky left irc: Ping timeout: 255 seconds
[02:22] <-- dreammaster left irc:
[03:22] <-- Mia left irc: Ping timeout: 255 seconds
[03:31] --> Tkachov joined #scummvm.
[03:32] #scummvm: mode change '+o Tkachov' by ChanServ!ChanServ@services.
[03:47] <-- simbalion` left irc: Quit: WeeChat 1.5
[04:33] <-- |Cable| left irc: Read error: Connection reset by peer
[04:34] --> |Cable| joined #scummvm.
[04:42] <-- Lightkey left irc: Ping timeout: 240 seconds
[04:55] --> Lightkey joined #scummvm.
[05:04] --> jamm joined #scummvm.
[05:04] <-- jamm left irc: Changing host
[05:04] --> jamm joined #scummvm.
[05:20] --> Begasus joined #scummvm.
[05:21] <Begasus> morning peeps
[05:24] --> Begas_VBox joined #scummvm.
[05:53] <Joefish_> morning
[05:53] Nick change: Joefish_ -> Joefish
[06:27] <Simei> morning :)
[06:29] --> frankyboy_ joined #scummvm.
[06:33] <-- LittleToonCat left irc: Remote host closed the connection
[06:35] <-- frankyboy_ left irc: Read error: Connection reset by peer
[06:36] --> frankyboy_ joined #scummvm.
[07:02] <-- jamm left irc: Ping timeout: 240 seconds
[07:23] --> t0by joined #scummvm.
[07:23] <-- t0by left irc: Changing host
[07:23] --> t0by joined #scummvm.
[07:23] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[07:23] <-- t0by left irc: Remote host closed the connection
[07:34] --> Henke37 joined #scummvm.
[08:12] --> waltervn joined #scummvm.
[08:12] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[08:43] <waltervn> morning
[08:45] --> Mia joined #scummvm.
[08:45] <-- Mia left irc: Changing host
[08:45] --> Mia joined #scummvm.
[08:50] --> m_kiewitz joined #scummvm.
[08:50] #scummvm: mode change '+o m_kiewitz' by ChanServ!ChanServ@services.
[08:58] --> jamm joined #scummvm.
[08:58] <-- jamm left irc: Changing host
[08:58] --> jamm joined #scummvm.
[09:18] --> borosky joined #scummvm.
[09:29] <-- ST left irc: Ping timeout: 240 seconds
[09:39] --> ST joined #scummvm.
[09:39] #scummvm: mode change '+o ST' by ChanServ!ChanServ@services.
[09:52] <-- ced117 left irc: Remote host closed the connection
[09:53] --> ced117 joined #scummvm.
[10:10] <-- frankyboy_ left irc: Remote host closed the connection
[10:12] --> _sev_ joined #scummvm.
[10:12] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[10:13] <-- _sev left irc: Ping timeout: 240 seconds
[10:14] --> WooShell joined #scummvm.
[10:17] <WooShell> good meowning =^.^=
[10:21] --> frankyboy_ joined #scummvm.
[10:26] --> Farmboy0 joined #scummvm.
[10:26] <-- Farmboy0 left irc: Changing host
[10:26] --> Farmboy0 joined #scummvm.
[10:27] <logix> wd
[10:27] <logix> oops
[10:48] <-- heroux left irc: Ping timeout: 246 seconds
[10:49] --> heroux joined #scummvm.
[10:52] <Joefish> I'm stuck with quite a big diff on one file. Is there a way to only apply part of a diff? I thought about manually editing the patch file generated by git format-patch but that's quite tedious..
[11:01] <wjp> git add -i
[11:01] <Joefish> of course.. I spend hours yesterday figuring out a better way to do that and google didn't help much. Now that I asked here and one search later I found what it wanted (git add -i)
[11:02] <Joefish> :)
[11:02] <Joefish> Is this well known or rather an obscure part of git?
[11:03] <Joefish> because it sure wasn't easy to find
[11:12] --> SylvainTV joined #scummvm.
[11:12] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[11:15] <-- Mia left irc: Ping timeout: 240 seconds
[11:54] --> t0by joined #scummvm.
[11:54] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[11:57] <t0by> o/
[11:57] <Joefish> \o
[11:57] <t0by> Joefish: hey there!
[11:57] <t0by> How is it going?
[12:01] <Joefish> okay I guess. Hit a few blocks in my project and now I'm stuck with a almost 900 lines diff I need split up. But it compiles again and going forward :)
[12:01] <Joefish> how about you?
[13:17] --> _sev|work joined #scummvm.
[13:17] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:21] <-- _sev|work left irc: Client Quit
[13:38] --> _sev|work joined #scummvm.
[13:38] <-- _sev|work left irc: Changing host
[13:38] --> _sev|work joined #scummvm.
[13:38] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:43] --> dreammaster joined #scummvm.
[13:43] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[13:43] <-- _sev|work left irc: Client Quit
[13:53] <t0by> _sev_:
[13:53] <t0by> [20:43] <_sev_> instead of calling processEvent(LEvent event, ScriptType st, int entityId)
[13:53] <t0by> [20:43] <_sev_> it should call processEvent(LEvent event)
[13:53] <_sev_> yes
[13:53] <t0by> I'm sorry, why are we losing the id?
[13:54] <_sev_> because take a look at the events diagram
[13:54] <t0by> Or rather, where do we find that again?
[13:54] <_sev_> the ID is dependent on the D version
[13:54] <_sev_> thus, you do it right inside of processEvent()
[13:54] <_sev_> e.g.
[13:55] <_sev_> you need to do something like this:
[13:55] <t0by> Arf?
[13:55] <_sev_> for mouseDown
[13:55] <_sev_> call primary handler (when...)
[13:56] <_sev_> if (events_disabled) stop. // you disable with dontPassEvent
[13:56] <_sev_> call score script, with the current score ID
[13:56] <_sev_> if (events_disabled) stop
[13:56] <_sev_> call cast member script with the cast ID
[13:57] <_sev_> if....
[13:57] <_sev_> call frame script with frame ID
[13:57] <_sev_> e.g. you have whole bunch of IDs involved
[13:57] <_sev_> err. second call is with the sprite ID
[13:57] <_sev_> which you're getting from the intersection
[13:58] <_sev_> t0by: is that clear?
[13:58] <t0by> No, I get that. However, I thought a primary motivation here was to hide the 3-ary processEvent method from client code, is that right?
[13:59] <t0by> I have the slight suspicion I misunderstood that entirely
[13:59] <_sev_> yes
[13:59] <_sev_> because what I described is for D4
[13:59] <_sev_> D3, D2 have simpler model
[13:59] <t0by> Please bear with me a second
[14:00] <-- SylvainTV left irc: Read error: Connection reset by peer
[14:00] <t0by> Okay. So the entry point for everything is to be the 1-ary processEvent, right? And that I can and should be able to obtain those IDs from inside there, right?
[14:00] <_sev_> yes
[14:01] <_sev_> then you have one place for those ifs
[14:01] <_sev_> I've started this work
[14:01] <_sev_> see events.cpp
[14:01] <t0by> I have it under my nose right now
[14:02] <_sev_> err
[14:02] <t0by> I reckon we need to mimic the handler chain diagram from Chapter 3 of the book (i.e. IMG_1670.JPG)
[14:02] <_sev_> lingo-events.cpp
[14:02] <t0by> Well, I have both :P
[14:03] <t0by> *However*, I'm not sure how to get those IDs if I am not given those from client code (note that calls appear also inside loops). What are you trying to tell me about those?
[14:03] <t0by> You can't mean those are hardcoded?
[14:03] <t0by> (Dependent on D version)
[14:04] <_sev_> you need to do exactly the same as it does now
[14:04] <_sev_> e.g.
[14:04] <_sev_> spriteId = currentFrame->getSpriteIDFromPos(pos);
[14:04] <_sev_> just do it once, generic and when needed
[14:05] <t0by> Oh. That was my original understanding then, but... I started having second thoughts as there is a bunch of those in tricky places.
[14:05] <t0by> e.g. void Frame::executeImmediateScripts()
[14:06] <t0by> How would you approach that?
[14:07] <_sev_> move whole thing into Lingo
[14:08] <t0by> Oh. Fine, then.
[14:08] <t0by> Thank you
[14:08] <_sev_> t0by: and there is no 100% assurance that this is the correct thing which we are doing
[14:09] <_sev_> t0by: in fact, as you see it
[14:09] <_sev_> _frames[_currentFrame]->executeImmediateScripts();
[14:09] <_sev_> thus, these are the frame scripts
[14:09] <_sev_> for the current frame
[14:10] <t0by> _sev_: as for that, I'm not touching semantics here
[14:10] <_sev_> maybe it is utterly incorrect, e.g. the subsequent call of kEventEnterFrame
[14:10] <_sev_> should actually do this thing
[14:12] t0by nods
[14:15] <t0by> thanks
[14:27] --> AJ_ joined #scummvm.
[14:31] <-- ajshell1 left irc: Ping timeout: 255 seconds
[14:49] --> LittleToonCat joined #scummvm.
[14:52] <TMM> I was thinking of maybe adding a sort of compatibility layer for ffmpeg's libavcodec/libavformat files to scummvm so we can easily import codecs from ffmpeg with minimal changes
[14:52] <TMM> would anyone have any objections to such a thing?
[14:54] <TMM> we could then replace our current codecs with stuff directly from ffmpeg too
[14:54] <t0by> _sev_: regarding earlier, I don't see a builtin symmetrical to dontPassEvents, is it our responsibility to re-enable those when we've reached the bottom of the fallthrough chain, right?
[14:59] --> GitHub36 joined #scummvm.
[14:59] <GitHub36> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHjIV
[14:59] <GitHub36> scummvm/master ec9351b Paul Gilbert: DEVTOOLS: Added extra German NPC data to create_titanic
[14:59] GitHub36 (GitHub36@192.30.252.40) left #scummvm.
[15:02] --> AJ__ joined #scummvm.
[15:04] <-- AJ_ left irc: Ping timeout: 246 seconds
[15:16] <t0by> (I'm naively assuming unstubbing amounts to a flag in DirectorEngine)
[15:32] <-- borosky left irc: Ping timeout: 246 seconds
[15:40] --> Littleboy joined #scummvm.
[15:40] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[15:46] --> borosky joined #scummvm.
[15:49] <-- Begasus left irc: Ping timeout: 240 seconds
[15:56] <-- dreammaster left irc:
[16:01] --> Begasus joined #scummvm.
[16:02] Nick change: _sev_ -> _sev
[16:02] <-- t0by left irc: Quit: t0by
[16:02] --> t0by joined #scummvm.
[16:02] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[16:04] <-- jamm left irc: Ping timeout: 240 seconds
[16:18] <-- AJ__ left irc: Quit: Leaving
[16:22] --> ny00123 joined #scummvm.
[16:24] <-- Henke37 left irc: Ping timeout: 240 seconds
[16:34] --> Strangerke_ joined #scummvm.
[16:36] <-- Strangerke left irc: Ping timeout: 255 seconds
[16:36] Nick change: Strangerke_ -> Strangerke
[17:02] <-- t0by left irc: Quit: t0by
[17:02] --> t0by joined #scummvm.
[17:02] <-- t0by left irc: Changing host
[17:02] --> t0by joined #scummvm.
[17:02] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[17:17] <snover> Does anyone have a copy of the German version of GK1 and could do a couple things for me? I am like 99% sure that https://bugs.scummvm.org/ticket/9791 is invalid but I would like it if someone else could independently verify this.
[17:19] <snover> CD German*
[17:20] <Harekiet> NEIN!
[17:22] <-- Tkachov left irc: Quit: Leaving
[17:23] t0by throws Harekiet off the blimp
[17:23] <t0by> No ticket.
[17:26] <-- Begasus left irc: Ping timeout: 255 seconds
[17:27] <logix> lol
[17:30] --> Henke37 joined #scummvm.
[17:37] --> Mia joined #scummvm.
[17:37] <-- Mia left irc: Changing host
[17:37] --> Mia joined #scummvm.
[17:38] --> Begasus joined #scummvm.
[17:44] <-- Begas_VBox left irc: Quit: Vision[0.9.8]: i've been blurred!
[17:58] <-- Begasus left irc: Ping timeout: 255 seconds
[18:10] --> Begasus joined #scummvm.
[18:14] --> Begas_VBox joined #scummvm.
[18:26] <-- frankyboy_ left irc: Ping timeout: 246 seconds
[18:31] <omer_mor> snover: I do
[18:32] <snover> omer_mor: could you send me complete hashes of your RESOURCE.MAP, ALT.MAP, RESOURCE.000, RESOURCE.ALT, and SIERRA.EXE?
[18:33] <omer_mor> 1 sec
[18:34] <omer_mor> 7884dc9c1b6922549e62afacf3c59a50 *ALT.MAP
[18:34] <omer_mor> 54c24e1b27d1da886ff85716dae58c6e *RESOURCE.000
[18:34] <omer_mor> 713e1985548aba06ed64b4f31dcab673 *RESOURCE.ALT
[18:34] <omer_mor> 373cc528dca3d006a488400e29e7c02a *RESOURCE.MAP
[18:34] <omer_mor> f6a72a5f74a23756571286c67bae5778 *SIERRA.EXE
[18:38] <snover> thanks
[18:38] <TMM> snover, do you have any negative opinions on me implementing a libavfomat/libavcodec compat layer into osystem? The idea is that we can import codecs from ffmpeg with almost no changes into scummvm.
[18:39] <TMM> snover, I'd want to add this layer, then move my mve player to it and then take the codecs we have scummvm that exist in ffmpeg also and import them from ffmepg, we can then sync them when we have bugfixes. Maybe cooperate a bit more with ffmpeg to get any enhancements we have into ffmpeg
[18:40] <TMM> snover, for the most part this would mean creating a couple of adapters for osystem
[18:41] <TMM> 's file abstractions, and create adapters, or implement, range-safe bytestream readers.
[18:41] <snover> TMM: i keep waffling back and forth on it
[18:42] <TMM> ffmpeg uses endian safe reading everywhere, it would just be a macro definition to use the osystem ones in those files etc
[18:42] <snover> definitely when you were talking about this the other day i thought about how scummvm will basically never get any security updates in codecs copied from upstream and modified
[18:42] <TMM> snover, full disclosure: I'm planning to implement more FMV games in the future, this is entirely self serving :P
[18:43] <snover> so from that perspective it would be good to make it easy to pull things from upstream (though of course someone would still need to pull things)
[18:44] <TMM> yeah, of course
[18:44] <TMM> but it'd be way less daunting to do so
[18:46] <TMM> I can do a first pass of this just enough to support MVE
[18:47] <TMM> If it is liked then I can convert the other formats
[18:47] <TMM> I don't know if we have any codecs that aren't in ffmpeg, if we do maybe I can port them *to* ffmpeg :)
[18:48] <snover> scummvms VMD support is better than ffmpegs
[18:55] <TMM> cool, maybe I should submit that to ffmpeg's then
[18:56] --> dreammaster joined #scummvm.
[18:56] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[18:56] --> girafe joined #scummvm.
[18:59] <TMM> snover, I'll just do a first pass for MVE, if it is liked I'll go on. I thinks submitting scummvm's vmd enhancements to ffmpeg is a good idea anyway
[19:04] <t0by> fuzzie: hi, I'm sorry, is there a Director movie that you know works with your viewer? (https://github.com/fuzzie/continuity)
[19:05] <t0by> Oh, nevermind, those from LIngo Workshop demos work fine
[19:12] <snover> TMM: that seems OK to me. i dont know if anyone else will end up having a more concrete objection. i am having a difficult time pinning anything down in my head today. all i can think of are largely personal objections (like my feelings about C/the ffmpeg API), with a rare abstract thought about how C and C++ are not entirely compatible.
[19:12] <bgK> TMM: please note that I would object to the replacement of ScummVM's QuickTime and Bink implementations with the ffmpeg ones
[19:13] <bgK> Myst uses QuickTime features that are not in ffmpeg (accurate edit lists) and were tricky to get working right
[19:14] <bgK> Myst III uses Bink seek support which is not in ffmpeg either, and was not easy to get working accurately either
[19:14] <bgK> replacing the docoders would require testing the games again
[19:14] <bgK> *decoders
[19:14] <snover> dohcoders
[19:14] <bgK> :)
[19:15] <DrMcCoy> Also, ScummVM's Bink decoder is based on ffmpeg code
[19:15] <DrMcCoy> Only cleaned up a lot
[19:16] <DrMcCoy> The VMD decoder in ScummVM is completely different from ffmpeg's one, though. I implemented IMD (Coktel Vision's earlier format) in ScummVM before then building it up to VMD, while ffmpeg never had an IMD decoder
[19:18] <DrMcCoy> I still have going through ffmpeg's VMD decoder and fixing it (or maybe even changing it completely to fit my own into it) in my TODO list
[19:19] <DrMcCoy> As for Bink, it took me a lot of time to clean up the code, I'd hate to see that time wasted, frankly. I do think it's far more readable now (but I'm biased, of course). I'd rather have fixes/improvements done to ffmpeg afterwards applied to that codebase
[19:20] <snover> DrMcCoy: hey! not to derail this conversation, but as long as we are talking about VMD, did you ever get a chance to look at that Lighthouse VMD sample?
[19:20] <DrMcCoy> snover: Argh, damn, nope, not really, sorry
[19:20] <DrMcCoy> snover: There's something funky going on, I'd have to dive into the game disasm, I guess
[19:21] <DrMcCoy> I didn't manage to make it work just by fiddling with the code, unfortunately
[19:21] <snover> no worries. i am under a pile of over 9000 bugs so itll probably be a bit before I go back to Lighthouse in any case.
[19:21] <DrMcCoy> (Back to the Bink decoder) On the other hand, where I do see room for improvements is the BitStream interface and especially the Huffman decoder. These are obviously very naive implementations, and not the fastest ones at that
[19:22] <DrMcCoy> The Huffman decoder can be improved by going the tree route, but that needs a BitStream decoder that can "peek", i.e. read without consuming the bits
[19:23] <DrMcCoy> Likewise, the BitStream decoder could be made faster by more intelligent caching
[19:23] <TMM> bgK, well, I'd implement those features in ffmpeg in case I'd go ahead with that of course. Part of the reason for this is to make sure all our stuff ends up in ffmpeg as much as it is to get bug/security features from ffmpeg into scummvm
[19:24] <DrMcCoy> I hacked something like that into xoreos, as a test (because I wrote an unrar library, and had written the code for that anyway), but for Bink itself, it wasn't really any speed-up at all, interestingly enough. Might help with other codecs, though
[19:25] <TMM> bgK, the idea is very emphatically not to just replace it without looking, or even replace all of the codecs at once
[19:25] <DrMcCoy> Or maybe that would show improvements for mobile systems?
[19:28] <bgK> bgK: be aware that it is a *lot* of work, not just importing the ffmpeg code making it match the quirks of ScummVM's video decoder API, regression testing the games -- and fixing them. For IMO very little benefit
[19:28] <TMM> DrMcCoy, if your code is more readable porting that back to libavcodec may be worth it. From talking to the ffmpeg developers they don't much care about the 'obscure video game formats' they seemed to imply that as long as no new security problems were introduced they don't give a crap
[19:28] <bgK> erm TMM
[19:29] <DrMcCoy> Well, the VMD format is very weird to begin with :P
[19:29] <DrMcCoy> It's a textbook Coktel Vision format
[19:30] <DrMcCoy> Very convoluted, with lots of weird stuff bolted on
[19:31] <DrMcCoy> Like optionally encoding frames as Indeo3, for example
[19:31] <TMM> if we have the libavformat/libavcodec split in scummvm that wouldn't be a big problem though
[19:33] <TMM> bgK, well, the benefit would be over the longer term, if we need to add more video codecs it'd be real easy to do so
[19:33] <TMM> bgK, and then there's the bugfix angle :)
[19:36] <bgK> TMM: are there many potentially useful codecs that are in ffmpeg but not in ScummVM?
[19:36] --> Poly-C joined #scummvm.
[19:37] <TMM> bgK, I can do an inventory of FMV games I'd like to support and their codecs if that's helpful for decision making
[19:38] <-- borosky left irc: Ping timeout: 260 seconds
[19:39] <snover> unfortunately that bugfix angle is a double-edged sword for code without unit testing
[19:39] <TMM> bgK, but the converse is certainly true, it'd be useful for the obscure media format community as a whole if it becomes easier to share scummvm improvements with ffmpeg :)
[19:40] <-- Polynomial-C left irc: Ping timeout: 240 seconds
[19:41] <TMM> err inverse, not converse
[19:41] <TMM> what the hell brain
[19:41] <bgK> TMM: making a list should help you as well decide if it's worth investing your time in it
[19:42] <TMM> well, my thought process right now is that I'm going to have to do this work at least once for the mve crap anyway
[19:42] <bgK> also I don't know if it's still true, but a few years back the ffmpeg API was a moving target
[19:42] <bgK> TMM: MVE is easy since there are no games using the decoder for now
[19:43] <TMM> that's true
[19:43] <TMM> I was referring more to whether or not to actually modify the codec or osystem to integrate it
[19:44] <TMM> if I modify the codec I now have two mve implementations I'm maintainer for :P
[19:48] <TMM> bgK, the last internal API change to ipmovie.c was in august 2014, and it was just a renaming of a logging function
[19:48] <TMM> bgK, I think the moving target for the internal apis aren't so bad
[19:49] --> GitHub9 joined #scummvm.
[19:49] <GitHub9> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHjlR
[19:49] <GitHub9> scummvm/master f827e27 Paul Gilbert: TITANIC: Fix crash clicking mouse when copyright dialog is showing
[19:49] GitHub9 (GitHub9@192.30.252.40) left #scummvm.
[19:51] <snover> they have a thing about their versioning/compatibility guarantees: https://ffmpeg.org/doxygen/trunk/index.html
[19:52] <snover> of course this is for public APIs only
[19:52] <TMM> snover, this is about the avcodec/avformat internal apis though, those don't seem particularly churny
[19:55] <snover> i am reading and communicating below a functional level today :)
[19:55] --> Strangerke_ joined #scummvm.
[19:58] <-- Strangerke left irc: Ping timeout: 240 seconds
[19:58] Nick change: Strangerke_ -> Strangerke
[20:03] --> borosky joined #scummvm.

[20:13] <-- exmensa left irc:
[20:15] <Lightkey> dreammaster: hey, why am I not surprised that's all you fixed? :-)
[20:15] <Lightkey> here is what you ignored: 494. should end with a comma as in the English version 516. Su\xDF""erst should be Superst I guess? 533. \xFC""r should start with an F 537. getStigt should be get\E4""tigt 550. allgemeinen should start with an A 553. Fr\xFCh""chte has a stray h after FC
[20:17] <Lightkey> I would add dashes to Mailing Listen and Flunker Finder because it hurts my eyes but up to you if you want to fix grammar errors in nouns
[20:21] <-- borosky left irc: Ping timeout: 240 seconds
[20:22] <dreammaster> Hey. Are we still talking about current? 494.. "wie man sie l\xF6""schen kann: \n"?
[20:35] <Lightkey> yes
[20:35] <Lightkey> <Lightkey> I don't understand your answer though, compare 491 with 413 and 494 with 415
[20:35] <Lightkey> <Lightkey> they are both ending with a comma in English
[20:44] <dreammaster> Okay, sorry. Now I'm starting to grasp the difference. This was actually correct in the original, but i guess it makes better sense with a comma than a colon. 516: No, it,s correct with \xDF..
[20:48] <dreammaster> 533: No, it's meant to start the line with \xFC and 'r'. 537: You're correct. 550: No, it's lowercase in the original. 553: You're correct.
[20:49] <dreammaster> As for the hyphens, sounds reasonable

[20:50] <Lightkey> I know 550 is lowercase in the original, grammar disagrees though
[20:51] <Lightkey> and how does 533 make sense, if there some F as a picture before that?
[20:51] <dreammaster> Alright then, I'll capitalize.

[20:53] --> borosky joined #scummvm.
[20:54] <dreammaster> Maybe you're right, and it's a misspelling.. do you speak German natively? Or you relying on Google Translate?
[20:54] <Lightkey> dreammaster: that line could have been Chinese just as well for me :-)
[20:55] <dreammaster> Okay, I'll leave it as is for now.. we can always easily change it if a German native can confirm the misspelling.
[20:56] <Lightkey> dreammaster: you need two? :-)
[20:57] <dreammaster> That just leaves, 533, which in the original is: 005AD170 20 64 65 6E 20 41 75 67 65 6E 2E 0A 0A 46 FC 72 den Augen...Fnr
[20:57] <dreammaster> So, you're correct about the missing F
[20:58] <dreammaster> Oh, so you are German fluent?
[20:59] <Lightkey> you want my passport or does /whois suffice? :p
[20:59] <dreammaster> No, that you are, then I'm happy to go ahead and change the superst as well
[21:00] <dreammaster> You know, there's that whole range of the STRINGS_DE array still waiting to be translated.. hint hint, nudge nudge, know what I mean. :)
[21:01] --> GitHub139 joined #scummvm.
[21:01] <GitHub139> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHjBS
[21:01] <GitHub139> scummvm/master 4433694 Paul Gilbert: DEVTOOLS: Further minor German fixes to create_titanic, thanks to Lightkey
[21:01] GitHub139 (GitHub139@192.30.252.42) left #scummvm.
[21:02] <logix> "Superst" might actually by "{ae}u{ss}erst" (the {..} are both supposed to be the corresponding umlaut chars)
[21:02] <Lightkey> dang logix

[21:04] <snover> no way. im sure that this was not a typo in the game, but just an extremely clever pun. absolutely. 100%.
[21:04] <Lightkey> dreammaster: what he said
[21:04] <logix> ah... E4 is a capital {Ae} in ISO 8859
[21:04] <logix> so, yeah
[21:05] <Lightkey> dreammaster: but you missed the missing F anyway
[21:05] <dreammaster> Which one I got the one on 533
[21:07] <_sev> people
[21:07] <_sev> look at this supercool stuff for gdb: https://github.com/vuvova/gdb-tools
[21:07] <Lightkey> dreammaster: wait, I don't see a change to 516 either in that commit
[21:07] <-- girafe left irc: Read error: Connection reset by peer
[21:09] <Lightkey> dreammaster: so if you haven't pushed it yet, change it to \xE4""u\xDF""erst
[21:09] --> GitHub157 joined #scummvm.
[21:09] <GitHub157> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHjRv
[21:09] <GitHub157> scummvm/master 7c65427 Paul Gilbert: DEVTOOLS: Further minor German fixes to create_titanic, thanks to Lightkey
[21:09] GitHub157 (GitHub157@192.30.252.40) left #scummvm.
[21:09] <TMM> _sev, oh, nice! I had written my own pretty printer, this is way better
[21:09] <dreammaster> Erk... just noticed your last line a moment too late, sorry :(
[21:10] <logix> if you want to correct the capitalization, you may change that to lowercase, as the following "erfolgreich", but you can argue for preservation of the original capitalization
[21:10] <_sev> TMM: the main feature for me is the arrow operator
[21:10] <Lightkey> dreammaster: and my three lines before that too? :p
[21:10] <_sev> TMM: so it can look up even recursively, the structures
[21:10] <TMM> _sev, oh, that'll save some typing
[21:11] <TMM> _sev, I cowardly use eclipse's gdb support for expanding nested structs :P
[21:12] <dreammaster> dreammaster: Okay, getting slightly confused here. :). What with the prevoius commit, and the further change to 516, what else would you like me to correct?
[21:12] <-- Begas_VBox left irc: Quit: Vision[0.9.8]: i've been blurred!
[21:13] <Lightkey> Lightkey: join the club!
[21:13] <-- Begasus left irc: Quit: Ex-Chat
[21:13] <Lightkey> dreammaster: no, only 516, per what logix said
[21:14] <dreammaster> Okay, good :)
[21:14] --> GitHub109 joined #scummvm.
[21:14] <GitHub109> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHjRY
[21:14] <GitHub109> scummvm/master fd51315 Paul Gilbert: DEVTOOLS: Yet another Superst correction in create_titanic
[21:14] GitHub109 (GitHub109@192.30.252.41) left #scummvm.
[21:17] <Lightkey> really through me off that Superst fit in that just as well, with all the exaggeration in the text, I'm starting to think that I'm just bad at puzzles
[21:18] <logix> Superst/Aeussert - through/threw - potato/potato
[21:18] <Lightkey> Aeusserst*
[21:18] <Lightkey> SCNR
[21:19] <logix> HA! right
[21:19] --> girafe joined #scummvm.
[21:20] <dreammaster> My gentle nudge about the text strings array needing translation is still open..
[21:22] <Lightkey> I will pass, this was a nice distraction from more important stuff, like.. cleaning up my room, so I could find my copy of Starship Titanic again :-P
[21:28] <dreammaster> No problem. In any case, it's looking less like German will be supported to begin with anyway. When I was figuring out the Sentence lists for each NPC script, it ended up looking like at least each NPC's process method is completely different
[21:28] <dreammaster> That seems indicative that there'd be substantial further changes needed. Not to mention some changes in the parser code.
[21:29] <-- Strangerke left irc: Ping timeout: 246 seconds
[21:29] <dreammaster> And I have the vage collection from when I looked at the German EXE months back that they changed the sound filenames for all the sounds. Which if true means all the sound calls would have to be tracked down at all
[21:30] <dreammaster> sorry, "vague recollection"
[21:38] <-- borosky left irc: Ping timeout: 260 seconds
[21:40] <-- chadj_ left irc: Ping timeout: 240 seconds
[21:41] <-- waltervn left irc: Read error: Connection reset by peer
[21:44] <-- ny00123 left irc: Quit: Leaving
[21:44] --> chadj joined #scummvm.
[21:58] <-- Farmboy0 left irc: Remote host closed the connection
[21:59] <-- demonimin left irc: Ping timeout: 240 seconds
[22:01] --> demonimin joined #scummvm.
[22:05] --> Stormkeeper joined #scummvm.
[22:13] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[22:22] <-- girafe left irc: Quit: Leaving
[22:22] --> borosky joined #scummvm.
[22:39] <-- WooShell left irc: Quit: If you understand or if you don't, if you believe or if you doubt - There's a universal justice, and the eyes of truth are always watching you.
[22:53] --> Strangerke joined #scummvm.
[23:11] <-- Stormkeeper left irc: Quit: KVIrc 4.9.2 Aria http://www.kvirc.net/
[23:12] <TMM> dreammaster, oh, I thought that titanic was all hardcoded logic?
[23:13] <dreammaster> It is. And that includes a set of classes for each NPC to implement all their custom response logic. I just call them names like "LiftbotScript" and "DoorbotScript"
[23:15] <TMM> ah I see
[23:15] <TMM> I can kind of see the appeal of developing a game that way actually
[23:15] <TMM> it's quite fleixble
[23:16] <dreammaster> Right. And given the language differences, it makes sense that the NPC script classes would need more than just minor alterations to handle the different syntaxes
[23:16] <snover> darn these incompatible human languages with different sentence structure.
[23:17] <snover> can you just write a german emulator instead and use the english code
[23:18] <TMM> #include <dosbox.h>
[23:18] <TMM> dooone
[23:18] <snover> more like #include <google-translate.h>
[23:19] <dreammaster> Hmmm... cme to that, I wonder if the default English script classes would be sufficient for someone playing the German version to finish the game. ie. doing only the needed conversation stuff to finish the game
[23:20] <dreammaster> If so, maybe we could technically treat it as supported, and a TODO for the future for it to be fully supported
[23:20] <dreammaster> I mean, as it is, I'm definitely not looking to fully support QSound, and we may end up with the problem with the few Indeo 4 transparency videos still being present.
[23:20] <snover> its unfair to my comedy career that you take my jokes and then figure out how to make them plausible.
[23:21] <snover> you monster.
[23:21] <dreammaster> So having the German version at least completeable, even if it doesn't support various arcane German questions might be nice
[23:21] <TMM> maybe you can just translate the game over in your current code rather than try to re the whole thing all over again?
[23:21] <dreammaster> :)
[23:22] <dreammaster> Well, for full support, it wouldn't be the whole thing. Hopefully just the NPC script class methods. And I do know the German parser has an extra word replacement method that also handles ecrtain German word suffixes,
[23:23] <dreammaster> But all the rest of the game logic remains the same. I think.
[23:24] <-- Strangerke left irc: Ping timeout: 260 seconds
[23:29] <snover> the qsound stuff seems like it might be gimmicky/primitive enough that it doesnt matter very much for the experience, but buggy video rendering sounds like a bigger problem
[23:31] <TMM> what is qsound?
[23:31] <snover> https://en.wikipedia.org/wiki/QSound this thing
[23:36] <snover> i wouldnt expect it to do a very good job at creating a realistic positional audio effect since it would have had to be fast enough to run on a Pentium 2, but maybe i am just too ignorant to have an opinion
[23:36] <dreammaster> From everything I've seen, they just did it a cheap and dirty way to have distint sounds without having to worry about manually specifying volume percentages.
[23:37] <dreammaster> Like when you first go into the Well area, there's a distant sound of water, wohich is basically just a reduction of the volume. So, yeh, I think my simulated volume reduction logic based on distance is quite adequate.
[23:37] <TMM> there's probably a cheap mostly-good enough way of doing that
[23:38] <TMM> especially since most people had shit speakers on their computer back then anyway ;)
[23:38] <snover> (or actually, on a 133MHz Pentium, since that was their recommended system requirement)
[23:38] <dreammaster> I agree though that ideally, the remaining Indeo 4 transparency videos should be fixed. Luckily, they're only used for the furniture in the SGT stateroom and the lever in the Arboretum. Off the top of my head.
[23:38] <dreammaster> So hence I'm still happy with going ahead with the testing period even if it isn't properly handled.
[23:40] <-- Joefish left irc: Ping timeout: 246 seconds
[00:00] --- Mon Jun 19 2017