ScummVM IRC Log - 28 May 2017

[Back to Index]

[00:22] <-- m_kiewitz left irc: Quit: technology isn't intrinsically good or evil. It's how it's used. Like the Death Ray.
[00:33] <-- WooShell left irc: Read error: Connection reset by peer
[01:44] <-- Dominus left irc: Ping timeout: 260 seconds
[01:45] --> Dominus joined #scummvm.
[01:51] --> dreammaster joined #scummvm.
[01:51] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[02:04] <snover> whee. gk2 has some halfway between sci2.1 and sci3 audio
[02:05] <dreammaster> Because of course, why not :P
[02:06] <-- nutron left irc: Read error: Connection reset by peer
[02:06] --> nutron joined #scummvm.
[02:15] <-- dreammaster left irc:
[02:18] --> nutron|w joined #scummvm.
[02:21] <-- nutron left irc: Ping timeout: 240 seconds
[02:28] <-- somaen left irc: Ping timeout: 260 seconds
[02:28] --> somaen joined #scummvm.
[02:47] --> Mataniko joined #scummvm.
[02:57] <Deledrius> >_<
[03:04] <snover> hey, Mataniko! how are things going?
[03:05] <Mataniko> not too bad
[03:05] <Mataniko> whats up
[03:05] <snover> just fixing a game-destroying GK2 bug
[03:06] <Mataniko> I got text to render
[03:06] <Mataniko> so there's that
[03:06] <Mataniko> but most of the text parsing code was already written
[03:06] <snover> nice! first pixels from a new subsystem is always exciting.
[03:06] <Mataniko> trying to decipher how these BMPs are stored
[03:06] <Mataniko> it looks like BMP files contain multiple frames of animation
[03:06] <snover> youre working on& the dynamix engine, right?
[03:06] <Mataniko> yeah
[03:08] <snover> sweet. i think someone once told me that some of the scaling code in SCI came from dynamix
[03:09] <Mataniko> must be in the later games
[03:10] <Mataniko> i got a stupid question
[03:10] <Mataniko> I have a Common:Array<struct>
[03:10] <snover> i have an incoherent response
[03:10] <snover> :)
[03:10] <Mataniko> and i'm pushing in a struct()
[03:10] <Mataniko> but my [] operator doesn't return the actual struct
[03:11] <Mataniko> _framesDefinition->push_back(FrameInfo());
[03:11] <Mataniko> _framesDefinition[i].width <-- doesn't like that
[03:12] <Mataniko> do i need to make Common:Array<FrameInfo &>?
[03:13] <snover> what is the compiler error?
[03:15] <Mataniko> width is not a member of Common::Array<Dgds::FrameInfo>
[03:17] <snover> so, sorry, i just noticed, you are using -> for that push_back, which means that _framesDefinition is not an array object
[03:20] <snover> what is your declaration of _framesDefinition? `Common::Array<Dgds::FrameInfo>* _framesDefinition` perhaps?
[03:21] <Mataniko> yeah, i was looking at an existing example in this project, it shouldn't be a pointer
[03:25] <snover> as far as the question about putting a reference in an array, as far as i know this isnt possible since references arent assignable. (and its probably not what you want anyway.)
[03:25] <Mataniko> yeah that's not what i want, i just declared the array wrong
[03:26] <snover> also, Deledrius, hi. didnt mean to ignore you. did you verify your kq7 games were copied correctly?
[03:27] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[03:42] <Mataniko> god vs2017 is a nightmare with scummv
[03:42] <Mataniko> m
[03:42] <snover> how so?
[03:42] <Mataniko> even without changing anything it goes to a standstill while hitting run
[03:43] <snover> dreammaster uses VS, maybe he knows something
[03:46] <Mataniko> i'm gonna move the project to my ssd
[03:47] --> Strangerke_ joined #scummvm.
[03:50] <-- Strangerke left irc: Ping timeout: 268 seconds
[03:50] Nick change: Strangerke_ -> Strangerke
[03:52] <Mataniko> ok that looks to be good
[04:41] --> GitHub17 joined #scummvm.
[04:41] <GitHub17> [scummvm] csnover pushed 3 new commits to master: https://git.io/vHcq4
[04:41] <GitHub17> scummvm/master 73fab1e Colin Snover: SCI32: Remove invalid interaction cursor spinloop in KQ7
[04:41] <GitHub17> scummvm/master ccf5665 Colin Snover: SCI32: Remove dead code
[04:41] <GitHub17> scummvm/master f940346 Colin Snover: SCI32: Minor cleanup to plane debug information
[04:41] GitHub17 (GitHub17@192.30.252.34) left #scummvm.
[04:45] <-- Cheeseness left irc: Remote host closed the connection
[04:49] --> Cheeseness joined #scummvm.
[05:23] <-- Lightkey left irc: Ping timeout: 258 seconds
[05:36] --> Lightkey joined #scummvm.
[05:40] <snover> bah.
[05:40] <snover> in the end, gk2 doesnt even call kDoAudioPreload, so this call will always just always return 0 and it would be fine.
[05:41] --> Begasus joined #scummvm.
[05:47] <-- eriktorbjorn left irc: Ping timeout: 272 seconds
[05:53] <snover> s/call will/call could
[05:53] <snover> (obviously not doing that though since it would probably break other games)
[05:55] <snover> gotta deal with some other edge cases too, like it looks like they started storing the original channel index (channels will shift around and end up in different indexes as they stop playing) and using that in some places
[05:57] <snover> they also seem to have reversed the channel insertion order, so a new channel is unshifted instead of pushed. i dont know offhand if there is any good reason to do this. it would seem to run counter to the behaviour of kDoAudioPreload, but then again, i dont recall seeing that call anywhere
[05:57] <snover> oh well. more work for tomorrow.
[06:01] <Deledrius> np snover, didn't feel ignored at all. :) Yes, I did verify it. Checked it against another (standalone) CD copy too, and they appear to be identical.
[06:01] --> eriktorbjorn joined #scummvm.
[06:01] #scummvm: mode change '+o eriktorbjorn' by ChanServ!ChanServ@services.
[07:42] <-- LittleToonCat left irc: Remote host closed the connection
[07:45] --> Begas_VBox joined #scummvm.
[07:47] --> t0by joined #scummvm.
[07:47] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[07:47] <t0by> Morning
[07:48] <t0by> Sime, _sev: I think that for the time being we can safely ignore multitexturing - i.e. the "true" branch of if(backdropExists)
[07:48] <_sev> of course
[07:48] <t0by> Simei: *
[07:48] <Simei> Morning
[07:50] <-- Harekiet left irc: Ping timeout: 240 seconds
[07:53] <Simei> https://www.irccloud.com/pastebin/qlXVczx1/
[07:54] <t0by> Hm?
[07:54] <Simei> _sev, t0by: I found in the code of loading backdrop(background) image, it can be non png image
[07:54] <t0by> Simei: yes, same goes for sprites
[07:54] <t0by> if you read the code you'll notice that png is only involved in spriteBankVersion > 3
[07:55] <t0by> (or something like that)
[07:55] <Simei> _sev, t0by: decide by the png_sig_cmp
[07:55] <Simei> _sev, t0by: So, what does it do here for non png image ?
[07:56] <t0by> IIRC PNGDecoder::loadSTream() returns true iff decoding was successful (and thus the stream *is* a png file)
[07:56] <t0by> Simei: reads raw bytes, line by line, pixel by pixel.
[07:56] <Simei> _sev, t0by: Ok, I see, else what use what to decode it ?
[07:57] <t0by> It's not encoded in the first place, it's a raw bitmap. Try bmp.h if the byte order matches (look it up), otherwise you may have to wrap it by hand?
[07:58] <t0by> (PROTIP: Probably someone has subclassed ImageDecoder for the right byte order already)
[07:59] <t0by> Sorry, I don't have the code in front of me now, I'm about to go grocery shopping
[08:00] <t0by> Simei: despite the fact that I said it's "not encoded" we still call it a BitmapDecoder. And indeed it's a form of coding, but you get my gist.
[08:01] --> Harekiet joined #scummvm.
[08:01] <Simei> _sev, t0by: Ok, I see. So, I'll look what is like in BitmapDecoder then.
[08:02] <t0by> I'll be back in 30 minutes
[08:03] <Simei> t0by: It's not a hurry. Take your time. Bon shopping :)
[08:34] --> Henke37 joined #scummvm.
[08:36] <Lightkey> https://twitter.com/simoncarless/status/868687656263204864 the future is here!
[08:57] --> Strangerke_ joined #scummvm.
[09:00] <-- Strangerke left irc: Ping timeout: 240 seconds
[09:00] Nick change: Strangerke_ -> Strangerke
[09:15] <TMM> anyone know of an engine that doesn't support launcher savegame loading?
[09:24] --> GitHub102 joined #scummvm.
[09:24] <GitHub102> [scummvm] hpvb closed pull request #952: GUI: Fully hide GMM save/load buttons if not supported by engine (master...hide_gmm_saveload) https://git.io/vH3zX
[09:24] GitHub102 (GitHub102@192.30.252.45) left #scummvm.
[09:24] --> GitHub70 joined #scummvm.
[09:24] <GitHub70> [scummvm] hpvb pushed 2 new commits to master: https://git.io/vHcZ9
[09:24] <GitHub70> scummvm/master dadf1db Willem Jan Palenstijn: GUI: Fully hide GMM save/load buttons if not supported by engine...
[09:24] <GitHub70> scummvm/master 379314c Hein-Pieter van Braam: Merge pull request #952 from wjp/hide_gmm_saveload...
[09:24] GitHub70 (GitHub70@192.30.252.41) left #scummvm.
[09:24] <TMM> dreamweb :)
[09:24] <TMM> (also I meant gmm)
[09:27] <wjp> dreamweb doesn't support gmm loading? How lazy of us
[09:30] <TMM> I think I did the merge wrong, it looks like scummvm doesn't normally merge from the github interface, sorry I added a gross 'merge pull request' commit to the history
[09:30] <wjp> we usually do that
[09:31] <TMM> oh
[09:31] <TMM> oh, I guess it's mostly snover that doesn't then
[09:31] <TMM> oh well
[09:40] --> GitHub165 joined #scummvm.
[09:40] <GitHub165> [scummvm] hpvb closed pull request #949: ENGINES: Allow detection entries to match on full paths (master...advanceddetector-fullpaths) https://git.io/vHTnZ
[09:40] GitHub165 (GitHub165@192.30.252.42) left #scummvm.
[09:40] --> GitHub2 joined #scummvm.
[09:40] <GitHub2> [scummvm] hpvb pushed 2 new commits to master: https://git.io/vHcnG
[09:40] <GitHub2> scummvm/master 159e5a9 Colin Snover: ENGINES: Allow detection entries to match on full paths...
[09:40] <GitHub2> scummvm/master d6645f3 Hein-Pieter van Braam: Merge pull request #949 from csnover/advanceddetector-fullpaths...
[09:40] GitHub2 (GitHub2@192.30.252.45) left #scummvm.
[09:43] --> jamm joined #scummvm.
[09:47] --> m_kiewitz joined #scummvm.
[09:47] #scummvm: mode change '+o m_kiewitz' by ChanServ!ChanServ@services.
[09:52] --> Littleboy joined #scummvm.
[09:52] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[09:58] --> Henke37b joined #scummvm.
[10:01] <-- Henke37 left irc: Ping timeout: 246 seconds
[11:05] --> Strangerke_ joined #scummvm.
[11:08] <-- Strangerke left irc: Ping timeout: 268 seconds
[11:08] Nick change: Strangerke_ -> Strangerke
[11:41] --> WooShell joined #scummvm.
[11:41] --> waltervn joined #scummvm.
[11:41] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[11:59] <WooShell> good meowning =^.^=
[11:59] <-- Henke37b left irc: Quit: ERR_SHUTDOWN
[12:06] <TMM> hi!
[12:09] <-- t0by left irc: Quit: t0by
[12:20] --> dreammaster joined #scummvm.
[12:20] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[12:37] <-- Strangerke left irc: Ping timeout: 240 seconds
[12:45] --> omer_mor_ joined #scummvm.
[12:48] <-- omer_mor left irc: Ping timeout: 268 seconds
[13:03] --> Strangerke joined #scummvm.
[13:12] <-- omer_mor_ left irc: Ping timeout: 245 seconds
[13:14] --> omer_mor joined #scummvm.
[13:25] --> omer_mor_ joined #scummvm.
[13:28] <-- omer_mor left irc: Ping timeout: 268 seconds
[13:47] --> t0by joined #scummvm.
[13:47] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[13:48] <t0by> Simei's blog, everyone: https://yinsimei.wordpress.com/2017/05/28/gsoc-week-1/
[13:50] <t0by> If only those Sierra employees kept so neat a blog...
[13:53] <dreammaster> Nice indeed. Though it's a WordPress site.. does that have a suitable feed for inclusion in Planet? And even if so, I'm not sure the table formatting will survive
[13:56] <t0by> um
[13:56] <t0by> planet is picking up her old posts, so.
[13:57] <dreammaster> Oh right, yes. It's already set up. Good. Just this latest post not yet appearing
[13:57] <criezy> I think I set it up to pick up the posts tacked with ScummVM.
[13:57] <criezy> This is the case of the first post but not the second one.
[13:58] <-- Begasus left irc: Ping timeout: 246 seconds
[13:58] <t0by> This one is
[13:58] <t0by> https://yinsimei.wordpress.com/tag/scummvm/
[13:58] <t0by> planet borked?
[13:58] <criezy> This can be changed to pick up all posts or those tagged with GSoC if needed
[14:00] <criezy> Also planet update itself once an hour. So you might need to be a bit patient.
[14:01] <criezy> Yes, the third post is also tagged with ScummVM. It's just the seond one that is not as far as I can see (the Git tips post).
[14:02] <criezy> Maybe we want to include all GSoC posts in planet and not just the ones tagged ScummVM?
[14:02] <Simei> :D
[14:03] <Simei> criezy: Ah, I think the git one is just for myself maybe.
[14:03] <Simei> It doesn't really related to scummvm
[14:04] <dreammaster> Ah, the new post has turned up. And the table formatting survived. Good
[14:05] <Simei> They are screenshots in fact :) I think table doesn't work well in Wordpress neither
[14:05] <dreammaster> Ah. That explains it
[14:07] <logix> the red wavy lines under some words might have been a hint :)
[14:08] <Simei> Ah, that's true. :)
[14:09] <t0by> I have a great april fool's idea for IETF.
[14:10] --> Begasus joined #scummvm.
[14:28] <_sev> t0by: planet is updated hourly
[14:30] <-- snw left irc: Ping timeout: 258 seconds
[14:35] --> snw joined #scummvm.
[14:55] <-- Harekiet left irc: Ping timeout: 240 seconds
[15:07] --> Harekiet joined #scummvm.
[15:13] --> Strangerke_ joined #scummvm.
[15:15] <-- dreammaster left irc:
[15:15] <-- Strangerke left irc: Ping timeout: 245 seconds
[15:15] Nick change: Strangerke_ -> Strangerke
[15:18] --> LittleToonCat joined #scummvm.
[15:30] <-- Strangerke left irc: Ping timeout: 260 seconds
[15:35] --> Henke37 joined #scummvm.
[15:41] --> ny00123 joined #scummvm.
[15:56] <-- t0by left irc: Quit: t0by
[16:06] --> t0by joined #scummvm.
[16:06] <-- t0by left irc: Changing host
[16:06] --> t0by joined #scummvm.
[16:06] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[16:08] <snover> TMM: im on the fence about that GMM change (PR#930)
[16:11] <snover> on the one hand, it doesnt look very good because there isnt enough space to display most game names without an ellipsis, and i would assume that a user knows what game they were playing
[16:12] <snover> the version number is probably more useful information than the game name, for games that are prepackaged with scummvm on e.g. GOG
[16:13] <snover> i was going to have an on the other hand but now im not sure i even have one of those.
[16:16] <-- Polynomial-C left irc: Quit: GNU/Linux, because I'd rather own a free OS than steal one that's not worth paying for.
[16:25] <-- Vampire0 left irc: Ping timeout: 268 seconds
[16:34] <-- jamm left irc: Read error: Connection reset by peer
[16:39] --> Vampire0 joined #scummvm.
[16:42] --> Strangerke joined #scummvm.
[16:43] <wjp> Let's maybe just close it. It was a gsoc PR made spontaneously since we have an ancient item on a todo page on the wiki somewhere
[16:47] <TMM> snover, wjp, yeah I kind of agree. I'd think that a user is more likely to know what game they are playing than what version of scummvm they're on
[16:49] --> GitHub140 joined #scummvm.
[16:49] <GitHub140> [scummvm] hpvb closed pull request #930: GUI: Display game title at the top of the GMM (master...master) https://git.io/vSnjF
[16:49] GitHub140 (GitHub140@192.30.252.42) left #scummvm.
[17:07] <-- Strangerke left irc: Ping timeout: 260 seconds
[17:07] <-- t0by left irc: Read error: Connection reset by peer
[17:09] --> t0by joined #scummvm.
[17:09] <-- t0by left irc: Changing host
[17:09] --> t0by joined #scummvm.
[17:09] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[17:48] <snover> sigh. i *always* read `test; setz; test; jz` backwards.
[17:57] <snover> wjp: its weird that the action command for ignore is none instead of ignore since it is printed as ignore
[17:58] <snover> also it would be nice if it were possible to filter away subops in the debugger itself. for example, I frequently want to watch all of the DoSound subops, except for DoSoundUpdateCues because it is noisy and doesnt tell me anything useful
[17:58] <snover> it doesnt appear that `bpk DoSoundUpdateCues none` & `bpk DoSound* log` does what i would hope
[17:59] <snover> i could postprocess the output with grep, except that i dont think i can do that from xcodes built-in terminal
[18:02] <wjp> right, that's a difference with the current logkernel. The new one only does positive matching, rather than a per-call yes/no flag
[18:04] <wjp> and since the breakpoint list is basically a list of independents action to perform on certain triggers, not entirely clear how to adapt that
[18:06] <wjp> more fancy patterns maybe; or just stick with the old setup
[18:06] <snover> break out of the loop over breakpoints entries when there is a match?
[18:08] <wjp> also an option, although then maybe I'd also make the action field a bitmap so you can mix them
[18:10] <snover> another option i suppose would be to add some additional keyword like "last" to the breakpoint command so you would write `bpk DoSoundUpdateCue ignore last` to stop searching when it matches
[18:11] <snover> (vs `bpk DoSoundUpdateCue ignore` which would continue searching for other matches)
[18:11] --> dreammaster joined #scummvm.
[18:11] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[18:12] <wjp> the rationale for ignore/none was to allow temporarily disabling breakpoints, by the way
[18:29] <wjp> I'm tempted to go with fairly ad hoc slightly more complicated patterns
[18:29] <wjp> DoSound*,!DoSoundUpdateCues or something
[18:30] --> ajax16384 joined #scummvm.
[18:30] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[18:31] <wjp> just to keep the model simple
[18:33] --> Mia joined #scummvm.
[18:33] <-- Mia left irc: Changing host
[18:33] --> Mia joined #scummvm.
[18:34] <snover> i dont have a strong preference one way or the other
[18:40] <Simei> _sev, t0by: Normally, do we rewrite ImageDecoder? Because in sludge, they use a different way to save and write the r g b value of a non png image
[18:40] <Simei> https://www.irccloud.com/pastebin/xPlmp7Iz/
[18:41] <Simei> Is there a byte reading function equal to this one in SeekableReadStream ?
[18:41] <t0by> Simei: if yoou are looking at a sufficientlyt weird format I think you can inherit ImageDecoder
[18:41] <t0by> um, why do you need it?
[18:41] <t0by> please link to the whole file ion github
[18:44] <wjp> that looks like readUint16BE()
[18:44] <Simei> _sev, t0by: https://github.com/yinsimei/scummvm/blob/newWIP1-1/engines/sludge/backdrop.cpp#L184
[18:45] <t0by> ew formatting
[18:45] <t0by> lemme see
[18:45] <Simei> wjp: Ah, ok, we are closer then
[18:46] <Simei> https://www.irccloud.com/pastebin/8BVWwqZJ/
[18:46] <t0by> Simei: I wouldn't touch savegames for now
[18:47] <t0by> I think that stuff is part of savegame functionality
[18:47] <Simei> _sev, t0by: Here, they read the 2 byte value, minus it by 32 if c & 32, and then that's the r, g, b value
[18:48] <Simei> No, it's backdrop for image reading
[18:48] <t0by> Huh?
[18:48] <t0by> A sec
[18:49] <Simei> Ah, sorry, at the wrong line
[18:49] <wjp> & 32? Haha, are they actually using the free bit in a 555 pixel format to do RLE?
[18:50] <wjp> (Of course I'm making assumptions about redValue/greenValue/blueValue here, and I could be completely misinterpreting this)
[18:52] <t0by> wjp: I'm squinting but I can't see RLE in there
[18:52] <Simei> This line, but same idea. https://github.com/yinsimei/scummvm/blob/newWIP1-1/engines/sludge/backdrop.cpp#L1096
[18:52] <wjp> t0by: the n?
[18:52] <Mataniko> I think Dynamix is using some weird RLE compression
[18:53] <Simei> No idea what it is :/
[18:53] <t0by> Simei: run length encoding, look it up on wikipedia
[18:54] <t0by> wjp: still can't see it
[18:54] <snover> t0by: if (c & 32) n = fgetc(fp) + 1; else n = 1
[18:56] <t0by> OOH
[18:56] <t0by> while (--n)
[18:56] <t0by> of course
[18:56] <snover> Mataniko: weird like how?
[18:56] <t0by> Simei: can you see it as well?
[18:56] <t0by> Anyway, you probably want to shove this into an ImageDecoder indeed.
[18:57] --> GitHub174 joined #scummvm.
[18:57] <GitHub174> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHcXU
[18:57] <GitHub174> scummvm/master 0be151d Paul Gilbert: TITANIC: Fix crash inserting photograph into star control
[18:57] GitHub174 (GitHub174@192.30.252.42) left #scummvm.
[18:57] <t0by> wjp: do we have anything that does RLE already
[18:57] <t0by> ?
[18:57] <dreammaster> MSRLE image decoder
[18:57] <dreammaster> Not sure if it's what you need specifically, though
[18:58] <Simei> Are there any customized Image Decoder example in other engines ?
[18:58] <t0by> Simei: yes. Protip: grep "extends ImageDecoder" engines/
[18:58] <t0by> A sec
[18:58] <snover> grep -R*
[18:58] <t0by> I'm searching for a simple one
[18:58] <wjp> uh
[18:58] <t0by> oh right
[18:59] <wjp> let's not do java :-)
[18:59] <t0by> -R
[18:59] <t0by> rats
[18:59] <t0by> sorry
[18:59] <t0by> ": public ImageDecoder"
[18:59] <dreammaster> How about engines\prince\mhwanh.h
[18:59] <Mataniko> If it uses the same format as Betrayal At Krondor, then it's checking for the control character being above or under 0x80 to decide if it's a compressed section
[18:59] <wjp> t0by: and/or Image::ImageDecoder
[18:59] <Mataniko> but that doesn't seem to be the case
[19:00] <wjp> t0by: but most RLE implementation are custom
[19:00] <t0by> Simei: engines/prince/mhwanh.cpp is a neat example, yes
[19:00] <t0by> dreammaster: thanks
[19:00] <wjp> "mhwanh"? wow
[19:00] <dreammaster> Mwa hah hah :)
[19:01] <t0by> wjp: this one certainly is
[19:02] <t0by> Simei: don't forget to replace the png stuff in loadLightmap & the like with your function and see if it works
[19:03] <t0by> This file is rather wacky anyway
[19:03] <snover> (mhwanh is apparently some format also known as HSI Raw used by Image Alchemy)
[19:06] <snover> Mataniko: i dont know how much cross-contamination there is between DGDS and SCI, but the SCI VGA view format uses 0x80 as a flag to indicate an RLE pixel, and pixel data comes from a separate area of the file
[19:07] <Simei> That one is good, thx dreammaster
[19:07] <Mataniko> that sounds similar
[19:07] <Mataniko> but that could be stuff in future dgds used in krondor
[19:07] <Mataniko> but not in the earlier stuff
[19:08] <Mataniko> krondor uses contro & 0x80 to determine if it's a RLE pixel or not
[19:08] <dreammaster> No problem. It's cool that the effort was taken to make a proper ImageDecoder descendent, rather than throwing together a loader method in one of the other classes. Good OO
[19:09] <Mataniko> does SCI use LZW compression then RLE?
[19:10] <Mataniko> or was RLE just added to DecompressorLZW
[19:11] <snover> for compressed views inside of resource volumes they did some crazy stuff
[19:11] <Simei> _sev, t0by: so now I'm about to write the custom ImageDecoder and replace things with PNGDecoder then. :)
[19:11] <t0by> Simei: finding what extends a given class is another thing a good IDE usually does for you.
[19:12] <t0by> Simei: yay :)
[19:12] <t0by> Simei: of course your ImageDecoder should be completely separate from PNGDecoder
[19:12] <t0by> you'd then first try to apply one, then the other.
[19:12] <Simei> t0by: ok, I see :/
[19:12] <t0by> Simei: what's with the :\ ?
[19:13] <t0by> Were you thinking of extending PNGDecoer?
[19:13] <Simei> how much you find a IDE is good
[19:13] <Simei> :)
[19:13] <t0by> Oh, I don't even use one.
[19:14] <Simei> No, just ImageDecoder like in prince ?
[19:14] <snover> Mataniko: you may look at kCompLZW1View and kCompLZW1Pic in engines/sci/decompressor
[19:14] <t0by> Sure, sure
[19:14] <t0by> That's right
[19:15] <Simei> t0by, _sev: Can't find a existing sludge game that use png as background yet. :/
[19:15] <t0by> Simei: make one :)
[19:15] <Simei> Ok
[19:15] <t0by> Simei: take Welcome.slg, convert the backdrop to png, update the script.
[19:16] <Simei> Ok, I see
[19:16] <snover> Mataniko: but then there is a separate decompression step for compressed views (in sci/graphics/view.cpp unpackCelData). IIRC this stuff only reconstructs the original view format from the packed resource volume format. by SCI32 they stopped doing this craziness and just compressed everything with STACpack
[19:17] <snover> er. where this stuff is the LZW view/pic compression
[19:18] <snover> and where STACpack is LZS
[19:19] <Mataniko> yeah this is not as complex
[19:19] <Mataniko> maybe if i stare at it really hard
[19:27] <-- Begas_VBox left irc: Quit: Vision[0.9.8]: i've been blurred!
[19:36] --> SylvainTV joined #scummvm.
[19:36] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[19:47] --> Strangerke joined #scummvm.
[19:47] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[19:54] --> GitHub177 joined #scummvm.
[19:54] <GitHub177> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHcMF
[19:54] <GitHub177> scummvm/master 38becee Paul Gilbert: TITANIC: Renamings of orientation related variables and methods
[19:54] GitHub177 (GitHub177@192.30.252.40) left #scummvm.
[19:57] --> Harekiet2 joined #scummvm.
[19:59] <-- t0by left irc: Quit: t0by
[20:00] --> t0by joined #scummvm.
[20:00] <-- t0by left irc: Changing host
[20:00] --> t0by joined #scummvm.
[20:00] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[20:01] <-- Harekiet left irc: Ping timeout: 258 seconds
[20:01] --> Harekiet joined #scummvm.
[20:04] <-- Harekiet2 left irc: Ping timeout: 246 seconds
[20:08] <-- Begasus left irc: Quit: Ex-Chat
[20:11] <-- t0by left irc: Quit: t0by
[20:14] --> GitHub121 joined #scummvm.
[20:14] <GitHub121> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHcDQ
[20:14] <GitHub121> scummvm/master 99ce2f6 Paul Gilbert: TITANIC: Renamings for photo surface
[20:14] GitHub121 (GitHub121@192.30.252.45) left #scummvm.
[20:16] --> t0by joined #scummvm.
[20:16] <-- t0by left irc: Remote host closed the connection
[20:17] <-- ajax16384 left irc: Read error: Connection reset by peer
[20:22] --> ny00123_ joined #scummvm.
[20:22] <-- ny00123_ left irc: Read error: Connection reset by peer
[20:22] --> ny00123_ joined #scummvm.
[20:23] <-- ny00123 left irc: Ping timeout: 255 seconds
[20:27] --> GitHub117 joined #scummvm.
[20:27] <GitHub117> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHcyC
[20:27] <GitHub117> scummvm/master 6587cac Paul Gilbert: TITANIC: Rename CStarControlSub8 to CPhotoCrosshairs
[20:27] GitHub117 (GitHub117@192.30.252.42) left #scummvm.
[20:42] <wjp> snover: how about this? https://github.com/wjp/scummvm/commit/dd4db8b02159fee2cea8822f8f6477ce7e8231bd
[20:43] <wjp> this makes the bpk argument a comma separated list of atoms, each of which can start with a ! and/or end with a * for negative matches and prefix matches, respectively
[20:43] <wjp> and the last matching atom in the comma separated list takes effect
[20:43] <wjp> so DoSound*,!DoSoundUpdateCues
[20:44] <wjp> or even crazier stuff like DoSound*,!DoSoundUpdate*,DoSoundUpdateCues
[20:47] --> girafe joined #scummvm.
[20:48] <snover> at first glance, it seems ok. curious to know if you think this can/should be generalised to work with non-kernel bps
[20:50] <wjp> now that you mention it, that does kind of make sense since selectors have this natural class prefix too
[20:50] <wjp> (class or object name)
[20:54] sh4rm4^bnc (sh4rm4@sabotage.tech) left #scummvm ("Leaving").
[21:00] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[21:16] --> GitHub134 joined #scummvm.
[21:16] <GitHub134> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHc9l
[21:16] <GitHub134> scummvm/master 8ad4c59 Paul Gilbert: TITANIC: Fix crash switching from starfield to photo display
[21:16] GitHub134 (GitHub134@192.30.252.45) left #scummvm.
[21:28] <-- waltervn left irc: Quit: Leaving
[21:30] <-- ny00123_ left irc: Quit: Leaving
[21:33] --> t0by joined #scummvm.
[21:33] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[22:04] <-- Dark-Star left irc: Ping timeout: 246 seconds
[22:06] --> Dark-Star joined #scummvm.

[22:18] --> GitHub70 joined #scummvm.
[22:18] <GitHub70> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vHcQQ
[22:18] <GitHub70> scummvm/master 30b5fc8 Paul Gilbert: TITANIC: Fix copying viewport _valArray data
[22:18] GitHub70 (GitHub70@192.30.252.41) left #scummvm.
[22:20] <t0by> Simei: hey, I'm about to crawl to bed
[22:20] <t0by> everything okay so far?
[22:23] <Simei> t0by: It's okay. :) Thx
[22:23] <Simei> Good night then !
[22:24] <t0by> Simei: done with the ImageDecoder?
[22:24] <-- 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.
[22:25] <Simei> Not yet, trying to display png read first. just around there. I think the image is too big, need to resize the window
[22:27] <-- TAS_2012v left irc: Ping timeout: 255 seconds
[22:28] <t0by> Mh
[22:28] <t0by> Simei: let's see your working dir please
[22:28] <t0by> Do the usual wip commit and branch
[22:31] --> TAS_2012v joined #scummvm.
[22:34] <Simei> https://github.com/yinsimei/scummvm/commit/e87985e5ce668a1fa16971e30336871e7fb724e9
[22:34] <t0by> Sweet, thanks
[22:34] <Simei> Euh, there is stilll an auto format thing in my editor
[22:35] <t0by> Arf.
[22:35] <Simei> I haven't added a lot code though :/
[22:35] <Simei> I need to disable it somewhere
[22:36] <Simei> https://github.com/yinsimei/sludge-test-games
[22:36] <t0by> Ah, super sweet, thanks
[22:36] <t0by> A sec
[22:36] <Simei> the png game, exactly the same but with png bg
[22:46] <t0by> oh crap, it's segfaulting
[22:46] <t0by> scummvm: backends/graphics/surfacesdl/surfacesdl-graphics.cpp:1368: virtual void SurfaceSdlGraphicsManager::copyRectToScreen(const void*, int, int, int, int, int): Assertion `buf' failed.
[22:47] <t0by> uh
[22:48] <-- m_kiewitz left irc: Quit: technology isn't intrinsically good or evil. It's how it's used. Like the Death Ray.
[22:49] <t0by> Loading is hit..
[22:49] <snover> t0by: werent you going to bed?
[22:49] <Simei> ? It doesn't have segfaulting here. doesn't work though.
[22:49] <Simei> :p
[22:49] <t0by> Simei: well, memory issues are unpredicatable like that.
[22:50] <t0by> snover: oh shut up, Mary Poppins.
[22:50] <snover> wah wah.
[22:51] <t0by> ...or should I say Jimi Hendrix?
[22:55] <snover> all i know is that reverse engineering when you are tired leads to questionable results&
[22:56] <snover> like thinking you know what is going on to break the grotto in GK2, but actually you read a test, setz, test, jz backwards again
[22:57] <snover> and then the music randomly stops playing for some reason
[22:57] <t0by> :)
[22:58] <t0by> I love you for what you are doing with GK2.
[22:58] --> Strangerke_ joined #scummvm.
[22:59] <t0by> Oh, damnit
[22:59] <t0by> No, nevermind
[22:59] <snover> yes, alwaysbody
[23:01] <-- Strangerke left irc: Ping timeout: 268 seconds
[23:01] Nick change: Strangerke_ -> Strangerke
[23:02] <t0by> This is weird.
[23:03] <snover> mlm

[23:05] <t0by> Simei: I enter if (!png.loadStream(*stream)) {
[23:05] <t0by> it's not at all a matter of too big or too small
[23:07] <t0by> Simei: as a side node you should use loadPng in *several* places, so I would not keep it backdrop specific, but disregard for the time being
[23:07] <t0by> *side note
[23:09] <t0by> Is openFileFromNum working as expected?
[23:10] <t0by> Simei: your commit d596702 is a mess of format-only changes btw.
[23:11] <t0by> (They are mostly in the right direction, which means that there was some stuff previously messed up)
[23:11] <-- SylvainTV left irc: Quit: User pushed the X - because it's Xtra, baby
[23:14] <t0by> uh
[23:20] <t0by> Simei: can you reproduce?
[23:22] <snover> hm. somehow this channel has a start time that is later than now&
[23:24] <dreammaster> We've gotta go back.. back to the future!
[23:24] <snover> i wish i had an rr recording of this
[23:25] <t0by> Simei: still there?
[23:27] <t0by> Oh lord
[23:27] <t0by> let me try something
[23:34] <t0by> Yes
[23:34] <t0by> Simei: I can't load at all.
[23:34] <t0by> Hypoteses:
[23:34] <t0by> 1. Seek is broken
[23:34] <t0by> 2. Weird file format but doubtful
[23:34] <t0by> 3. Some padding somewhere?
[23:35] <t0by> Simei: try to replicate that result and let me know if it's only me
[23:35] <t0by> then try to compare the seek offsets used by the original and your tree
[23:35] <t0by> consider a simple printf
[23:35] <t0by> (Anyway in general you should probably start using a debugger)
[23:36] <t0by> Simei: note that my build of WelcomePng has a different md5
[23:36] <t0by> (but it's good, I guarantee)
[23:36] <t0by> good night
[23:36] <t0by> _sev: ^
[23:40] <-- t0by left irc: Quit: t0by
[00:00] --- Mon May 29 2017