#exult@irc.freenode.net logs for 12 Mar 2010 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[00:16:35] <Dominus> how annoying /files/zip/libminizip.la was pulling in the oggvorbis files
[00:17:08] <Dominus> Libraries that this one depends upon.
[00:17:08] <Dominus> dependency_libs= bla bla bla...
[00:17:16] <Dominus> suuuuurrrre :)
[00:28:34] <-- Malignant has left IRC (Quit: ChatZilla 0.9.86 [Firefox])
[00:34:12] <Dominus> colourless before I go to bed, static building of your branch works fine (after some randomly hacking makefiles so they don't all include the oggvorbis stuff)
[00:34:37] <Dominus> no need for me making the frameworks, since I can just use the static libs I got
[00:36:25] <Dominus> good night all, I'm hurrying away before marzo got something for me to test and I'll be up another four hours :)
[00:36:29] <-- Dominus has left IRC (Quit: Leaving.)
[01:07:31] --> Malignant_Manor has joined #exult
[01:07:41] --- Malignant_Manor is now known as Malignant
[03:44:32] <-- Malignant has left IRC (Quit: ChatZilla 0.9.86 [Firefox])
[03:45:05] --> Malignant_Manor has joined #exult
[04:44:15] --> Marzo1 has joined #exult
[04:44:15] <-- Marzo has left IRC (Disconnected by services)
[04:44:18] --- Marzo1 is now known as Marzo
[04:47:36] --> julien- has joined #exult
[04:49:47] <Malignant_Manor> Is none of the new stuff supposed to be savable with ES?
[05:18:02] <Marzo> They should all be saveable with ES
[05:18:13] <Marzo> If they aren't, it is a bug
[05:32:45] <Malignant_Manor> They aren't saving for me. I tried the field, infinite reagent, and force usecode.
[05:33:49] <Malignant_Manor> I need to reboot for search to recognize the .cc extension.
[05:33:56] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
[05:38:33] --> Malignant_Manor has joined #exult
[05:43:11] <Marzo> I committed the fix
[05:44:07] <Marzo> You can try and see that not only you are not being able to save the new stuff, you are not being able to save any of the text-based shape data
[05:44:30] <Marzo> All due to a single letter that somehow got left when I copied the file over
[05:45:33] <Marzo> (you will probably have to manually delete shapevga.o and shapewrite.o if you don't want to make a clean build)
[05:47:38] <Marzo> Anyway, good night
[05:48:14] <Malignant_Manor> good night
[05:52:08] <-- Marzo has left IRC (Ping timeout: 258 seconds)
[06:31:33] <-- Rottingbeef has left IRC ()
[08:37:18] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
[09:57:04] --> Dominus has joined #exult
[09:57:04] --- ChanServ gives channel operator status to Dominus
[11:26:36] --> shazza has joined #exult
[11:59:09] <-- Dominus has left IRC (Quit: Leaving.)
[12:01:56] --> Dominus has joined #exult
[12:01:56] --- ChanServ gives channel operator status to Dominus
[12:54:32] <-- Colourless has left IRC (Quit: casts improved invisibility)
[12:57:54] <-- Kirben has left IRC ()
[13:53:08] <-- shazza has left IRC ()
[14:02:08] --> artaxerxes has joined #exult
[14:02:20] --- artaxerxes is now known as Artaxerxes
[14:02:35] <Artaxerxes> haven't been here in _YEARS_
[14:02:47] <Artaxerxes> how's everyone?
[14:04:05] <Artaxerxes> it's been so long, it seems I'm not even recognized by exultbot
[14:04:12] <Artaxerxes> exultbot: help
[14:04:22] <Artaxerxes> exultbot: list
[14:04:25] <Artaxerxes> exultbot: time
[14:05:22] <Dominus> hey ho Artaxerxes
[14:05:26] <Artaxerxes> ?seen wjp
[14:05:26] <exultbot> wjp is right here!
[14:05:30] <Dominus> ?seen Artaxerxes
[14:05:30] <exultbot> artaxerxes is right here!
[14:05:31] <Artaxerxes> ?list
[14:05:35] <Artaxerxes> ?help
[14:05:38] <Dominus> ?logs
[14:05:38] <exultbot> Logs are available at http://log.usecode.org/exultlog.php
[14:05:44] <Artaxerxes> that was it!
[14:06:10] <Dominus> and exultbot DOES recognize you, look at the logs :)
[14:06:17] <Artaxerxes> I've been incredibly busy with Translation
[14:06:50] <Artaxerxes> it remembered my colour!
[14:06:53] <Artaxerxes> cool
[14:07:02] <Dominus> I see you've been already at implementing features that were only talked about :)
[14:07:13] <Artaxerxes> ;-)
[14:07:25] <Artaxerxes> it gives the extra motivation to do it.
[14:07:57] <wjp> hi
[14:08:00] <Artaxerxes> Who knows, maybe I'll implement them myself in Exult if noone wants to try
[14:08:04] <Artaxerxes> hey wjp
[14:08:21] <Artaxerxes> although Dominus, you were talking about a feature freeze?
[14:08:43] <Dominus> yes, kind of, if you followed the ML, we are talking about releasing
[14:08:59] <Artaxerxes> I still get the ML, but it slipped my mind
[14:09:17] <Dominus> especially the kinds of features that might upset the normal game play
[14:10:00] <Dominus> but you *could* do it like Colourless and start a new branch
[14:10:16] <Artaxerxes> what's in Colourless' branch?
[14:10:22] <Dominus> though you should palaver with Marzo, he might have already some ideas
[14:10:35] <Dominus> Colourless throws out SDL_Mixer dependency
[14:11:09] <Artaxerxes> that's the library that gave me the most grief in Zaurus
[14:11:22] <Dominus> :)
[14:12:24] <Dominus> it gives me kind of grief when building a static universal binary for OS X. There is no way around me packaging the SDL + SDL_Mixer framework in the Exult.app
[14:12:34] <Dominus> which adds another 10MB overhead
[14:12:47] <Artaxerxes> btw, the Zaurus port cannot progress since Exult requires gcc 3 or 4 and the zaurus only have gcc 2.96.something
[14:13:37] <Dominus> yes, you mentioned it on the ML
[14:14:23] <Dominus> with heavy backporting, could it be made to compile?
[14:14:42] <Dominus> I'm not even sure how much interest there still is for the Zaurus, though
[14:15:05] <Artaxerxes> true enough
[14:16:16] <Dominus> Unfortunately we don't know how many times the port was downloaded since we only offer your packages on our webspace
[14:17:14] <Artaxerxes> the forum is a good indicator of how much demand there is for Zaurus, which is, pretty close to nil
[14:18:03] <Dominus> Since I'm going to the US and will most likely buy an ipad, I would be much more interested in an Exult port for that :)
[14:18:36] <Artaxerxes> I didn't know you were the Mac maintainer Dominus
[14:18:49] <Dominus> I'M NOT!!!
[14:18:52] <Dominus> by choice :)
[14:18:54] <wjp> (he is :-) )
[14:19:27] <Dominus> Last june I switched to Mac and since then cleaned up stuff with heavyhelp from wjp and Marzo
[14:19:59] <Dominus> the problem are my enormous coding skills, or I would be much more of help :)
[14:20:01] <Artaxerxes> is there still a dependence on Darwin?
[14:20:32] <Dominus> see, you are already asking me questions, I don't know what to make off :)
[14:20:47] <Dominus> BUT no, AFAICT
[14:21:18] <Dominus> with the gcc suplied by Apple you can build pretty much everything
[14:21:57] <Dominus> stuff just needs some patches and if you use MacPorts or Fink as an installer it's quite easy to build stuff
[14:22:33] <Dominus> theoretically with MacPorts (and Fink I guess) you can even switch to newer gcc versions than the one Apple supplies
[14:30:11] <-- Dominus has left IRC (Quit: Leaving.)
[14:30:18] --> Dominus has joined #exult
[14:30:18] --- ChanServ gives channel operator status to Dominus
[14:30:35] <Dominus> just testing my stupid… problem
[14:31:39] <Dominus> hmm, I thought this strange character problem I have when typing … could be resolved by switching codepage
[14:32:25] <-- Dominus has left IRC (Client Quit)
[14:32:32] --> Dominus has joined #exult
[14:32:32] --- ChanServ gives channel operator status to Dominus
[14:32:36] <Dominus> maybe... now
[14:32:54] <Dominus> yeah this might havehelped it...
[14:33:01] <Dominus> ...
[14:37:19] <Dominus> needed to switch encoding to ASCII instead of UTF-8 for exultbot to see my three dots as three dots :)
[14:54:43] --> Marzo has joined #exult
[14:57:32] <Dominus> Hi Marzo, now I get some warnings on my OS X 10.6 (gcc 4.2.1) on data_utils.h
[14:57:35] <Dominus> http://pastebin.com/H87tFTB5
[14:57:52] <Marzo> brb
[14:57:53] <Dominus> though only warnings
[15:01:01] <Marzo> Back
[15:03:53] <Dominus> and something else I jsut noticed when playing the SI intro scene, the rain seems to concentrate very much on the Avatar and when you run, it seems like you are outrunning the rain :)
[15:03:58] <Artaxerxes> Marzo: I have a test file if you are interested in trying the translation textmsg
[15:04:28] <Marzo> Send it in
[15:05:16] <Artaxerxes> will do. but I need to know something that wasn't too clear from the forum post:
[15:05:42] <Artaxerxes> the translation must contain ALL the strings in usecode or just the translated ones? I can do it either way very easily.
[15:06:34] <Artaxerxes> currently the textmsg contains all the strings
[15:07:17] <Marzo> The way I intend to do it, it can have only the translated ones
[15:07:39] <Marzo> (need to make a branch and write all the code; I already have all the working in my head)
[15:08:03] <Artaxerxes> can I play with your head ? ;-)
[15:08:08] <Marzo> Exult will look for the translation of a string and use it if found -- or default to original
[15:08:13] <Marzo> hehe
[15:08:17] <Dominus> Iron Maiden: Can I play with Madness?
[15:08:58] <Artaxerxes> the way I've created the textmsg, it takes the translated string if done otherwise it uses the original one
[15:09:12] <Dominus> http://www.youtube.com/watch?v=xSZbbTjM0Es (<- the iron maiden video)
[15:09:12] <Artaxerxes> but it's a walk in the park for me to change it to your requirements
[15:10:02] <Artaxerxes> done
[15:10:07] <Marzo> I will fix the way that textmsg.txt/text.flx/exultmsg.txt are done post release
[15:10:12] <Artaxerxes> where can I push this file?
[15:10:29] <Dominus> pastebin?
[15:10:38] <Marzo> You can probably use your client to sent it to me
[15:10:44] <Artaxerxes> it's copyrighted material, no?
[15:10:45] <Marzo> (or pastebin, if it is text)
[15:11:23] <Marzo> Instead of requiring a full patch as Exult does now, Exult will allow a partial patch like it does for nearly every other text data
[15:12:02] <Marzo> So that you will have to output only the translated text
[15:14:14] <Artaxerxes> how do I transfer files with irssi?
[15:15:35] <Artaxerxes> maybe I should start small...
[15:15:46] <Artaxerxes> let me copy/paste here (4 lines)
[15:15:50] <Artaxerxes> %%section usecode
[15:15:50] <Artaxerxes> 0403,0093:'"Quelle traversée nous venons de faire,
[15:15:50] <Artaxerxes> 0403,00AF:! Quand nous sommes passés entre les piliers de serpent, j'aurais juré que nous volions!"
[15:15:52] <Marzo> At least in my client, I can right-click on your name and select 'send file'
[15:15:53] <Artaxerxes> %endsection
[15:16:14] <Artaxerxes> I'm in remote session through SSH to my home machine from work... ;-)
[15:16:21] <Marzo> I think that a better output would be along the lines of:
[15:16:38] <Artaxerxes> (accents failed to pass btw)
[15:16:39] <Marzo> %%section 0x0403
[15:16:56] <Marzo> 0x0093:'"Quelle traversée nous venons de faire,
[15:17:28] <Artaxerxes> ohhh, I misunderstood your post in the forum then.
[15:17:35] <Marzo> Or, for mods, it could also be
[15:17:40] <Marzo> %%section functionname
[15:17:40] <Artaxerxes> no pb, that's easy to fix
[15:18:04] <Marzo> This has the advantage of grouping all of the text for any function into a single heading
[15:18:22] <Marzo> Exult can easily read the whole section when loading the function
[15:18:37] <Marzo> And set up the translation map
[15:18:37] <Artaxerxes> so to sum up, the format is:
[15:18:46] <Artaxerxes> %%section functionname
[15:18:55] <Artaxerxes> offset:string
[15:18:57] <Marzo> (or number)
[15:18:59] <Artaxerxes> %%endsection
[15:19:01] <Marzo> Aye
[15:19:30] <Artaxerxes> or number?
[15:20:00] <Marzo> functionname or functionnumber
[15:20:05] <Marzo> in %%section
[15:20:21] <Marzo> For the original games, it will have to be the function number
[15:20:28] <Artaxerxes> ahhh got it
[15:20:40] <Artaxerxes> I have almost no exposure to mods
[15:21:21] <Marzo> Considering that there are only four of them that actually released anything, it is no surprise
[15:22:01] <Marzo> UCC generated usecode has a symbol table with function names
[15:23:02] <Marzo> Since usecode function numbers in mods are dynamically generated on compilation, using function numbers would not work (except for some shape and NPC functions)
[15:23:58] <Artaxerxes> here is a new sample based on your specs:
[15:24:03] <Artaxerxes> %%section 0x0403
[15:24:03] <Artaxerxes> 0x0093:'"Quelle traversée nous venons de faire,
[15:24:03] <Artaxerxes> 0x00AF:! Quand nous sommes passés entre les piliers de serpent, j'aurais juré que nous volions!"
[15:24:06] <Artaxerxes> %endsection
[15:24:19] <Artaxerxes> this lack of accents in irssi is annoying
[15:24:57] <Dominus> I do see them, though
[15:25:03] <Artaxerxes> really?
[15:25:06] <Marzo> So do I
[15:25:11] <Artaxerxes> here they are replaced with ?
[15:25:28] <Dominus> exultbot: doesn't like them though
[15:26:05] <Marzo> Does Tisane output any sort of quoting around the strings or is that single quote in 0x0093 a stray?
[15:26:16] <Artaxerxes> reminds me of http://www.bash.org/?244321
[15:26:19] <Marzo> (by the way, the numbers do not need to be in hex)
[15:27:00] <Artaxerxes> My mistake for the quote
[15:27:17] <Artaxerxes> the single quote shouldn't be there
[15:27:52] <Artaxerxes> I actually record the funcnumber and offsets in hex in my db
[15:28:19] <Marzo> Well, it makes no difference either way
[15:28:22] <Marzo> :-)
[15:28:27] <Artaxerxes> the beauty with your solution Marzo, is that they will never be an issue with 32 bit usecode
[15:28:40] <Marzo> I would not be so sure
[15:28:53] <Marzo> A while back, a guy was making a mod
[15:29:27] <Marzo> In a function, he managed to crash UCC -- it needed 32-bit header extension due to the *code*
[15:29:38] <Artaxerxes> right
[15:29:59] <Artaxerxes> I wonder if there will be some sort of conflict between mods and translators somewhere down the line
[15:30:07] <Marzo> I told him about loops, which helped shrink it down, then fixed ucc to handle it correctly
[15:30:45] <Marzo> The translation of mods can be done in one of two ways, as far as I know
[15:31:09] <Marzo> One is the way the U7 itailan team did it -- directly translating the usecode file
[15:31:23] <Marzo> The other would be translating the usecode source files and recompiling
[15:31:52] <Artaxerxes> would there be a way to use this newfangled translation textmsg feature for mods?
[15:32:02] <Marzo> When this new translation interface becomes finished, something like Tisane could do it too
[15:32:19] <Marzo> It shouldn't be hard
[15:32:35] <Marzo> I just have to fix the way Exult handles textmsgs first
[15:33:05] <Marzo> Right now, it is an all-or-nothing feature
[15:33:47] <Marzo> I plan to make Exult load from Exult_xx.flx, then patch/load from 'static', patch from 'patch' then patch from 'patch/translate'
[15:34:24] <Marzo> (or somesuch; I haven't yet decided where the translations are going)
[15:35:10] <Marzo> One thing I know is that you should write a file called 'usecode.txt' in the format described above
[15:36:33] <Artaxerxes> I was wondering about the filename
[15:36:50] <Artaxerxes> used to be usecode.msg. Now it's usecode.txt
[15:36:53] <Artaxerxes> thx!
[15:37:56] <Artaxerxes> would it make sense to put the translated text.flx in there as well?
[15:39:27] <Marzo> The text.flx should go into a file called textmsg.txt
[15:41:19] <Artaxerxes> that could be an improvement I could make, as I currently ignore text.flx altogether
[15:41:48] <Artaxerxes> maybe the export could create both usecode.txt and textmsg.txt and zip them for delivery or something like that
[15:41:50] <Marzo> This file has 3 sections: '%%section shapes' for shape names
[15:42:02] <Marzo> (the first 0x400 messages in text.flx)
[15:42:52] <Marzo> '%%section msgs' start from 0x400 and end before 0x500
[15:42:59] <Artaxerxes> odd. I though the first 149 were useless chunk shape descriptions
[15:43:26] <Marzo> And '%%section miscnames' are from then on
[15:43:37] <Artaxerxes> from http://si-french.sourceforge.net/HOWTO.php, I had:
[15:43:42] <Artaxerxes> * 0-149: chunk shapes description. Not really important to translate, since it won't show anywhere.
[15:43:46] <Artaxerxes> * 150-1667: items and phrases. You want to translate all of them.
[15:43:48] <Artaxerxes> * 1670-2571: cheat screen words. No need to translate that since it won't be used by Exult (Exult provides its own cheat screen).
[15:44:04] <Artaxerxes> (relevant to Silver Seed)
[15:44:37] <Marzo> Oh, I am not saying that the first 150 messages are useful
[15:44:44] <Marzo> They are in the section, that is all
[15:45:29] <Artaxerxes> ok
[15:45:31] <Marzo> The messages in the 0-1023 range should go into the shapes section of textmsg.txt
[15:45:35] <Marzo> That is all
[15:45:47] <Artaxerxes> will do
[15:45:52] <Marzo> The file splits into sections because Exult allows shapes >= 1024
[15:47:27] <Artaxerxes> brb
[15:47:36] <Marzo> Messages from 1024-0x500 must have 0x400 subtracted and put into the msgs section
[15:48:45] <Marzo> The 'miscnames' are for things such as frame names, and start at 0x500 and should have 0x500 subtracted from their indices
[15:49:12] <Marzo> If you want your head to explode, you can take into consideration the differences between SI and SS in Tisane, like Exult does
[15:49:28] --> matt_o has joined #exult
[15:50:11] <Artaxerxes> back
[15:50:23] <matt_o> welcome back :)
[15:51:27] <Marzo> Dominus: by the way, I finished adding 16-bit and 32-bit versions of the 8-but scalers
[15:51:37] <Artaxerxes> well, I would create the textmsg.txt automatically from parsing the text.flx, so it shouldn't matter, should it?
[15:51:39] <Dominus> yay
[15:51:54] <Marzo> That is when it would matter, yes
[15:51:57] <Artaxerxes> matt_o: thx!
[15:51:59] <Marzo> :-)
[15:52:10] <Dominus> Marzo: should I test it first or will you put it directly into code?
[15:52:27] <matt_o> I think I was last here maybe 2 or 3 years ago. It's good to see the banner still flying atop Lord British's castle :)
[15:52:32] <Artaxerxes> you mean the textmsg.txt will have different ranges based on whether it's SS or SI?
[15:52:41] <Artaxerxes> matt_o: same for me
[15:52:49] <Marzo> When Exult reads text.flx (or exultmsg.txt, for that matter) it remaps some text strings from original SI into their SS names
[15:52:53] <Artaxerxes> matt_o: just joined today after multiple years of silence
[15:53:05] <Dominus> welcome matt_o, the cause of my documentation job!!!
[15:53:07] <Marzo> Er... SI *indexes*, not names
[15:53:08] <matt_o> I just fired up x-chat and it remembered all my old channels :)
[15:53:12] <matt_o> Dominus, me? :o
[15:53:21] <Dominus> you!!!
[15:53:25] <Marzo> But Exult performa no such translation of indexes when reading from textmsgs.txt
[15:53:25] <matt_o> what did I do?
[15:53:38] <Artaxerxes> Marzo: I'm confused
[15:53:57] <Dominus> matt_o: http://exult.sourceforge.net/faq.php#thanks
[15:54:01] <Marzo> To put it simply: textmsgs.txt uses the same indices whether it is SI or SS
[15:54:08] <matt_o> lol!
[15:54:20] <matt_o> I'm honored.. and guilt-ridden.. all at the same time
[15:54:42] <Marzo> text.flx and exultmsg.txt don't -- some items at the end have different locations in text.flx in SI versus SS
[15:54:56] <Dominus> matt_o: on the very old jackchaos forum, you had the old opensource answer for me when I wrote something like "all these questions turn up again and again, there should be a FAQ!"
[15:55:10] <Marzo> It only starts at the miscnames portion: 0x500 and up
[15:55:20] <Dominus> your answer: then do it yourself
[15:55:33] <Artaxerxes> Dominus: and the rest is history
[15:56:41] <Dominus> since then I've become the bane of the forum!!! (even though I try hard to be less grumpy, I really do)
[15:57:17] <Artaxerxes> Marzo: I think I understand what you mean.
[15:57:33] <Marzo> I am writing a text document for simplicity
[15:57:52] <Artaxerxes> once extracted from text.flx, the indices must be modified
[15:58:13] <Marzo> Yes
[15:58:27] <Artaxerxes> that's a shame
[15:59:11] <Artaxerxes> what's exultmsg.txt anyways?
[15:59:56] <Artaxerxes> http://exult.svn.sourceforge.net/viewvc/exult/exult/trunk/data/exultmsg.txt?revision=5178&view=markup
[16:00:53] <Artaxerxes> what in there needs to be changed?
[16:02:51] <Artaxerxes> Marzo, since the user doing the translation provides the text.flx, should I care about those indices?
[16:03:24] <Artaxerxes> After all, if he's translating SS, it will use SS's indices. If he translated SI, he will use SI's indices
[16:03:33] <Marzo> Artaxerxes: if you want things to display correctly, yes
[16:04:09] <matt_o> Dominus, hehehehe... jackchaos! wow that brings back memories!
[16:04:11] <Marzo> The problem is in using SI indices -- Exult deals internally with the SS indices for simplicity in the coding
[16:04:39] <Marzo> The textmsg.txt file is read as if it contained the SS indices already
[16:07:40] <Artaxerxes> a successful translation would include: conversations (usecode), descriptions (text.flx), graphics (spellbook and others, shapes.vga) and intro/conclusion (exultmsg.txt). Am I missing anything?
[16:08:12] <Marzo> None that I can think of
[16:08:31] <Artaxerxes> I am done with usecode, I can give a shot at text.flx and exultmsg.txt, but for graphics, no web interface is gonna cut it.
[16:08:42] <Marzo> shapes.vga probably doesn't have anything in it to translate, though, but gumps.vga does
[16:08:54] <Artaxerxes> yes, sorry.
[16:09:03] <Artaxerxes> although, no, it does
[16:09:09] <Artaxerxes> alphabet
[16:09:21] <Marzo> Artaxerxes: here it is: http://pastebin.com/XHqvvp3y
[16:09:29] <Artaxerxes> it needs to be extended if the translator's language contains accents
[16:09:32] <Marzo> The alphabet is in fonts.vga
[16:09:36] <Marzo> :-)
[16:09:40] <Artaxerxes> of course. I'm getting old
[16:11:22] <Marzo> In SI, some shapes in sprites.vga may also need translation
[16:13:14] <Marzo> If you are going to implement the index remapping for original SI, maybe you could ask which game is the text.flx from when the user is importing
[16:13:21] <Marzo> (in Tisane, I mean)
[16:13:46] <Artaxerxes> unless Tisane creates a new text.flx?
[16:14:15] <Artaxerxes> then Exult would look to see if it's SI or SS and apply the relevant changes?
[16:14:15] <Marzo> Or that
[16:14:33] <Marzo> This would work for the original games
[16:14:44] <Artaxerxes> suppose you have a text.flx in the patch dir, would Exult use that one instead?
[16:14:53] <Artaxerxes> (sorry been a long time since I played with conf paths)
[16:14:58] <Marzo> When I fix the handling of text, yes
[16:15:20] <Marzo> Right now, it wouldn't work for mods
[16:15:56] <Artaxerxes> so, the way I see it, there are two options:
[16:15:58] <Marzo> Mods supply a textmsg.txt written by ES which makes Exult ignore everything else
[16:16:23] <Marzo> (which is why I need to fix it...)
[16:16:30] <Artaxerxes> either I parse the text.flx provided by the translator and offer an interface to translate ALL its strings, then export a newly crafter text.flx
[16:17:32] <Artaxerxes> or I parse the text.flx provided by the translator and determine which text.flx it is and generate a textmsg.txt file in the %%section format above, catering to the version of text.flx so that Exult can load it properly
[16:17:47] <Artaxerxes> (after it's been translated)
[16:18:02] <Marzo> As I said: for now, yes
[16:18:26] <Artaxerxes> cool
[16:18:32] <Artaxerxes> I'm fine with that
[16:19:13] <Marzo> After the release, all of this will change; I am holding off since there is the potential to break things in subtle ways
[16:19:23] <Artaxerxes> next step for me would be to parse the translator's config file to automatically fetch the usecode and text.flx instead of requiring it to be in the home dir
[16:19:45] <Artaxerxes> although, that might not work since Tisane requires a webserver
[16:20:30] <Artaxerxes> anyways, I have to return to work
[16:20:44] <Artaxerxes> nice chatting with all of you. It's been too long
[16:20:59] <Marzo> Hehe, welcome back :-)
[16:21:02] <Artaxerxes> and thx so much for all your help
[16:21:20] <Marzo> Np
[16:21:43] <Artaxerxes> I will finish up the help page and publish it under si-french.sf.net (which needs MAJOR cleanup)l
[16:21:45] <-- Artaxerxes has left IRC (Quit: leaving)
[17:05:15] --> Marzo1 has joined #exult
[17:05:15] <-- Marzo has left IRC (Disconnected by services)
[17:05:21] --- Marzo1 is now known as Marzo
[19:50:57] --> Malignant_Manor has joined #exult
[20:38:20] <Malignant_Manor> Looking at the logs, there's been more that 4 with releases. Keyring, SI Fixes, TFL, minor U6 coversion, Alun's interactions (if I remember the name somewhat right), and SI BG Merge. There's also the beginnings of a new game, dev_game.
[20:39:07] <Malignant_Manor> Of course, Alun's lives on pretty much only in your mods.
[21:01:54] <-- Malignant_Manor has left IRC (Remote host closed the connection)
[21:03:37] --> Malignant_Manor has joined #exult
[21:11:18] <Marzo> Now, that is optimization: Scale2x will be a lot faster on Exult in 16-bit and 32-bit than it would have been if I had gone with the naïve approach I used at first
[21:12:18] <Marzo> Under extreme load, the naïve Scale2x approach resulted in ~80% usage in one core, but with the optimized version, it uses only ~50%
[21:28:18] <Malignant_Manor> What resolution is that though? The current Scale2x doesn't use all that much cpu even compared to point.
[21:31:10] <Marzo> 320x200; as I said, under *extreme* load
[21:31:29] <Marzo> That means doing things so it is forced to redraw everything all the time
[21:34:19] * Dominus is looking forward to this
[21:44:56] <Malignant_Manor> Marzo, do you even have Windows anymore?
[21:45:38] <Marzo> I still have my Windows laptop, use it for a lot of things
[21:46:18] <Marzo> I don't do much dev work with it because it is at least 4-8 times slower for compiling things
[21:49:30] <Malignant_Manor> I was just wondering about Windows specific issues.
[21:50:39] <Malignant_Manor> The current Mac development condition isn't nearly that good.
[21:53:15] <Marzo> I just committed the 16-bit and 32-bit scalers to SVN
[21:53:15] <Dominus> huh?
[21:54:54] <Malignant_Manor> Dominus finds problem with Mac->Marzo finds something that may work->Dominus reports whether it works or not
[21:58:41] <Dominus> ah you mean that the Windows development condition is not that good for Windows as it is for Mac
[21:59:39] <Malignant_Manor> Windows development is better since Marzo has access to Windows.
[22:00:21] <Malignant_Manor> He can actually test things.
[22:05:12] <Malignant_Manor> If I remember correctly, interlaced actually looks nicer now.
[22:06:35] <Marzo> Really? I could not see any differences
[22:06:55] <Marzo> (comparing them side-by-side, that is)
[22:10:28] <Malignant_Manor> I'm not sure as I never really used it.
[22:10:36] <Malignant_Manor> It doesn't look nice.
[22:10:45] <Malignant_Manor> It looks muddied.
[22:11:19] <Malignant_Manor> Maybe I'm thinking of when I forced it to 1x
[22:11:48] <Malignant_Manor> Did you change the fullscreen check?
[22:12:03] <Marzo> No, not yet
[22:13:09] <Marzo> Dominus: if you want to try out Colourless' branch with all the updates from the trunk revision, you can try using this: http://www.orcaware.com/svn/wiki/Svnmerge.py
[22:13:21] <Marzo> (you need to have Python installed)
[22:13:47] <Malignant_Manor> Maybe this driver is different than the other I had installed. It doesn't list bellow 640x400 but allows fullscreen toggle.
[22:13:50] <Marzo> You can expect 4 conflicts
[22:14:13] <Malignant_Manor> The check was using the unscaled size.
[22:15:13] <Marzo> (run this and wait: svnmerge.py init; svnmerge.py merge -F;)
[22:15:30] <Marzo> (that should have been ; ) not a smiley
[22:16:53] <Dominus> YES!!!!
[22:17:02] <Dominus> works FINE now in fullscreen!!!!
[22:17:16] <Dominus> did I put enough ! ?
[22:17:17] <Marzo> To be honest, the interlaced algorithm can probably go away; to work correctly, it should alternate which rows are drawn and which ones aren't, and draw 60 times per second
[22:17:20] <Dominus> !!!!!!!!
[22:19:23] --> Cahaan has joined #exult
[22:20:21] <Malignant_Manor> Would that actually provide a speed increase?
[22:21:13] <Marzo> No
[22:21:25] <Marzo> It would drastically reduce it
[22:21:42] <Marzo> But the image would not be so dark
[22:23:21] <Malignant_Manor> Again, is the interlaced scaler really worth it?
[22:24:08] <Dominus> as for the svnmerge.py, I think I'm going to skip it for now. a modified exult.cfg plays nicely with the no-sdl_mixer branch and I'm likely to fiddle with something that I shouldn't right now :)
[22:25:24] <Dominus> don't think the interlaced scaler is something most people need. but who knows...
[22:26:16] <Dominus> building with X11 support and will try whether it still crashes with the 8bit scalers
[22:26:27] <Dominus> aehm, the former 8bit only scalers
[22:36:45] <Dominus> yeah!!! again!!!
[22:36:49] <Dominus> works fine now
[22:38:56] <Dominus> Malignant_Manor: the way development works for Mac now is waaaayyyyyyy better than it worked before, when only Fingolfin was responsible and Mac people had no "champion for their cause" :)
[22:40:12] <Dominus> this way, me reporting and Marzo fixing, me checking feels just like it usually did on Windows, too
[22:40:15] <Malignant_Manor> Someone should donate OSX to Marzo and clone him 2 times.
[22:40:54] <Marzo> Dominus: I doubt you will be able to test the 8-bit scalers
[22:41:11] <Dominus> yeah :)
[22:41:40] <Malignant_Manor> "IN QUAS XEN"
[22:48:57] <Malignant_Manor> All 8 bit scaling still works if I force it.
[22:49:04] <-- Cahaan has left IRC ()
[22:50:05] <Marzo> Which is why I said "Dominus: I doubt you will be able to test the 8-bit scalers" :-)
[22:50:28] <Dominus> I see no need to force that :)
[22:54:41] <Marzo> (currently, the 8-bit scalers are a hold-over -- there is probably some portable device out there that can play Exult but needs 8-bit -- which probably won't be used by anyone)
[22:56:23] <Dominus> I can see that this might be a good reason for keeping them for that reason
[22:57:28] <Marzo> Which is why I kept them :-)
[22:58:01] <Dominus> so need to scratch that 8bit scaler issue from the FAQ, only the fullscreen one remains :)
[22:58:28] <Malignant_Manor> With this patch, the forcing bpp doesn't even need to be added to the cfg automatically.
[22:58:43] --> Kirben has joined #exult
[22:58:43] --- ChanServ gives channel operator status to Kirben
[22:59:09] <Dominus> yes, much better to have good defaults not have people edit their cfg file
[22:59:42] <Malignant_Manor> Well, the problem was using 8 bits reported that only 8 bits of color could be displayed.
[23:00:48] <Malignant_Manor> That is an 8 bit scaler in fullscreen to fullscreen 16/32 scaler.
[23:02:00] <Malignant_Manor> It might be good to have it addable for the very slight chance it might not work on some setups.
[23:02:52] <Dominus> as I understood if it doesn't work it will use 8bit again
[23:03:26] <Dominus> OpenGL doesn't display in fullscreen but I don't care enough :)
[23:03:50] <Malignant_Manor> It no longer works on your machine?
[23:03:58] <Marzo> OpenGL is still too buggy all around
[23:04:48] <Dominus> yeah OpenGL worked a couple of revisions ago in fullscreen somewhat I think but stopped working on one of the fixes for it
[23:04:49] <Malignant_Manor> You helped one issue with the using more than 8 bits by default.
[23:05:09] <Dominus> but even when it was working it had issues
[23:05:29] <Dominus> Good call to declare it not ready for release :)
[23:07:25] <Malignant_Manor> Change imagewin line 243 to
[23:07:26] <Malignant_Manor> int hwdepth = 32;
[23:16:23] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
[23:16:46] <Dominus> I was just about to report that this did make openglscaler worse :)
[23:21:39] <Dominus> actually not worse just didn't change it at all. does opengl scaler for you (whoever reads that and feels like testing it) also crap out on the journey onward, red moongate screen?
[23:25:22] <Marzo> Dominus: out of random curiosity, what happens if you set hwdepth to 24?
[23:25:59] <Dominus> i can tell you exactly, I need to do a massive recompile :)
[23:26:45] <Dominus> (actually no this time, oh well :)
[23:28:06] <Dominus> something is borgged
[23:28:43] <Dominus> on switching to fullscreen the screen goes white for a second or so (doesn't flash, it stays a bit white)
[23:28:47] <Dominus> then black
[23:29:31] <Dominus> and when I "walk" around the screen is neatly spit in four squares which change colour (the whole square)
[23:30:08] <Dominus> the colors are all in the brownish part of the spectrum
[23:30:41] <Dominus> when I switch back to windowed mode (I've set the res to 640x480 to be able to switch)
[23:30:48] <Dominus> the screen stays black
[23:33:06] <Dominus> and starting a game does not work at all (in windowed), when I click journey onward game window gets black and after a second I get a spinning mouse cursor and the CPU usage gets higher and higher
[23:34:38] <Dominus> oh, found a bug in the Audio Gump
[23:35:52] <Dominus> When you disable all Audio in the gump, all buttons disappear but the "digital music" enable/disable button (and of course the Audio enable/disable button)
[23:37:09] <Dominus> and when you close the gump and open it again that extra button is gone :)
[23:43:01] <Dominus> marzo, scale2x crashes exult with a segmentation fault on 640x480 in fullscreen
[23:43:52] <Marzo> Dominus: does it also crash without fullscreen? Also, is that 640x480x2 or 320x240x2?
[23:43:57] <Marzo> Also: fix committed for the audio gump issue
[23:44:19] <Dominus> 640x480x2
[23:44:55] <Dominus> and starts to crash when going to 512x384
[23:45:12] <Dominus> in fullscreen. 320x200, 400x300 do work
[23:45:28] <Dominus> I'll test windowed now
[23:45:42] <Marzo> In windowed mode, it doesn't seem to crash
[23:45:51] --> Malignant_Manor has joined #exult
[23:46:20] <Dominus> crashes in windowed for me, too (512x384)
[23:47:06] <Dominus> if that helps:
[23:47:07] <Dominus> 0 exult 0x00000001000b9030 void Scale2x_noblur<unsigned char, unsigned int, Manip8to32>(unsigned char*, int, int, int, int, int, int, unsigned int*, int, Manip8to32 const&) + 1296
[23:47:08] <Dominus> 1 exult 0x00000001000b8598 Image_window::show_scaled8to32_2x_noblur(int, int, int, int) + 152
[23:47:08] <Dominus> 2 exult 0x000000010005ba63 Game_window::clear_screen(bool) + 67
[23:47:08] <Dominus> 3 exult 0x0000000100040950 ExultMenu::run() + 48
[23:47:08] <Dominus> 4 exult 0x000000010003b6d2 Init() + 2146
[23:47:19] <Dominus> part of the error report
[23:47:32] <Dominus> I can do a debug built if you want
[23:47:37] <Marzo> Please do
[23:47:53] <Dominus> heavy?
[23:50:25] <Marzo> Aye
[23:52:43] <Dominus> gdb exult, run, bt
[23:52:46] <Dominus> right?
[23:53:32] <Dominus> http://pastebin.com/cDDavi5Z
[23:53:36] <Malignant_Manor> OpenGL crashes for me if too large when journey onward.
[23:53:40] <Dominus> if the answer is yes
[23:53:51] <Malignant_Manor> but my hardware should support it.
[23:55:16] <Marzo> Does point or (say) hq2x crash at that size?
[23:55:29] <Marzo> And what is the maximum resolution your monitor can handle?
[23:55:49] <Dominus> point did not crash, I had that res set and ran through the scalers
[23:56:08] <Dominus> 1920x1200
[23:57:11] --> Colourless has joined #exult
[23:57:11] --- ChanServ gives channel operator status to Colourless
[23:57:50] <Dominus> hq2x does not crash at that size nor at 800x600
[23:57:58] <Dominus> neither in window or fullscreen
[23:58:50] <Malignant_Manor> Mine supports 1280x1024, hq2x didn't crash at 640x480 x2
[23:59:13] <Malignant_Manor> OpenGL does at journey onward, I can switch it in game though.
[23:59:26] <Dominus> point also has no problem