#exult@irc.freenode.net logs for 22 Dec 2016 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[00:01:55] --> KnightCaptain has joined #exult
[00:34:35] <-- KnightCaptain has left IRC (Ping timeout: 250 seconds)
[02:35:29] <-- Dominus has left IRC (Ping timeout: 250 seconds)
[02:36:31] --> Dominus has joined #exult
[02:36:31] <-- Dominus has left IRC (Changing host)
[02:36:31] --> Dominus has joined #exult
[02:36:31] --- ChanServ gives channel operator status to Dominus
[03:12:38] --> KnightCaptain has joined #exult
[03:23:58] <KnightCaptain> Marzo, I'm working on a sed script to make understanding the output of UCXT easier to read for SI.
[03:46:58] <KnightCaptain> Dominus, I have a Mac @ work I can use. I'd love to see the build in action.
[04:38:10] --> azeem_ has joined #exult
[04:42:01] <-- azeem has left IRC (Ping timeout: 250 seconds)
[07:29:06] <-- Rottingbeef_ has left IRC ()
[08:25:37] <-- Lightkey has left IRC (Ping timeout: 245 seconds)
[08:33:13] <KnightCaptain> The sed script conversion for SI has been emailed. I found a lot of cool stuff in doing this.
[08:33:48] <KnightCaptain> It partially converts conversation into something closer to Usecode C so converting code from that will be easier.
[08:34:10] <KnightCaptain> I also identified a lot of functions, even if they aren't immediately useful.
[08:34:20] <KnightCaptain> Hopefully this will be useful to others...
[08:34:22] <KnightCaptain> Cheers!
[08:38:19] --> Lightkey has joined #exult
[08:50:10] <Dominus> KnightCaptain sounds very useful
[08:51:07] <Dominus> I'll have an OS X build as soon as someone else can watch this sick child next to me...
[09:00:45] <KnightCaptain> How'd you do it in the video?
[09:08:01] <Dominus> I changed the code but my playcopy of the source is a mess ;)
[09:08:22] <KnightCaptain> I really appreciate it!
[09:08:39] <KnightCaptain> If it is easier a video of Fawn with fog would rock.
[09:09:33] <Dominus> I can do either. I guess a video would be easier for you ;)
[09:09:57] <Dominus> With clouds or without?
[09:10:26] <KnightCaptain> At the risk of being a petulant poster from Korea - Both!
[09:12:16] <KnightCaptain> The sed script takes 90 seconds on my laptop. It's hundreds of commands. :D
[09:16:58] <Dominus> Ok, I'll see about making the videos later
[09:47:17] <Marzo> I have an old VB program that does something similar
[09:47:31] <Marzo> The issue is that some functions have odd structure
[09:49:35] <Marzo> For example: in FoV, Lord British's function is irreducible (in graph theoretical sense), and so it can't be represented in a nice way; and there is at least one more function I know of which has an odd conversation structure which defeated my old program
[09:49:47] <Marzo> There are probably others, these are the ones I know about
[09:50:25] <Marzo> And this is unfortunately a problem which even sed can't fix: you need an actual grammar for it
[09:55:26] <Marzo> And since making a program that understands the structure of the code at a deep level is exactly what a disassembler does anyway, it is why I intend to rewrite UCXT to give proper output right away instead of making another program that does much of the same just to improve readability
[11:04:19] <KnightCaptain> Are my findings thus far of any use to that project?
[11:13:16] <Dominus> video coming up in a few seconds....
[11:14:36] <Dominus> https://youtu.be/3xV0wEIu47o
[11:15:01] <Dominus> it has a few clouds, mostly clear so I left it like this...
[11:15:16] <Dominus> real fog weather would need much more work IMO
[11:16:07] <Dominus> -quick palette change when stepping inside (don't know if that would trigger under other "roof" areas like an overpass in fawn)
[11:16:21] <Dominus> - slow clouds
[11:16:51] <Dominus> slow clouds vanishing with the palette at the end of fog weather
[11:32:02] <KnightCaptain> That is amazing!!!
[11:35:57] <KnightCaptain> To expand on my idea a bit, I am going to have the fog be noxious (smelling, at least) in Fawn. Clearing it in different sections will require beating specific monsters made by Mad Iolo. Slaying each monster will set a flag, and that will be used to turn off the fog effect on that part of town.
[11:36:25] <KnightCaptain> So I will be calling this palette by Usecode eggs rather than straight weather eggs.
[11:37:07] <KnightCaptain> If I can code it, I want to put those flytrap monsters from the Beta in the game and on DEATH they would transform back into potted plant shapes.
[11:38:27] <KnightCaptain> I've got a very primitive function process setup already that takes away the killEveryone function and replaces it with something that will enable the Banes' takeover.
[11:39:36] <KnightCaptain> Coming up with a worthy plot and the details will take more work, of course. :|
[12:04:25] <Marzo> KnightCaptain: this seems less like a case for fog and more like a case for a screen-sized translucent sprite repeatedly created over the screen
[12:05:06] <Marzo> Because the fog effect fades in and out slowly
[12:09:23] <Marzo> And more: you clear the fog in one part of town; move to one with fog; return while the fog is starting to the part you just cleared; fog will be there
[12:09:39] <Marzo> And using fog weather will conflict with other weather effects
[12:10:20] <Marzo> So: you go to a foggy part of town; fog starts and covers the screen; you activate a weather egg for rain; result: rain starts, fog goes away
[14:07:22] <Dominus> this SDL2 scale 1x bug is driving me nuts. Everything is fine when bpp is forced thus bypasses "if (!paletted_surface && !force_bpp)" in imagewin
[14:07:40] <Dominus> something in that code's ifdef SDL2 is messing up...
[16:03:01] --> Rottingbeef has joined #exult
[17:38:31] <KnightCaptain> I don't need other weather effects. The other weather eggs in Fawn could be removed via Usecode (like the Fawn Tower cleanup) and then re-placed via some 'victory over Mad Iolo' usecode.
[17:39:14] <KnightCaptain> The problem with doing a full-screen sprite is that even the existing ones don't fully cover my resolution. :(
[17:47:57] <-- KnightCaptain has left IRC (Quit: Nettalk6 - www.ntalk.de)
[20:01:34] <-- Rottingbeef has left IRC ()
[20:04:51] --> Rottingbeef has joined #exult
[20:55:50] <Dominus> help me Colourless, you are my only hope!
[20:55:52] <Dominus> :)
[22:20:01] <wjp> What's broken?
[22:27:48] <Dominus> Heya wjp
[22:27:54] <Dominus> https://sourceforge.net/p/exult/bugs/1923/
[22:28:47] <Dominus> SDL2 game size set to auto, scaler 1 will distort our graphics immensely
[22:29:47] <Dominus> Unless you force bpp in exult.cfg to either 8,16 or 32
[22:30:09] <Dominus> Because then it skips the troublesome code at https://github.com/exult/exult/blob/fb407b7c66f6dfaeafc4bfd74bc2cfd08285c86a/imagewin/imagewin.cc#L566
[22:31:15] <Dominus> So something is wrong or missing in the ifdef SDL2 code section
[22:50:33] <-- Marzo has left IRC (Ping timeout: 250 seconds)
[23:13:01] <wjp> I see
[23:26:51] <wjp> yeah, that's a bit broken
[23:27:15] <wjp> I'm a bit confused by the point of this code path
[23:35:11] <wjp> ok, this code path was just completely broken by SDL2
[23:36:16] <wjp> the SDL1 version asks for a 8bpp surface here, but the SDL2 code asks for whatever bpp the desktop is using
[23:36:30] <wjp> (but then the rest of exult still thinks it got an 8bpp surface)
[23:41:08] <wjp> don't know SDL2 well enough to immediately see how to fix it
[23:43:00] <wjp> maybe we should just pass this case through the scaler system anyway so that we use the point scaler as a 8bpp->32bpp converter