03:34:10 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 08:02:25 <04dracoomega> So, I was a bit confused when trying to add Mystic to the character creation screen that the category/category_priority fields in the job yaml didn't seem sufficient. And you have to manually edit a structure in newgame.cc anyway for anything to show up. But after a bit more investigation, it looks like the job groupings that are automatically generated from job yaml are never used. And have probably never been used since the 08:02:26 code to generate them was added 2 years ago. It generates a parallel data structure (and errors out if you don't give it the information to do so) but then this structure is never actually referenced by the game. Which is actually somewhat fortunate, since the data it generates would crash if it ever was used. The coordinates provided for layout are wildly incorrect (it tries to place Zealots in the 15th column of the table, for instance) and 08:02:26 as-written can't be dynamically correct since background selection (unlike species selection) puts multiple categories into the same column and would need code to space them properly as the first category grows or shrinks, but the header generation uses fixed coordinates (somewhat defeating the purpose of doing automatic layout, even if it worked, since you'd still have to edit this manually on the regular.) 08:02:28 <04dracoomega> %git 6925c6b 08:02:28 <04Cerebot> NormalPerson7 {dolorous} * 0.32-a0-1694-g6925c6ba18: Generate job groups on the character selection screen from yaml files (2 years, 6 months ago, 31 files, 141+ 130-) https://github.com/crawl/crawl/commit/6925c6ba18bc 08:19:34 <09hellmonk> huh 08:30:44 <06dolorous_84348> Yikes. 08:40:24 <04dracoomega> I guess I will figure out how to fix this while I'm at it 08:42:23 <04dracoomega> (I'm not sure the width parameter of this is used by anything either, for the last 7 years, since the UI framework used for newgame menus changed) 09:28:11 <04dracoomega> Well, that's done and working now 09:33:00 -!- OnASnoop1hone is now known as OnASnoopPhone 09:33:42 <04dracoomega> Hmm... actually, it seems there's a minor makefile issue. If the groupings change such that job-data.h is regenerated, compilation doesn't seem to get the menu that newgame.cc needs to be recompiled until you build a second time afterward 09:38:00 <04dracoomega> Even adding it to the part of the makefile that says # Headers that need to exist before attempting to compile cc files doesn't seem to actually work 09:39:53 <04dracoomega> A lot of this makefile stuff is well outside my ability to read, though >.> 09:40:12 most of it needs to be replaced with something sane (fsvo) 09:40:33 I know makefiles pretty well and … it's even more of a mess than most makefiles 09:40:56 at one point someone was working on translating it to cmake, I think 09:41:10 Did you ever get to the bottom of why make -j sometimes mangles the tilesheets? 09:41:24 cmake has its own problems but it's still be a big improvement over the current ad hoc mess 09:43:22 <04dracoomega> Pinkbeast: I wasn't aware it had anything to do with -j but rather a similar thing of like "First compilation after switching to a tiles build from a non-tiles build doesn't do things right, but a second compilation will.". I don't know the details, though. 09:44:03 <04dracoomega> I am definitely not the person to spearhead any of that, I'm afraid >.> 09:44:18 dracoomega> oh! Well, that's handy to know in and of itself. I always thought it was -j because I curse and build a second time without -j and then it works, but ofc that's consistent with what you write 09:45:09 <04dracoomega> I suspect it's another manifestation of a similar "Some other file doesn't get flagged as dirty and needing regeneration until after it has been used in the current compilation" that seems to be affecting job-groups.h here too 09:45:21 <04dracoomega> If I had to guess 09:49:15 Kind of moot, if I'm feeling clever it's the memory leak in skill-menu.cc I'll have a crack at (I got Simon Tatham to explain a bunch of things to me so maybe I'm less completely confused) 09:58:53 <04dracoomega> Yeah, I don't feel like I'm making any useful progress hitting this repeatedly >.> 10:00:16 <04dracoomega> Of course, this isn't data that is changed very often, so it'll will be rare to cause an issue, but I greatly dislike the incorrectness that a compilation can sometimes not reflect the data it was supposedly compiled with 10:07:55 <09gammafunk> I would love if we could go through with a conversion to cmake 10:08:26 <08nicolae> more like pee make 10:08:28 <09gammafunk> I don't know build systems well at all, but it's not hard to see what a nightmare our Makefile is 10:09:10 <09gammafunk> now your comment is saved in the irc logs for all of eternity! your professional reputation is ruined! 10:09:43 <08nicolae> hell yes 10:10:01 <09gammafunk> one thing about any hypothetical conversion to cmake is that we'd have to deal with cao probably not having it nor it being available without installing a local version, but that's standard CAO problem stuff 10:21:47 all build systems suck. sadly, that's because pretty much any nontrivial build sucks 10:22:19 (nontrivial meaning it isn't just compiler all-source-files) 10:24:09 and I know crawl's makefile mess all too well because I modified it to use macports for its dependencies a decade back… ugh. ncurses in particular was a horrible mess because it kept insisting on using the ancient macos one and the "easy" workarounds broke the build completely 10:25:08 also an introduction to why `-isystem` is sometimes necessary 10:30:43 <09gammafunk> geekosaur: I assume you'd not be willing to revive that cmake conversion PR or otherwise undertake the conversion yourself? 10:30:51 <09gammafunk> "....how hard can it be?" 10:31:46 <06dolorous_84348> Much like how "should" is one of Those Words. As in, "This should work, but..." 10:32:37 we'd have to see how healthy I can be and how much time I have available around a bunch of other stuff that's a couple months behind due to health issues 😦 10:38:30 <09gammafunk> no worries/rush, just want to mention that such work would be greatly appreciated if you're able 10:39:14 <09gammafunk> speaking of work, I really need to merge that one pcre library update PR that the i18n developer put together 10:39:57 <09gammafunk> I'm not sure if anyone else is taking on merging that whole i18n project. It's definitely outside my expertise, but the pcre library update part is probably straightforward enough 12:16:31 New branch created: pull/5419 (1 commit) 13https://github.com/crawl/crawl/pull/5419 12:16:31 03Rosstin02 07https://github.com/crawl/crawl/pull/5419 * 0.35-a0-1027-g40daa1be6e: new welcome messages 10(2 minutes ago, 1 file, 26+ 0-) 13https://github.com/crawl/crawl/commit/40daa1be6e0c 13:09:02 03Rosstin02 07https://github.com/crawl/crawl/pull/5419 * 0.35-a0-1028-g891cca7fcc: welcome messages refer to the Orb of Zot by its full name 10(43 seconds ago, 1 file, 13+ 13-) 13https://github.com/crawl/crawl/commit/891cca7fcc88 15:45:54 Unstable branch on underhound.eu updated to: 0.35-a0-1026-g18991dcc64 (34) 23:45:19 Monster database of master branch on crawl.develz.org updated to: 0.35-a0-1026-g18991dcc64