01:08:14 -!- elliptic has quit [Ping timeout: 240 seconds] 01:10:12 I like the unborn deep dwarves. 01:10:16 -!- elliptic has joined ##crawl-dev 01:18:51 -!- Napkin_ is now known as Napkin 01:52:17 -!- Zaba has quit [Ping timeout: 264 seconds] 02:02:26 -!- st_ has quit [Ping timeout: 250 seconds] 02:02:44 -!- Zaba has joined ##crawl-dev 02:06:29 -!- valrus has quit [Remote host closed the connection] 02:15:06 -!- st_ has joined ##crawl-dev 02:15:44 kilobyte: This mildewed gcc 4.2 doesn't like the #pragma GCC diagnostic ignored "-Wstrict-overflow" in viewgeom.cc 02:15:45 greensnark: You have 1 message. Use !messages to read it. 02:16:19 I'm adding Wno-unknown-pragmas to cflags; any objections? 02:16:31 * greensnark always builds with -Werror and warnings are annoying 02:16:47 ... oh dear. 02:17:00 !messages 02:17:01 (1/1) casmith789 said (10h 30m 19s ago): AM is still broken on http://crawl.akrasiac.org/scoring/players/casmith789.html - thinking it is on Am doesn't show AM games at all in the table. 02:17:03 I think I need to write a shunt to fix something 02:17:31 this will be intersting 02:17:48 !tell casmith789 Can you be more specific about the Am/AM problem. I don't understand what's broken 02:17:48 greensnark: OK, I'll let casmith789 know. 02:18:18 greensnark: in the winning characters table 02:18:30 and the games played table too, I guess 02:18:38 Oh, got it 02:18:57 or not 02:19:05 Which bright spark screwed up the abbreviation of AM as Am anyway 02:19:39 Not sure. 02:20:44 greensnark: people were complaining about DG and DS not being Dg and Ds, also 02:20:49 Arcanemarksman 02:20:57 demi-god, demon-spawn 02:21:07 Yeah, those words have implied hyphens 02:21:21 You could go either way on those, but Arcane Marksman is pretty clearly two separate words 02:21:25 yeah, I agree 02:21:33 Also it shouldn't be Marksman at all 02:21:48 We go to excruciating pains to avoid human centrism in skil titles and then blow it in a class name :P 02:21:56 hm 02:22:02 okay, so I can see branch mimics fucking up travel and overview 02:22:18 but I don't see any reason why they would fck up future placement of the branch stairs 02:22:38 due: I assume that valrus just had an orc entrance hidden behind a secret door, they can be quite hard to find 02:22:45 yes, that is more likely 02:22:53 -!- st_ has quit [Ping timeout: 264 seconds] 02:22:56 i will write a fix for save files that have broken overview and travel cache shit 02:23:36 apparently there's no such thing as TAG_MAJOR_VERISON. 02:23:40 those verizon bastards! 02:25:23 TAG_MAJOR_VENISON 02:28:25 "mmm, dear" 02:29:59 -!- st_ has joined ##crawl-dev 02:36:36 -!- galehar has joined ##crawl-dev 02:40:09 hi 02:46:01 moin 02:46:08 h izaba 02:51:43 hmm. 02:51:47 -!- casmith789 has joined ##crawl-dev 02:55:49 Erm, wow, what happened to dolls? 02:56:00 Mine is hideously borken. 02:59:17 @??twister 02:59:18 unknown monster: "twister" 02:59:21 Oh, ok 02:59:24 I will update it when I push labrats. 02:59:58 Does the twister have two faces 03:00:02 do lab rats replace the altered rats in that one vault? 03:00:23 yes 03:01:59 -!- stabwound has quit [Ping timeout: 240 seconds] 03:02:21 -!- st_ has quit [Ping timeout: 260 seconds] 03:04:26 03due * r8e670811def5 10/crawl-ref/source/enum.h: Wrap MONS_UNUSED_XXX in #if-preprocessor blocks. 03:04:31 03due * r363d84146c1a 10/crawl-ref/source/enum.h: Dump the monster enum comments. 03:04:31 Oh, push did work. 03:04:32 03due * r6e6bef613321 10/crawl-ref/source/ (25 files in 5 dirs): Implement laboratory rats, using coolio's tiles as base. 03:04:32 03due * r0013d85e6deb 10/crawl-ref/source/dat/des/variable/mini_monsters.des: Place laboratory rats in the laboratory vault! 03:07:26 -!- st_ has joined ##crawl-dev 03:12:40 greensnark: Where is grd() unmarshalled per level? 03:12:59 Oh, tag_construct_level. 03:13:07 tag_read_levle, nm. 03:14:18 * greensnark is afraid due is planning mischief on a grand scale. 03:14:32 Due Fawkes day 03:14:35 :) 03:15:32 minf_mimic_feat? 03:16:09 Ohhh. 03:17:59 * greensnark <3 zsh completion that doesn't have to complete from start-of-word. 03:18:06 Does bash have a similar feature 03:18:23 Because I have so many directories where I know the directory includes the word "foobar" but not at the start 03:20:21 greensnark: Is there a guarantee that tag_read_you_dungeon has been called before tag_read_level? 03:20:26 I have a bunch of todo tiles in the little tile toolbar thingy 03:20:40 or where can I look for this? 03:20:41 due: Yes, character file is always loaded before level file 03:20:45 okay, good 03:20:58 Because until the character file is loaded the game doesn't know what level the character is on 03:21:10 yeah 03:21:11 s/file/chunk/ since everything is a chunk in the cs container now 03:21:25 I thought that was the case, but I didn't wnat to make Incorrect Assumptions. 03:21:58 greensnark: yesterday, I detached the new command tab jpeg did. 03:22:12 hopefully, this will motivate her in finishing the missing tiles 03:22:23 galehar: Cool, I was just wonder if my build was broken in some way :) 03:22:24 and it will be optionable very soon 03:22:38 I don't often look at tiles so it's hard for me to notice if something is actually broken :D 03:22:52 greensnark: We really just need to kidnap Denzi, co, coolio, purge, etc, and lock them in a basement somewhere. 03:22:56 Er, galehar. 03:23:22 Interesting, tiles looks much better on my big monitor than on my laptop 03:23:30 due: I'm not sure it's necessary. tiles production is great now! 03:23:55 greensnark: it does 03:23:57 greensnark: I think the viewport is scaled down on low res 03:24:00 galehar: Well, yes :) 03:24:25 huh 03:24:57 Ohhh, that's right 03:25:04 galehar: Yeah, and the fonts in particular seem to suffer; they were grainy and icky on my laptop, but pretty ok on my desktop 03:25:06 I have to do level_id cur_level = level_id::current(); 03:25:13 ---^ That is confusing. 03:25:41 due: Why is that confusing? 03:26:06 greensnark: Because it... is? 03:26:23 The current level should be stored in a global variable or an instance of level id globally available. 03:26:47 Calling a static function of the level_id class to find the current level id is totally contrary to you.where_are_you for branch type, absdepth0 for depth, etc 03:27:40 It's actually perfectly ok :) 03:27:51 Well, it's okay :) 03:28:08 But surely there is some object-oriented rule against it? 03:28:09 It's better to use that function than to peek at you.where_are_you and friends directly 03:28:48 There are object-oriented rules against everything 03:28:55 greensnark: Why not have level_id player::current_level() instead? 03:29:01 I'm not so much disagreeing with having a function returning it 03:29:15 Just disagreeing with the function being part of the level_id class as opposed to the player class. 03:29:57 In my mind, the data should be segregated: stuff relevant to the player goes in the player class; level_id should actually care where you currently are, it should just do stuff with its members in instances. 03:30:01 Anyway, it's a pointless quibble 03:30:04 That I refuse to fixate on. :) 03:31:35 Erm, shouldn't actually care, rather. Hooray for contradictions! 03:33:36 I can do stair_level[stair_level.find(level_id)] if I check stair_level.find(level_id) != stair_level.end, yes? 03:34:17 Or... not. 03:35:19 Fucking stackoverflow has such a hardon for boost. 03:36:00 -!- Kurper has quit [Remote host closed the connection] 03:37:53 due: "stuff relevant to the player should go in the player class" <-- arguable 03:38:04 greensnark: well, yeah, player is massive :( 03:38:17 Yeah, and object modeling is never so simple 03:38:25 alas :( 03:38:35 Player being affected by cloud :: should that be part of player or of cloud? etc. 03:39:27 player-cloud class! 03:39:36 player is floating around on a clout. 03:39:45 Well, on a cloud. I'm sure you must have clout to float on clouds. 03:40:22 -!- st_ has quit [Ping timeout: 250 seconds] 03:40:31 Okay. What's ffaster? i = blah; i < blah.size(); ++i || i = blah.begin(); i < blah.end(); ++i)? Is there a specific difference/ 03:40:32 -!- upsy has joined ##crawl-dev 03:40:32 -!- st_ has joined ##crawl-dev 03:42:09 You shouldn't be worrying about performance at that level 03:42:19 The difference is likely to be negligible 03:42:34 So it's just a syntactic choice? 03:44:34 Well, the second form works for most containers 03:44:57 The first form will be efficient for vectors, not so much for others 03:45:40 I'd use for (int i = 0, size = v.size(); i < size for the first form instead of checking size every time through the loop, but again the compiler will probably lift that anyway 03:46:30 Ah, makes sense. 03:46:48 greensnark: instead of -Wno-unknown-pragmas we might use -Wno-strict-overflow; any case where the compiler would assume it wrongly would be our fault 03:47:39 the warning suppressed is "assuming no signed overflow occurs when simplifying a + b > a" 03:49:30 elliptic: Do you have a save with corrupted travel info for branch stair mimics? 03:50:03 -!- herself is now known as superself 03:50:13 -!- superself is now known as herself 03:50:51 herself: :p 03:50:52 due: not sure what you mean by corrupted travel info... just ctrl-o showing a branch entrance that was really a mimic? 03:51:04 elliptic: yeah 03:51:17 I last saw that a couple games ago, sorry 03:51:46 I wouldn't have expected it to be that hard to replicate, though? 03:51:52 kilobyte: OH NO YOU SAW MY TRUE FORM 03:51:58 ;-p 03:52:01 elliptic: No, probably not, I suppose 03:59:27 due: actually my current game seems to have the problem 03:59:35 didn't realize it until just now :) 04:00:17 due: any objections to splitting DNGN_TREE to DNGN_SWAMP_TREE (possibly DNGN_MANGROVE or such...)? 04:00:29 but crypt shows up on ctrl-o and G and I haven't actually gotten to it yet 04:00:39 it is indeed easy to reproduce. Go to any level with valid branch, create stair mimic until you get a branch entry one, kill it and check c-O 04:01:31 well, here's my save with it in case anyone wants it: http://crawl.develz.org/saves/dumps/hyperbolic-f442e38-110105-1000.tar.bz2 04:04:19 kilobyte: what part of my code gives problem for merging ench-split? reskilling? 04:04:46 that and counting the accumulated exp 04:04:52 mostly the former 04:05:26 oh, just found another one: during reskilling, you hold skill enums, they'd break too 04:06:01 ooh, ench split is happening now? 04:06:05 it's of course nothing that can't be worked around 04:06:11 sure 04:06:33 which sort of split is it, again? I seem to remember some odd cases, at least 04:06:49 but if you think the ench split is good enough, it might be merged without that work 04:08:11 -!- jpeg_ has joined ##crawl-dev 04:08:16 hi! 04:08:18 well, we can merge ench-split when we bump major 04:08:22 hi jpeg 04:08:48 monky: protective/enhancing auras on living beings (Charms) vs enchanting everything elxe (Hexes) 04:08:51 I can't compile tiles because the labrat offset function is unknown 04:08:52 jpeg: meow! 04:09:11 galehar, kilobyte: hey! 04:09:30 I think we should put a todo list on the wiki for the major bump: reduce elf, merge ench-split, sort and cleanup enum.h,.. 04:09:34 monky: I think Haste is the only non-self case currently 04:10:03 or on a branch 04:10:06 so Invis is in hexes such as not to break Enchanters? 04:10:18 03j-p-e-g * r276ceb58ec4a 10/crawl-ref/source/ (tags.cc tileview.cc): Clear env.tile_names before loading/creating a level. 04:10:45 gah, hopefully _now_ it finally works as intended 04:11:11 good 04:11:36 right now Charms, but it's so important to Enchanters I think it'd be better dual-schooled (if we keep the current theme of the new schools) or perhaps even we can shoehorn it into Hexes (but that'd require some explanations) 04:11:53 thanks for all this jpeg. playing tiles in trunk will be much better now! 04:12:12 galehar: up to now, it's mostly been causing crashes :( 04:12:27 shit happens ;) 04:12:37 but it's for the better! 04:12:43 -!- stabwound has joined ##crawl-dev 04:12:53 galehar: it was supposed to be a simple fix, but I guess it's always like that 04:14:06 due: I don't know how the labrat offset is supposed to work 04:14:30 due: so I don't know how to fix the tiles build 04:14:34 also what happens to projected noise, and speaking of it, it's pretty really broken 04:16:25 whats worng? 04:16:40 huh. 04:16:52 jpeg_: it shoould be tile_offset, I thought I fixed it before I pushedit 04:17:28 gimme a sec and I'll fix it 04:17:40 tile_offset_ can't be found either 04:17:56 is it supposed to be generated in the tiledef? 04:18:03 no, it's in ghost.cc 04:18:11 but I don't understand, I'm compiling tiles builds without issue 04:20:27 that's crazy 04:20:34 apparently I managed to alter the commit and then revert myself 04:21:05 meh,git 04:21:06 fixed 04:21:07 quite a feat :) 04:22:22 pushing now 04:22:30 (or at least I will once it finishes rebasing) 04:23:05 03due * rc405a85b101c 10/crawl-ref/source/rltiles/dc-mon/unique/jory.png: Use coolio's tile for Jory. 04:23:07 03due * r2d9497dd2310 10/crawl-ref/source/dat/des/variable/large.des: Jory's Tower of Silence vault. 04:23:10 03due * r679089538a8d 10/crawl-ref/source/tilepick.cc: Fix tiles compilation. 04:23:18 And apparently I had other commits not pushed. Meh. 04:23:55 I'll hopefully have a fix for the branch stair thing shortly. 04:23:59 Just testing to make sure it actually works. 04:26:01 Jory's tile looks awesome 04:26:13 Yes, I agree 04:26:39 Hm, if I have tiles in windowed mode and I alt-tab between it and my IRC, the logo and "loading maps" is just a black square. 04:26:43 -!- galehar has quit [Ping timeout: 265 seconds] 04:26:44 (Where IRC overlayed it.) 04:27:24 -!- galehar has joined ##crawl-dev 04:28:00 wb 04:28:56 -!- stabwound has quit [Ping timeout: 240 seconds] 04:32:02 -!- Mu_ has joined ##crawl-dev 04:32:40 -!- ortoslon has joined ##crawl-dev 04:33:03 Napkin, due, kilobyte: please update windows builds 04:33:22 ASSERT(found) in dungeon.cc at line 7105 failed 04:33:26 ortoslon: not yet 04:33:32 ortoslon: I've still got a few more fixes to push 04:33:46 moin guys 04:33:48 jpeg_, it couldn't find a nearby staircase for some reason 04:33:52 sometimes, when trying to wizmode-jump to the Temple from another branch 04:33:52 ah ok 04:34:26 (and once, when wizmode-jumping from the Hells back to the Dungeon) 04:34:32 jpeg_: now that I have detached the command tab, you have to add the missing tiles! 04:34:52 galehar: bah, it's on the wiki 04:35:07 galehar: players will be unhappy with the question marks and come up with tiles of their own :D 04:35:08 jpeg_, I really don't know why that would happen, though 04:35:31 the third pass in dgn_find_nearby_stair looks like something that is unlikely to fail 04:35:55 if it only happened for the Temple, it might be related to #2355 04:35:59 (for normal levels, anyway) 04:35:59 jpeg_: good point. I'll wait for missing tiles before adding an option to opt-out of it :) 04:36:01 but it doesn't look like that is th case 04:36:35 Zaba: the odd thing is, sometimes it works even from distant branches (e.g. Shoals -> Temple) 04:36:41 galehar: are e/in-vokable abilities in tabs yet? 04:36:56 Zaba: sometimes, it fails even if the Temple entrance is in plain sight and the map has already been generated 04:37:22 Zaba: once I got the same error trying to climb out of a wizmode-entered Temple back to a dungeon level I might not have visited before 04:37:26 jpeg_: co has suggested recolouring vampires away from blue skin and has volunteered to do so if accepted. what do you think? 04:37:35 ortoslon: no. There's only a button to pull up the screen for abilities 04:37:38 Vampires have blue skin? 04:37:44 greensnark: Greeny/blue. 04:37:47 ortoslon: no, there are so many of them 04:38:02 jpeg_, seems to be quite easy to reproduce, even 04:38:11 jpeg_: less than spells :) 04:38:17 ortoslon: I really don't want to have that many question marks at the same time, so that'll happen once we got the tiles 04:38:24 good 04:38:24 ortoslon: you volunteering? ;) 04:38:37 http://crawl-ref.git.sourceforge.net/git/gitweb.cgi?p=crawl-ref/crawl-ref;a=blob;f=crawl-ref/source/rltiles/dc-mon/undead/vampire_knight.png;h=f0a5951689b26a1ab28d2463e78fac3304b064e0;hb=master 04:38:43 i can't draw << good enough excuse? :) 04:38:44 vampire knight doesn't 04:38:47 jpeg_, I'll look into it.. 04:38:55 http://crawl-ref.git.sourceforge.net/git/gitweb.cgi?p=crawl-ref/crawl-ref;a=blob;f=crawl-ref/source/rltiles/dc-mon/undead/vampire_mage.png;h=d9a60f6fec8ae3e8dde69ffcf5e0c81f3560b03a;hb=master mage does, as does plain vampire http://crawl-ref.git.sourceforge.net/git/gitweb.cgi?p=crawl-ref/crawl-ref;a=blob;f=crawl-ref/source/rltiles/dc-mon/undead/vampire.png;h=a3f5cbedacc821c4208b11017a7fa7398dbad033;hb=master 04:38:57 due: yes, I always thought they looked a bit sickly 04:38:59 also, I think we have to stop adding tabs. In the smallest res (800x480), the inventory is only 4 lines high, and the tab buttons fill the heigth. 04:39:17 jpeg_, oh, I know what the issue is.. 04:39:24 jpeg_, it's looking for DNGN_STONE_STAIRS_DOWN_I in the temple 04:39:25 galehar: well, we can remove existing tabs 04:39:26 spell and ability tab could be combined 04:39:33 jpeg_, that is quite very wrong, huh. 04:41:23 galehar: skill tabs aren't absolutely necessary, either, now we've got the fancy menu 04:41:47 galehar: replacing that with an ability time is fine with me 04:41:55 s/time/tab 04:42:30 jpeg_: ok. We can also use an option to define priority and put as much as fit 04:43:05 hm. 04:43:21 galehar: that, too 04:44:04 03greensnark * r82297df3abd1 10/crawl-ref/ (8 files in 3 dirs): Mac makefiles to create universal binary app bundles. 04:44:05 03greensnark * ra788123b3833 10/crawl-ref/source/contrib/sdl: Update SDL contrib to build on Mac with -arch ppc -faltivec with the 10.4u SDK. 04:47:28 jpeg_, basically, it happens because _wizard_go_to_level compares absolute depths in order to find out what staircase it should be looking for on the destination level 04:47:30 galehar: the tab order is not customizable yet, is it? 04:47:46 Zaba: meaning? 04:47:48 no 04:47:51 but it will be soon 04:47:58 galehar: :D 04:48:34 galehar: and of course I mean "priority", rather than "order", that's not entirely the same 04:49:00 jpeg_, suppose you're on absdepth of 18 in shoals, and temple is on absdepth of 8 04:49:25 jpeg_, then _wizard_go_to_level makes it look for a down staircase in temple, because you're coming from below 04:49:27 Zaba: it's trying to place downstairs? 04:49:39 wow... 04:50:12 how odd 04:50:35 surely it would check branch_length != 1 for thoe instances? 04:50:44 I've recently cleaned up dgn_find_nearby_stair's third pass, and one consequence is that it now never picks places that are in vaults. Temples are always encompass vaults, so now this issue has surfaced. 04:51:09 the new assertion popup is gorgeous, but it does have the downside that if I click it away quickly I can't look up the error anymore is not printed to the shell 04:51:12 jpeg_: My plan is to have an option to define the priority of the statcol. Default will be "minimap, inventory, commands, spells, gold_turns, monsters" 04:51:12 Will that impact on encompass levels that can be placed in the main dungeon? 04:51:15 i.e. it is not printed to the shell 04:51:18 jpeg_: It's not? 04:51:29 due, those likely will have down staircases 04:51:35 due, it would impact Slime:6, for example 04:51:37 jpeg_: Huh! I thought I had it set up to print as well. I'll tweak it. 04:51:50 due, but it's only an issue that arises when using wizard interlevel travel 04:51:57 Ahh. 04:52:00 ...I hope. 04:52:10 jpeg_: Things will be put int the order. minimap will always be at the top, and inventory at the bottom. detached tab will be in the order of the priority. 04:52:12 Okay, shim written to fix the mimic branch stairs fuckup. 04:52:41 galehar: gold_turns more important than monsters? 04:53:00 galehar: cool! 04:53:04 jpeg_: If you have time to think about it: do you think we should add metal wall tile to "tile improvement" page? 04:53:17 due: what do you mean? 04:53:32 due: I know the glass walls are problematic, but metal? 04:53:39 Hm. 04:53:52 It could just be that I have a completely metal level. :) 04:54:06 It just looks very new and shiny. :) 04:54:28 completely metal levels are possible indeed :P 04:54:36 due: ah, so you're basically asking for variation tiles 04:54:41 jpeg_: yeah! 04:54:42 due: those are always welcome 04:54:43 all of Dis, for example :P 04:54:55 jpeg_: it's just that the monster tab size is flexible, so adding it last seems a reasonable default. 04:55:37 galehar: does using the gold_turns option actually steal a line from the tabs? 04:55:51 -!- stabwound has joined ##crawl-dev 04:55:59 jpeg_: Also, I'm thinking swamp might be improved by decreasing the amount of red in the tiles. I'm goin gto do a mockup, but I'd appreciate your opinion, obviously. 04:56:02 (Floor tiles.) 04:56:11 galehar: ah, so the gold_turns doesn't flicker in and out depending on how monsters are on the screen 04:56:24 jpeg_: no 04:56:33 galehar: no? 04:56:43 due: should I keep that cell.swamp_tree_water code, or assume the water can be based on the feat in a simpler way? 04:57:08 jpeg_: the layout is redone only when spells are memorised. The monster tab size isn't resized when new monsters arrive. 04:57:40 galehar: oh, shows what I know :p 04:57:59 kilobyte: Hm, so long as feat_is_watery returns true for the new feature type, it can probably go 04:58:22 jpeg_: currently, setting the gold_turns option does steal a line from the tabs. With my new priority system, it will only steal a line from the monster tab ( or not if you change the option) 04:59:47 galehar: i see 04:59:59 due: feel free to add whatever tiles you feel are needed to the wiki 05:00:01 due: when testing the wizmode-travel bug I also teleported to the Dwarven Halls 05:00:14 Dwarvn Halls is 1-level too, yeah? 05:00:23 due: and the walls really clash with the limestone floor, it looked really awful 05:00:30 I haven't been 05:00:49 Swamp's tiles at least look very much like ... tiles, I guess. 05:01:20 due: a lot of tiles do because the border is slightly darker 05:01:38 due: I still haven't decided whether that's a good thing (the game _is_ grid-based after all), or not 05:01:49 I'm not too sure myself. 05:02:13 shoals look the least tiley 05:02:23 they also look worst 05:02:25 The ones we currently use for the dungeon balance this very well; at first glance, they don't appear to be too much like a grid, but you can see it on closer inspection. 05:02:29 might be a coincinence though 05:02:38 *coincidence 05:03:22 Some tiles type should give water tiles and such curved irregular outlines for Shoals/Swamp 05:03:25 It would be *really* nice if we could do water tiles in Swamp as .... I dunno, puddles. Single tiles are puddle of water in the centre; two in a row, a sort of trench of water in the middle, etc. 05:03:31 greensnark: <3 05:03:38 Wesnoth does that really well, btw 05:03:48 That's actually one of the games I was thinking of. 05:03:50 Try starting Wesnoth, open the map editor and place some water 05:03:57 That and another one... can't remember which one it is. 05:03:59 The tiles are readjusted as you draw <3 05:04:13 Must be Wesnoth then, because that's what I was thinking about. Odd! 05:04:52 HoMM also did that 05:05:18 and yes, that would be really nice 05:06:23 ortoslon: what do you mean, shoals looks the worst? 05:06:48 to me 05:07:14 ortoslon: obviously :) 05:07:17 greensnark: Can you double-check that the shim I've written in my latest push isn't hideously broken? 05:07:26 I've tested it and it works! 05:07:28 03due * rc3d461f50669 10/crawl-ref/source/ (dgn-overview.cc tag-version.h tags.cc): Fix #3146: branch stair mimics breaking overview, ^G travel. 05:07:30 But I am overly cautious :D 05:07:30 ortoslon: but could you be more specific about _what_ you don't like? 05:07:41 ortoslon: is it the layout, or the tiles, or the way they interact? 05:07:41 jpeg_: the history of limestone floor is as follows: when I merged the branch, it specified "limestone" for floor, making me think it is needed -- so I made a set of floors, testing them on D:1 05:08:07 jpeg_: it's hard to parse: items are hard to see on sandy terrain, too much curved edges in water 05:08:07 Oh... my. 05:08:15 jpeg_: but it turned out blue_anna merely copied the Tomb set and made a mistake writing the tile code :p (he plays console only) 05:08:21 Actually, I like the limestone floor, but it looks very... real. 05:09:06 jpeg_: even autopickup green border is harder to see in shoals 05:09:11 yeah, I think I did a quite decent job with them, but they indeed don't fit well with some wall types 05:09:39 kilobyte: to be more specific, they don't fit with wall types that have a lot of detail themselves 05:09:43 The shoals ground doesn't set off green very well, no. 05:09:46 s/types/tiles 05:09:56 We really need cliff tiles for Shoals, but I've no idea how they could be drawn. 05:10:16 The stone looks... really oddly out of place. (Actually, I find stone almost always looks out of place except for some maps.) 05:10:47 ortoslon: hmm, true, the autopickup border is hard to see on light floor tiles 05:11:24 ortoslon: that could be improved, I guess 05:11:29 also low contrast between sandy floor and stone walls 05:11:34 perhaps limestone could be made much darker? 05:12:00 of course, Dwarf needs the wall type to be changed, but that was obvious from day 1: come on, skulls!?! 05:12:18 ortoslon: also, do the curved edges bother you aesthetically or do you think they make moving around more difficult? 05:12:34 the latter 05:12:43 ortoslon: in general, if you've got any specific complaints, feel free to add them to the tiles improvement page 05:13:15 are these complaints specific enough? 05:13:35 ortoslon: sure, now you've clarified them 05:13:45 due: ortoslon does have a point about curved water borders making movement more difficult 05:13:52 due: that would also apply to Swamp pools 05:14:01 yes 05:14:17 i'd rather have my tiles squarish 05:14:24 for the record, i have grid on in wesnoth :) 05:14:33 Ick 05:14:59 I always used grids in homm, myself 05:15:23 if we do make the tiles less "tiled", we absolutely need to add a grid option 05:15:45 I ... was about to say, do we add a grid option? 05:16:09 due: currently it's really only an issue for Shoals 05:16:13 yes 05:16:21 due: so it'd be a shoals_grid option for now :) 05:16:48 due: might be more distracting than useful in other branches 05:16:51 03zaba * r943bb48f6f3b 10/crawl-ref/source/wiz-dgn.cc: Try harder to pick correct staircase in wizard interlevel travel. 05:16:53 single 1 pixel black border? I've no idea how you could draw it though. 05:16:59 (not to forget ugly) 05:17:05 * due not an SDL expert. 05:17:19 (If I was I might come up with a way to do background ripple effect from the agrd centre for liquefaction :D) 05:17:40 also "items are hard to see on sandy terrain": the worst case is probably a needle or a dart 05:17:41 due: the easiest solution is to add an overlay icon with two lined edges 05:17:49 due: put together, it forms a grid 05:17:57 oh, interesting. 05:18:22 ie, a north and east lined overlay? 05:18:27 ortoslon: stones/bullets too, I sometimes mistake those for background rocks 05:18:34 due: yes 05:18:49 due: that's where the current tiles' tiled look comes from 05:19:38 due: and here I thought you already had the liquefaction tiles drawn and ready 05:19:45 jpeg_: I did. 05:19:53 jpeg_: It looked absolutely terrible in-game :( 05:20:14 due: oh, that happens sometimes 05:20:23 Per-square ripple is unfortunately not the way to dgo. 05:20:33 due: do you think the tile is salvageable? 05:20:45 No, unortunately not. 05:20:46 i.e. upload it and let someone else improve it :) 05:21:03 It was just the simplest of ripple effects in a dark, transparent colour. 05:21:13 i see 05:21:17 I tried light colour, dark colour... inverse, etc. Even tried animating it, but that just made it look... odd. 05:21:29 so that would have to have neighbours in account as well 05:22:22 jpeg_, I've fixed travelling to temple and slime:6, at least 05:22:46 what I'd *like* to do would be to somehow draw the floor tiles onto an SDL canvas of some description, then apply a ripple effect onto it. 05:22:49 Zaba: slime 6? 05:23:02 -!- elliptic has quit [Ping timeout: 240 seconds] 05:23:07 jpeg_, it was also problematic when coming from a deeper absdepth 05:23:16