00:00:02 I discovered in the midst of this that there are a few portals that aren't transparent (abyss, hell, pan) that probably should be, because they were throwing a lot of vetos in that layout. 00:00:21 There are occasions when disabling the large Zot entries doesn't help a huge amount, actually. 00:00:29 WebTiles minimap mouse support issues by Medar 00:01:19 Grunt: my theory is D:27 is a combination of loads of different factors; so non-transparent vaults will certainly not be helping 00:01:21 If I remember correctly, what helped the most was telling D:27 not to place a primary vault (possibly so it spends a lot less time trying to rotate the huge zot entry vaults into place). 00:01:29 Anyway, sleep time for me now. 00:02:01 -!- CampinSam has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 00:02:03 Grunt: yeah, it also carries on placing loads of other random floating vaults even if there's a primary vault, _even if that primary vault is encompass!_ (which is when you get really slow times) 00:02:44 quick: what's the most generic item 00:02:47 bh: that sounds cool, yeah you need a bit of room to explore it 00:02:50 ontoclasm: Bread ration 00:02:57 good choice 00:03:06 What did I just answer, anyway? 00:03:16 "what should detected items look like" 00:03:18 Haha 00:03:31 Don't they have their own tile? 00:03:42 yes but it's confusing 00:03:48 (And if that one is getting replaced, shouldn't it be with something that doesn't look like some SPECIFIC item?) 00:03:49 it's a pile of stuff 00:04:12 bh: i kind of want to try making an infinite level without morphing using the same techniques but loads of extra layers of detail 00:04:16 Maybe an outlined item with a ? mark on it or something? 00:04:20 ontoclasm: are you gonna stick a question mark on it? 00:04:20 well if i do a bread ration it'll be greyed out with a '?' on it 00:04:28 Ha 00:04:31 mumra: makes sense 00:04:33 :smug: 00:05:15 Unstable branch on crawl.s-z.org updated to: 0.12-a0-2965-gf1e848c (34) 00:05:43 https://dl.dropbox.com/u/1000017/crawl/unseen_armour.png 00:05:49 for example 00:06:49 The problem with a greyed out gauntlet is that it just looks like normal steel or something 00:06:55 hm 00:07:07 maybe outlines are the way to go 00:07:09 I know the original is more colorful, but it still looks like that one could be normal coloration 00:07:45 greyed out bread is probably pretty easily discernable? 00:07:50 it is very hard to draw "an item you can't identify" 00:07:52 Yes, probably morespo 00:07:57 ufo 00:08:13 ontoclasm: A treasure chest! :P 00:08:31 but one of them is "unseen weapon / ammo" 00:08:41 Oh, there are really more than one type of detected item? 00:08:44 yes 00:08:46 I thought it was just the one thing 00:09:11 your confusion is understandable 00:09:18 I get amulets mixed up with unseen items 00:09:28 -!- bonghitz has quit [Remote host closed the connection] 00:09:34 -!- SetecAstronomy has quit [Quit: Page closed] 00:09:35 really i just want to make them big green triangles with ITEM written on them 00:09:36 in particular, I think an amulet is a still-unseen space 00:09:47 that's probably not a bad plan 00:10:56 ontoclasm: That doesn't seem inherantly bad 00:13:02 mumra: Well, it looks like it can't place the stairs because the level is all rock walls 00:13:12 mumra: Or at least as much of the level as is fitting in the console buffer 00:14:09 that has to mean either 1) the code is erroring so the level doesn't render or 2) it's getting overwritten after rendering for some unknown reason 00:15:05 ontoclasm: http://morbiusjkromwell.files.wordpress.com/2010/08/chalk_outline.jpg 00:15:06 the layout has its own internal debug mode which will spit out the floor plan as it's rendering 00:15:42 mumra: Well, it's not running into errors enough to terminate things up to that point, at least 00:15:49 How does one access this internal debug mode? 00:16:10 it's _VAULTS_DEBUG in dlua/v_layouts.lua 00:16:16 DracoOmega: try again with a bigger console window? 00:17:09 SamB: Well, I could, but instead of trying to spit out the whole grd, this time I just had it tell me if any tile was non-rock (and none seem to be) 00:17:22 oh 00:18:31 * SamB wonders how hard it would be to allow resizing the console while Crawl is running 00:18:33 mumra: Where is the output for this supposed to show up, anyway? 00:18:46 normally console 00:18:55 Well, I set it to true, ran it again, and see nothing 00:19:19 hmm, something must stop it during that script 00:19:43 -!- dcssrubot791 has quit [Ping timeout: 264 seconds] 00:19:57 I am wondering if the layout lua is getting properly called at all here? 00:20:20 Since I am printing OTHER messages to the console with printf and such, so it's not surpressing that generally 00:21:20 DracoOmega: wait, you're in tiles, so why don't you have decent scrollback on your console anyway? 00:21:53 SamB: I think this lua script that's failing to generate the map right is console-only 00:22:31 DracoOmega: yes i remember now, i found this weirdness as well; it was really hard to found out what the veto was because it didn't show up in console 00:22:41 -!- SamB has left ##crawl-dev 00:22:54 i can't remember how i got around it but it was really obscure what did or didn't print 00:23:04 and i don't have a debugger set up yet 00:23:05 and i don't have a debugger set up yet 00:23:34 -!- SamB has joined ##crawl-dev 00:23:38 whoops ... 00:23:52 anyway, WTH!?!? 00:24:50 !tell Grunt I added transparent to those entry vaults, but D:27 still has massive problems. 1m33s vs e.g. 0m12s for D:10 !! 00:24:51 mumra: OK, I'll let grunt know. 00:25:40 This is kind of annoying 00:25:45 03mumra 07* 0.12-a0-2966-g7b82cf1: Make some entry vaults transparent 10(15 minutes ago, 3 files, 6+ 5-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=7b82cf118930 00:25:56 DracoOmega: the all-rock thing makes sense because while i was developing the layout i'd sometimes get that random floor square error when e.g. i'd broken something that stopped anything rendering 00:26:24 -!- bh has quit [Quit: quit] 00:26:25 the question is, *why* is it all-rock? either the lua isn't running or the map is somehow getting reset after it runs 00:26:40 Yes, and the annoying thing is that if debug output from it won't show up in console, then um.... 00:26:44 It makes it hard to tell 00:27:48 could you output stuff to the diagnostics channel? 00:27:50 i don't really understand the codepath of mapstat but ... is it possible to actually make it run in a slightly more normal environment so you can see each layout as it renders? 00:28:08 mumra: Well, that would help if I understood in what way its environment was different, but I don't 00:28:16 * SamB namedrops kilobyte 00:28:37 SamB: Well, this seems to bypass most of the game entirely 00:28:44 So you don't get normal messages, even debug ones 00:28:52 oh 00:29:09 Which is why I am using printf instead of mpr or something 00:29:13 Since that at least shows up 00:29:25 DracoOmega: thinking about it, it'd b fairly trivial to write a wizmode command that did ^R numerous times and told you the total time at the end 00:29:40 DracoOmega: or in your case output monster stats to a file? 00:29:51 Hmmm... maybe that would be easiest 00:29:57 But it would still be nice to know why this is broken 00:30:02 yes, definitely 00:30:03 Because that might imply something else that is broken 00:30:23 possibly it's the whole idea of bypassing most of the game ;-P 00:30:46 It's possible! But I don't really understand what it's doing in detail. I was just tediously tracing down to the first point at which the builder complained 00:30:58 i'm wondering if there's less memory available to lua ; it's very likely that V is using more memory than any other lua layout 00:31:35 Wouldn't that throw some other kind of error, if it ran out of memory? 00:31:39 Rather than just doing nothing? 00:31:57 Since I assume that rock wall is the default state for a level to be in, unless you modify it 00:32:02 yup 00:32:38 check the lua docs? 00:33:07 but lua handling is pretty weird about a few things i've noticed and it wouldn't surprise me if it was just shutting down without ceremony at the slightest sign of trouble 00:33:26 const int LUA_CHUNK_MAX_SIZE = 512 * 1024; 00:33:29 that's in dlua.h 00:33:35 maybe it needs bumping? 00:33:56 nah 00:33:58 if that was the problem then it wouldn't work in-game either though? 00:34:01 that's something else 00:35:42 lua chunks are code blocks yeah 00:35:55 that should be plenty 00:36:15 512kb should be enough for anyone! 00:36:45 I mean, a wizmode command for this would be nice anyway, since otherwise I need to recompile in console debug just to be able to run it. But still! 00:37:31 hmm, there's something called _clua_panic 00:38:35 maybe i should redraw the zot gates :( 00:38:49 I really don't think that will make this work 00:40:14 ontoclasm: They seem okay to me. At least, a lower priority than some things (though don't like me stop you!) 00:42:45 DracoOmega: good point about recompile, that bothered me too 00:43:19 DracoOmega: also for all we know, some vetoes on other levels could be getting caused by the same problem and skewing the results, it just happens to trigger every time on V levels 00:43:22 -!- Mandevil has quit [Quit: Leaving] 00:43:32 Yes, this is one of the reasons I want to find the cause here 00:43:33 so running in a proper environment is all-round a more realistic simulation 00:43:38 Since it seems plausible it is having spillover effects 00:43:51 Oh, that is also true 00:43:53 okay, so it looks like each lua interpreter gets 6 MiB 00:44:03 I thought you were saying that the side-effects of this might persist into the normal game in more restricted circumstances 00:44:09 (Which may still be true, I guess) 00:44:38 DracoOmega: well, also that; you're right, we really have no idea 00:45:06 the line to edit would be: 00:45:08 #define CLUA_MAX_MEMORY_USE (6 * 1024) 00:45:11 in clua.h 00:45:25 But it's not clua, it's dlua, yes? 00:45:54 they're not entirely separate things 00:46:00 well, dlua is a CLua ... 00:46:19 I do not think I would have named things thus, but I'm not sure about that 00:47:32 It is? I thought clua was for stuff like autofight and such 00:47:43 Admittedly I am not that familiar with the lua end of things 00:47:53 yes, that's why I hope I'd have named the CLua class something less confusing 00:48:20 Some progress: it is picking proper layout vaults to place, but giving them a size of (0,0) 00:48:33 -!- Stelpa has quit [Remote host closed the connection] 00:48:35 Which means that it doesn't draw anything to the grid 00:48:37 there is a CLua called clua which is used for config stuff 00:48:49 Oh, CLua and clua are different? 00:48:54 there is another called dlua that is used for dungeon stuff 00:49:05 yes, C++ is case-sensitive dontcha know 00:49:22 This is true! But I didn't know that you were being there :P 00:49:32 The c in clua is for 'client', yes? 00:50:11 dunno, clua.cc and clua.h are missing the comment at the top summarizing their purpose 00:50:21 -!- Stelpa has quit [Changing host] 00:50:52 also, most of what's in clua.h seems to apply to both 00:50:56 the clua folder is client lua, it's stuff that you can customise per user account like autofight 00:51:03 dlua is internal dungeon generation and stuff 00:51:23 however Clua is the c++ library that provides the interface between C++ and Lua and the virtual machine and whatever 00:51:29 it could stand for config lua, too 00:51:33 maybe 00:51:36 Oh, so the C there is for the language 00:51:49 dunno! 00:51:55 possibly 00:52:00 there are no comments about that 00:52:06 i hadn't realised that, but it seems likely 00:52:31 -!- Kalir has quit [Changing host] 00:52:52 -!- dtsund has joined ##crawl-dev 00:54:35 okay, I'm confused. CLua is pulling in some files from dlua/ :-( 00:55:24 hang on, Clua or clua? 00:55:50 CLua::init_lua() 00:55:53 Yes, clearly calling it clua was a mistake 00:56:21 clua can certainly use includes from dlua, it just won't have access to the same api i.e. when it comes to stuff the player shouldn't have access to like the dungeon grid 00:56:33 it's kind of confusing though 00:59:43 -!- Krakhan has quit [Ping timeout: 245 seconds] 01:03:43 Unstable branch on crawl.develz.org updated to: 0.12-a0-2966-g7b82cf1 (34) 01:03:45 -!- st_ has quit [Ping timeout: 245 seconds] 01:04:23 Okay, my mistake. The size of 0 is normal for this layout, it seems 01:04:35 What is NOT normal is that it is not placing any subvaults when run via place-population 01:04:44 Well, 'subvaults' 01:04:47 Rooms 01:06:30 strange normals we have around here 01:06:36 -!- _dd has joined ##crawl-dev 01:07:00 -!- eeviac has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 01:07:08 Well, presumably the size is 0 since it doesn't use normal painting logic? 01:07:14 And just does magic in lua instead 01:07:28 But the magic won't happen T.T 01:07:51 well, it seems like the lua magic ought to also set the size to something more representative of the actual size, or something 01:08:28 dunno 01:09:01 Well, the rooms have normal sizes 01:09:09 The layout itself doesn't seem to, since the layout is not drawn 01:10:59 -!- ussdefiant has quit [Ping timeout: 252 seconds] 01:11:47 But it does make me increasingly think the lua for the layout just isn't getting called at all here 01:11:59 If no attempts to place rooms seem to be happening at all 01:17:42 Yes, it never loads anything aside from builder_layout_vaults.dsc 01:19:46 gah, you aren't supposed to use hugeterm for *coding*! 01:19:49 just playing 01:20:33 Windows builds of master branch on crawl.develz.org updated to: 0.12-a0-2966-g7b82cf1 01:20:54 Ah 01:20:56 I think I have an idea 01:20:58 "if not crawl.game_started() then return end 01:20:59 " 01:22:05 -!- ontoclasm has quit [Read error: Connection reset by peer] 01:22:28 yeah, bypassing the game would tend to make that not work ;-P 01:23:00 why the heck are there so many of those? 01:23:26 -!- ontoclasm has joined ##crawl-dev 01:24:01 mumra: any idea what that's for? 01:24:12 Well, that fixes the problem 01:24:26 really?? 01:24:32 Seems like 01:24:35 I commented them all out and it runs 01:24:35 it actually calls that further in 01:24:43 oh you commented them _out_ 01:24:54 Well, yes 01:24:59 Just to see if that really was the problem 01:25:07 The script is now running (slowly) 01:26:20 Why is that check there, anyway? Is there some purpose it actually serves? 01:26:25 i'm not sure where i got crawl.game_started originally, most layouts have if not crawl.is_validating() instead 01:26:40 but i locally tried changing it to that and it didn't seem to fix the mapstat issue 01:28:29 Well, I will try running mapstats once this is done 01:29:46 i reckon i copied and pasted from _somewhere_ and ended up thinking that line was important 01:30:15 is_validating is definitely needed because you don't want to run the whole map generation during validation (which takes place when crawl first starts) 01:30:38 i'm trying mapstats now 01:31:50 -!- magicpoints has quit [Ping timeout: 255 seconds] 01:34:41 awesome, that works 01:34:41 Ugly things: most common enemy on V:2 by far, it seems 01:34:43 Kinda sad 01:35:44 don't we have the powah to fix that now? 01:36:31 Yes, though I had hesitated to make too many population distribution changes so close to an ostensible release 01:36:36 well spotted; i really shouldn't have missed that 01:36:45 Certainly there are many things that can and should be done for 0.13, though 01:37:02 is anyone going to complain if V gets less ugly? 01:37:10 Oh, I might do that, depending :P 01:37:18 But there are tons of stuff that could use doing in multiple places 01:37:31 At the very least, I am going to do more stats and tweak new V stuff before release 01:38:21 What was the mapstat command, again? Now that that run is done 01:38:30 well i've got mapstat running here 01:38:40 worryingly it's much slower than i though 01:38:54 Yes, this bit was slow enough that I wondered if the 50 vetos might be quicker, to be honest 01:38:58 Since they were getting vetoed really early 01:39:21 the command is: time ./crawl --mapstat Vaults:1 01:39:42 damn, it's _worse_ than D:27 :( 01:40:11 Doesn't entirely surprise me (sorry to say) 01:40:43 i know, it's actually other people that convinced me V wasn't as bad as D:27 01:40:50 it's funny how people's perceptions work 01:40:53 Heh 01:42:14 Hmmm... mine just seems to be stuck on 13 out of 100 01:43:03 Oh, there it goes again 01:45:34 does anyone else get this thing where checkwhite makes git thinks every single file has been changed? 01:46:13 once the commit completes it's realised that actually only 1 file changed but the commit takes ages while it figures this out ... 01:47:01 I don't believe I have used checkwhite (or know how to, for that matter) 01:47:22 But my IDE happily strips trailing whitespaces whenever I save anything anyway 01:47:27 (Not so good for lua, though) 01:48:22 mumra: on which OS are you doing this again? 01:48:29 Windows 01:48:36 Oh, huh. It occurs to me now that perhaps this was slowed down on my end by those debugging printfs that were still printing in the background, maybe >.> 01:48:40 that might be the EOL thing, then 01:48:56 DracoOmega: well, it is really slow 01:49:04 Yes, I know, but this might have made it worse :P 01:49:10 What time did you get for yours? 01:49:21 SamB: i disabled autocrlf and added a pre-commit hook today, per the git config instructions i was previously unaware of 01:49:26 Actually, that's probably meaningless, since I have no idea how our computers compare 01:49:31 now when i try to commit it tells me to run checkwhite 01:49:44 DracoOmega: it took over 2 minutes here 01:49:50 and this isn't exactly a slow computer 01:49:50 mumra: where are these instructions? 01:49:59 mumra: Took over 7 here 01:50:05 (And neither is this) 01:50:10 docs/develop/git/config.txt 01:50:10 I have disabled all those now, and am rerunning 01:50:27 -!- ontoclasm has quit [Read error: Connection reset by peer] 01:50:51 DracoOmega: this is a quadcore gaming laptop i got towards the end of last year. of course only 1 of those cores will be used for level generation but still, it's pretty nippy 01:51:03 Oh, well I'd say that definitely beats mine 01:51:07 compiling crawl is actually pretty fast with -j 01:51:21 I mean, mine was technically a quadcore gaming desktop, too :P 01:51:23 But it's much older 01:51:41 03ontoclasm 07* 0.12-a0-2967-gc3a52c9: Detected item and feature tiles 10(78 minutes ago, 5 files, 0+ 0-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=c3a52c912a23 01:51:41 03ontoclasm 07* 0.12-a0-2968-gff19bbb: Pandemonium portal tiles 10(5 minutes ago, 6 files, 5+ 1-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=ff19bbb6acd6 01:52:06 -!- ontoclasm has joined ##crawl-dev 01:52:14 thanks internet 01:52:38 Free round trip to the wonderful world of disconnects 01:53:24 The unseen stuff looks quite good 01:53:41 Also, does it mean that this stone arch with a question mark will replace shop tiles? :P 01:53:52 it always has 01:53:55 or i mean 01:53:58 -!- ophanim has quit [Read error: Connection reset by peer] 01:53:59 since that tile existed 01:54:27 Has it REALLY been that long since I saw it in action, then? =/ 01:54:40 I thought it was new, since it was in the commit here, but I guess you just updated it? 01:54:41 at least a couple of months 01:55:03 yeah, it was based on the super-old arch tile 01:55:15 so i replaced it with the new(er) arch 01:55:16 Since I have started coding for Crawl more regularly, I have almost stopped PLAYING Crawl >.> 01:55:48 yeah, it doesn't help that i suck at crawl 01:55:58 i am highly unlikely to see that pan exit ever 01:56:00 You're not so bad, I think! 01:56:10 Well, you've won, yes? 01:56:21 If you can win, you can see a Pan exit 01:56:24 yeah, but not with any regularity 01:56:36 !won 01:56:37 ontoclasm has won 8 times in 715 games (1.12%): 1xCeSt 1xDESu 1xDsCj 1xHOFE 1xMfIE 1xMiBe 1xOpWz 1xTrHe 01:56:48 That's really not that bad 01:56:53 !lg * / won 01:56:54 11262/2169267 games for *: N=11262/2169267 (0.52%) 01:57:01 More than twice the average! 01:57:32 mumra: Okay, not too much better that time. 6m58s 01:57:33 mumra: well, it looks like checkwhite *does* modify the files 01:58:01 so I guess it makes sense that git would have to re-read them to see they're not changed 01:58:23 D:1 takes a lovely 27s by comparison >.> 01:59:17 well i've managed to more than halve the time here by fiddling with some options 01:59:28 but it's still not good enough 01:59:28 Reducing the max number of attempts? 01:59:37 yep 01:59:42 and also reducing the total 01:59:47 # of rooms actually palced 01:59:54 Well, I hope that doesn't make the maps too much smaller, noticably 02:00:00 I do quite like the layouts as-is 02:00:01 27 is just too many, it completely fills the map, so it's very hard to find the last few spaces 02:00:59 * SamB tries to figure out what's so hard if it fills the map ... 02:01:11 Well, it needs to find room to place the rooms 02:01:17 (Or is there wiggle room?) 02:01:18 And if there's not much room, it fails to find space repeatedly 02:01:24 And thus many more iterations 02:01:33 (As I understand it) 02:01:33 once a map is placed it can't be moved 02:01:59 -!- WalkerBoh has quit [Remote host closed the connection] 02:02:14 also trying to place large maps can be slow because it checks every square until one is blocked 02:03:16 I realize it's a little late at this point, but I wonder what speedup would be gotten by doing those block checks in C++ instead of lua 02:03:19 with these settings the rooms are certainly noticably less packed ... *although* in some respects some of these actually look more interesting 02:03:41 DracoOmega: lots certainly 02:03:50 but there's a lot more going on than that ... 02:04:02 Yes. Know of any good lua profilers? :P 02:04:10 oh, the rooms are all different sizes 02:04:11 for one thing there are certain lua callbacks that get passed into the checking process (and even more so with the new version) 02:04:22 SamB: And shapes 02:04:31 Non-rectangular ones and all 02:04:36 now it's certainly possible to send lua callbacks up to a c++ function and so forth but this will all add some overhead 02:05:12 and we'd need to store some overall state in between calls to the c++ placement function too 02:05:48 it would all be possible ... but basically it was easier to prototype and design in lua because i kept changing stuff and it's actually a really flexible language 02:06:16 Oh yes, I do understand 02:06:38 and doesn't have those pesky long compilations? 02:06:55 WebTiles minimap doesn't show friendly/peaceful monsters properly by Medar 02:07:05 But this may be the kind of heavy-lifting code that makes sense to try to port back 02:07:16 If not for 0.12, then at some point (maybe) 02:07:18 yeah exactly, nor all the impossible-to-understand compilation failures that i'm finally starting to understand now i'm messing with c++ a bit more ;) 02:08:16 * SamB <- using a Pentium II ... 02:08:32 That's... kind of impressive, in a way 02:08:34 -!- Moredread has quit [Remote host closed the connection] 02:08:38 DracoOmega: exactly, now i'm happy that i have an algorithm that does everything ... the only problem is v3 is even more dependent on callbacks 02:08:55 on the other hand v3 has some features that will help optimise things better 02:09:15 Well, tonight is the closest I've ever gotten to knowing how the dlua layer works :P 02:09:18 (Which is 'not very') 02:09:43 The Crawl codebase is so massive that you can spend ages with it and still have tons of it that is completely foreign to you 02:10:14 Monster and spell code feels so much more familiar for some funny reason :P 02:10:27 ;-) 02:10:50 So, on the minimap items get higher priority than stairs. One dart and you can't use minimap to locate stairs anymore. 02:11:32 * SamB isn't positive he actually fixed the inter-game recall bug ... 02:11:36 I'm wondering whether I should come up with more sensible priority order, or also add an option for tweaking that. 02:11:46 SamB: Really? The fix sounded simple 02:11:58 Dungeon Sprint "Red Sonja", banished into abyss, going further into abyss causes crash by Fergy 02:12:04 I'll check after I die next 02:12:16 with, you know, wizmode ... 02:13:03 DracoOmega: so ... one option is to place less rooms per floor but add, say, another 3 floors to V :P 02:13:18 Haha 02:13:20 (It's possible that I had some orcs wander off and didn't realize how many there were, but ...) 02:13:22 I suspect there might be opposition 02:13:40 wasn't the reason it got shortened mainly because of the increased number of monsters due to newvaults? 02:13:48 I think people wanted to shorten it anyway? 02:14:08 5 does feel like a nice number 02:14:41 I don't think it's slow enough to be unreleasable as-is, to be honest, if that's what you're thinking 02:14:50 Since only people running stats need to generate 100 levels at once 02:15:46 i slightly worry about servers during the tournament, but it's not like people will be in V constantly causing level generation 02:16:01 Indeed. And I mean servers have been running it all this time, already 02:16:03 yeah, they have to DO stuff too 02:16:26 and maybe it will do the players some good to wait for the computer for a change 02:16:34 haha 02:18:10 okay, I guess it did work 02:18:27 because I just died and then fired up wizmode and I've got no dudes 02:18:42 it only ever seems to take about 1s on the servers 02:19:51 so I guess some of my dudes wandered off when I'd thought they'd died 02:23:34 Well, that is goodf 02:23:38 On both counts, I guess! 02:24:00 on the other hand, I did die to a hydra after that, so ... 02:24:28 Okay, that is perhaps not so good :) 02:24:58 but at least I don't have to feel like I was cheating or anything :-) 02:25:35 wtf angels dont follow me to the next level of a zig? 02:25:45 Nothing does 02:25:53 but its stairs :| 02:25:53 Hostile or friendly 02:26:08 It functions more like a portal entry here, admittedly 02:26:37 Being unable to drag hostiles through is pretty important to ninjaing floors, though 02:31:18 -!- syraine has quit [Ping timeout: 245 seconds] 02:37:44 -!- Zifmia has quit [Ping timeout: 256 seconds] 02:37:47 Yeah, I was afraid something might be off and it seems so 02:38:26 Vault sentinels and convokers appear to have become about 4-5 times less common on V:2 than before the mon-pick merge 02:38:59 huh, and kilobyte tried to be so careful 02:39:10 It looked strange to me that they were not eligable to spawn on V:1-2 in the new data set, though kilobyte said that was the same as before 02:39:19 Baring ood spawns, I mean 02:39:21 I guess the mapstat breakage is why it didn't work out? 02:40:16 But they seem to have gone from 2-2.2% of monsters to 0.34-0.41% 02:43:32 DracoOmega: how do you feel about the number of walls with big doors or multiple doors? 02:43:41 Wow, for that matter, yaktaur captains have fallen from 2.76% to 0.37% according to this 02:43:48 That's... very significant 02:43:54 -!- Svendre has quit [Ping timeout: 264 seconds] 02:43:57 wow 02:44:19 Stuff of that magnitude seems unlikely to be a statistical anomoly 02:47:00 I am suspecting there may be some systemic issue here 02:52:01 %git ge3ef79a 02:52:01 Could not find commit ge3ef79a (git returned 128) 02:52:36 drop the g 02:52:38 it's not hex 02:53:17 Minimap priority tweaks by Medar 02:56:59 Stone giants are also around 4 times less common 02:57:14 More yaks, though! 02:57:44 -!- SkaryMonk has quit [Quit: Leaving.] 02:58:00 (Not yaktaurs. Plain yaks) 02:58:20 what about death yaks 02:58:22 Which have gone from not spawning on V:2 at all, to being the 3rd most common creature there, according to this 02:59:20 -!- simmarine has quit [Ping timeout: 260 seconds] 03:01:49 All in all, this seems fairly bad 03:02:56 About the last thing Vaults needs is fewer humanoids and more yaks :P 03:14:14 03mumra 07* 0.12-a0-2969-g18ba91c: Fix an erroneous Lua call preventing mapstat for Vaults 10(89 minutes ago, 1 file, 2+ 9-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=18ba91c90358 03:14:14 03mumra 07* 0.12-a0-2970-gc5845cd: Tweak performance parameters in Vaults 10(34 minutes ago, 2 files, 7+ 7-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=c5845cd05462 03:14:14 03mumra 07* 0.12-a0-2971-g059b507: Make multiple doors and big doors less likely in Vaults; restore windows 10(25 minutes ago, 1 file, 10+ 4-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=059b5071c6b5 03:14:51 Weren't windows specifically removed? 03:15:58 At least I seem to recall requests that they not be added 03:16:40 -!- blackcustard has quit [Read error: Connection reset by peer] 03:17:44 -!- rast has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 03:19:30 oh right 03:19:39 i don't remember that 03:19:59 -!- substitute has quit [Ping timeout: 245 seconds] 03:20:01 someone should have added a comment about it 03:20:02 when i traced back to where the code was, it looked completely accidental that they were removed 03:20:26 You don't remember that? I suppose I could be misremembering 03:20:49 I don't mean that some other dev removed them, but that when the layout was still in flux and getting feedback, it was suggested they were a bad idea 03:20:58 And then they didn't show up in the final version and everyone assumed it was intentional 03:22:23 -!- SkaryMonk has joined ##crawl-dev 03:22:46 %git 9abe3f94dc7529080b5da34d2ba915ed68990120 03:22:47 03mumra * 0.12-a0-2271-g9abe3f9: Hyper - refactor, tidy, implement a couple TODOs 10(5 weeks ago, 3 files, 61+ 76-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=9abe3f94dc75 03:22:58 that's the commit where they were removed 03:23:33 it looks like i simply meant to refactor the decorator code, not remove them 03:25:36 The problem with windows is that it tends to interfere with autoexplore, and the stuff on the other side is often not immediatly reachable anyway 03:25:42 The serial vault that does this is much maligned 03:26:12 well vaults windows are almost always right next to a door aren't they? 03:26:20 Are they? 03:26:26 The ones I saw were not, but it was an old version 03:27:28 they're usually in the same wall as a door 03:27:48 but they're not guaranteed to be right next to it (although the walls aren't that big, so they'll always be near it) 03:28:13 i had one implementation when they were either side of the door but it's just random here 03:28:48 well i can easily disable them again now, or should i wait for feedback? 03:29:11 Admittedly I haven't looked (still futzing with monster gen stuff) 03:29:36 But I was just recalling the original conversation (or at least trying to) 03:29:54 May as well wait for feedback, I guess, since there are other layout changes here 03:30:33 Also, I MAY have found the source of this bug 03:30:56 -!- SkaryMonk has quit [Ping timeout: 246 seconds] 03:30:57 The generation ranges for Vaults seem to have been calculated as though V:1 was absolute depth 16, instead of 21 (which it is) 03:31:13 -!- SkaryMonk has joined ##crawl-dev 03:31:48 Meaning that every floor has effectively become 5 shallower 03:32:12 nobody said anything about the windows in either of the mantis tickets; so if anything was said on irc i forgot about it (but there was a lot going on at that point and everything was in constant flux!) 03:32:24 Yes, it was on irc 03:32:58 logs or it didn't happen 03:33:03 hehe 03:33:12 it's a simple matter to disable them again anywayu 03:33:13 Might have been in ##crawl =/ 03:33:21 When people were termcasting it 03:34:12 yep 03:34:20 i'll get some new feedback off elliott anyway 03:34:20 I was watching, too! :P 03:34:44 I will pull and rebuild and have a look now 03:34:48 gtg now anyways ; goodnight folks 03:34:51 Goodnight! 03:35:16 !tell kilobyte I think you accidentally made Vaults generate monsters 5 levels shallower than before the merge 03:35:16 DracoOmega: OK, I'll let kilobyte know. 03:36:32 -!- dcssrubot555 has quit [Read error: Connection reset by peer] 03:37:56 oh, and thanks for joining up 03:39:59 -!- mumra has quit [Ping timeout: 256 seconds] 03:40:35 -!- Ashenden_ has quit [Quit: Leaving] 03:41:03 Oh dear... I am getting lua errors and monsters in walls in Vaults now 03:41:22 Lots of monsters in walls, in fact 03:41:29 Some in empty space, even 03:42:46 Wow, the positioning of these here looks like a vault placed monsters, then the vault was erased or something 03:42:50 And the monsters are all neatly arranged in the void 03:57:02 !tell mumra I think you broke things. Vaults is giving me lua errors and regularly placing monsters and items in walls or in the middle of nowhere (see: http://imgur.com/HydrZHj which looks like a vault was placed and then erased or something) 03:57:02 DracoOmega: OK, I'll let mumra know. 04:06:57 -!- ontoclasm has quit [Quit: Leaving.] 04:14:17 -!- yogaFLAME has quit [] 04:14:30 -!- alefury|2 has joined ##crawl-dev 04:18:48 -!- Melum has quit [Ping timeout: 252 seconds] 04:25:53 -!- jeanjacques has joined ##crawl-dev 04:31:31 -!- DracoOmega has quit [Quit: Leaving] 04:36:53 -!- bmfx has quit [Read error: Operation timed out] 04:40:57 -!- Kalir has quit [Quit: CADENZA-CLASS CLOCKWORK KNIGHT--OPERATION TERMINATED] 04:49:39 -!- mumra has joined ##crawl-dev 04:49:54 -!- frtuip has quit [Client Quit] 04:55:30 -!- dtsund has left ##crawl-dev 04:56:37 !tell DracoOmega thanks, fixed 04:56:38 mumra: You have 1 message. Use !messages to read it. 04:56:38 mumra: OK, I'll let dracoomega know. 04:56:42 !messages 04:56:42 (1/1) DracoOmega said (59m 40s ago): I think you broke things. Vaults is giving me lua errors and regularly placing monsters and items in walls or in the middle of nowhere (see: http://imgur.com/HydrZHj which looks like a vault was placed and then erased or something) 04:57:43 -!- peepsalot has quit [Quit: Leaving] 05:00:26 03mumra 07* 0.12-a0-2972-g5d36b40: Fix Vaults Lua error 10(4 minutes ago, 1 file, 1+ 1-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=5d36b40055ec 05:01:35 testing is good! so good! :P 05:01:44 heh :) 05:02:06 i did test, just didn't see that really obvious lua error in the message window 05:02:18 not being tired when testing is good too :) 05:06:38 -!- dcssrubot623 has quit [Read error: Connection reset by peer] 05:08:06 Can I get a specific draconian colour in wiz mode? 05:08:13 -!- alefury|2 has quit [] 05:09:23 &Myellow draconian 05:09:24 -!- Vizer_ has quit [Quit: bye] 05:09:40 or did you mean as a player ? 05:10:52 yep 05:11:01 got lucky anyway, so don't need it anymore :P 05:11:08 ah cool 05:11:41 you could try whatever the command is to change your species and type foo draconian 05:11:56 or if that doesn't work keep changing to draconian until you got the right colour i guess 05:12:21 oh you can change species 05:12:22 thanks 05:14:46 -!- DainDwarf has quit [Changing host] 05:16:05 -!- Pepe[netbook] has quit [Ping timeout: 256 seconds] 05:31:28 -!- SkaryMonk1 has joined ##crawl-dev 05:32:00 -!- SkaryMonk has quit [Ping timeout: 257 seconds] 05:37:16 -!- minmay has quit [Quit: Leaving] 05:37:21 -!- minqmay has quit [Quit: Leaving] 05:37:33 -!- syraine has quit [Ping timeout: 245 seconds] 05:45:56 Losing a level that lead to gaining permanent flight will make it impossible to stop flying by Medar 05:47:18 -!- MaxFrost has quit [Ping timeout: 256 seconds] 05:56:58 -!- clinew has quit [Quit: leaving] 06:01:11 -!- absolutego has quit [Quit: leaving] 06:04:34 -!- GuraKKa has quit [Quit: Leaving.] 06:05:24 -!- GuraKKa has quit [Read error: Connection reset by peer] 06:06:12 -!- bonghitz has quit [Remote host closed the connection] 06:08:42 -!- ebarrett has quit [] 06:10:45 -!- TacoPrincess has quit [Quit: WeeChat 0.4.0] 06:19:03 -!- BoredOne has quit [Quit: Page closed] 06:25:34 -!- elliott_ has joined ##crawl-dev 06:25:37 -!- elliott has quit [Disconnected by services] 06:25:44 -!- elliott_ is now known as elliott 06:27:03 you aren't really a dev until you've broken things. Keep it up mumra :) 06:27:03 galehar: You have 1 message. Use !messages to read it. 06:27:08 !messages 06:27:08 (1/1) |amethyst said (8h 51m 38s ago): Just changed all the foo:quote to foo:quote to fix #6805. I didn't see any occurrences in the translations, but you might want to verify that (and check your scripts) 06:27:29 !seen amethyst 06:27:30 Sorry galehar, I haven't seen amethyst. 06:27:38 oh right, the sneaky one 06:27:41 do you get combo points for removing something and breaking the game in one commit? 06:27:43 galehar: You should have dev achievements 06:27:49 I think !seen |amethyst works now 06:27:51 !seen |amethyst 06:27:51 I last saw |amethyst at Fri Mar 22 03:47:20 2013 UTC (7h 40m 31s ago) saying 'I was trying to think of ideas for a Barthes roguelike (it is s-z.org after all) but all I could come up with was "The van hits! You die."' on ##crawl-dev. 06:27:57 galehar: thanks! 06:28:07 elliott: thanks 06:28:35 what about tell? Does it works with the | now? 06:29:17 I believe so 06:30:03 -!- syraine has quit [Ping timeout: 245 seconds] 06:30:06 !tell |amethyst thanks for fixing links. I'll remove them from transifex anyway or they get translated which only serves to break the link. 06:30:07 galehar: OK, I'll let |amethyst know. 06:36:43 -!- dcssrubot996 has quit [Read error: Connection reset by peer] 06:37:32 -!- edlothiol has joined ##crawl-dev 06:39:46 -!- yogaFLAME has joined ##crawl-dev 06:45:02 -!- bonghitz has quit [Ping timeout: 255 seconds] 06:49:21 kober (L24 DsFi) (D:21) 06:52:26 -!- PsyMar has quit [Read error: Connection reset by peer] 06:53:08 -!- yogaFLAME has quit [Ping timeout: 255 seconds] 06:58:18 -!- dcssrubot599 has quit [Ping timeout: 264 seconds] 06:58:35 -!- s951 has joined ##crawl-dev 07:00:58 -!- ystael has quit [Ping timeout: 245 seconds] 07:02:51 -!- MarvinPA has joined ##crawl-dev 07:05:08 -!- Cerepol has quit [Ping timeout: 246 seconds] 07:07:18 -!- Blazinghand has quit [Ping timeout: 276 seconds] 07:10:04 -!- Gastrox has quit [Quit: Page closed] 07:14:20 -!- edlothiol has quit [Ping timeout: 260 seconds] 07:15:54 -!- bonghitz has quit [Ping timeout: 252 seconds] 07:31:02 -!- SkaryMonk1 has quit [Ping timeout: 256 seconds] 07:31:35 -!- SkaryMonk has joined ##crawl-dev 07:32:54 -!- magicpoints has quit [Ping timeout: 245 seconds] 07:33:39 !messages 07:33:40 (1/1) DracoOmega said (3h 58m 23s ago): I think you accidentally made Vaults generate monsters 5 levels shallower than before the merge 07:33:58 !tell DracoOmega I see my error: the data was generated before Vault's absdepth was adjusted from 16 to 19. I don't see that 21 you're talking about, though -- do I miss something? 07:33:59 kilobyte: OK, I'll let dracoomega know. 07:38:22 !tell bh huh, the lightningform -> lightningForm thing stopped happening for me so i assumed somebody had fixed it 07:38:23 MarvinPA: OK, I'll let bh know. 07:38:51 -!- syraine has quit [Quit: Page closed] 07:39:19 !tell bh it looks like "Form" is the up-to-date branch now, so could you delete "form"? it looks like msysgit can't tell the difference between the two at all, so if i do anything it will probably just break things further :P 07:39:19 MarvinPA: OK, I'll let bh know. 07:40:59 !tell bh oh, capitalised Form is the up-to-date one after all, oops 07:40:59 for extra screw-up, name a branch "ð–šð–“ð–Žð–ˆð–”ð–‰ð–Š" (as the "unicode" branch used to be named), it kills msysgit dead 07:41:00 MarvinPA: OK, I'll let bh know. 07:41:13 wow i mistyped again 07:41:15 poor bh 07:42:31 !tell bh hi please ignore all previous messages, i am terrible and cannot think or type apparently. "lightningform" is the working, compilable branch. :( 07:42:31 MarvinPA: OK, I'll let bh know. 07:42:36 Windows has some support for unicode > 16 bit, but it's mostly lip service, resulting in all kinds of fun (it uses UTF-16 with surrogates internally) 07:43:07 MarvinPA: perhaps name it lightning_form to avoid problems with deleting? 07:44:27 -!- yogaFLAME has joined ##crawl-dev 07:45:18 the problem now seems to be that i can't even check out the lowercase version of the branch, though :( 07:46:46 can you rename remote branches? i feel like the chances of me just messing up git things here is dangerously high :P 07:46:48 on Windows or somewhere else? 07:46:53 on windows, yeah 07:47:11 rename no, push then delete yes 07:48:43 the top commit of lightningForm is from Mar 9, lightingform Mar 16, is the former indeed more recent? 07:48:57 doh, a merge commit only 07:49:30 yeah, the original problem that started all of this was that i did that merge commit because the branch wasn't compilable before 07:49:48 and when i pushed it, i accidentally pushed to lowercase lightningform 07:50:02 dactions.cc: In function ‘void apply_daction_to_mons(monster*, daction_type, bool)’: 07:50:04 dactions.cc:160:39: error: ‘mi’ was not declared in this scope dprf("going hostile: %s", mi->name(DESC_PLAIN, true).c_str()); 07:56:14 -!- yogaFLAME has quit [Remote host closed the connection] 08:14:52 !tell bh I renamed the branch to lightning_form, to work around name conflicts on Windows, and fixed a few bugs (didn't build in debug mode, was based on a broken master, tiles showed an untransformed player instead). I hope this won't break whatever extra code you have on top of it. 08:14:52 kilobyte: OK, I'll let bh know. 08:16:51 -!- bonghitz has quit [Ping timeout: 276 seconds] 08:16:56 -!- jeanjacques_ has joined ##crawl-dev 08:16:57 nice, thanks 08:17:32 New branch created: lighting_form (4 commits) 08:17:33 03bh 07[lighting_form] * 0.12-a0-2964-g89e5fea: Electric form 10(10 weeks ago, 10 files, 87+ 2-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=89e5feaebfd3 08:17:33 03bh 07[lighting_form] * 0.12-a0-2965-g3ce7c39: Blinkbolt 10(13 days ago, 10 files, 77+ 0-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=3ce7c39c3631 08:17:33 03bh 07[lighting_form] * 0.12-a0-2966-gff59c3d: Limit Blinkbolt 10(13 days ago, 3 files, 11+ 6-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=ff59c3d1fb63 08:17:33 03kilobyte 07[lighting_form] * 0.12-a0-2967-g87d305f: Use the electric golem tile for electric form for now. 10(7 minutes ago, 1 file, 1+ 0-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=87d305f7e52d 08:18:57 -!- jeanjacques has quit [Ping timeout: 248 seconds] 08:18:57 -!- jeanjacques_ is now known as jeanjacques 08:25:52 -!- Zermako has quit [] 08:27:08 -!- dcssrubot512 has quit [Read error: Connection reset by peer] 08:36:29 -!- edlothiol has joined ##crawl-dev 08:37:00 -!- kek has joined ##crawl-dev 08:37:17 -!- BoredOne has quit [Quit: Page closed] 08:39:40 -!- Melum has quit [Quit: Was eaten by a grue.] 08:42:29 -!- ussdefiant has joined ##crawl-dev 08:47:16 -!- Taco_Princess has quit [Quit: Leaving] 08:50:33 -!- edlothiol has quit [Ping timeout: 245 seconds] 08:52:36 Mac daily builds are still supposed to be done by greatzebu, right? 08:52:52 they're apparently stuck since 2013-01-07 09:03:26 -!- Wehk has joined ##crawl-dev 09:06:41 -!- mikee_ has joined ##crawl-dev 09:08:12 -!- st_ has joined ##crawl-dev 09:27:40 -!- inspector071 has quit [Read error: Operation timed out] 09:27:58 -!- Datul has quit [Ping timeout: 245 seconds] 09:30:51 -!- SkaryMonk has quit [Ping timeout: 260 seconds] 09:30:53 -!- SkaryMonk1 has joined ##crawl-dev 09:31:15 -!- dagonfive has quit [Read error: Connection reset by peer] 09:33:18 -!- xxx has quit [Quit: Page closed] 09:34:24 -!- ussdefiant has quit [Ping timeout: 264 seconds] 09:40:40 -!- yogaFLAME has joined ##crawl-dev 09:57:14 -!- dcssrubot737 has quit [Read error: Connection reset by peer] 10:00:16 Q: should a player moving between Pan levels ever land in a vault? 10:00:40 It looks like this was, at some point, not supposed to happen, but that checking was disabled by some unrelated bugfixes. 10:00:50 what if the level is an encompass vault? 10:00:53 think of hilarious hellion isle deaths 10:01:03 i think it's fine as-is, yeah 10:01:18 I like it because it is good for taking a pan portal to be dangerous 10:01:20 also it's sometimes helpful 10:01:25 i landed in hellion island and survived by the skin of my teeth, it was awesome, i would hate for other people to never have that experience 10:01:27 like if you end up near the rune you can just pick it up and teleport 10:01:34 i like this development philosophy - "we should keep it because it's hilarious" =P 10:03:31 -!- nht has quit [Quit: leaving] 10:05:54 -!- nooodl has joined ##crawl-dev 10:13:54 Pan levels change tiles after saving by KiloByte 10:16:56 -!- s951 has quit [Quit: leaving] 10:17:25 Grunt: landing in vaults is good 10:18:07 I don't really care one way or the other about hellion island, but landing in the middle of one of the unique pan lord vaults is quite good 10:18:14 I'm mainly asking because there's still a comment indicating the original behaviour in dungeon.cc. 10:18:34 I'll just remove said comment if we currently intend that behaviour. 10:22:42 the hellion isle tends to kill most characters in 1-2 rounds (unless you're on the edge) so this is the one vault that deserves some thinking 10:22:54 pan lord vaults are not that hard to at least teleport from 10:24:01 -!- DainDwarf has quit [Quit: Dain's default quit message] 10:27:21 one funny thing about hellion island was that it was used as an example to explain no_rtele_into, but doesn't actually have no_rtele_into 10:27:45 i dont know whether the documentation got changed or if that is still a thing 10:29:04 03Grunt 07* 0.12-a0-2973-g8f37593: Replace an outdated comment. 10(9 minutes ago, 1 file, 2+ 2-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=8f375938078b 10:29:04 03MarvinPA 07* 0.12-a0-2974-g8d78cb9: Pan vault tweaks 10(8 minutes ago, 1 file, 18+ 21-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=8d78cb9d2f3c 10:46:32 -!- Kellhus has quit [Quit: Page closed] 10:47:00 -!- Wehk_ has joined ##crawl-dev 10:49:35 -!- MaxFrost has quit [Ping timeout: 260 seconds] 10:50:02 -!- Wehk has quit [Ping timeout: 272 seconds] 10:54:22 03MarvinPA 07* 0.12-a0-2975-gfb53cde: Update the options guide for rod autopickup 10(5 minutes ago, 1 file, 2+ 1-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=fb53cdece5ea 10:54:45 books are on by default, too 10:54:57 er 10:55:03 i did an amethyst 10:55:06 * Grunt applauds MarvinPA pulling an |amethyst :b 10:55:20 :P 10:56:29 maybe it was important enough to make sure I see it twice? 10:57:26 ??? 11:00:28 -!- madreisz has quit [Ping timeout: 245 seconds] 11:02:20 -!- ystael has quit [Ping timeout: 256 seconds] 11:04:49 -!- edlothiol has joined ##crawl-dev 11:05:48 -!- rast has joined ##crawl-dev 11:06:47 i cant seem to find the CDO putty key 11:07:00 http://crawl.develz.org/wordpress/howto links to the CAO one instead... 11:07:26 they use the same key 11:08:35 it says Server refused our key 11:08:38 under putty 11:10:19 hm 11:10:28 you need to use the putty key 11:10:28 ??cdo 11:10:29 cdo[1/4]: Crawl server (also running development versions), located in Germany, crawl.develz.org, telnet port 345 or ssh port 22, ssh-username: crawl, ssh-key necessary: http://crawl.develz.org/cao_key (openssh) or http://crawl.develz.org/cao_key.ppk (putty) 11:10:33 are you sure you used the putty format key? 11:10:54 -!- elliptic has quit [Quit: Leaving] 11:14:37 -!- meowfelid has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 11:14:59 yeah 11:17:44 -!- SkaryMonk1 has quit [Quit: Leaving.] 11:27:19 -!- dcssrubot919 has quit [Read error: Connection reset by peer] 11:28:05 -!- Nakatomy has quit [Ping timeout: 255 seconds] 11:32:08 -!- tophat has quit [Ping timeout: 245 seconds] 11:32:30 -!- yogaFLAME has quit [Ping timeout: 276 seconds] 11:33:48 -!- inspector071 has quit [Ping timeout: 245 seconds] 11:35:59 More Thrall poetry on mantis \o/ 11:38:06 -!- ontoclasm has joined ##crawl-dev 11:39:16 -!- yogaFLAME has joined ##crawl-dev 11:39:32 is thrall a dev yet 11:39:53 Alliance Throw auto pickup by Thrall 11:44:10 it's like a poem 11:44:27 <+Grunt> More Thrall poetry on mantis \o/ 11:44:42 great minds 11:44:48 >.> 11:47:33 -!- simmarine has joined ##crawl-dev 11:47:50 we established already, Grunt and ontoclasm are the same person 11:52:18 * ontoclasm flees in terror 11:53:15 Lemuel the Annihilator (L14 TeFi) ASSERT(smc) in 'mon-util.cc' at line 1190 failed on turn 2338. (D) 11:55:26 !lm Lemuel sprint crash -log 11:55:27 2. Lemuel, XL14 TeFi, T:2338 (milestone): http://crawl.akrasiac.org/rawdata/Lemuel/crash-Lemuel-20130322-165314.txt 11:55:55 oh wow, it really is some kind of poem 11:55:58 -!- myrmidette has quit [Ping timeout: 245 seconds] 11:56:28 should give it to some highschool students to interpret 11:56:34 see what they come up with 11:56:37 -!- simmarine has quit [Quit: Leaving] 11:57:15 -!- simmarine has joined ##crawl-dev 11:59:17 -!- Mandevil has quit [Quit: leaving] 12:04:58 get them to interpret it in the form of dance 12:07:25 -!- EvilGrin has quit [Quit: Leaving] 12:07:58 -!- SetecAstronomy has quit [Ping timeout: 245 seconds] 12:08:01 Stable (0.11) branch on crawl.akrasiac.org updated to: 0.11.2-11-g28dacbe 12:08:06 -!- rast has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 12:09:13 greatzebu says the server he has been doing nightly Mac builds has died 12:09:34 so I'm afraid something needs to be done to Mac trunk builds 12:09:53 too bad my mac is too old to be useful :/ 12:10:40 if all else fails, trunk build can be dropped, as obsolete trunk is worse than no trunk (as it doesn't receive bug fixes) 12:11:01 greatzebu has a laptop he can do manual builds, but that's no good for nightlies 12:12:34 zarath (L17 MiFi) ASSERT(smc) in 'mon-util.cc' at line 1190 failed. (D (Sprint)) 12:15:20 -!- rast has joined ##crawl-dev 12:18:15 zarath (L18 MiFi) ASSERT(smc) in 'mon-util.cc' at line 592 failed. (D (Sprint)) 12:22:02 Unstable branch on crawl.akrasiac.org updated to: 0.12-a0-2975-gfb53cde (34) 12:30:26 Mac nightlies are down by KiloByte 12:32:15 -!- yogaFLAME has quit [Ping timeout: 260 seconds] 12:40:32 03kilobyte 07* 0.12-a0-2976-g0e85153: Axe an oh-so-informative comment. 10(3 hours ago, 1 file, 0+ 1-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=0e851530634a 12:40:32 03kilobyte 07* 0.12-a0-2977-gbbb7fc7: Debucklerize Louise's large shield tile. 10(4 minutes ago, 1 file, 0+ 0-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=bbb7fc7714a6 12:42:14 -!- bmfx has quit [Ping timeout: 252 seconds] 12:44:01 zarath (L20 MiFi) ASSERT(smc) in 'mon-util.cc' at line 2575 failed. (D (Sprint)) 12:44:07 -!- yogaFLAME has joined ##crawl-dev 12:49:45 zarath (L20 MiFi) ASSERT(smc) in 'mon-util.cc' at line 2575 failed. (D (Sprint)) 12:50:33 -!- zarath has quit [] 12:50:47 !lg zarath crash -log 12:50:48 Malformed argument: crash 12:50:48 er 12:50:52 !lm zarath sprint crash -log 12:50:53 4. zarath, XL20 MiFi, T:3009 (milestone): http://dobrazupa.org/morgue/zarath/crash-zarath-20130322-174944.txt 12:51:00 -!- Sombrero_Mott has quit [Ping timeout: 272 seconds] 12:54:53 -!- Pepe[netbook] has quit [Read error: Connection reset by peer] 12:55:44 03Grunt 07* 0.12-a0-2978-gdf5c2b2: Vaultsvaultsvaultsvaultsvaults. 10(4 minutes ago, 4 files, 242+ 0-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=df5c2b20e0a9 12:56:23 -!- alefury has quit [Quit: ChatZilla 0.9.85 [Iceape 2.0.11/20121208085021]] 12:57:24 -!- dcssrubot878 has quit [Read error: Connection reset by peer] 12:57:29 -!- rast has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 13:01:28 -!- Kalir has quit [Changing host] 13:05:05 -!- xxx has quit [Quit: Page closed] 13:05:10 -!- mumra has quit [Ping timeout: 252 seconds] 13:09:35 -!- yogaFLAME has quit [Ping timeout: 260 seconds] 13:09:59 -!- Taco_Princess has quit [Ping timeout: 245 seconds] 13:13:23 -!- inspector071 has quit [Ping timeout: 245 seconds] 13:20:09 -!- yogaFLAME has joined ##crawl-dev 13:20:11 -!- Scherzo has quit [Quit: Page closed] 13:26:06 Donald rN infoleak? by Swiss 13:27:15 -!- Blazinghand has quit [Read error: Connection timed out] 13:47:14 -!- Aryth has quit [Ping timeout: 256 seconds] 13:47:30 -!- yogaFLAME has quit [Ping timeout: 256 seconds] 13:50:06 -!- yogaFLAME has joined ##crawl-dev 14:07:03 -!- Taco_Princess has quit [Changing host] 14:11:17 -!- _D_ has quit [Quit: leaving] 14:12:33 -!- inspector071 has quit [Ping timeout: 245 seconds] 14:15:54 Train (L27 HuWz) ASSERT(feat > DNGN_UNSEEN) in 'dbg-scan.cc' at line 583 failed. (Abyss:3) 14:16:35 -!- yogaFLAME has quit [Ping timeout: 252 seconds] 14:17:58 -!- Datul has quit [Ping timeout: 245 seconds] 14:20:25 does autoexplore really need to stop when "you sense a monster nearby" but it's just one of your own dudes? 14:22:46 -!- yogaFLAME has joined ##crawl-dev 14:25:54 Okay, let's have a look-see... 14:26:46 03dolorous 07* 0.12-a0-2979-g9b96b1d: Fix Lua error on startup: mini_monsters.des needs init_hm_decor_walldepth. 10(7 minutes ago, 1 file, 12+ 0-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=9b96b1da7aab 14:26:46 03dolorous 07* 0.12-a0-2980-ga9f55e5: Synchronize a comment. 10(5 minutes ago, 1 file, 1+ 1-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=a9f55e5109dc 14:26:48 Orc:4 is unreachable without taking hatches (if so) by absolutego 14:27:30 -!- dcssrubot829 has quit [Read error: Connection reset by peer] 14:28:43 -!- kek has quit [Quit: Kicked by KickServ] 14:30:43 I mean, we don't actually seem to reveal if the detected monster is friendly or not ... 14:30:52 Tenaya (L14 DsAK) ASSERT(_valid()) in 'ray.cc' at line 194 failed. (WizLab) 14:30:56 ... but it's kind of annoying 14:31:41 i thought that changed a little while ago and you can distinguish between friendly and hostile detected monsters now? 14:32:14 i think friendly detected monsters used to interrupt running and autotravel every turn or something and then it was fixed, i don't recall exactly though 14:36:10 I remember friendly detected monsters showing up in a different colour than hostile detected monsters. 14:36:16 really? 14:37:26 x doesn't give me any info on detected monsters so it's not obvious to me, since mostly I only detect my orcs 14:37:33 they show up as a green glyph, yeah 14:37:57 oh but not with antennae 14:38:02 ah 14:38:04 maybe the fix only applied to ash? 14:39:19 -!- dcssrubot727 has quit [Read error: Connection reset by peer] 14:41:58 Crash in Wizard's Laboratory after Corrupt by Tenaya 14:43:21 ok, finally some time 14:43:32 what's left to do for mumra? 14:43:47 funny thing is, I don't see anything about this in the code 14:44:35 oh, wait 14:45:24 friendly sensed monsters have their own monster_type for some reason 14:45:47 -!- blabber has joined ##crawl-dev 14:48:10 -!- yogaFLAME has quit [Ping timeout: 260 seconds] 14:50:30 so, is this *supposed* to be just for Ash worshipers, or should it also work for those with antennae? 14:52:42 -!- Yllodra has quit [Ping timeout: 276 seconds] 14:53:35 seems like it should work with antennae too, since it's just meant to reduce annoyance 14:54:35 -!- Flun has quit [Quit: Page closed] 14:56:05 -!- yogaFLAME has joined ##crawl-dev 15:00:17 -!- dtsund has joined ##crawl-dev 15:02:36 -!- rchandra has quit [Ping timeout: 256 seconds] 15:02:52 -!- Nightbeer has quit [Quit: Verlassend] 15:04:11 should I go for if (mon->friendly()) ... else if (you.religion ...) ..., or if (you.religion ...) ... if (mon->friendly()) ... ? 15:06:50 -!- Taco_Princess has quit [Read error: Connection reset by peer] 15:07:05 with the former, I need to reformat the Ash condition onto two lines and add braces to that body ... 15:10:30 -!- rchandra1 is now known as rchandra 15:11:41 -!- GuraKKa has quit [Read error: Connection reset by peer] 15:12:26 -!- MarvinPA_ has joined ##crawl-dev 15:13:51 -!- Mindiell has quit [Quit: Quitte] 15:15:27 -!- MarvinPA has quit [Ping timeout: 276 seconds] 15:15:27 -!- MarvinPA_ is now known as MarvinPA 15:21:28 -!- absolutego has quit [Quit: leaving] 15:22:18 -!- yogaFLAME has quit [Ping timeout: 264 seconds] 15:26:40 -!- yogaFLAME has joined ##crawl-dev 15:26:47 -!- simmarine has quit [Read error: Connection reset by peer] 15:30:06 Napkin: I think we needed you to improve is Mantis foo 15:30:07 galehar: You have 2 messages. Use !messages to read them. 15:30:10 !messages 15:30:11 (1/2) |amethyst said (7h 47m 38s ago): what if some other language wants to have a real translation where English has just a link? Maybe there's some Lithuanian folk tale about a swamp dragon hide... 15:30:24 !messages 15:30:25 (1/1) |amethyst said (7h 47m 31s ago): err, a real quote rather 15:31:03 seen |amethyst 15:31:07 !seen |amethyst 15:31:08 I last saw |amethyst at Fri Mar 22 12:50:14 2013 UTC (7h 40m 54s ago) saying '!tell BoredOne *your' on ##crawl. 15:31:50 !tell |amethyst I'm implementing another way for contributers to submit new quotes. Will document when finished 15:31:51 galehar: OK, I'll let |amethyst know. 15:32:45 !tell mumra Idea re D:27 of the moment: http://sprunge.us/QYPN (the second part being more relevant) 15:32:46 Grunt: OK, I'll let mumra know. 15:36:59 -!- simmarine has joined ##crawl-dev 15:37:01 -!- bonghitz has quit [Remote host closed the connection] 15:37:34 -!- Vizer has joined ##crawl-dev 15:38:14 -!- magicpoints has quit [Ping timeout: 246 seconds] 15:43:06 -!- blackcustard has joined ##crawl-dev 15:44:07 -!- Taco_Princess has quit [Read error: Connection reset by peer] 15:46:10 -!- Pthing has quit [Read error: Connection reset by peer] 15:46:49 -!- Psyknux has quit [Quit: My work is end \(^o^)/] 15:47:44 -!- edlothiol has quit [Remote host closed the connection] 15:48:14 -!- Chozo has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 15:51:40 -!- yogaFLAME has quit [Ping timeout: 245 seconds] 15:54:39 -!- yogaFLAME has joined ##crawl-dev 15:59:03 -!- Krakhan has quit [Changing host] 15:59:06 -!- Nightbeer has quit [Client Quit] 16:01:37 -!- ToastyP has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 16:02:15 Blar, I'm having my traditional "can't compile for Windows anymore after an OS and/or hardware change" issues 16:03:21 I can compile for Linux just fine, but attempting to use mingw32 to compile for WIndows gives: make: *** No rule to make target `contrib/install/include/pcre.h', needed by `glwrapper.o'. Stop. 16:04:56 -!- DracoOmega has joined ##crawl-dev 16:06:17 dtsund: did you update the submodules? 16:06:27 Yeah. 16:06:43 ...hm, git reset --hard fixed the problem. I don't know why it did, but it did. 16:07:56 !tell kilobyte Yes, oops. 3 levels. I guess I got the 21 from looking at V:2 (which is depth 21) 16:07:57 DracoOmega: OK, I'll let kilobyte know. 16:08:05 -!- nht has quit [Quit: leaving] 16:08:21 Though those 3 levels seem to make a very dramatic difference to monster composition at this point 16:08:48 Though I believe I have adjusted things to match the original intention now, anyway 16:12:31 -!- Taco_Princess has quit [Ping timeout: 245 seconds] 16:12:55 03DracoOmega 07* 0.12-a0-2981-g5319588: Properly rescale Vaults monster depths to the new system 10(11 minutes ago, 1 file, 106+ 126-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=5319588b3908 16:13:02 Bah, it only seems to have fixed it for console. 16:13:20 Tiles is complaining about missing SDL.h 16:13:49 make clean? 16:15:35 -!- spriseris has quit [Ping timeout: 272 seconds] 16:15:47 -!- dcssrubot416 has quit [Read error: Connection reset by peer] 16:16:06 No dice. 16:16:51 so, you really haven't got an SDL.h in there? 16:16:56 Nope. 16:17:24 What's weird is that I copied my build directory directly from a machine where it did work 16:17:37 maybe you messed up the submodule dance? 16:18:11 This is entirely possible 16:18:38 -!- anidude has quit [Read error: Connection reset by peer] 16:20:32 -!- ark__ has joined ##crawl-dev 16:22:33 -!- inspector071 has quit [Ping timeout: 245 seconds] 16:22:45 Hm, I do have an SDL.h in contrib/sdl/include/, perhaps copying it to contrib/install/include/SDL/ will make it work 16:22:56 This plan cannot possibly fail 16:25:27 ...nope. 16:26:08 -!- Serfuzz has quit [Remote host closed the connection] 16:27:02 dtsund: do you have a contrib/install ? 16:27:10 I mean, did you? 16:27:15 maybe you should just delete that ... 16:28:12 I do 16:29:21 Doing that makes console builds fail too. 16:29:33 (Yes, I did do a git submodule update) 16:29:34 -!- Namey has quit [Quit: Leaving] 16:31:39 -!- ApsychicRat has quit [Client Quit] 16:32:03 -!- Mrmini231 has quit [Quit: Page closed] 16:39:25 oh, make clean-contrib ? 16:39:31 -!- Sombrero_Mott has joined ##crawl-dev 16:40:46 -!- Sorbius_ has quit [Read error: Connection reset by peer] 16:41:38 Nope. 16:42:09 -!- ark__ has quit [Quit: Leaving] 16:42:15 -!- substitute has quit [Ping timeout: 240 seconds] 16:42:26 * dtsund downloading a fresh pull 16:45:22 03SamB 07* 0.12-a0-2982-g3b129d7: Don't interrupt autoexplore for friendly detected monsters, even w/o Ash 10(14 minutes ago, 2 files, 10+ 4-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=3b129d7c2e51 16:45:59 -!- ZChris13 has joined ##crawl-dev 16:55:34 Nope, a clean pull misses pcre.h. 16:55:41 -!- Cethron has quit [] 16:59:50 huh 17:04:13 Hm. I think what I wound up doing last time was downloading the Stone Soup stable tarball and just jamming contribs/ from that into place. 17:07:51 you do remember the submodule dance is needed after a fresh clone, right? 17:10:27 What is the dance, aside from git submodule update? 17:11:01 I can never remember, which is why I call it the dance 17:11:24 * dtsund curses this build system 17:12:01 rats, you're not a dev so I can't count that as a vote for autoconf 17:12:10 -!- clouded_ has joined ##crawl-dev 17:13:37 I think I may have needed "git submodule update --init" last time I did it? 17:14:16 Hm, that actually seems to be doing something 17:14:54 -!- st_ has quit [Ping timeout: 256 seconds] 17:15:03 -!- SetecAstronomy has quit [Ping timeout: 245 seconds] 17:15:11 -!- santiago_ has quit [Ping timeout: 246 seconds] 17:16:20 -!- clouded_ is now known as st_ 17:17:35 -!- SaintWacko has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/] 17:25:31 -!- Palyth has quit [Remote host closed the connection] 17:27:59 -!- SkaryMonk has joined ##crawl-dev 17:29:39 -!- rast has joined ##crawl-dev 17:30:36 -!- Arkaniad has quit [Ping timeout: 260 seconds] 17:31:26 -!- Staplegun has quit [Read error: Connection reset by peer] 17:34:13 -!- SetecAstronomy has quit [Ping timeout: 245 seconds] 17:39:08 -!- Vizer has quit [Quit: bye] 17:40:09 -!- Vizer has joined ##crawl-dev 17:40:55 -!- Vidiny has quit [Quit: [19:23] don't they have refill machines in ikea [19:24] bet those muricans took their vases to those straight away after paying] 17:42:48 Finally seems to be working. 17:45:53 -!- dcssrubot252 has quit [Read error: Connection reset by peer] 17:50:08 -!- Cerepol has quit [Ping timeout: 255 seconds] 17:55:43 -!- yogaFLAME has quit [] 17:57:17 -!- eeviac has joined ##crawl-dev 17:57:48 -!- ontoclasm has quit [Quit: Leaving.] 17:58:26 -!- inpho has quit [Quit: inpho] 18:00:09 -!- SamB_ has joined ##crawl-dev 18:00:40 -!- the_glow1 has quit [Ping timeout: 260 seconds] 18:01:48 -!- SamB has quit [Read error: Connection reset by peer] 18:10:23 -!- mumra has joined ##crawl-dev 18:10:41 -!- absolutego has quit [Quit: leaving] 18:10:52 -!- Scherzo has quit [Quit: Page closed] 18:10:57 mumra: So, I had been thinking to try porting some of the inner loops of the Vaults generator to C and see how it ran, but then it occured that this might be wasteful when a new version is coming down the pipe and I don't know what parts of that are rewritten :P 18:12:40 ah 18:12:41 mumra: You have 1 message. Use !messages to read it. 18:12:45 !messages 18:12:45 (1/1) Grunt said (2h 39m 59s ago): Idea re D:27 of the moment: http://sprunge.us/QYPN (the second part being more relevant) 18:13:33 -!- dtsund has quit [Ping timeout: 276 seconds] 18:14:53 DracoOmega: the problem is that significant portions of the inner loop are now modular callbacks; so a layout has more flexibility to choose what are acceptable places to place draw rooms 18:15:14 -!- edlothiol has joined ##crawl-dev 18:15:44 hmm, you can probably see this on my fork 18:15:47 More complicated than 'fits here'? (This is what it currently does basically, yes?) 18:15:55 yes 18:16:15 Can you give me a quick example or two of what other sort of parameters you'd use to determine placement validity? 18:16:21 for instance one placement strategy is "place this anywhere you like" (this is typically used for the first room that is placed) 18:17:06 so ... sometimes rooms can be placed inside rock -- carving rooms into existing solid terrain 18:17:14 sometimes rooms can be placed in open areas 18:17:18 -!- Blazinghand is now known as | 18:17:22 i.e. on existing floor space 18:17:24 -!- | is now known as Blazinghand 18:17:43 in current Vaults you see both of these strategies side by side 18:18:03 but the new version allows you to go "here i will only be placing rooms in open space" 18:18:05 Hmmm... well 'open space is okay' and 'rock is okay' seem like they could be flags or booleans or something, without needing callbacks. (I realize callbacks are more flexible, but I am wondering how much this flexibility is used) 18:18:21 this actually makes the algorithm more efficient because you're eliminating potential coords where rooms can be placed 18:19:05 (note : i've just been for a few drinks and come back to the office to pick up my laptop so i'm not entirely fit to explain everything!) 18:19:11 Haha 18:19:13 Fair enough :P 18:19:40 I have been generating maps and running stats and such 18:20:22 another placement strategy is : subdivide the empty space into a random grid, then place a room anywhere you like in each grid cell 18:20:38 if anything, the thing to optimise might be actually rendering the rooms / vaults 18:21:13 -!- jeanjacques has quit [Quit: I love my HydraIRC -> http://www.hydrairc.com <-] 18:21:16 i need to actually figure out which parts of the code are slowest 18:21:17 Well, if reducing number of room attempts cut time in half, that does point to room placement attempts being a large part of the overall runtime 18:21:30 multicolored troll hides by minmay 18:21:39 it didn't reduce the time anything like as much as i thought it should have 18:21:46 so that points to something else being slow 18:21:49 Oh, was it not actually half? 18:22:01 does lua have profiling 18:22:04 i need to perform proper whatever-its-called to work out which bits of code are slowest 18:22:08 that's the word, profiling ;) 18:22:31 do we need a profiling framework that can do Lua and C[++] at the same time? 18:22:33 DracoOmega: it reduced it by more than half. but that was by reducing the total iterations by way more than half! 18:22:35 Some of the layouts I've been getting seem unpleasantly sparse, but I wonder if sometimes ones looked like that anyway, and now I'm reading too much into it? 18:22:46 like the total iterations should be many times less but it still only reduced the time by half 18:22:52 Wish there was a way to do ABX testing here >.> 18:23:47 some of the new layouts will be inherently faster - doing a grid subdivide then placing rooms in the grid cells you don't need to poll random placements at all, you can work out exactly where to place a room first time 18:24:01 so i'm thinking the best way for V is to adapt it like that instead 18:24:15 since i'm now seeing lots of ways that grids can be usedf 18:24:17 I'm not sure I like the sounds of that, appearance-wise 18:24:25 -!- Nightbeer has quit [Ping timeout: 258 seconds] 18:24:25 -!- us17 has quit [Ping timeout: 258 seconds] 18:24:29 But perhaps it is more flexible than it sounds 18:24:33 it can look exactly the same 18:24:50 Well, the more grid-like ones in Vaults currently are generally less interesting, I think 18:24:57 because when subdividing the grid, you know what zones are next to each other, so you can place rooms adjoining -- 18:25:21 actually i think layouts can be more reliably interesting this way 18:25:37 but the current reductions do certainly make things look a bit sparse sometmies 18:25:44 -!- _dd has quit [Ping timeout: 256 seconds] 18:25:51 I actually would rather it take longer and look better, personally ^^; 18:25:58 If that is the tradeoff at hand 18:26:07 Perhaps others will disagree 18:26:24 well i agree but it depends how slow some people's computers are 18:26:28 well there's server load and stuff to consider? 18:26:39 well, maybe more reliably efficient algorithms will give more reliably interesting results, who knows 18:26:39 -!- edlothiol has quit [Ping timeout: 245 seconds] 18:26:40 like if it can take 5-10secs to generate a level on someone's slow computer then that's way too long 18:26:59 elliott: server load isn't an issue, it takes less than a second to generate a V level on s-z 18:27:10 and it's not like people are constantly in vaults and generating new levels there 18:27:14 elliott: Server load is irrelevant, I think, given that it takes about 2 seconds a level, at most 4 times a whole game (and few people make it that far) 18:27:17 s/vaults/Vaults 18:28:00 SamB_: it would be good to profile the Lua at least (i've been pointed to some stuff that can do that, although it's easy to do some fairly simple profiling of specific areas by hand) 18:28:09 mumra: Well, 5-10 seconds is slow in comparison to most levels, but it's also hardly a protracted loading time in most contexts for something that happens at most 4 times a game 18:28:41 Like, I'm all for improving this, but I wouldn't want to sacrifice too much to do so 18:28:44 10 seconds would seem like the game had crashed if it happened to me 18:29:02 the values can definitely be upped a bit more i think 18:29:07 Well, if you're on a computer slow enough to take that long, you're probably used to that 18:29:30 aww, -mapstat only works with DEBUG_DIAGNOSTIC 18:29:31 but actually it's only certain layouts that seem sparse, particularly the grid levels. the maze layouts look more interesting like this is my feeling. 18:29:58 nonsense, if the game crashed, you would have a shell prompt 18:30:06 Well, not 'crashed', but 'froze' 18:30:32 yeah don't hold me to specific words i'm having enough trouble typing :) 18:30:50 ;-P 18:30:52 And if that is the case, perhaps the room totals could be tweaked on a per-layout basis? 18:30:58 i'm not actually that drunk, just really tired too 18:31:12 I actually don't know the backend well enough to identify the base layout in a lot of cases, aside from a couple obvious ones 18:31:52 DracoOmega: they can already be tweaked easily on a per-layout basis, i just changed the global config, but each layout can pass in its own config table to override specific globals 18:32:04 does dgn even get notified which layout is called what? 18:32:14 (by your code, I mean) 18:32:23 SamB_: It shows up in the normal generation debug output, yes 18:32:30 I just hadn't been looking too closely :P 18:32:30 SamB_: ues 18:32:35 yes i mean 18:32:49 I didn't see where it set/added to the layout string 18:33:05 there aren't that many base layouts (and actually with the old settings a lot of them often ended up looking the same, they tend to be slightly more distinct now ...) 18:33:50 SamB_: 18:33:51 name = string.lower(name) 18:33:51 name = string.gsub(name," ","_") 18:33:51 e.layout_type(name) 18:33:58 the name is passed in from the layout 18:34:09 this is in v_layouts.lua 18:34:58 -!- ZChris13 has quit [Ping timeout: 252 seconds] 18:35:07 DracoOmega: further regarding optimisation -- currently it has a list of all potentially eligible coords. it picks one and tries to place the room there 18:35:23 the way this should work instead is first pick a random _room_ out of all the ones that have been placed 18:35:30 and then find a way to attach to it 18:35:54 Yes, this would probably help 18:35:57 that room could be an open space, even the initial open space that started off the level, because in the new version that's just considered as a room instead 18:36:20 whereas in trunk it's a separate code path that paints the initial floor 18:36:39 -!- Pacra___ has joined ##crawl-dev 18:36:44 -!- Pacra has quit [Read error: Connection reset by peer] 18:37:02 -!- Pacra___ is now known as Pacra 18:37:28 I am wondering about what of this makes sense for 0.12 and what should wait for 0.13 now. I mean, we were hoping to freeze before the end of this month, after all. 18:37:47 -!- eith has quit [Ping timeout: 246 seconds] 18:38:07 Which is not much more than a week at most here :P 18:41:11 ...I may have just set a new low in Lua black magic in this vault I'm working on right now. 18:44:51 -!- ldf has quit [Read error: Connection reset by peer] 18:49:14 -!- bonghitz has quit [Remote host closed the connection] 18:50:12 -!- Koolguydude has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 18:52:44 -!- Chozo has quit [Remote host closed the connection] 18:52:55 -!- rast has quit [Ping timeout: 245 seconds] 18:53:33 -!- rast has joined ##crawl-dev 18:53:58 -!- MarvinPA_ has joined ##crawl-dev 18:54:36 -!- ontoclasm has joined ##crawl-dev 18:54:53 -!- Cerepol has quit [Read error: Connection reset by peer] 18:57:59 -!- MarvinPA has quit [Ping timeout: 256 seconds] 18:57:59 -!- MarvinPA_ is now known as MarvinPA 18:58:41 -!- rast has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 19:01:21 -!- scummos has quit [Quit: Konversation terminated!] 19:04:29 -!- alefury has joined ##crawl-dev 19:08:17 http://sprunge.us/TKWh 19:08:30 (ignore the WEIGHT: <_<) 19:14:56 -!- Swagmaster has quit [Quit: Page closed] 19:15:57 -!- dcssrubot112 has quit [Read error: Connection reset by peer] 19:18:54 -!- Nightbeer has quit [Read error: Connection reset by peer] 19:20:58 don't do that please 19:21:07 people (and I) already hate the ranger ending 19:21:07 -!- blabber has quit [Remote host closed the connection] 19:23:11 -!- tren has quit [Quit: Leaving.] 19:23:33 is w still deep water and W shallow? 19:23:44 or was that (finally) swapped at some point? 19:24:04 there's no reason to swap it 19:24:06 It's still like that. 19:25:07 well, w being deep and W being shallow is really counterintuitive, which i think is a great reason to swap it 19:25:19 but i guess its too much work to be worth it :/ 19:26:42 so in that case the island thing is even worse than i first thought, because its deep water with land bridges and not shallow water 19:27:39 i still cant read what monsters are actually placed, though 19:28:16 ranger ending is terrible though 19:28:21 -!- blackcustard has quit [Ping timeout: 245 seconds] 19:31:07 I've been meaning to fix it 19:31:11 -!- inspector071 has quit [Ping timeout: 252 seconds] 19:31:52 the MarvinPA way? 19:32:27 no 19:32:30 :( 19:32:36 by removing all the crossbows 19:32:48 so just the centaurs are still ranged? 19:33:13 Gem (L14 DsFi) ASSERT(smc) in 'mon-util.cc' at line 2575 failed. (D (Sprint)) 19:33:16 tbh i dont mind the ranged orcs so much, what bothers me is that i have to get rid of the fucking crossbows so they dont make more ranged orcs 19:33:36 or rather, the ammo 19:34:12 the ranged orcs are kind of annoying, but i could tolerate them 19:34:59 -!- Yllodra has quit [Ping timeout: 245 seconds] 19:39:00 -!- DracoOmega has quit [Quit: Leaving] 19:46:46 -!- blackcustard has joined ##crawl-dev 19:50:25 * Grunt gets off the phone... 19:50:36 -!- kaiserfro has quit [Quit: Page closed] 19:51:29 The idea there goes something like this: there are four islands placed throughout the level, each with a shop on it. 19:52:04 One is a "boss" island and the other three are normal islands. Most of the time (3 chances in 4), each island has an orc knight and 5 orc warriors, and the boss island has an additional warlord and knight. 19:52:27 ...on that small chance, you get a small number of cyclopses and ettins instead, and the boss island has a stone giant. 19:52:38 (All of those generate normally in Orc, just on a small chance.) 19:52:58 ...by small number, I mean 1 ettin and 2 cyclopses on each island. 19:53:18 The normal version is probably more fair without the crossbows anyway <_< 19:53:30 (and it means I don't need to write out all the normal armour and weapon logic) 19:53:48 -!- Kellhus has quit [Ping timeout: 245 seconds] 19:53:56 -!- yogaFLAME has joined ##crawl-dev 19:57:06 I think I am going to rework the layout a bit to spread out the water a bit more so the stores are more directly accessible... 20:00:05 http://sprunge.us/UOYa 20:00:11 (that's almost legible now!) 20:02:33 -!- Zermako has quit [] 20:02:41 -!- faz has quit [Quit: Leaving] 20:10:32 -!- ebarrett has quit [] 20:17:01 I would drop the water and just have each shop have a lot of dudes next to it, and variety in knights/sorcerers/priests/trolls/ogres/giants 20:19:41 simpler, they can swarm you better, and most of the things in orc will not fight well in shallow water 20:22:47 -!- blackcustard has quit [Ping timeout: 256 seconds] 20:25:47 -!- blackcustard has joined ##crawl-dev 20:26:21 -!- mumra has quit [Remote host closed the connection] 20:26:43 -!- LordLovebone has quit [Ping timeout: 245 seconds] 20:29:30 -!- Duke- has quit [Remote host closed the connection] 20:29:48 -!- Scherzo has quit [Quit: Page closed] 20:34:25 -!- Kellhus has quit [Quit: Page closed] 20:35:56 -!- dtsund has joined ##crawl-dev 20:43:03 03kilobyte 07* 0.12-a0-2983-g09df719: Unbreak Vaults branch OODs. 10(51 minutes ago, 2 files, 2+ 2-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=09df719c0f71 20:43:03 03kilobyte 07* 0.12-a0-2984-g119402f: Simplify some Vaults monster distributions. 10(34 minutes ago, 1 file, 19+ 19-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=119402f64b01 20:43:03 03kilobyte 07* 0.12-a0-2985-g25201df: Go through normal item creation for hide drops. 10(19 minutes ago, 1 file, 16+ 24-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=25201dfa28d0 20:46:02 -!- dcssrubot633 has quit [Read error: Connection reset by peer] 20:49:06 -!- alefury has quit [] 20:49:41 ...and the other idea I had: http://sprunge.us/BKNE 20:49:42 Grunt: You have 1 message. Use !messages to read it. 20:50:05 the weight seems rather inflated 20:50:06 kilobyte: You have 1 message. Use !messages to read it. 20:50:11 kilobyte: ignore that :b 20:50:31 too small to be any reliable as a testing one, too :p 20:50:38 -!- bh has joined ##crawl-dev 20:50:42 hai 20:50:42 bh: You have 5 messages. Use !messages to read them. 20:50:47 oh god. I must have broken something 20:50:50 !messages 20:50:50 (1/5) MarvinPA said (13h 12m 28s ago): huh, the lightningform -> lightningForm thing stopped happening for me so i assumed somebody had fixed it 20:50:54 !messages 20:50:54 (1/4) MarvinPA said (13h 11m 35s ago): it looks like "Form" is the up-to-date branch now, so could you delete "form"? it looks like msysgit can't tell the difference between the two at all, so if i do anything it will probably just break things further :P 20:50:56 sorry, should use a proper letter :þ 20:51:05 s/use/have used/ 20:51:08 !messages 20:51:08 (1/3) MarvinPA said (13h 10m 9s ago): oh, capitalised Form is the up-to-date one after all, oops 20:51:11 !messages 20:51:11 * kilobyte meows at bh. 20:51:12 (1/2) MarvinPA said (13h 8m 40s ago): hi please ignore all previous messages, i am terrible and cannot think or type apparently. "lightningform" is the working, compilable branch. :( 20:51:12 st_: I don't really see the appeal of a plain floorvault (so to speak). 20:51:19 !messages 20:51:19 (1/1) kilobyte said (12h 36m 27s ago): I renamed the branch to lightning_form, to work around name conflicts on Windows, and fixed a few bugs (didn't build in debug mode, was based on a broken master, tiles showed an untransformed player instead). I hope this won't break whatever extra code you have on top of it. 20:51:33 bh: you didn't break anything important :) 20:51:35 kilobyte: I don't have any extra code 20:51:47 Grunt: I don't see the appeal of water for no reason 20:51:50 good or bad xom powers: xom pushes items away from you, xom pulls items toward you 20:53:26 the appeal of floorvault: simplicity is good, fitting in is good 20:53:27 I wish xom action selection were simpler. Instead of this big if-else tower, we could compute a chance for each action and then sample 20:53:33 ??floorvault 20:53:34 I don't have a page labeled floorvault in my learndb. 20:54:23 my favourite orc:4 vault is orc_open (or whatever I called it) 20:55:09 because it is purely "hey here is a big encounter have fun" 20:57:32 -!- yogaFLAME has quit [] 21:00:04 -!- WalkerBoh has quit [Remote host closed the connection] 21:00:13 -!- dcssrubot486 has quit [Read error: Connection reset by peer] 21:03:12 bh: why did you make imp/etc frequency go up, but ancient lich down, by abyssal depth? 21:03:22 seems reversed to me 21:03:41 UP, // linearly from near-zero to 100% at the top 21:03:41 DOWN, // linearly from 100% at the bottom to near-zero 21:04:23 ...the way that reads to me is that "UP" biases towards shallower levels, and "DOWN" biases towards deeper levels. 21:04:33 (This is the opposite of what I would expect given those names, but...) 21:05:39 in a range of 1-5, UP goes from 20% to 100%, DOWN from 100% to 20%, as one would expect 21:05:54 So it is the comment that reads funny to me. 21:05:56 I guess the word "top" is misleading here 21:08:28 end? 21:08:31 Perhaps something like "... to 100% increasing with depth". 21:10:11 -!- johnny0_ has joined ##crawl-dev 21:10:13 -!- Zephryn has quit [Read error: Connection reset by peer] 21:10:32 sounds better 21:12:31 -!- clinew has quit [Ping timeout: 245 seconds] 21:12:57 -!- johnny0 has quit [Ping timeout: 256 seconds] 21:13:24 03kilobyte 07* 0.12-a0-2986-g78e652b: Use start/end instead of top/bottom in the documentation of mon-pick distribution. 10(7 minutes ago, 1 file, 2+ 2-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=78e652b297eb 21:14:43 -!- ontoclasm has quit [Quit: Leaving.] 21:15:03 -!- SetecAstronomy has quit [Ping timeout: 245 seconds] 21:18:29 03kilobyte 07* 0.12-a0-2987-g713f552: Use even less ambiguous wording (Grunt) 10(8 minutes ago, 1 file, 1+ 1-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=713f5527d862 21:21:32 -!- HellTiger has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/] 21:21:33 -!- scummos has quit [Read error: Operation timed out] 21:27:01 hmm, seems the sensed monsters have descriptions even though 'x' won't show them 21:29:14 and that the one for "friendly sensed monster" isn't always correct anymore, since I made antennae distinguish those too 21:29:36 -!- Kalir has quit [Ping timeout: 245 seconds] 21:30:26 I did not know that fire crabs had a miniscule chance of spawning on Lair:8. 21:30:31 * Grunt goes to do up a lair ending... 21:33:15 -!- mumra has joined ##crawl-dev 21:33:37 -!- st_ has quit [Read error: Connection reset by peer] 21:34:00 -!- st_ has joined ##crawl-dev 21:34:24 kilobyte: yeah, what Grunt said 21:36:12 -!- Cerepol has quit [Ping timeout: 264 seconds] 21:44:29 -!- bh has quit [Quit: bbiab] 21:46:06 -!- blackcustard has quit [Ping timeout: 276 seconds] 21:47:56 -!- blackcustard has joined ##crawl-dev 21:50:01 -!- Wehk__ has joined ##crawl-dev 21:53:19 -!- Wehk_ has quit [Ping timeout: 258 seconds] 21:53:49 -!- SamB_ is now known as SamB 21:54:33 ... I think I've created a monstrosity. 21:55:06 -!- Kalir has quit [Changing host] 21:55:20 http://sprunge.us/egBU 21:55:40 -!- nooodl has quit [Ping timeout: 272 seconds] 21:57:21 * SamB wonders if we need a better license policy for images 21:58:37 -!- ussdefiant has joined ##crawl-dev 21:59:58 grunt why do you make vaults 22:00:02 is it just to annoy people 22:04:05 -!- Fhqwhgads_ has quit [Remote host closed the connection] 22:06:13 If by "annoy people" you mean something along the lines of "offering people some challenge and variety above and beyond killing the same monsters in the same basic layouts while in the same area", then yes. 22:07:20 -!- ystael has quit [Ping timeout: 260 seconds] 22:08:13 -!- guest23951 has joined ##crawl-dev 22:08:41 I would never ever clear this vault 22:09:13 -!- LordLovebone has quit [Ping timeout: 245 seconds] 22:09:56 If I say up front that it's "a monstrosity", that's generally a sign that I think it's terrible. :b 22:10:43 -!- guest23951 has left ##crawl-dev 22:13:54 -!- Luterac has quit [Quit: Page closed] 22:19:33 -!- Poncheis has quit [Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]] 22:20:42 -!- raskol has quit [Quit: Konversation terminated!] 22:23:15 -!- crate has quit [Quit: Leaving] 22:33:13 -!- simmarine has quit [Read error: Connection reset by peer] 22:33:14 -!- simmarine_ has joined ##crawl-dev 22:35:35 -!- Arkaniad|Laptop has quit [Ping timeout: 256 seconds] 22:39:02 -!- mumra has quit [Ping timeout: 256 seconds] 22:41:14 -!- santiago has quit [Ping timeout: 258 seconds] 22:47:06 -!- bhaak has quit [Read error: Operation timed out] 22:54:02 -!- s951 has quit [Quit: Leaving] 22:57:15 -!- dashdotdam0 has quit [Ping timeout: 256 seconds] 22:57:19 -!- bhaak has joined ##crawl-dev 22:57:41 -!- Kalir has quit [Ping timeout: 256 seconds] 23:02:29 -!- varmin has quit [Remote host closed the connection] 23:02:36 -!- Kalir has quit [Changing host] 23:02:37 -!- varmin has joined ##crawl-dev 23:03:47 * SamB thought Psyche's dagger of chaos was supposed to be pre-IDed 23:03:59 SamB: nope, because she doesn't always get chaos. 23:04:04 Sometimes she gets distortion. 23:04:06 -!- dtsund has quit [Ping timeout: 276 seconds] 23:04:09 oh 23:06:11 Debian builds of the trunk branch on crawl.develz.org updated to: 0.12-a0-2987-g713f552 23:09:14 Lemuel the Annihilator (L14 TeFi) ASSERT(smc) in 'mon-util.cc' at line 2575 failed on turn 2163. (D) 23:10:07 about weapons of chaos, what about a slight chance for them to generate normally? 23:11:22 the funniest brand, even if not the wisest thing to use 23:11:52 * Grunt envisions D:1 kobolds with chaos weapons... 23:13:45 * kilobyte sics a D:1 kobold with distortion against Grunt. 23:13:53 they are not that rare 23:13:59 !lm . abyss.enter min=xl 23:14:01 34. [2013-01-17 20:33:22] SGrunt the Chiller (L1 MfIE) is cast into the Abyss! (kobold) (D:1) 23:14:02 xom will create some for you probably always 23:14:06 kilobyte: been there, done that :b 23:15:35 interesting that curare depends on item level, while distortion is constant 23:15:46 why does it use a separate function, even? 23:16:15 -!- Sovek has quit [Quit: Not that there is anything wrong with that] 23:17:34 Curare? ...let's find out. 23:17:56 makeitem.cc _got_distortion_roll() 23:18:21 Oh, distortion. 23:20:23 -!- bh has joined ##crawl-dev 23:20:33 -!- Arkaniad|Laptop_ has quit [Read error: Operation timed out] 23:20:56 sometimes I wish henzell would just tell me if I have messages on join. 23:21:14 Lemuel the Devastator (L12 TeFi) ASSERT(smc) in 'mon-util.cc' at line 592 failed on turn 1870. (D) 23:21:53 -!- santiago_ has quit [Ping timeout: 258 seconds] 23:22:39 man. maybe we could make crawl crash less. 23:23:03 I think it was enne who opined that the project is essentially untestable. 23:23:14 That one's pretty consistent. 23:23:24 kilobyte: 23:23:29 %git c5f1e22 23:23:29 03greensnark * rc5f1e2290930: Make random distortion-branded weapons much rarer to correspond to their huge potential to Mess Up Life (yes, we can discuss this further if necessary :-)). 10(6 years ago, 1 file, 13+ 7-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=c5f1e2290930 23:23:32 %git b69356a 23:23:33 03greensnark * rb69356ae0558: Make distortion weapons unmarked, restore old chances of generation (well, spears and staves now have a slightly lower chance of distort, but I will live with that). 10(6 years ago, 1 file, 20+ 12-) 13http://git.develz.org/?p=crawl.git;a=commitdiff;h=b69356ae0558 23:23:50 (those are three commits apart, incidentally) 23:24:24 -!- Wester has quit [Read error: Connection reset by peer] 23:27:21 Grunt: do you think it would be a waste of time to pull apart _xom_is_bad and _xom_is_good? 23:27:50 In what sense? 23:27:58 -!- jday_ has quit [Quit: Leaving] 23:28:53 -!- Arkaniad|Laptop_ has quit [Quit: I've killed the senator.] 23:29:01 I think the whole if (tension && one_chance_in(foo)) ladders are really gross and an impediment to modifying the code 23:36:30 -!- MaxFrost has quit [] 23:43:02 -!- yogaFLAME has joined ##crawl-dev 23:46:03 -!- Thalfon has quit [Quit: Page closed] 23:47:10 -!- santiago has quit [Ping timeout: 258 seconds] 23:55:56 -!- eith has quit [Ping timeout: 246 seconds] 23:58:50 -!- Wehk__ has quit [Ping timeout: 260 seconds] 23:59:49 -!- santiago_ has quit [Ping timeout: 258 seconds]