00:09:59 The build passed. (0.22.1 - 8f2143f #10270 : gammafunk): https://travis-ci.org/crawl/crawl/builds/429386536 00:49:08 advil: did you get as far as figuring out how to make windows builds in Travis available as releases? 00:57:01 advil: hrm, it seems that the ghost data isn't getting copied correctly for the debian packages 00:57:27 I can't see my ./crawl/saves/dat folder getting populated with it and it appears to be missing in /usr/share/crawl/dat 00:58:03 Regenerating level. 00:58:05 No ephemeral ghost files for this level. 00:58:08 Trying to save ghosts. 00:58:10 Could not find any ghosts for this level to save. 00:58:13 in wiz mode when I reloaded d:3 until a ghost vault placed 00:58:53 yeah I got a clone ghost 00:59:07 let me try with the windows installer through wine 01:03:17 hrm, not seeing ghost data here either 01:05:47 advil: it looks to me like the bones files are just not getting installed into the packages or the windows zips or installers 01:06:10 the download step went fine but maybe we're not copying those files where they need to go? 01:06:30 http://crawl.develz.org/release/0.22/ 01:06:35 files I've built thus far 01:07:22 er 01:07:37 oh I haven't copied these over 01:07:38 one sec 01:08:33 yeah let me try this again 01:08:34 the debs definitely aren't working but maybe the windows files are fine 01:09:20 ok yeah great I do see them 01:09:29 so just have to figure out why the debs done have them 01:41:22 if Gretell is going to be absent from ##crawl in the forseable future I could add CDO to Postquell's announcement filter 01:43:55 Jorgrell is also not registered with nickserv it looks like 01:49:08 Gretell might return, but it depends on whether we get a host for the updated Chei 01:50:04 just have to apply those join order changes to gretell since it's using the same perl irc modules 02:34:02 ok, I'll hold off for now but it's a possibility if anyone comes asking around for cjr or cdo announcements in ##crawl 02:48:46 Monster database of master branch on crawl.develz.org updated to: 0.23-a0-175-g97aacdb 03:12:05 Stable (0.22) branch on crawl.beRotato.org updated to: 0.22.1-0-g8f2143f 04:28:15 -!- amalloy is now known as amalloy_ 04:37:59 figured out why the debian package had no ghosts, it's because my build chroot didn't have wget 04:38:13 so hopefully that's now installed via pbuilder and this will work 05:13:38 gammafunk, hmm, does the build download things from the Internet? :( 06:08:30 |amethyst, did you find out where the games missing on CAO scores web pages went? Or should I just get on with my life. :) 06:13:46 Or maybe I'm confused and they shouldn' be there, but AFAICR they were there before 18-09-14 and then disappeared? 06:14:00 (for reference, compare http://crawl.akrasiac.org/scoring/players/tas2012.html with !lg tas2012 v=0.23.0-a0 s=src) 06:22:57 jilles: yes, sadly it does have to do that 06:23:18 thankfully I was able to build the packages, but I had to use a hack 06:24:02 after figuring out that I didn't have wget in chroot, then that I didn't have networking in chroot, then that I didn't have certificates in chroot for wget to be happy with the github url 06:24:20 I found that the url has a component that requires git to check the tag to get a version 06:25:03 we'll need to find a way to pass that info to the Makefil since for debian package building we're not going off of the repo 06:25:12 I manually modified the url in the Makefile, which I guess is also fine as a workaround 06:25:25 but after all that, the bones showed up in the built debian packages and work fine 06:25:44 I'll finish up releasing this later 06:39:39 <|amethyst> TAS_2012v: hm, I think the games are there, if you look at the counts in the "Games Played" table 06:40:30 <|amethyst> TAS_2012v: I think it's that the "recent games" table is going by the order in which games were imported into the database, not the timestamp on the games 06:42:09 Hey guys - I haven't really been paying attention to the android builds in awhile, but today I checked on them - turns out they've been failing for awhile now. This is the build error: https://pastebin.com/1V0jc5zf 06:43:27 <|amethyst> miserium: looks like android-project/jni/src/Android.mk hasn't been updated with changes to the list of source files 08:04:00 |amethyst, ah, I see, the more recently imported 0.22 games pushed the more recently played 0.23 games off the table, thanks for the explanation! 09:24:26 Stable (0.22) branch on underhound.eu updated to: 0.22.1-0-g8f2143ff1b 09:39:39 !tell gammafunk thanks for dealing with all that, would it just be easier to manually install the bones files? that's not very hard 09:39:39 advil: OK, I'll let gammafunk know. 09:40:26 in fact I could add makefile support for itr 10:06:26 03advil02 07* 0.23-a0-176-g3888083: Clear trapping nets when a monster moves via gell's (minmay) 10(4 days ago, 1 file, 8+ 1-) 13https://github.com/crawl/crawl/commit/3888083293a3 10:06:26 03advil02 07* 0.23-a0-177-gdbc91a4: Add a way of using a local bones distribution for make install 10(4 minutes ago, 2 files, 7+ 1-) 13https://github.com/crawl/crawl/commit/dbc91a4fd46c 10:06:59 forgot to push that gell's fix before release, sigh 10:09:13 didn't android builds break when we updated SDL? 10:10:29 I guess that build process isn't using contribs 10:13:55 !tell gammafunk dbc91a4fd46c may give you a better way to install the bones dist if you're building in that kind of chroot setting 10:13:56 advil: OK, I'll let gammafunk know. 10:15:47 hm how do other build rules that use git describe handle the case where you aren't building in a git repo? is that what util/release_ver is for? 10:24:14 looking at this I suspect the best solution for debian is to do something during the package-source step 10:46:10 03advil02 07* 0.23-a0-178-g440b721: Add bones downloading to the `make source` step 10(4 minutes ago, 1 file, 12+ 0-) 13https://github.com/crawl/crawl/commit/440b721736ce 10:47:28 !tell gammafunk also 440b721736ce, sorry for not figuring this out before release; I didn't really understand how debian packaging worked. May be worth manually using a makefile with these changes to build source/debs for this release? 10:47:28 advil: OK, I'll let gammafunk know. 10:50:12 possibly I should have just bit the bullet and commited binary files to the repo 10:50:27 I'm not sure this is much more elegant 11:10:41 Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-178-g440b721736 (34) 12:21:57 Stable (0.21) branch on underhound.eu updated to: 0.21.2-0-g371e54c1e4 13:02:08 Stable (0.22) branch on crawl.akrasiac.org updated to: 0.22.1-0-g8f2143f 13:07:42 Unstable branch on crawl.akrasiac.org updated to: 0.23-a0-178-g440b721 (34) 13:29:53 Changes.txt no longer being updated since 07/31 13https://crawl.develz.org/mantis/view.php?id=11680 by silenth 13:30:16 I wonder if something went wrong with cron generally on cdo 13:30:25 that's around when windows builds stopped 14:12:36 <|amethyst> advil: hm, in the past that has happened (changes.txt not updating and new builds not happening) when CDO's local commits have a conflict with trunk 14:14:08 ah, that might explain it. Though I think it's still updating trunk for playing? 14:14:10 &versions 14:14:16 !versions 14:14:29 CAO: 0.23-a0-178-g440b721, CBRO: 0.23-a0-175-g97aacdb, CDO: 0.23-a0-151-gf6d1ed2, CJR: 0.23-a0-178-g440b721736, CPO: 0.23-a0-178-g440b721, CUE: 0.23-a0-175-g97aacdb306, CWZ: 0.23-a0-136-g722f094, CXC: 0.23-a0-175-g97aacdb, LLD: 0.23-a0-175-g97aacdb306 14:15:09 %git f6d1ed2 14:15:09 07advil02 * 0.23-a0-151-gf6d1ed2: Add cross-compiling to travis 10(11 days ago, 4 files, 16+ 1-) 13https://github.com/crawl/crawl/commit/f6d1ed2717bd 14:15:13 <|amethyst> okay, I just fixed the merge conflict 14:15:33 oh, there was one? what was it? 14:15:41 thanks! 14:16:09 was it the version it is currently stuck on, or something earlier? 14:16:15 <|amethyst> CDO has its own cdo-install target, which was right next to the clean target 14:17:21 <|amethyst> and the addition of clean-bones-dist to the latter messed up the context enough that it couldn't apply that diff 14:17:27 aha 14:18:51 <|amethyst> okay, rebuilding trunk, let's see if this works (and if it installs the updated changelog, I think it will) 14:19:49 thanks, didn't even know you had access for this 14:21:28 <|amethyst> yeah, I don't have root (only nap.kin does) but I do have access to the crawl home directory 14:22:50 gammafunk has access to something but maybe not that directory 14:26:23 looks like changes.txt updating is fixed at least 14:26:56 |amethyst any ideas about the dead nightly building? Is it built in the same place? 14:27:12 I think it broke earlier than I added the bones stuff 14:27:21 so it's probably not the same issue 14:27:53 oh I just misread the first commit 14:27:53 not updated yet 14:31:11 <|amethyst> hmm 14:32:06 <|amethyst> error removing the bones 14:32:23 <|amethyst> (in clean-bones-dist) 14:35:07 <|amethyst> !tell Napkin on CDO somehow ~/source/crawl-git/crawl-ref/source/dat/dist_bones and beneath got the wrong ownership.. I renamed it to dist_bones_old, but can't remove it 14:35:08 |amethyst: OK, I'll let napkin know. 14:35:25 <|amethyst> advil: oh, just noticed that clean-bones-dist will fail if you run it twice in a row 14:35:32 <|amethyst> advil: it should probably use -rf, not just -r 14:35:42 ah 14:36:23 isn't it already using -f? 14:36:35 `RM = rm -f` 14:36:44 <|amethyst> oh, good 14:37:10 on os x it works twice in a row 14:37:24 <|amethyst> yeah, that's fine, I missed that 14:37:41 honestly I sort of wish I had just commited the tarbell to the utils directory instead of all this complicated downloading stuff 14:40:07 <|amethyst> hm, that still didn't update changes.txt 14:40:53 <|amethyst> I haven't looked at the relevant sections of the Makefile, but why commit a tarball instead of just the individual bones files? 14:42:28 or that 14:43:43 <|amethyst> Yeah, I think putting the distributed bones under version control might not be a terrible idea 14:43:53 <|amethyst> definitely a lot nicer than a tarball, anyway 14:46:50 |amethyst when you have some time don't forget that there's some Cheibriados fixes awaiting you too :) 14:47:20 <|amethyst> what's the repo again? 14:47:30 https://github.com/crawl/cheibriados/ 14:54:04 -!- The topic of ##crawl-dev is: Crawl Development | Logs: http://s-z.org/crawl-dev/ | People with +v have commit access. | Please keep general Crawl-related chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to a pastebin service, please. 14:54:12 <|amethyst> well, not registered now, but 14:54:25 <|amethyst> %git HEAD 14:54:25 07advil02 * 0.23-a0-178-g440b721: Add bones downloading to the `make source` step 10(4 hours ago, 1 file, 12+ 0-) 13https://github.com/crawl/crawl/commit/440b721736ce 14:54:54 <|amethyst> yeah, still unregistered, but it runs at least 14:55:21 <|amethyst> I guess I need to /nick cheibriados here then message nickserv 15:03:29 -!- The topic of ##crawl-dev is: Crawl Development | Logs: http://s-z.org/crawl-dev/ | People with +v have commit access. | Please keep general Crawl-related chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to a pastebin service, please. 15:04:00 <\amethyst> hm, also not joining ##crawl 15:04:00 \amethyst: You have 2 messages. Use !messages to read them. 15:04:15 -!- \amethyst is now known as |amethyst 15:06:22 JoinAfterAuth should print some messages to stdout if it's working 15:06:36 <|amethyst> it printed the message to stderr 15:06:43 <|amethyst> but didn't actually join the channel 15:06:47 hm 15:07:45 it doesn't look like it's actually authed 15:08:14 whois isn't giving me "Cheibriados is logged in as..." 15:08:24 <|amethyst> ah 15:10:35 -!- The topic of ##crawl-dev is: Crawl Development | Logs: http://s-z.org/crawl-dev/ | People with +v have commit access. | Please keep general Crawl-related chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to a pastebin service, please. 15:10:45 -!- The topic of ##crawl is: Play Dungeon Crawl Stone Soup online now! Type ??servers for instructions. | http://crawl.develz.org | FooTV: http://termcast.develz.org - ??footv for instructions | See also ##crawl-offtopic 15:10:50 <|amethyst> It was authing, but I hadn't verified the account yet 15:10:55 ahh 15:10:56 <|amethyst> %git 15:10:56 07advil02 * 0.23-a0-178-g440b721: Add bones downloading to the `make source` step 10(4 hours ago, 1 file, 12+ 0-) 13https://github.com/crawl/crawl/commit/440b721736ce 15:10:58 <|amethyst> yay! 15:11:02 nice! 15:11:04 <|amethyst> works in ##crawl too 15:11:17 <|amethyst> Thanks for the fixes and the help! 15:11:37 you're welcome 15:11:42 I learned about moose 15:12:06 for some reason my Dev one wasn't announcing commits in ##crawl even though it should have, so next time someone commits something I'll have to check if that was a config error or something I actually broke 15:12:27 <|amethyst> oh 15:12:33 <|amethyst> only some commits get announced in ##crawl 15:12:42 yeah, even the ones that are supposed to 15:12:54 I might just have had the wrong one set as the main channel? 15:13:37 but basically, I tried to change it so that it would never announce in ##crawl-dev and ##crawl unless its nick is Cheibriados 15:13:45 <|amethyst> ah 15:13:57 and I don't have perfect confidence in what I did 15:14:00 <|amethyst> "main" channel is just whichever is listed first 15:14:17 <|amethyst> where is that logic? 15:14:26 <|amethyst> or is it only local, not in git? 15:14:37 it's in https://github.com/crawl/cheibriados/commit/7fa86d43f9d08b6ea399093b42c4715db37ed61e, so if this is a problem one could probably just revert this commit 15:15:25 the key thing is the change from using before to around, so that it can intercept calls 15:15:57 -!- ldlework[MIA] is now known as ldlework 15:16:59 it definitely talks *somewhere*, I was testing it in a different channel ##crawl-bot 15:17:16 <|amethyst> seems reasonable enough, and that wouldn't affect only announcements anyway 15:17:34 <|amethyst> I guess we'll see what happens when there's a commit 15:18:00 yeah 15:18:07 hm I wonder if I have something small to test it with 15:31:15 %git 15:31:15 07advil02 * 0.23-a0-178-g440b721: Add bones downloading to the `make source` step 10(5 hours ago, 1 file, 12+ 0-) 13https://github.com/crawl/crawl/commit/440b721736ce 15:31:34 03advil02 07* 0.23-a0-179-gcb8baf8: Don't crash when doing describe-all with a buggy net in los 10(4 minutes ago, 4 files, 7+ 19-) 13https://github.com/crawl/crawl/commit/cb8baf84dfb7 15:31:45 hm 15:31:50 oh there we go 15:31:58 looks like it works! 15:32:42 !tell gammafunk Cheibriados is up and running with auth, so you can point the other bots back to it rather than CheibriadosDev 15:32:42 advil: OK, I'll let gammafunk know. 15:37:00 oh, I should have said in that commit message, but the motivation behind not crashing is that there's no way to fix a buggy net up, so it's a holdover until I or someone add fixup code (and imo the way to fix buggy nets up is to rewrite nets) 16:10:43 Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-179-gcb8baf84df (34) 16:25:26 -!- ProzacElf_ is now known as ProzacElf 16:26:45 The build was broken. (master - cb8baf8 #10273 : advil): https://travis-ci.org/crawl/crawl/builds/429725676 16:45:12 03advil02 07* 0.23-a0-180-ga5275e7: Fix the build 10(23 seconds ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/a5275e7ff7f3 16:53:47 Wheeee, just posted my species proposal! Once again huge thanks to you guys and specially to |amethyst for your help. <3 17:10:47 Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-180-ga5275e7ff7 (34) 17:14:45 -!- amalloy_ is now known as amalloy 17:31:50 If you want to test a github repo you made to make sure it works, what is the best way to do that? 17:44:50 YokkasoMB: clone it locally and build it or otherwise run the code from the clone 17:45:30 advil: thanks for that Makefile change; I managed to build the packages by hard-coding the url and getting wget+networking to work in the chroot, but using a local file is probably the best way 17:45:39 so I can try that next release 17:46:05 I think we might want to make package-source include said local bones_dist dir if it doesn't already 17:46:11 that way the source files would just have it 17:46:18 in fact I might try that and repackage source 17:46:41 in which case I could also rebuild the packages with that approach to see if it works, since building the debs doesn't take long 17:46:52 I'll remap the bots to use chei later today 17:53:23 The build was fixed. (master - a5275e7 #10274 : advil): https://travis-ci.org/crawl/crawl/builds/429756630 18:21:34 Unstable branch on underhound.eu updated to: 0.23-a0-180-ga5275e7ff7 (34) 18:44:07 YokkasoMB: nice work on the species 18:44:31 did you see the faerie dragon proposal from Realz a few months ago? 18:44:57 I did 18:45:21 It was the impetus for making the yokkaso 18:45:32 There is 100% a niche waiting to be filled for a starter caster 18:46:07 I get why the devs rejected it, and hopefully addressed the concerns expressed over FDs 19:01:32 alexjurkiewicz: anything glaring about the species you think should be altered or changed? 19:02:15 I think a lot of people insisting on this idea of "easy spellcaster" have some weird notions of what spellcasters do 19:02:26 for instance "most spellcasters end the game in ice or fire dragon scales"? 19:03:09 Yes duvesa already called me on it 19:04:05 I got carried away by the symmetry of the 7-11-15 ENC to demonstrate the difference sturdy frame would make >_> 19:06:26 the MP refunding thing does just feel like an overly complex way to just give them more starting MP 19:07:32 Complicated to understand for players? I don't mind it being a complicated way to increasing their MP as long as it encourages good play. 19:08:39 I mean, it is a bit complicated to understand if we're coming from a new player perspective, yes, but it also seems like something that's not necessarilly better than just starting with a bunch of MP or maybe just systematically lowering spell costs 19:09:16 as little to check as possible is good; with this proposed mutation you do have a harder time keeping track of your MP due to how it works 19:11:07 Yeah there are problems with it, I don't like the idea of unconditionally increasing the mana pool (or reducing spell costs to the same effect) 19:12:03 The best example of a "learn spellcasting" race I've seen so far is probably Gnoll; they start with a lot of MP and general durability, a lot of spell levels, and you have no skilling to think of (arguably autotraining is taking care of that for new players though) 19:12:51 I had sort of thought that if you wanted to help players into spellcasting you'd need some way to really simplify the complicated part of spells that takes a long time to learn: how to use spells effectively 19:13:18 That is the main problem isn't it. 19:13:20 it's true that MP availability is a signficant part of this, but spells are just pretty complicated to use 19:14:24 I think pubby's idea is the best approach to this: remove skilling (just one skill, train it to make spells better) and limit spell choice 19:14:41 yeah that's one angle for doing things 19:14:55 YokkasoMB: have you looked at pubby's ogremagic / oni species? 19:15:00 I think, well hope, that if players got a "bonus" for being at high health in a fight that would lead them to think more proactively about their positioning and incoming damage. 19:15:27 ebering: Not yet, I'll go find it now 19:19:42 I don't know that I like that approach too much tbh (maybe I'm just biased?) 19:21:19 There are 2 ways they used to teach morse code, first way was with using all the letters dots and dashes at a slow speed, then improving the speed. The other method was teaching at full speed one letter at a time until they had a full alphabet. 19:21:53 I think the oni solution is leaning a lot more towards the second teching philosophy 19:22:31 "Maybe I'm biased" says the person who's ircname is the same as their species proposal??? :) 19:22:47 *whose 19:22:52 shhhh don't bring that up it's no biggie 19:23:03 -!- gammafunk is now known as REMOVEFOOD 19:23:20 ^ lol 19:23:20 maybe we should remove food? just a thought 19:26:07 I I'm trying to think of the right phrasing to use for this so forgive me if it sounds dumb 19:27:33 I think learning when and what magic skills to train is nearly as important as learning when and where to use the spells you learn, I think Oni does a good job of breaking that learning into two halves, but I don't know if thats a good thing? 19:28:53 This is always the issue though, making casting more new player accessible feels... odder than it should 19:29:30 -!- REMOVEFOOD is now known as gammafunk 19:30:04 Yeah I think this is sort of the core problem with the idea of easy to learn spellcasting in crawl; spellcasting is just very complicated, and you can only simplify a few aspects 19:30:53 there is a people who strongly want "mibe of spellcasting" to be a thing, which I'm not convinced it should be, but there could still be some interesting species idea in there ultimately 19:31:10 I'm going to find time to take a look at how to use pubby's idea in the near term 19:31:51 I don't really love this idea of relatively complicated mp cost reduction; maybe ER mutation is a good route and maybe something with MP together with pubby's idea 19:32:05 you could ask MarvinPA what he thinks about your proposal in terms of what works and what doesn't 19:32:24 he's another one who's good at giving honest feedback 19:33:21 we do tend to go off of a consensus where enough devs who care about the idea are happy enough that a thing can go forward 19:33:59 and that can take a while to get going (gnolls are a pretty funny example of this; wu jian is another, but we're not fully happy with where wu is right now) 19:34:58 s/there is a people/there are a lot of people/ 19:37:05 i had a glance at the thread, i do still kinda feel like the existing races (DE, Dr, Dg) do already fill a lot of the niches that most of the recently-proposed "easy magic species" are aiming for 19:37:42 That's fair enough, I know that the devs obviously have a lot more knowledge about the game than me 19:38:36 like, Dr in particular i think do work nicely in terms of a kinda easy magic species that's got some good innate durability 19:45:27 Might just be a perception issue then, or perhaps rather, new players lack of understanding the signifigance of relative aptitudes, I don't think people would naturally gravitate towards a species with a -1 spellcasting and 0 all spells species for a caster. 19:46:52 but if it does the job, it does the job you know? 19:53:49 I keep coming back to just making an advanced/extended spellcasting tutorial as all the other options to integrate easier spells into the main game seem... unsavoury. But then there MUST be some elegant way to teach players magic/spellschools I haven't thought of... 20:10:43 A question for developers: https://crawl.develz.org/tavern/viewtopic.php?f=8&t=25759 20:12:00 that sounds like an oversight, yeah 20:13:45 it'd probably be best for ddoor to have the same guaranteed turn 20:17:51 I think there's space for multiple species that "teach spellcasting" in different ways 20:18:35 eg there is space for Oni, Gnoll and FD/Yo to all be newbie-friendly in different ways 20:20:36 gammafunk: did you ever give an opinion on FD with the only unusual point being -1 MP? Eg none of the defensive mutations or flight 20:21:36 I think you want to be careful about that multiple ways idea 20:21:50 it's going to justify adding a lot of species in an already-long species list 20:22:16 but I do agree that you don't want a bunch of "spellcasting species" all trying to be that in the same way 20:23:27 I mean if you make FD without a lot of the other weirdness but with -1 MP cost, isn't it just a very slightly tweaked Draconian at that point? 20:23:31 right. You don't want three species that are all 'easy caster' (eg Minotaur 1, Minotaur 2, and Minotaur 3). I don't think Oni/Gnoll/-1MP have that problem though 20:23:50 It's more like Mi/Gr/VS 20:24:39 you want to be clear about what problem you're trying to address with a given species; is it MP availability? other species do address this pretty OK (Gnoll, DE, Og) 20:24:54 -!- flappity_ is now known as flappity 20:25:58 Yeah, it's important that species don't fill the same gameplay niche 20:26:12 chequers one thing I would suggest to you in general is try winning some mages in DCSS first; you've said in the past that they're not really your style 20:26:31 !lg chequers mage won 20:26:31 No games for chequers (mage won). 20:27:24 so I think you tend to have views about their design that's in-part skewed by your tendancy to play melee things exclusively 20:28:27 I don't profess to be good at crawl or mages. And you're right that -1MP makes magic "more like melee" in that it allows (sorta) infinite tab 20:30:40 give Dg really high innate mp regen 20:30:41 to me, the mutation changes how the player can use magic in a way that Draconian or Deep Elf doesn't. There's a qualitative difference between -1MP and any other species, and I think a solidly fun gameplay niche as well 20:32:47 !kw mage 20:32:48 i think high MP regeneration works better than -1 MP cost, since with -1 MP cost you should cast a level 1 spell every turn instead of waiting 20:32:48 Keyword: mage => AE|Cj|EE|--FE|IE|Ne|Su|VM|Wz 20:33:26 -1MP has a problem with level 1 spells; I'm not a huge fan of making those cost nothing. I can't say I feel super strongly about that, but I know of a couple other devs who probably don't like the idea as well 20:33:38 but aside from those, just having a lot of MP is pretty similar to -1 MP 20:34:08 How about new_cost = max(1, cost - 1) 20:34:22 so aside from the level 1 spell thing it doesn't feel that distinct; mp regen might be a better way to get the same idea, though 20:34:31 yeah I guess you could make it conditional 20:34:50 but have you tried playing as a gnoll mage alex? 20:34:57 you really got an obscene amount of MP 20:35:05 yeah if you're going to do mp cost reduction i like coming at it from the opposite way: all spells cost half their original cost, or just 1 MP, or whatever 20:35:23 that way the effect is bigger for higher-level spells, instead of bigger for lower-level spells 20:35:54 yeah that's sort of a nice benefit of that approach in terms of it staying relevant 20:36:03 and as a player i would rather spam bolt of fire than spam flame tongue, so an MP gimmick that further incentivizes me to do the latter is not appealing 20:37:03 probably halving the cost is the sane way to go, I guess the only strangeness is 7mp spells cost the same as 6mp spells (or as 8mp spells if you round up) 20:37:10 could do half + round random, so that sometimes level 1 spells don't use mp 20:37:40 yeah the level breakpoints in that are gross 20:37:48 vehumet flashbacks 20:38:24 YokkasoMB: did you consider making the species not for newbies and changing the MP cost reduction so it increases inversely with HP? More try-your-luck approach 20:38:45 hm one could also get exactly the same effect by doubling mp, and then there's no rounding problem 20:38:52 I guess there's longer waiting 20:38:57 double mp regen too 20:38:58 we did it 20:39:01 heh 20:39:26 have to remember to make spirit shield deduct twice the normal amount of MP though 20:39:37 yeah 20:39:41 and mana drain 20:39:52 gimmicks that only come into play at low HP have the problem that they only come into play in a very small percentage of fights 20:39:53 half + coinflip() sounding easier :) 20:39:54 see sanguine armour 20:58:07 alexjurkiewicz: I hadn't thought about that, am a hugh fan of skilled brinksmanship in games (red tearstone ring in dark souls FTW!) 20:58:30 Originally the species was for noobs but if the devs like any of the ideas they are free to run with them 20:59:18 Maybe reverse the proportion and remove the 50% cap so any % less than 100% and you might benefit 21:11:56 FR: species that gets faster with lower hp, action delay is multiplied by (hp/mhp) 21:12:09 as hp approaches 0 your speed goes to infinity! 21:12:32 hmm 21:12:49 maybe make all actions take twice as long to begin with 21:13:54 looking forward to the 1 hp zot:5 runs 21:14:43 I guess it's not possible for an action in crawl to take between 0 and 0.1 time though 21:14:59 only with randround 21:17:38 Fond memories of https://paragonwiki.com/wiki/Defiance (Original Defiance) 21:20:15 once you're in that range actions have an increasing chance of taking no time at all 21:20:50 There's some sort of hilarious TRJ one-shotting in there 21:45:44 could make for a hilariously OP but risky species 21:46:30 even if you take 0 time the monster just needs to already have enough energy to take an action? 21:46:37 idk how energy works 21:47:27 More seriously a god (etc) that lets you build up to a magnificent carnival of carnage already exists in Usk 21:47:59 Specifically one where no opponent can do anything. 21:48:03 yeah usk is fun 21:56:11 advil: would it make more sense to have the makefile change in HEAD~2 just untar the dist_bones.tar.gz and have the check in HEAD~3 simply check that dat/dist_bones exists? 21:56:26 if I'm reading this right, the source both has the tgz file and the untarred contents 21:57:01 I think I should just revert all this complicated stuff and put dat/dist_bones in the repo 21:57:05 haha 21:57:15 I was kind of thinking them same :) 21:57:35 *especially* when I got to the point where I realized how not having a git repo broke the debian build 21:58:00 that way cdo will get it too. It won't interfere with other servers' bones files because it only draws on dist_bones when the regular bones directory has nothing in it 21:58:02 that's another sort of problem with our current build 21:59:26 yeah I guess just having the dist_bones dir and files in the repo is fine 21:59:31 they're not large files or anything 21:59:46 I also noticed how we have bones data for branches that currently can't have ghosts 21:59:50 but that's not a big deal 22:03:44 let me go move those bots over to chei 22:08:20 yes, that was semi-intentional in case someone wants to write vaults that place ghosts in those branches 22:09:11 yeah it wouldn't be hard to adapt existing vaults for pan, at least; hell and tomb would need more specialized vaults that integrate into the encompass maps 22:09:35 not sure we can really do ghosts in the abyss at all, but other portals could potentially have ghost vaults using subvaulting for the encompass maps, at least 22:10:09 zigs are the one branch were we could consider relaxing the rule about ghost vaults maybe, at least past a certain depth 22:10:20 all ghost zig floor 22:10:27 B E L I E V E 22:11:24 did anyone ever prototype dpeg's alternate zig layout ideas that have collected dust on the wiki? 22:11:49 no I haven't seen them 22:11:49 (devwiki) 22:12:11 basically designed to murder players. I think the premise was the player lands in the middle 22:12:18 and the exit is somewhere on the edge good luck don't die 22:12:41 advil: fyi all my bots are moved back to Cheibriados, so you can feel free to stop your copy if you want to 22:16:41 https://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:dungeon:portal_vaults:ziggurats 22:16:47 Ghosts should ideally occur on the ziggurat level they were created on. If that is not feasible, they should only be generated after some fixed depth, e.g. Zig:10. Urgent 22:17:05 Urgent (2011) 22:17:05 :) 22:18:29 as long as it's fixed by 2020 it's fine 22:18:41 though I couldn't find the dpeg thing I was thinking of 23:52:18 aidanh: seems that the ! toggle on monster descriptions is broken in webtiles 23:52:21 or rather not shown 23:52:31 it's shown and works on console 23:52:40 this is the toggle that switches between quote and description