00:04:03 |amethyst: according to a comment in skills.cc where it actually uses an int64_t, overflowig is maybe a risk? 00:04:12 *overflowing 00:06:30 <|amethyst> yeah, but that is a rescaled number, not the original value 00:06:42 <|amethyst> also, int -> unsigned doesn't give you a lot as far as overflow goes 00:06:56 <|amethyst> hence int64_t there 00:07:26 <|amethyst> but e.g. player.experience can never be large enough to need to be unsigned int itself 00:07:42 <|amethyst> (on 32-bit-int systems) 00:08:57 yes, true 00:09:18 <|amethyst> if it could be larger, I would argue it should be an int64_t instead 00:09:27 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-314-gb3271ec (34) 00:10:23 <|amethyst> honestly it should really be an experience_points 00:11:58 <|amethyst> but doing a change of *that* scope gives me the willies a bit 00:18:20 is using unsigned int soley because a value shouldn't be negative a kind of "classic" programming mistake? 00:18:53 I sort of reasoned that it would make a lot of sense if values that shouldn't be negative were always declared unsigned 00:18:53 <|amethyst> at least in C++ 00:19:39 <|amethyst> gammafunk: one problem is, what if you take such a thing that is 0, and subtract one from it? 00:19:49 <|amethyst> gammafunk: obviously you shouldn't be doing that 00:19:56 <|amethyst> but unsigned doesn't prevent you from doing that 00:20:07 <|amethyst> instead, you get a value that looks like a valid one 00:20:18 yeah, I see what you mean 00:20:18 <|amethyst> whereas with a signed int, it's obviously a bad value, because it's negative 00:20:29 that's sort of counter-intuitive 00:20:52 that declaring something as unsigned actually makes it less likely you'll catch certain errors 00:21:05 I guess you could put a ceiling on values 00:21:24 <|amethyst> in which case you lost the advantage of that extra bit 00:22:03 right, I had been imagining that going unsigned is not a "more space" issue but a "prevent more errors" issue 00:22:10 but you're saying it actually won't prevent more errors 00:22:14 <|amethyst> exactly 00:22:54 <|amethyst> also, if you overflow an int, the compiler/runtime/CPU is allowed to crash your program. It usually doesn't on most architectures, but it's allowed to 00:23:06 <|amethyst> if you overflow an unsigned, that is guaranteed to wrap around to zero 00:24:07 Icon for Possessable status (Yred Enslave Soul) 13https://crawl.develz.org/mantis/view.php?id=11239 by Ultraviolent4 00:24:07 <|amethyst> so theoretically unsigned hides errors in that way as well 00:24:26 I don't quite understand the need for a silence icon 00:24:33 since silence is already dispalyed in tiles 00:24:53 did you know ru can silence enemies? 00:24:55 in terms of the "floor halo" 00:25:03 oh that sort of silence 00:25:04 i certainly never knew 00:25:08 zin might also do that 00:25:09 <|amethyst> it's silence and not mute? 00:25:15 right, those are different 00:25:30 so maybe that tile should be called a mute tile? 00:25:41 unless crawl also calls that silence (mistakenly) 00:25:48 yep, i didnt realise that was different, my bad 00:25:52 lets see 00:26:51 <|amethyst> your tilepick.cc change seems like it would only make it apply to monsters like silent spectres 00:27:12 <|amethyst> that radiate silence 00:27:17 <|amethyst> (that's what MB_SILENCING means) 00:27:19 yes, theres a status called mute 00:27:29 sorry, i didnt realise those were different things 00:28:49 it comes from ench_mute which is what zin does 00:29:03 well that one makes sense to have a tile for 00:29:20 im trying to find what ru does 00:29:31 <|amethyst> gammafunk: I like how we have species_undead_type(species) == US_HUNGRY_DEAD so many places rather than species == SP_GHOUL 00:29:40 hah 00:29:43 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.21-a0-314-gb3271ec (34) 00:29:54 someone knew that hungry undead was such a cool concept 00:30:01 <|amethyst> in case we decide to add another species that works like ghouls 00:30:02 that there would soon be many more hungry undead species 00:30:04 <|amethyst> :) 00:30:17 this is giving me inspiration for a vault! 00:30:25 gammafunk_hungry_hungry_ghouls 00:30:31 <|amethyst> gammafunk: other than that, which can be changed easily enough, _banned_combination could just be turned into a data structure 00:30:41 ah, good point 00:30:54 it was a somewhat horrible looking boolean conditional change before that 00:31:00 ugh, apparently ru's retaliation is actually antimagic and not silence 00:31:03 <|amethyst> yeah, I'm looking at the diff 00:31:28 I swear I've seen ru do some weird mute-like effect 00:31:57 <|amethyst> gammafunk: there's a vault 00:31:59 oh, apparently not 00:32:20 wow lasty 00:32:21 message = " loses " + mons->pronoun(PRONOUN_POSSESSIVE) 00:32:21 + " magic into the devouring truth!"; 00:32:24 the flavor 00:32:34 Lasty: the devouring truth?! 00:33:00 ok well i guess that icon on makes sense as mute 00:33:09 the game works nothing like i thought it did xD 00:33:12 |amethyst: yeah but I wanted it to reference hungry hungry hippos and somehow have a similar layout 00:33:26 hrm 00:33:55 <|amethyst> hm 00:34:33 ru's aura description maybe shouldnt say that it silences enemies 00:35:06 <|amethyst> ah 00:35:08 <|amethyst> it was changed 00:35:09 aha 00:35:14 <|amethyst> %git 737b7ca6 00:35:14 07Lasty02 * 0.16-a0-3339-g737b7ca: Switch Ru's muting effects to antimagic 10(2 years, 9 months ago, 1 file, 10+ 10-) 13https://github.com/crawl/crawl/commit/737b7ca64572 00:35:24 else if (power > 10 && mons_can_be_blinded(mons->type)) 00:35:24 { 00:35:24 mprf(MSGCH_GOD, "You focus your will and blind %s in retribution!", 00:35:24 mons->name(DESC_THE).c_str()); 00:35:24 mons->add_ench(mon_enchant(ENCH_BLIND, 1, act, power+random2(100))); 00:35:27 <|amethyst> and probably that message, and gammafunk's memories, were not updated 00:35:28 for Ru's passive 00:35:32 Ru can blind monsters 00:35:37 not an apoc effect apparently 00:35:49 ah and that commit from |amethyst explains it, changed in 0.16-a 00:36:06 for the apoc thing 00:36:32 03advil02 07* 0.21-a0-315-g9255a14: Only calculate ash bonus for ash characters 10(62 seconds ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/9255a1437d5a 00:37:55 i feel like using that icon for antimagic would just confuse people because it looks like spells cant be cast 00:38:07 ah well 00:38:43 that bug was the crash bug...turns out the ash skill functions don't return 0 if you aren't ash, which is what (for some reason) I had assumed 00:39:21 well an icon for anti-magic seems fine 00:39:34 <|amethyst> gammafunk: I notice you disabled Sac Arcana but not the other skill sacrifices? 00:39:35 I don't know if there's an existing analog for that 00:39:44 |amethyst: yeah....we need to figure out Gn + Ru 00:39:52 as in a consistent approach 00:40:18 <|amethyst> my thought is: you lose the skill, it stays at zero for all intents and purposes, but XP is still directed into it uselessly 00:40:59 yes, that would be pretty reasonable 00:41:05 <|amethyst> there's the question about what to do with the bonus piety 00:41:13 oh, Lasty had talked to me about that 00:41:21 his thought was 00:41:32 don't give the bonus piety for the invested skill 00:41:38 but then don't train the skill either 00:41:58 so the lack of bonus piety makes the sac worth less which balances the xp benefit 00:42:04 but maybe your idea is the better one 00:42:09 <|amethyst> I don't think it balances it very well 00:42:19 03advil02 07* 0.21-a0-316-g7453ecb: For real disable training target UI for gnolls 10(25 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/7453ecba3e14 00:42:28 yeah 00:42:28 the only confusing part is the interface, wrt your idea 00:42:36 i made a new note on that submission 00:42:37 how do you indicate that the skill is not usable but still being trained? 00:42:46 advil: thanks for that fix 00:42:55 <|amethyst> gammafunk: + and grey it out 00:43:01 <|amethyst> gammafunk: or, don't indicate it 00:43:10 <|amethyst> gammafunk: just have the training %s not add up to 100 00:43:17 |amethyst: well I feel it's kind of confusing for players that way 00:43:34 or at least, they should really understand what's going to happen when they make that sac wrt the xp situation 00:43:40 <|amethyst> gammafunk: IMO, only confusing for players trying to minmax their gnoll skilling :P 00:43:49 <|amethyst> that may be true 00:43:58 |amethyst: well, no, I'm just saying that they shouldn't have to guess as to the effect 00:43:59 <|amethyst> so they know that it's not minmaxing to do that 00:44:06 I agree we don't want them to try to minmax 00:44:15 but it should be clear that they can't, I guess I'm saying 00:44:26 maybe it's a problem for the sac description 00:44:30 <|amethyst> hm 00:44:38 just indicate it for gnolls specifically 00:44:40 <|amethyst> I think that should be more in the gnoll description or the gnoll UI 00:44:41 <|amethyst> hm 00:44:45 <|amethyst> maybe that 00:44:48 yeah 00:44:48 you could have it show as 0 and color it red or something 00:45:02 <|amethyst> that by itself doesn't explain anything though 00:45:06 too bad strikeout can't work in console 00:45:12 set at 0 and still show the e.g. 3% 00:45:15 <|amethyst> yeah 00:45:20 yeah that's what I meant 00:45:22 <|amethyst> I think that's probably the most reasonable 00:45:32 <|amethyst> or at least the easiest thing 00:45:39 well that sounds good to me and better than conditionally excluding a lot of sacrifices 00:46:01 <|amethyst> hm, I'm not sure how I feel about showing the % in the first place though 00:46:06 <|amethyst> for gnolls 00:46:09 oh 00:46:15 yeah, and a related issue 00:46:20 the level up message spam :/ 00:46:23 I guess to be really kind you could give them a warning on the sac option 00:46:42 right, just want them to know what's going to happen before they choose the sac 00:46:45 I thought about the levelup spam for 5 minutes this morning and gave up 00:46:58 ??impossible 00:46:58 devteam[5/27]: sadly that bug is literally impossible to fix by which i mean i tried like once and then gave up 00:46:58 it would all be easier if it weren't for the oddities of training 00:47:08 pretty much 00:47:26 I was vaguely thinking about trying to make the skill levels not have that stochastic randomness 00:47:42 since at some point in a game everything is randomly distributed over +- .2 00:47:42 and then we could maybe have smarter messages 00:47:45 yes, that would be great...I'll warn you it's hard 00:47:56 yeah the other vague part of my thoughts where how you'd do it 00:48:02 lol 00:48:14 I looked into it earlier in my skill branch 00:48:26 <|amethyst> implement it completely differently: 00:48:34 <|amethyst> disable all skills for gnolls but one 00:48:38 the problem is the code has like directly evolved from victory dancing code via lamarckian-style evolution or something 00:48:54 |amethyst: only one skill for the entire game, chosen randomly at like XL7? 00:48:56 <|amethyst> a new skill, disabled for everyone else, called 'gnollitude' (or 'gnolledge', thanks geekosaur) 00:49:03 the draconians of skill! 00:49:09 well the thing about that which is a problem 00:49:10 <|amethyst> which does 100% crosstraining with all other skils 00:49:21 we're probably going to want to have different apts for some skills :/ 00:49:31 <|amethyst> for Gnolls? 00:49:32 maybe that's something we should really weigh carefully though 00:49:33 yes 00:49:35 <|amethyst> hm 00:49:48 in order to balance it? 00:49:51 right 00:50:28 melee skills, maybe spellcasting; but we might want to take a harder look at that if we could simplify life with one apt 00:50:42 also how do gnolls get titles 00:50:48 right now they're the stochastic title race 00:50:53 <|amethyst> they should have their own titles 00:51:03 elliptic had suggested making invocations have higher apt 00:51:03 <|amethyst> or always get invo titles 00:51:07 <|amethyst> :) 00:51:15 yeah that too, just rig it to invo 00:51:30 their title should always be "the gnoll" 00:51:36 and thats it 00:51:46 after all this I now have much more empathy for the person who designed vampires 00:51:52 <|amethyst> also their job and name 00:51:54 <|amethyst> hm 00:52:42 <|amethyst> oh 00:53:17 <|amethyst> there is another concern beyond stochastic skills re things happening not all at the same time 00:53:20 <|amethyst> jobs 00:54:23 right, we have to award starting xp into the relevant schools for gameplay reasons 00:54:31 so those always have a bit higher level 00:54:42 and yeah if you have one skill that doesn't work so hot 00:54:47 <|amethyst> gammafunk: do we have to? 00:54:59 <|amethyst> gammafunk: I mean, if you spread that XP out 00:55:11 what would starting level be with +8? 00:55:20 <|amethyst> hm, good questions 00:55:26 <|amethyst> s/s$// 00:55:33 maybe we don't actually but we might make the worst D:1 species of all time 00:55:35 <|amethyst> but 8.4 conjuration on a starting character seems kind of excessive 00:55:45 <|amethyst> right now they might be the best D:1 species 00:56:00 <|amethyst> for jobs that use their starting skills 00:56:37 well even then, tr, dd, sp all use their starting skills 00:56:38 <|amethyst> it wouldn't have to be the same amount divided evenly 00:56:54 there's steep competition for best, but yeah balance wise there's probably currently an issue 00:57:17 <|amethyst> you could for example just give all gnolls enough XP to raise all skills to (say) 3 00:57:20 yes 00:57:23 <|amethyst> or 4 if that feels too weak 00:57:39 I'm not sure, maybe we can get away with not having different apts, more play-testing will help 00:58:08 Training Target Level deactivates when you gain experience during Okawaru's Heroism 13https://crawl.develz.org/mantis/view.php?id=11240 by inipro 00:58:18 it is a very attractive idea in a few different ways 00:58:20 ugh 00:58:21 <|amethyst> gammafunk: I'm not sure how much I actually believe this, but: 00:58:46 <|amethyst> gammafunk: if we'd be inclined to change gnolls' apt for a skill, that's likely an indication that the skill itself needs to be rebalanced 00:59:21 |amethyst: in no way are crawl's current skills balanced, so I certainly agree 00:59:23 <|amethyst> but saying that doesn't take into account synergy of skills 00:59:43 !apt hu 00:59:43 Hu: Fighting: 0, Short: 0, Long: 0, Axes: 0, Maces: 0, Polearms: 0, Staves: 0, Slings: 0, Bows: 0, Xbows: 0, Throw: 0, Armour: 0, Dodge: 0, Stealth: 1, Shields: 0, UC: 0, Splcast: -1, Conj: 0, Hexes: 0, Charms: 0, Summ: 0, Nec: 0, Tloc: 0, Tmut: 0, Fire: 0, Ice: 0, Air: 0, Earth: 0, Poison: 0, Inv: 1, Evo: 0, Exp: 1!, HP: 0, MP: 0 00:59:44 but crawls skills are basically not balanced "by design" 00:59:53 <|amethyst> gammafunk: in fact we deliberately unbalanced spellcasting and... what amalloy said 01:00:24 well I'm not sure I understand what you're suggesting I guess 01:00:29 Hu have a good stealth apt because stealth is a bad skill, i assume. so we want it to train faster than other things for all species across the board 01:00:39 i mean, not "all" 01:00:45 <|amethyst> once upon a time, spellcasting cost more XP to gain a level than other skills 01:00:46 but the "baseline" for stealth is +1, not 0 01:00:49 <|amethyst> and invo and stealth less 01:00:53 i think stealth is legit 01:00:57 <|amethyst> we changed that by folding it into aptitudes 01:01:01 but i'm also in a minority there 01:01:08 we're certainly not going to reasonably approach gnolls by saying 01:01:35 "well let's rework any schools where it would be bad if gnolls had the same amount of skill in this school compared to most others" 01:01:52 since that's proposing we go rebalance summoning, rebalance necromancy, rebalance ranged, ... 01:02:00 but maybe that's not what you're suggesting 01:02:01 <|amethyst> right, I definitely don't think it's a workable approach 01:02:17 <|amethyst> note that I said "I'm not sure how much I actually believe this" :) 01:02:29 oh, well I actually believe it! 01:02:36 <|amethyst> but it should at least be considered 01:02:37 wait, do so many races have obscene tloc apts because it's trash other than cblink and passage? =p 01:03:02 tloc is insane imo, a great school 01:03:04 <|amethyst> most insane aptitudes are pure flavour 01:03:21 yeah, i know 01:03:21 yeah I don't think we can't really consider it in a sense because we first have to agree that we want to make all crawl skills roughly the same level of strength 01:03:23 and then we'd have to agree how 01:03:31 the great spell in tloc is not cblink, but blink 01:03:31 the "=p" was meant to indicate sarcasm 01:03:31 in the case of changing the stealth apt it was much simpler 01:03:41 i really underuse blink 01:05:10 although, honest question here, are any of the high level tloc spells besides cblink good? 01:05:32 because apport and passage and pproj are all fantastic 01:05:38 |amethyst: also don't tell elliptic that mf having +5 polearms is flavor! 01:05:46 and all require about 5 levels of training to use reliably 01:05:47 +4 01:05:51 oh is it +4? 01:05:55 well it's still insane 01:05:56 !apt mf 01:05:57 Mf: Fighting: 1, Short: 2, Long: 1, Axes: -2*, Maces: -2, Polearms: 4, Staves: -2, Slings: -2, Bows: -2, Xbows: -2, Throw: 0, Armour: -3*, Dodge: 3, Stealth: 2, Shields: 0, UC: 1, Splcast: -1, Conj: -2, Hexes: 0, Charms: 1, Summ: 0, Nec: -2, Tloc: -2, Tmut: 3, Fire: -3*, Ice: 1, Air: -2, Earth: -2, Poison: 1, Inv: 1, Evo: 0, Exp: 0, HP: 0, MP: 0 01:06:09 <|amethyst> gammafunk: why is it polearms rather than, say, axes? 01:06:11 malign gateway is a summon spell, so 01:06:21 ProzacElf: imo the stuff above level 7 is all basically irrelevant because you only put that much xp into a school if you've won anyway 01:06:42 but malign gateway is pretty good, dispersal and disjunction are both okay 01:06:50 |amethyst: if they had +4 axes you'd not call it an insane apt? 01:07:21 if teleport other still exists it is probably the worst tloc spell 01:07:41 yeah, but you can make summoning do most of the heavy lifting on gateway anyway 01:07:54 gell's is worse imo 01:08:00 ugh 01:08:02 oh shit gells exists 01:08:05 rip me 01:08:06 i strongly dislike gell's 01:08:11 it got changed 01:08:14 tloc is a weird mix of really good spells and garbage 01:08:21 i also don't care for force lance 01:08:35 yeah, it's better than it was 01:08:35 but it's still not good 01:08:35 imo 01:08:35 psh, force lance is a decent conj spell 01:08:43 force lance is not very good but I'd not call it garbage 01:08:55 I'm not sure there are really many outright bad tloc spells 01:08:56 maybe i haven't really given it a chance 01:09:03 but it was just trash at first 01:09:03 maybe gell is 01:09:13 ProzacElf: the damage it deals is decent, and it gets monsters to break contact with you so you can flee 01:09:29 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-316-g7453ecb (34) 01:09:32 certainly you could call dispersal/disjunction garbage because they're so high level but their effects are pretty strong 01:09:34 i'll admit that i've had aliches and the like hit me with force lance 01:09:42 and been like "whoa, that was impressive" 01:09:56 <|amethyst> gammafunk: hm, bikeshedding question... 01:10:02 but there's always kind of asymmetry between monster spells and player spells 01:10:09 <|amethyst> gammafunk: should gnoll be moved to the 'Simple' column? 01:10:13 so i'm never sure if it's going to be that cool for me 01:10:17 I was wonderin that myself 01:10:20 <|amethyst> gammafunk: since it is "Simple" and not "Easy" 01:10:24 seems pretty simple in a lot of ways 01:10:36 i really think that it's a problem to wait for the final iteration of gnoll 01:10:38 normal HP, good stats 01:10:55 unless you guys are certain that it's going to stay, and stay this way 01:11:13 we should probably let them go through any more tweaks, but I agree they would possibly be a good fit for simple 01:11:15 <|amethyst> gammafunk: oh, skill help doesn't seem to work for gnolls 01:11:17 but i guess it does fit the definition of "simple" if you want it to be correct for now 01:11:19 <|amethyst> gammafunk: on the m screen 01:11:25 right, someone else mentioned that problem as well 01:11:44 more skill menu rigging is needed, I guess 01:11:58 I will try to update the plan with gnoll items 01:12:30 <|amethyst> also bikeshedding, I do think the species name should be changed 01:12:59 -!- mibe_ is now known as mibe 01:13:12 <|amethyst> since they are nothing at all like the monsters, and I wouldn't want to give new players false impressions about one or the other 01:14:15 +1 01:14:34 yeah, they have fangs 1 and strong nose 01:14:46 so there is some flavor connection there mutation-wise 01:15:16 but it's not exactly easy to explain how their central idea is "gnoll-like" 01:15:34 mikee recommended homonculus but I don't even know what those are 01:16:11 I think he just wanted the Ho prefix 01:16:17 !makewords ho 01:16:20 <|amethyst> err 01:16:27 <|amethyst> there's a problem with that... 01:16:27 hoAE hoAK hoAM hoAr hoAs hoBe hoCK hoCj hoEE hoEn hoFE hoFi hoGl hoHu hoIE hoMo hoNe hoSk hoSu hoTm hoVM hoWn hoWr hoWz 01:16:49 oh 01:16:51 nm 01:16:58 he wanted some other prefix 01:17:32 what on earth prefix did he even want? regardless I don't know if that's a good basis and suggestions welcome 01:17:54 there is some awkwardness with that I guess 01:18:05 I guess it's just another species rename for sequell 01:18:15 <|amethyst> I'd use Hm if that were the name, but no words... 01:18:48 |amethyst surprising against the Hm species! 01:19:01 ??Hmm 01:19:01 Hmm ~ hm ~ |amethyst[2/24]: <|amethyst> hm 01:19:01 *surprisingly 01:19:20 <|amethyst> HmDoh 01:19:43 makings me think of "thinking putty" 01:19:55 ...but you can hmm along? 01:20:16 <|amethyst> !send geekosaur hidden markov models 01:20:16 Sending hidden markov models to geekosaur. 01:20:19 Unstable branch on crawl.develz.org updated to: 0.21-a0-316-g7453ecb (34) 01:21:37 gnolls -> demigods? demigods are very skilled. I'm sure this has already been suggested and dropped 01:21:56 demigods already exist, johnstein 01:21:57 what do you rename demigods then 01:22:07 replace! 01:22:15 no thanks 01:22:23 what am i supposed to 15 rune in the tournament if you do that 01:22:24 <|amethyst> 'halflings' 01:22:34 <|amethyst> since they don't have a real monster to compete with 01:22:38 I know some people don't like both of those species but 01:22:43 just keep them gnolls imo, and remove halfling 01:22:45 a bunch of people (including devs) do 01:22:57 I think I have to fix this oka bug tomorrow, I've been confounded by the fact that actor::skill is one of those overloaded crawl functions where the monster and player version have nothing to do with each other 01:22:59 more so for Dg than for Ha maybe, but I think some devs also like Ha 01:23:03 heh 01:23:18 !lg johnstein dg|ha won s=char 01:23:19 One game for johnstein (((dg || ha)) won): HaBe 01:23:25 hasn't even played them... 01:23:42 !lg . dg 01:23:43 3. johnstein the Bolt Thrower (L3 DgWn), slain by Crazy Yiuf (a +4 quarterstaff of chaos) on D:2 (uniq_crazy_yiuf_cottage) on 2015-04-27 07:04:32, with 57 points after 1870 turns and 0:09:52. 01:23:52 <|amethyst> advil: you shouldn't need to do anything to the monster one should you? 01:23:58 <|amethyst> advil: oh, you're adding a parameter 01:24:01 yeah 01:24:15 guilty! 01:24:15 I guess I just add it to the monster one 01:24:18 I liked Ha 01:24:19 <|amethyst> advil: you can make your version with whatever parameters, then have the existing one call yours 01:24:28 <|amethyst> advil: or that 01:24:44 ah, your way is better 01:25:20 <|amethyst> advil: hm, why not just use real? 01:25:38 it's kind of annoying 01:25:42 <|amethyst> advil: oh, I guess you want Ash to affect the targets 01:25:44 I want to factor in crosstraining and ash bonuses 01:25:50 but not ash penance, or oka 01:26:02 <|amethyst> I guess , bool temp 01:26:06 yep 01:26:11 <|amethyst> in which case it probably is reasonable to add it to monsters 01:26:18 <|amethyst> but three bools in a row :( 01:26:28 I know 01:26:34 <|amethyst> maybe make a bitfield enum? 01:28:51 well, that's definitely too much refactoring to do tonight 01:31:33 maybe I'll just do this quick fix :-( 01:32:01 the joys of finally merging your branch 01:32:11 now you get to fix bugs with added time pressure 01:32:27 it's not even fully live yet 01:32:34 just on cao and cjr I think 01:32:41 oh 01:32:44 &versions 01:32:50 I think cbro has updated 01:32:55 maybe that's a bit later 01:32:57 CAO: 0.21-a0-309-g57ddf66, CBRO: 0.21-a0-297-ga3d34c2, CDO: 0.21-a0-286-g6fcecc0, CJR: 0.21-a0-316-g7453ecb, CPO: 0.21-a0-316-g7453ecb, CUE: 0.21-a0-311-g9ae2175, CWZ: 0.21-a0-277-geaec956, CXC: 0.21-a0-297-ga3d34c2, LLD: 0.21-a0-142-gdd972a7a06 01:33:00 cpo will also have it 01:33:04 rebuilds every 15 mins 01:33:28 I think my shit is almost set up to do releases finally, just gotta push some things and re-tag or w/e 01:33:32 looks like cue also has 01:33:36 maybe it will work!!!!! 01:33:43 hellmonk: how did you enable the builds to be saved btw 01:33:53 or otherwise avail on your github 01:34:11 also did you figure out how to do a crosscompile? 01:34:57 if you needed help with that it's just some mingw packages you need and some specieal make targets and make args for the build command 01:35:35 yeah I ended up doing a bunch of roundabout nonsense to get a secure key so travis can upload things to my releases page 01:36:18 idk if it will work or not bc I have it setup to do it on a tagged release and I havent retagged yet 01:36:58 *tagged commit 01:37:05 you mean an annotated tag? 01:37:23 or signed tag maybe, we only use just annotated ones 01:39:18 I dont think it has to be annotated 01:39:55 maybe it does, idk 01:39:59 maybe it's fine with a lightweight tag 01:40:17 I'll figure it out this weekend I guess 01:43:38 03advil02 07* 0.21-a0-317-gd0d0d88: Fix skill target behavior for heroism (11240) 10(87 seconds ago, 7 files, 28+ 16-) 13https://github.com/crawl/crawl/commit/d0d0d88cf025 01:46:20 Skill target? 01:47:14 Does that pop open the skill menu when you hit a skill level you chose? 01:47:33 no, it stops training it 01:47:51 if you're only training one thing it'll pop open the skill menu 01:48:02 that is a good feature imo 01:48:06 I suppose doing that always could be an alternate possibility 01:48:21 cant wait to spend 10 hours trying to merge it 01:48:28 haha 01:48:35 I have had something like that in my rc for a while 01:48:47 I think I stole it from elliptic 01:49:01 yeah, there's an rc that elliptic/gammafunk put together 01:49:09 hellmonk: it might not be so bad because I suspect you haven't touched most of the places where there were bigger changes 01:49:12 good to hear it's finally implemented fully 01:49:56 the vast majority of effort on that patch was just trying to get some kind of reasonable text input across build targets 01:54:15 I feel like the skill target UI currently requires an awful lot of keys 01:54:24 maybe that's something that could be improved over time, and it's a great feature to have 01:55:03 as people use it more maybe we can find ways to streamline it some 01:55:16 the weapon delay skill target hotkey is cool 01:55:21 wasn't expecting that to exist 01:57:32 Windows builds of master branch on crawl.develz.org updated to: 0.21-a0-316-g7453ecb 02:09:20 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-317-gd0d0d88 (34) 02:39:05 How do I access these new features? Where's the commit? 02:45:29 examine weapon to see the set skill target thing 02:45:44 go to m screen and hit ! to go to skill targets to set them directly 02:54:30 Monster database of master branch on crawl.develz.org updated to: 0.21-a0-317-gd0d0d88 02:59:52 Nice, thanks! 03:22:16 Unstable branch on crawl.beRotato.org updated to: 0.21-a0-317-gd0d0d88 (34) 06:11:25 d2r123 (L9 DrTm) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (D:8) 08:18:08 -!- neizenel is now known as nikheizen 08:19:10 -!- nikheizen is now known as neizenel 08:39:13 gammafunk: agreed re lots of keys 08:39:30 I thought about ways of adding something from the main screen but they didn't seem very satisfactory 08:39:55 (though one thing that would be easy to do is something that just asks you for a skill name and a target in the message window) 08:40:33 the skill menu is sort of already a house of cards, but I don't yet have any great ideas for more radical restructurings to it 08:41:29 one thing I expect to see now is rc files that auto-set targets, one idea I had is something that does it in at least some cases automatically on wield 08:42:02 %git 08:42:02 07advil02 * 0.21-a0-317-gd0d0d88: Fix skill target behavior for heroism (11240) 10(7 hours ago, 7 files, 28+ 16-) 13https://github.com/crawl/crawl/commit/d0d0d88cf025 08:42:15 !lm crash x=src,cv 08:42:16 11. [2017-05-07 09:18:22] [src=cbro;cv=0.18] Crash the Ducker (L4 DsGl of Qazlal) became a worshipper of Qazlal on turn 1348. (D:4) 08:42:36 !lm * crash x=src,cv 08:42:36 18210. [2017-09-29 10:11:25] [src=cao;cv=0.21-a] d2r123 the Fighter (L9 DrTm of Ashenzari) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed on turn 8559. (D:8) 08:44:06 &versions 08:44:20 CAO: 0.21-a0-309-g57ddf66, CBRO: 0.21-a0-317-gd0d0d88, CDO: 0.21-a0-316-g7453ecb, CJR: 0.21-a0-316-g7453ecb, CPO: 0.21-a0-317-gd0d0d88, CUE: 0.21-a0-311-g9ae2175, CWZ: 0.21-a0-277-geaec956, CXC: 0.21-a0-314-gb3271ec, LLD: 0.21-a0-142-gdd972a7a06 08:44:47 ??rebuild 08:44:47 rebuild[1/2]: http://crawl.akrasiac.org/rebuild/ http://underhound.eu:81/rebuild/ http://crawl.berotato.org/crawl/rebuild/ http://crawl.xtahua.com/rebuild/ https://crawl.jorgrun.rocks/rebuild/ Bug |amethyst or Nap.Kin for CDO. Use your powers wisely. 08:54:17 <|amethyst> advil: oh, instead of you_worship(GOD_ASHENZARI) that should probably be have_passive(passive_t::bondage_skill_boost) 08:54:51 ah 08:55:09 <|amethyst> advil: since ash doesn't give the boost if you have too little piety, or if you're under penance 08:56:02 I think in the piety case ash_skill_point_boost will do the calculation correctly, but maybe not penance 08:57:09 <|amethyst> advil: it looks like it would add at least one skill point 08:57:15 hmm 08:57:26 do you know why ash_skill_point_boost wouldn't check this stuff? 08:57:45 <|amethyst> since every factor there (except for species_apt_factor) is greater than or equal to 1 08:57:55 <|amethyst> advil: probably because the caller already checked that 08:58:04 <|amethyst> now that it's used multiple places, maybe it would make sense 08:58:56 yeah, I think it should return 0, I was just too tired last night to work through that 09:03:20 well, it's hard to avoid duplicating code putting the check there, so I guess that's why it isn't there 09:03:29 without rewriting this stuff more drastically 09:05:52 Unstable branch on crawl.akrasiac.org updated to: 0.21-a0-317-gd0d0d88 (34) 09:23:16 Wrong messaging for Maurice stealing part of a stack 13https://crawl.develz.org/mantis/view.php?id=11241 by Yermak 09:27:58 |amethyst: have you ever noticed segfaults if you spam a lot of dprfs? 09:28:12 <|amethyst> advil: no, but 09:28:14 and not segfaults that produce a crash dump either 09:28:29 <|amethyst> advil: I don't run fulldebug very often at all 09:28:36 ah, that's mostly what I run 09:28:45 I've been noticing this lately 09:29:03 something weird in the messaging code maybe 09:32:19 <|amethyst> I can't think of any reason it wouldn't also happen when spamming mprf, but I guess we don't do that to such a huge extent 10:06:08 drock42 (L5 FoHu) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (D:4) 11:00:06 !lm * crash x=src 11:00:06 18211. [2017-09-29 14:06:01] [src=cbro] drock42 the Chucker (L5 FoHu of Wu Jian) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed on turn 3653. (D:4) 11:00:11 hm 11:00:52 !lm * crash -log 11:00:52 18211. drock42, XL5 FoHu, T:3653 (milestone): http://crawl.berotato.org/crawl/morgue/drock42/crash-drock42-20170929-140601.txt 11:02:05 I hate skill math 11:19:22 this time it really is double, I think 11:23:26 plus the wrong assumption that training targets are always integer, which is false for some shields cases 11:26:23 item-specific targets that is 11:34:31 03advil02 07* 0.21-a0-318-g40bfade: Handle ash boosts better in item desc (|amethyst), + cleanups 10(28 minutes ago, 5 files, 79+ 53-) 13https://github.com/crawl/crawl/commit/40bfade87494 11:34:31 03advil02 07* 0.21-a0-319-g17ab366: Remove some doubles, and fix a wrong assumption about shields 10(3 minutes ago, 2 files, 25+ 17-) 13https://github.com/crawl/crawl/commit/17ab366a4688 11:35:16 &versions 11:35:18 ??rebuild 11:35:18 rebuild[1/2]: http://crawl.akrasiac.org/rebuild/ http://underhound.eu:81/rebuild/ http://crawl.berotato.org/crawl/rebuild/ http://crawl.xtahua.com/rebuild/ https://crawl.jorgrun.rocks/rebuild/ Bug |amethyst or Nap.Kin for CDO. Use your powers wisely. 11:35:39 CAO: 0.21-a0-317-gd0d0d88, CBRO: 0.21-a0-317-gd0d0d88, CDO: 0.21-a0-316-g7453ecb, CJR: 0.21-a0-317-gd0d0d88, CPO: 0.21-a0-317-gd0d0d88, CUE: 0.21-a0-311-g9ae2175, CWZ: 0.21-a0-277-geaec956, CXC: 0.21-a0-314-gb3271ec, LLD: 0.21-a0-142-gdd972a7a06 11:36:43 I'm just going to rebuild cbro and cao, hopefully the shield bug is obscure enough (just applies to species with shield skill penalty modifiers I think) that it won't have a huge impact 11:37:33 stupid question: will the rebuild continue if I close these windows? 11:47:31 Unstable branch on crawl.akrasiac.org updated to: 0.21-a0-319-g17ab366 (34) 11:48:53 Unstable branch on crawl.beRotato.org updated to: 0.21-a0-319-g17ab366 (34) 12:01:40 Vall (L23 HuFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Vaults:3) 12:01:54 Vall (L23 HuFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Vaults:3) 12:03:20 haldagan (L27 GnFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Pan) 12:03:51 augh for real 12:04:14 !lm * crash x=src 12:04:15 18214. [2017-09-29 16:03:20] [src=cue] haldagan the Tortoise (L27 GnFi of Cheibriados) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed on turn 81569. (Pan) 12:04:20 !lm * crash x=src -2 12:04:21 18213/18214. [2017-09-29 16:01:54] [src=cao] Vall the Warrior (L23 HuFi of Okawaru) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed on turn 52222. (Vaults:3) 12:04:45 !lm * crash x=src -2 -log 12:04:46 18213/18214. Vall, XL23 HuFi, T:52222 (milestone): http://crawl.akrasiac.org/rawdata/Vall/crash-Vall-20170929-160154.txt 12:06:02 ok, probably the shields thing again 12:08:20 !lm * crash x=src -1 -log 12:08:21 18214. haldagan, XL27 GnFi, T:81569 (milestone): https://underhound.eu:81/crawl/morgue/haldagan/crash-haldagan-20170929-160320.txt 12:08:58 ouch, that one had just killed cerebov 12:09:22 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-319-g17ab366 (34) 12:09:57 %git 9ae2175 12:09:57 07gammafunk02 * 0.21-a0-311-g9ae2175: Unbrace 10(19 hours ago, 1 file, 0+ 2-) 13https://github.com/crawl/crawl/commit/9ae21755b45d 12:11:14 cue is a bit behind, but I can't rebuild there 12:13:51 ok, yeah, the cerebov one is an older version of the bug that is also fixed 12:22:10 Vall (L23 HuFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (D:14) 12:47:49 haldagan (L27 GnFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Pan) 12:48:26 !lm Vall crash -log 12:48:27 5. Vall, XL23 HuFi, T:54645 (milestone): http://crawl.akrasiac.org/rawdata/Vall/crash-Vall-20170929-162209.txt 12:49:56 haldagan (L27 GnFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Pan) 12:52:43 haldagan (L27 GnFi) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Pan) 13:01:27 Nattefrost (L20 KoBe) ASSERT(xp >= 0) in 'player.cc' at line 2702 failed. (Shoals:4) 13:01:42 -!- yesno_ is now known as yesno 13:29:36