00:02:01 <Gretell> Unstable branch on crawl.develz.org updated to: 0.11-a0-2537-gab6943a (33)
00:13:15 <Gretell> Windows builds of master branch on crawl.develz.org updated to: 0.11-a0-2537-gab6943a
00:41:50 -!- phunktion has joined ##crawl-dev
00:42:46 <phunktion> hi all
00:47:20 -!- Patashu[Zzz] is now known as Patashu
01:04:36 -!- phunktion has quit [Ping timeout: 245 seconds]
01:24:15 -!- eeviac_ has quit [Quit: ChatZilla 0.9.88.2 [Firefox 13.0/20120601175215]]
01:46:53 -!- ussdefiant has quit [Ping timeout: 246 seconds]
02:21:51 -!- alefury has joined ##crawl-dev
02:24:13 -!- phunktion has joined ##crawl-dev
02:45:07 -!- Nomi has quit [Ping timeout: 264 seconds]
02:54:54 <alefury> crate suggested making oka heroism a passive and increasing piety decay, because with the current cost it makes sense to use it before any even remotely interesting fight
02:55:21 <alefury> any opinions?
02:56:13 <evilmike> I'd rather increase the cost of heroism so you can't use it every fight
03:03:44 -!- dtsund has quit [Ping timeout: 244 seconds]
03:04:00 <alefury> well yeah, thats another option
03:04:46 <alefury> i think faster piety decay would be interesting, though, and this would be one way to introduce it for a single god
03:05:56 <|amethyst> IMO faster piety decay would make more sense for Trog than Okawaru
03:06:05 <|amethyst> WHY YOU NO KILL????!!!
03:06:14 <alefury> imo it would make sense for lots of gods (see tavern)
03:06:24 <evilmike> trog already makes it fairly hard to get to ******
03:06:37 <evilmike> I mean, you can do it, but it's normal to be at *****. most of the game
03:06:49 <alefury> https://crawl.develz.org/tavern/viewtopic.php?p=65951#p65951
03:06:51 <alefury> and the following
03:08:12 <alefury> i think piety fluctuating pretty strongly both with time and ability use would be neat, because having a piety clock that actually hurts if you ignore it would help with a lot of crawl's problems
03:08:40 -!- ToBeFree has joined ##crawl-dev
03:09:21 -!- SoulOfTheInterne has joined ##crawl-dev
03:12:08 -!- ToBeFree has quit [Read error: Connection reset by peer]
03:12:24 -!- SoulOfTheInterne has quit [Read error: Connection reset by peer]
03:13:11 <evilmike> the game doesn't even have a piety clock
03:14:34 <alefury> for some gods it does
03:14:59 <alefury> trog piety decays fairly fast already, its especially noticeable after gifts
03:15:23 <evilmike> the ones that accept corpse sacrifices have higher piety decay. I don't know why, I assume it's because there are more ways to get piety for them
03:15:52 <alefury> could also be for the same flavor reasons i mentioned on tavern
03:16:07 <alefury> they want you to kill dudes all the time
03:16:34 <|amethyst> okawaru is about precision, though
03:17:11 <alefury> well, he is now. and him wanting to precisely kill dudes all the time still works.
03:17:20 <|amethyst> heh
03:17:27 <evilmike> I think okawaru and trog are the same, with piety decay (both gift, as well)
03:17:31 <evilmike> trog has more expensive abilities though
03:18:26 <alefury> heh, and you think they are still too cheap (im not really arguing there)
03:19:14 <evilmike> well, I think a lot of abilities are too cheap. Some are closer to where they should be
03:19:22 <evilmike> I wouldn't increase the cost of trog's hand, for example
03:19:28 <evilmike> heroism is too cheap though (1 piety)
03:20:19 -!- ainsophyao has quit [Remote host closed the connection]
03:20:49 <|amethyst> what about Berserk?
03:21:08 <mikee_> piety cost for berserk? =P
03:21:19 <evilmike> oh wait, it's not quite 1, there's a formula
03:21:28 <evilmike> comment says random2( (piety_cost + 1) / 2 + 1 )
03:21:31 <|amethyst> I guess having to rest off the Slow is enough of a piety cost
03:21:58 <|amethyst> evilmike:  so 0.5
03:22:13 -!- ainsophyao has joined ##crawl-dev
03:22:31 <alefury> |amethyst: actually the comment was misquoted
03:22:34 <alefury> i think
03:22:46 <evilmike> oh, yeah
03:22:47 <alefury> i read it as 1d3 for heroism
03:22:52 <evilmike> there's a "+" before that
03:23:22 <|amethyst> so 1d2
03:23:23 <alefury> oh wait, 1d2
03:23:40 <alefury> i always forget random2 goes to argument-1
03:25:25 <|amethyst> you remembered the zero though
03:25:48 <alefury> yeah, no problems on that end :)
03:27:20 <|amethyst> random2() has always struck me as a funny interface
03:29:22 <|amethyst> I mean, it's perfect for picking an item from an array
03:29:34 <|amethyst> but then it gets used in damage formulas and stuff
03:31:30 <|amethyst> at least we don't fudge the dice rolls based on the player's luck
03:34:47 <alefury> that can actually be pretty cool (binding of isaac, again)
03:34:58 <alefury> not for crawl though
03:36:01 <alefury> luck influences which room layouts (and i think enemy sets) you get there, also some less interesting stuff
03:36:22 -!- ToBeFree has joined ##crawl-dev
03:36:31 <alefury> in the expansion theres a challenge where you start with minimum luck, and its really noticeable. and fucking hard.
03:37:23 <|amethyst> I was referring to another well-known game
03:38:03 <|amethyst> where it is quite pervasive
03:39:32 <|amethyst> well, I guess 54 instances isn't "pervasive"
03:39:40 <|amethyst> still :P
04:02:27 <Cheibriados> Merfolk don't fully autoexplore shoals anymore (https://crawl.develz.org/mantis/view.php?id=5856) by absolutego
04:11:04 -!- syllogism has joined ##crawl-dev
04:11:35 -!- evilmike has quit [Read error: Connection reset by peer]
04:14:10 -!- absolutego has joined ##crawl-dev
04:37:13 -!- st_ has joined ##crawl-dev
04:55:04 -!- edlothiol has joined ##crawl-dev
05:00:02 <Gretell> OSX builds of master branch on crawl.develz.org updated to: 0.11-a0-2537-gab6943a
06:06:59 -!- blabber has joined ##crawl-dev
06:27:26 <CIA-50> 03edlothiol * rc009cb19910b 10/crawl-ref/source/webserver/static/scripts/client.js: Fix lobby idle time updating in webtiles.
06:27:36 <CIA-50> 03edlothiol * r6f2e67e1cc4e 10/crawl-ref/source/ (3 files in 2 dirs): Implement old_blood rendering in webtiles.
07:22:05 -!- blabber has quit [Quit: leaving]
07:34:36 -!- ToBeFree has quit [Remote host closed the connection]
08:15:49 -!- kilobyte has quit [Ping timeout: 265 seconds]
08:21:36 -!- kilobyte has joined ##crawl-dev
08:37:26 <Gretell> Grildrak (L12 DESu) ERROR in 'mon-movetarget.cc' at line 109: ZotDef: monster it failed to pathfind to (40,26) (the Orb) (D:1 (ZotDef))
08:41:30 -!- absolutego has quit [Quit: leaving]
08:51:51 -!- Nomi has joined ##crawl-dev
09:06:28 <alefury> can random jelly spawns please be disabled for 0.11?
09:07:00 -!- Patashu is now known as Patashu[Zzz]
09:07:29 -!- MarvinPA has joined ##crawl-dev
09:16:41 -!- phunktion has quit [Ping timeout: 245 seconds]
09:23:32 <elliptic> alefury: I don't really see the point
09:24:03 <alefury> making people stash in lair is silly
09:24:15 <elliptic> jelly spawns have nothing to do with that
09:24:39 <alefury> sure they do. please remember that not everyone is a good player :P
09:24:44 <elliptic> many monsters can use many items
09:25:12 <elliptic> and you can't be sure that you've killed all the jellies on a level even if you've "cleared" it anyway
09:25:50 <elliptic> how does this have anything to do with being a good player? orcs pick up and use up potions, wands, etc
09:26:15 <alefury> just dropping stuff in a corner is generally totally fine, right?
09:26:34 <alefury> but people dont realize that, because they are deathly afraid of losing items.
09:26:42 <alefury> stuff being picked up is not permanent, stuff being eaten is.
09:27:09 <elliptic> "and use up"
09:27:43 <elliptic> are you planning to post a big sign saying "jellies don't respawn"? since most players will have no way of knowing this
09:29:53 <alefury> people learned about "having" to stash on lair:2, people will unlearn it too
09:29:55 <elliptic> people can figure out on their own that lair doesn't have orcs, jellies, etc... you are talking about adding in something spoily
09:30:14 <elliptic> not to mention hacky
09:30:29 <alefury> oh well
09:32:53 -!- Grunt has joined ##crawl-dev
09:34:10 -!- headzone has quit [Ping timeout: 276 seconds]
09:35:32 <alefury> hm, i think heroism helps identify items. good? bad?
09:36:07 <elliptic> bad IMO, but kilobyte is the person who knows how it is supposed to work
09:42:21 -!- syllogism- has joined ##crawl-dev
09:42:46 -!- ussdefiant has joined ##crawl-dev
09:43:38 -!- syllogism has quit [Ping timeout: 240 seconds]
09:49:57 -!- headzone has joined ##crawl-dev
09:50:56 -!- ussdefiant_ has joined ##crawl-dev
09:51:38 -!- ussdefiant has quit [Ping timeout: 240 seconds]
09:51:39 -!- ussdefiant_ is now known as ussdefiant
10:01:22 -!- Wensley has joined ##crawl-dev
10:03:41 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.11-a0-2539-gc009cb1 (33)
10:35:49 -!- ussdefiant has quit [Ping timeout: 248 seconds]
10:42:45 -!- Nomi has quit [Ping timeout: 265 seconds]
10:51:57 -!- SkaryMonk1 has joined ##crawl-dev
11:24:54 -!- G-Flex has joined ##crawl-dev
11:28:01 <Gretell> PrincessTileah (L27 VpAs) ASSERT(branch != NUM_BRANCHES && depth != -1) in 'travel.cc' at line 2983 failed. (D:32)
11:28:30 -!- blabber has joined ##crawl-dev
11:40:20 -!- mikee_ has quit [Quit:  HydraIRC -> http://www.hydrairc.com <- In tests, 0x09 out of 0x0A l33t h4x0rz prefer it :)]
11:48:50 -!- Xiberia has quit [Ping timeout: 260 seconds]
11:54:01 -!- lodestone has joined ##crawl-dev
11:59:44 -!- Dixie has joined ##crawl-dev
12:04:20 -!- lodestone has quit [Quit: lodestone]
12:27:55 -!- Keskitalo has quit [Ping timeout: 260 seconds]
12:28:22 -!- Keskitalo has joined ##crawl-dev
12:35:35 <kilobyte> alefury, elliptic: there was no instant identify when Heroism was added, yeah.  And the idea behind new autoid assumes skills don't generally cheaply increase.
12:35:59 <elliptic> kilobyte: the autoid should use the unboosted skill then
12:36:13 <kilobyte> it can be argued both ways, as followers of a god of battle may have a passive clue about what weapons are good, but yeah
12:36:22 <kilobyte> not sure what with Ashenzari
12:49:02 <CIA-50> 03kilobyte * rc1762c3f1e3a 10/crawl-ref/source/misc.cc: Don't count Heroism nor Ashenzari when identifying weapons by skill.
12:49:12 <CIA-50> 03kilobyte * r5de740afa3f7 10/crawl-ref/source/dat/database/ (rand_all.txt rand_arm.txt rand_wpn.txt): An imperial buttload of randart name parts, by jeffqyzt.
12:54:36 -!- kek has joined ##crawl-dev
13:12:58 -!- eeviac_ has joined ##crawl-dev
13:13:19 -!- aleksil has joined ##crawl-dev
13:14:42 <eeviac_> ??silver
13:14:43 <Henzell> silver[1/1]: +5% damage per mutation (up to *1.75), 1.75x vs chaotic things (marked as such in their description). This means shapechangers, mutaters, ugly things, abominations, Tiamat, Killer Klowns, chaos spawns.
13:20:34 <ChrisOelmueller> !learn edit silver s,changer,shifter
13:20:35 <Henzell> silver[1/1]: +5% damage per mutation (up to *1.75), 1.75x vs chaotic things (marked as such in their description). This means shapeshifters, mutaters, ugly things, abominations, Tiamat, Killer Klowns, chaos spawns.
13:22:27 -!- Textmode has joined ##crawl-dev
13:36:38 -!- dtsund has joined ##crawl-dev
13:59:58 -!- eeviac_ has quit [Quit: ChatZilla 0.9.88.2 [Firefox 13.0.1/20120614114901]]
14:06:16 -!- ToBeFree has joined ##crawl-dev
14:18:07 <CIA-50> 03MarvinPA * r41df555bbb78 10/crawl-ref/source/view.cc: Change lantern of shadows effect to allow distinguishing LOS boundary and monster colours
14:26:00 <|amethyst> galehar:  sorry I'm being a little slow... still working on the constriction implementation change... I have save compat worked out, and just need to handle monster_info and do some testing
14:26:13 <Grunt> |amethyst, what's changing?
14:26:14 <|amethyst> galehar:  then the gameplay changes should be straightforward
14:27:14 <|amethyst> galehar:  I haven't really given a lot of thought to the 'held' stuff... I've added the enum and the actor member, but it's currently only used for constriction (and even then, not really "used")
14:27:51 <|amethyst> Grunt:  implementation-wise, the constriction array (using mindexes) is turning into a map (using mids)
14:28:01 <Grunt> Aha.
14:28:05 <Grunt> Sounds sensible.
14:28:50 <|amethyst> Grunt:  gameplay-wise, there will be damage formula changes, blink (but not teleport) prevention, and "holding" (constriction without damage)
14:29:09 <ChrisOelmueller> people will be incredibly sad if this does not make it into 0.11, and they also will complain if it does
14:30:01 <|amethyst> blink won't be completely prevented--it will count as N escape attempts (probably N=2)
14:30:21 <|amethyst> then there are some mimic changes that depend on this
14:30:30 <Eronarn> the blink change is still real dumb
14:30:31 <Grunt> What kind of mimic changes?
14:31:39 <|amethyst> Grunt:  plain mimics will deal no damage with constriction, and ravenous will do damage but less
14:31:50 <Grunt> Aha.
14:32:03 <Grunt> "The door mimic grabs you!"
14:32:25 <Grunt> I'm curious - what's the rationale for the blink change?
14:32:43 <|amethyst> you'll have to ask galehar
14:32:58 <|amethyst> It was always intended that you can't simply blink away from constrictors
14:33:38 <|amethyst> that was removed because of bugs, but a simpler implementation (without the "along for the ride" stuff, etc) could fix that
14:33:50 <Eronarn> the rationale is that there isn't one, it's just there because some people don't like blocking and some don't like not blocking
14:34:00 <Eronarn> so it's just a stupid compromise that makes no sense at all
14:34:20 <|amethyst> I'm not sure about that making it into 0.11
14:34:58 <ChrisOelmueller> i don't think you should release with the current mimics
14:35:29 <ChrisOelmueller> not allowing blink would be something that needs more testing than there can happen now, probably
14:35:38 -!- Dixie has quit [Ping timeout: 240 seconds]
14:35:47 <|amethyst> ChrisOelmueller:  that was my thought
14:36:25 <Grunt> Blocking blink altogether doesn't make very much sense, in my view.
14:36:37 <Grunt> What is blinking? You're suddenly disappearing and reappearing in another position.
14:37:18 <Eronarn> there's some rationale for it with statues and stuff blocking blinking
14:37:31 <Eronarn> but unless we make monsters start blocking blinking, it's not a strong one
14:37:44 <|amethyst> I believe kilobyte has a similar objection
14:38:30 <Eronarn> what's really awful though is having it count as an escape attempt
14:40:25 -!- Twinge has quit [Ping timeout: 252 seconds]
14:42:34 <|amethyst> I'm not entirely convinced that "you can blink through non-constrictors, why not constrictors?" is all that good an objection---after all, blink takes your inventory with you, but blinking across a floor item doesn't move it
14:43:06 <|amethyst> really I think it should be evaluated in terms of gameplay, not magic realism
14:43:11 <Grunt> I'm not sure how that argument would even come up in the first place.
14:43:34 -!- Twinge has joined ##crawl-dev
14:43:44 <Grunt> You don't blink "through" anything; you disappear suddenly and reappear somewhere nearby.
14:43:49 <ChrisOelmueller> well i am clearly biased because i found "coming along for a ride" absolutely disgusting and not fitting into crawl
14:43:53 <Grunt> ...perhaps by means of spatial distortion.
14:44:05 <Grunt> (Hence Bend Space.)
14:44:22 <ChrisOelmueller> but what are advantages over the current implementation, game-play wise?
14:45:33 <ChrisOelmueller> everything that tries to hurt spellcasters with easy access to blink also hurts melee even more, since blink scrolls (a valuable item) are currently often consumed to escape
14:45:59 <|amethyst> I could see the argument for making it apply only to the spell
14:46:09 <ChrisOelmueller> and you're taking even that last "guaranteed" instant escape , making teleport your best bet in a hairy situation
14:46:19 <ChrisOelmueller> i wouldn't special-case either
14:46:22 <elliptic> currently a L2 spell and relatively common randart ability is a guaranteed escape
14:46:26 <elliptic> this isn't very good
14:46:34 <ChrisOelmueller> (not that blink at L2 is a great spell to have at all)
14:46:49 <ChrisOelmueller> i would be very fine with only having scblink and cblink available
14:47:01 <ChrisOelmueller> but yeah, that's not the topic. :)
14:47:52 <|amethyst> maybe that could be it... controlled and semicontrolled let you out, uncontrolled is just N escape attempts
14:48:00 -!- evilmike has joined ##crawl-dev
14:48:20 <Grunt> I don't buy the "N escape attempts". I don't think it makes any sense.
14:48:26 <elliptic> why not?
14:48:29 <elliptic> you are trying to pull away
14:48:32 <elliptic> you might succeed
14:48:48 <Grunt> So, blinking is "you're pulling away with N times as much force"?
14:50:48 <Grunt> The mechanics of casting a blink spell (or evoking it from an artefact) work rather differently than attempting to use brute force to pull away from something crushing you to death.
14:51:02 <elliptic> do they really
14:51:09 <elliptic> glad to know you are such an expert on crawl realism
14:52:37 <Grunt> I don't necessarily expect realism, but I do at least expect consistency.
14:52:52 <Grunt> ...and having a spell behave completely differently just because there's a monster next to you isn't very consistent.
14:53:18 <evilmike> this doesn't strike me as an inconsistency
14:53:26 <evilmike> if there's a problem, we just write it into the spell description
14:54:10 <elliptic> personally I have no experience with the physics of translocation spells and how they distinguish between the scrolls in your pocket and the tentacle wrapped around you
14:54:20 <elliptic> that might just be me though!
14:54:29 <evilmike> you're really missing out
14:55:43 <elliptic> Grunt: it isn't just because "there's a monster next to you"
14:56:06 <|amethyst> just forbid casting and equipment swaps while constricted, let those already wearing a +blink artefact get out... problem solved
14:56:30 <Grunt> I do think "no casting while constricted" makes sense.
14:56:40 <Grunt> It would be hard to wave your arms around while being crushed to death, after all.
14:56:44 <|amethyst> (I think that would be too harsh)
14:56:55 <evilmike> I think crawl's casting has no physical requirement, it's just words
14:56:57 <elliptic> Grunt: what makes you think that casting involves waving your arms around?
14:57:14 <|amethyst> also, you can swing your sword around
14:57:41 <|amethyst> (and if you remove that too, might as well just use melee paralysis)
14:59:13 <Grunt> It might not be the case for all spells, but are monsters casting spells not consistently described as "gesturing", "waving [their] arms", or what have you.
14:59:50 <evilmike> you probably could get away with making blink just count as 1 escape attempt
14:59:57 <evilmike> it's not *that* hard to escape by moving away, now
15:00:20 <|amethyst> a gesture could be like a mudra... and arm-waving only happens with airstrike
15:00:51 <Grunt> Sure it is, if you happen to be, say, a Sp or DE or something else with low strength.
15:01:21 <Eronarn> it's good to have mechanics that hurt spellcasters more than melee
15:01:22 <Eronarn> 'let's block blink and teleport' is just an attempt to shoehorn that into constriction, because it happens to be a newer mechanic
15:02:02 <elliptic> Grunt: looking at monspeak.txt, most of the time gesturing isn't mentioned
15:02:16 <Grunt> That's because you want to be looking in monspell.txt, not monspeak.txt.
15:02:48 <elliptic> I still don't see many gestures
15:03:00 <Eronarn> also does using a large weapon still reduce spellcasting success? see also shield and armor... it's pretty clear that crawl spells are intended to be like D&D ones, where you do need freedom of movement to some degree
15:03:04 <elliptic> Eronarn: no
15:03:34 <elliptic> "intended to be"
15:03:48 <Eronarn> 'are often directly copied from', is that better
15:04:06 <elliptic> I'd like to know who it was who intended this
15:04:10 <elliptic> and why we should care
15:04:23 <MarvinPA> crawl realism arguments, the most constructive arguments of all
15:04:27 <elliptic> obviously we could flavor things like that if we wanted
15:04:48 <elliptic> but we certainly aren't locked into doing so
15:04:58 <ChrisOelmueller> MarvinPA: constrictive*
15:05:10 <MarvinPA> ho ho
15:05:32 <Eronarn> constriction blocking spellcasting is wrong anyways, i think it should just be a penalty
15:06:03 <elliptic> a penalty to casting success? that does sound more plausible
15:06:18 <Grunt> That actually makes more sense than blocking it altogether.
15:07:15 <Eronarn> a while ago what i had been thinking of for armor was a flat miscast rate, rather than adjusting your miscast chance, because that penalizes even very good spellcasters 
15:07:18 <Eronarn> i think that same rationale applies here
15:07:21 <Eronarn> you shouldn't be able to cast magic dart  or blink 100% of the time when constricted even if you are an archmage
15:07:42 <Grunt> ...in much the same sense that even an archmage could not cast those spells 100% of the time while wearing plate mail.
15:07:50 <Eronarn> it makes any anti-caster drawback really be 'anti-weak-caster'
15:08:11 <|amethyst> I like that it's feasible to be an armoured caster
15:08:26 <|amethyst> having a (say) 20% fail rate on all your spells would kind of stop that
15:08:32 <elliptic> Eronarn: I'm not necessarily opposed to the flat miscast rate idea but I think a hypothetical constriction casting penalty should be the same type as penalty as all the other penalties we have
15:08:44 <elliptic> s/type as penalty/type of penalty/
15:09:06 <Eronarn> |amethyst: you could go with very small penalties, really; even single digits is harsh
15:09:21 <Eronarn> also armor skill, etc.
15:09:35 <|amethyst> Eronarn:  single digits isn't harsh for bad casters, though
15:10:14 <Eronarn> elliptic: i think the current penalty is not wel designed, there is no easy way to communicate to players what affects it. also there is the possibility for having two different effects: like constriction based failures get a different message, no miscast effect, no lost MP
15:10:30 <Eronarn> that would make them feel rather different
15:10:46 <elliptic> Eronarn: this sounds like just making the system more complicated
15:10:48 <Eronarn> |amethyst: right, that was the point - you could have low level casters in plate, even with low skill, if they were willing to have (say) a 10% failure rate
15:11:19 <Eronarn> but high level casters would have to actually think about whether they want dragon armor affecting their spellcasting rates
15:11:26 <elliptic> you have to think about it now
15:11:43 <Eronarn> not to the same degree that you do at low spellcasting skill
15:11:44 <|amethyst> Eronarn:  you're making it hurt the low-level spells more than the high ones, though
15:11:51 <elliptic> Eronarn: sure you do
15:12:36 <elliptic> unless "high level casters" means "all skills in the 20s", heavy armour penalties are noticeable
15:12:40 <Eronarn> elliptic: putting on armor at low skill can bring you from okay to deadly miscast, instead of from great to less great
15:12:56 <elliptic> no, because you are casting different spells
15:12:57 <|amethyst> Eronarn:  that's not a matter of just skill though
15:13:04 <elliptic> it takes different spells from great to less great
15:13:08 <elliptic> in both cases
15:13:34 <|amethyst> Eronarn: currently, at lower skill you have to decide "Is this armour worth making my L3 spells go from 1% to 11%"; and at higher skill, "Is it worth making my L7 spells go from 1% to 11%"?
15:13:39 <|amethyst> what elliptic said
15:14:04 <elliptic> or 5% to 20% or whatever
15:14:34 <Eronarn> the issue being that you don't have many spell levels to fall back on at low level
15:14:35 <|amethyst> anyway, I've got to go now... have a fruitful discussion
15:15:24 <elliptic> the current system isn't great and I'm open to changing it, I just would rather make it simpler than just add constriction as an extra component that works different from everything else
15:16:59 <Eronarn> possibly just make it work the same as nets
15:17:01 <|amethyst> !tell galehar  I made a few comments to you about my constriction progress in the scrollback (around 14:26 in the online logs, 19:26 UTC).  Also, should the escape forulae be tweaked along with the damage formulae?
15:17:15 <Eronarn> until we actually overhaul spellcasting
15:17:23 <Henzell> |amethyst: OK, I'll let galehar know.
15:25:53 -!- elliptic has quit [Ping timeout: 246 seconds]
15:27:14 -!- elliptic has joined ##crawl-dev
15:35:53 -!- MarvinPA_ has joined ##crawl-dev
15:37:37 -!- MarvinPA__ has joined ##crawl-dev
15:39:05 -!- MarvinPA has quit [Ping timeout: 252 seconds]
15:39:05 -!- MarvinPA__ is now known as MarvinPA
15:41:08 -!- MarvinPA_ has quit [Ping timeout: 240 seconds]
15:41:34 -!- ainsophyao has quit [Remote host closed the connection]
15:42:38 -!- aleksil has quit [Ping timeout: 240 seconds]
15:42:53 -!- yazz has joined ##crawl-dev
15:43:20 -!- yazz has left ##crawl-dev
16:00:29 -!- frogbotherer has joined ##crawl-dev
16:15:17 -!- blabber_ has joined ##crawl-dev
16:15:43 -!- blabber has quit [Ping timeout: 264 seconds]
16:18:08 <stabwound> wow, the webtiles version is pretty impressive, i must say
16:18:16 <stabwound> it's come a long way since when it was first released
16:18:48 <stabwound> almost makes me wish you could play it with ascii instead :s
16:19:56 <stabwound> does it only work with CDO or is there a version for CAO?
16:20:37 <Wensley> stabwound: you can play it with ascii, but last I checked it was a little funky
16:20:43 <Wensley> cdo only
16:20:46 <stabwound> ah
16:21:00 <Wensley> there's a button somewhere to turn on ascii mode
16:21:53 <Wensley> aha!
16:21:56 <Wensley> stabwound: press f10
16:23:16 <stabwound> cool, that does work although the font is pretty funky looking
16:23:51 <stabwound> or maybe it's not actually a font
16:25:11 <edlothiol> stabwound: the font is whatever is Monospace on your system
16:28:10 <elliptic> edlothiol: last I checked, glyph mode in webtiles didn't have the branding that console has (e.g. sleeping monsters are blue)... any chance of getting that?
16:36:56 <kilobyte> |amethyst: if you have any problems with save compat, there is hardly any reason to waste time there
16:37:41 <kilobyte> the only downside would be a trunk game or two not being able to transfer to a version with some bugfix
16:38:00 <kilobyte> nothing else cares, as 0.11 is not compatible with 0.10 anyway
16:39:39 <edlothiol> elliptic: hm, sure, I'll look into it
16:40:35 <evilmike> speaking of webtiles, there's a new (trunk) webtiles server
16:40:45 <evilmike> it's probably worth a blog post at least
16:41:00 <edlothiol> I don't think the server is quite ready for that
16:41:43 <kilobyte> we really need to think about an infrastructure to add new temporary servers for the tourney
16:42:35 <kilobyte> at least, doing some planning
16:44:45 <kilobyte> elliptic: do you have any plans when exactly the tourney would be?
16:45:25 <elliptic> I haven't really thought about the next tourney at all yet
16:47:11 <kilobyte> I'm asking mostly because it'd be good to know when a feature freeze should be
16:49:29 <kilobyte> I had the impression that the tourney is set for the second half of August, which means a feature freeze in the beginning of July, and a full freeze not long after
16:53:32 <elliptic> looking at the calendar, Aug 4-19 or Aug 11-26 would be the options in August with the two-week schedule starting on a saturday and finishing on a sunday
16:54:51 <elliptic> Aug 18 - Sep 2 would be reasonable also presumably
17:10:13 <CIA-50> 03galehar * rb8e867a99dd4 10/crawl-ref/source/ (godprayer.cc religion.cc stash.cc): Make god_likes_items return false for GOD_NO_GOD instead of error.
17:10:23 <CIA-50> 03galehar * r754e3deff954 10/crawl-ref/source/stash.cc: Only visit changed stashes if needed (#5849).
17:15:04 <Keskitalo> Hnygh, haven't had time for the tutorial, unlikely that I'll have any tomorrow either.. bnext week then!
17:15:29 <kilobyte> I'm fixing the merfolk in Shoals not autoexploring bug, and I noticed that beogh_water_walk() doesn't allow autoexploring
17:16:19 <galehar> you mean it only autoexplore islands, not the sea
17:16:19 <Henzell> galehar: You have 2 messages. Use !messages to read them.
17:16:29 <galehar> !messages
17:16:30 <Henzell> (1/2) faze said (1d 5h 8m 37s ago): I saw you made show_gold_turns and game_turns default to true. How about show_inventory_weights = true while you're at it? :)
17:16:44 <kilobyte> technically, it is correct (you can possibly run out of piety), but it seems bad in a vast majority of cases.  Any ideas?  Perhaps it should act as good enough if you're on the next piety step?
17:17:01 <galehar> show_inventory_weights should be auto be default
17:17:03 <galehar> !messages
17:17:15 <Henzell> (1/1) |amethyst said (1h 59m 53s ago): I made a few comments to you about my constriction progress in the scrollback (around 14:26 in the online logs, 19:26 UTC).  Also, should the escape forulae be tweaked along with the damage formulae?
17:17:38 <kilobyte> galehar: your commit 35eb27d2 stopped merfolk from autoexploring parts with too low sea floor, making them not notice islands
17:18:21 <galehar> I'm not sure that's a bad thing
17:18:22 <kilobyte> galehar: a fix seems trivial, I'd exclude everyone with permanent water walk from it
17:18:30 <galehar> are items ever generated below sea level?
17:19:13 <kilobyte> no, but the player receives no feedback why some of identically looking deep water gets explored and some does not
17:19:27 <galehar> Seems good that autoexplore doesn't waste its time stupidily exploring the sea
17:19:39 <galehar> yeah, that's true
17:20:33 <galehar> yeah, it's probably better to let permanent flyers/swimmers to explore the sea then
17:21:21 <kilobyte> the logic is in feat_is_traversable_now(), it looks good to me
17:21:36 <kilobyte> except for the Beogh issue, which is not restricted to just Shoals
17:23:10 -!- blabber_ has quit [Quit: leaving]
17:23:26 <kilobyte> (and +10 to show_inventory_weights, of course)
17:23:28 -!- blabber has joined ##crawl-dev
17:24:03 <galehar> feat_is_traversable_now looks good to me too
17:31:47 <galehar> |amethyst: can you push your thing to a branch?
17:32:03 <CIA-50> 03kilobyte * r12d28aa9d876 10/crawl-ref/source/travel.cc: Allow merfolk (and others with unlimited water traversal) to fully explore Shoals.
17:32:42 <kilobyte> if you're doing to constriction what I am thinking about, keeping compat would be quite a bit of work
17:32:54 <kilobyte> (not sure if you're doing that kind of cleanup)
17:32:58 <galehar> |amethyst: I've read the backlog. I'm sorry, but your few comments have been drowned in a blink/constriction balance argument
17:33:13 <galehar> I mean I've read some of the backlog...
17:34:06 <galehar> kilobyte: well, I've talked with |amethyst of the same implementation we've talked before. And he said he's done save compat.
17:34:49 <galehar> |amethyst: about escape attempts, no need to touch the formula. This one has already been adjusted and it's working fine now
17:35:57 <galehar> about constriction blocking blink, I think it won't be a too harsh nerf, because you'll still be very likely to escape with 2 blinks. 3 at most.
17:37:37 <kilobyte> which against tentacled monstrosities can take all of your maxhp
17:38:03 <galehar> the damage formula: dam * (1 + stepdown(time, 5) / 5)
17:38:23 <galehar> which means reduce base constriction damage of high HD monsters
17:38:37 <galehar> like tentacled monstrosities
17:39:47 -!- ToBeFree has quit [Read error: Connection reset by peer]
17:44:16 <kilobyte> if you're standing on a pile that contains poisonous, contaminated and clean chunks, confirm_butcher=never will happily go with useless or bad chunks even if there's another, just a single turn older, good corpse below
17:45:44 <kilobyte> any reasons to not make it use the "best" corpse instead?  ("Best" defined by the same formula as drop_spoiled_chunks().)
17:46:05 <galehar> yeah, it should prioritize quality over freshness
18:04:37 <frogbotherer> kilobyte: any chance you could tell me what screen you were looking at to break the unicode stuff?
18:06:57 -!- blabber has quit [Quit: leaving]
18:08:45 <kilobyte> frogbotherer: the manual viewer, with ../docs/crawl_manual.txt overwritten with https://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-demo.txt
18:09:19 <frogbotherer> ah! so you deliberately set out to break it :)
18:09:21 <kilobyte> (the manual is autogenerated, so the file needs to have a timestamp newer than the source or make will rebuild it)
18:10:00 <frogbotherer> the way i've written it, there's a tradeoff between the total number of unique characters that can be displayed in one go and the amount of memory it uses
18:10:13 -!- eeviac_ has joined ##crawl-dev
18:10:15 <frogbotherer> because it still stuffs each glyph into part of a texture
18:10:18 <kilobyte> this file has nothing fancy: no combining characters, only five CJK ones
18:10:49 <frogbotherer> i'll have a look (i'd set the hard limit to 255 unique glyphs on the screen at once, tho)
18:10:57 <kilobyte> bumping the number above 256 somewhat helps, but the "animation" is still there
18:11:23 <kilobyte> lemme rebase away my wip parts and push complete fixes
18:12:38 <frogbotherer> yeah, it's cos it's trying to write two glyphs into the same part of the texture in the same draw of the screen
18:13:30 <frogbotherer> i'll repeat what you've done and see if there's a way to make it smarter
18:14:08 <frogbotherer> (also: nothing fancy? this has got ethiopian and braille in it!) :)
18:14:10 <CIA-50> 03evilmike * rc824640e261a 10/crawl-ref/source/dat/des/serial/ (6 files): Tag most serial vaults with luniq_serial.
18:14:13 <CIA-50> 03kilobyte 07unicode-tiles * r68da026a57dd 10/crawl-ref/source/fontwrapper-ft.cc: Don't let debugging printf spew invalid characters.
18:14:13 <CIA-50> 03kilobyte 07unicode-tiles * r71a6d884bb97 10/crawl-ref/source/fontwrapper-ft.cc: Formatting.
18:14:14 <CIA-50> 03kilobyte 07unicode-tiles * re8cd0829e3d5 10/crawl-ref/source/fontwrapper-ft.cc: A pixel = 1px, 4 bytes.
18:14:14 <CIA-50> 03kilobyte 07unicode-tiles * r47acb2b5b0c2 10/crawl-ref/source/fontwrapper-ft.cc: Fix a crash on drawing out of a character's bounds.
18:14:24 <CIA-50> 03kilobyte 07unicode-tiles * rb06106206de2 10/crawl-ref/source/fontwrapper-ft.cc: Fix a potential integer overflow.
18:14:29 <kilobyte> I am almost sure the droppings come from cropping to bmp->width×bmp->height rather than max_width×max_height
18:15:17 <kilobyte> "nothing fancy" as: all glyphs are plain wcwidth() == 1 stuff
18:15:24 <frogbotherer> i had so much trouble getting the blasted font metrics to work properly, so you're probably right
18:15:37 <frogbotherer> bear in mind i've got next to no experience in any of this stuff :D
18:15:54 <kilobyte> tested with MONOSPACED_FONT="/usr/share/fonts/truetype/ttf-dejavu/DejaVuSansMono.ttf"
18:16:09 <kilobyte> I have it neither :(
18:17:29 -!- _dd has quit [Ping timeout: 248 seconds]
18:17:29 <kilobyte> also, no idea what to do with CJK.  DejaVu has the benefit of being exactly Bitstream Vera, except with extra glyphs -- so no one can complain about looks.
18:18:42 <kilobyte> I looked at fontconfig/pango before, but that's way too complex to handle anytime soon
18:19:00 <frogbotherer> i spent some time learning japanese in the past, so i can tell you how the language works
18:19:14 <frogbotherer> but not much about how to render it properly :D
18:19:32 <kilobyte> two fonts, one regular, one with CJK?  Having local tiles broken for CJK in 0.11 (no regression...)?
18:20:12 <kilobyte> I know what to do in console, including writing my own terminal interpreter, but not the slightest clue about rendering it graphically.
18:20:28 <frogbotherer> font metrics are easy: all the glyphs are square
18:21:03 <frogbotherer> and the unicode stuff will Just Work
18:21:34 <frogbotherer> i think it's just a case of getting a font with the right set of characters in it
18:22:04 <frogbotherer> what did the guy who did the Korean port do?
18:22:13 <kilobyte> can't ship it with Crawl (would triple the download size for even a too ugly to live font)
18:22:43 <kilobyte> IIRC a hard-coded font that supports Korean but not much more
18:29:00 <frogbotherer> had a quick look at how Wesnoth does it - it looks like they use separate western and CJK fonts, with the latter being a stripped down character set borrowed from Android
18:31:27 <frogbotherer> http://forums.wesnoth.org/viewtopic.php?f=7&t=35745 has some of the details in it
18:40:08 <CIA-50> 03evilmike * r780937c527a4 10/crawl-ref/source/dat/des/branches/lair.des: Reduce the weight on wad's wood vaults.
18:40:19 <kilobyte> perhaps something more radical: what about always using system fonts?
18:40:23 -!- kek has quit [Remote host closed the connection]
18:41:10 <kilobyte> I think only WinXP and lower lack a decent one, and all have working ones for selected locales
18:42:29 <frogbotherer> so long as it's not me that has to test it :D
19:15:13 -!- syllogism- has quit []
19:16:57 <blmarket> Using one good fonts which is monospace, have a good alphabet glyphs and CJK characters then we will be fine... not many set of fonts are needed.
19:24:26 -!- Patashu[Zzz] is now known as Patashu
19:30:25 <CIA-50> 03kilobyte 07unicode-tiles * rf79dc8e539c9 10/crawl-ref/source/fontwrapper-ft.cc: Fix droppings of old glyphs appearing on a new glyph's edges.
19:33:37 -!- Pacra has quit [Ping timeout: 240 seconds]
19:34:33 -!- edlothiol has quit [Read error: Connection reset by peer]
19:35:14 <kilobyte> frogbotherer: do you have an idea what the memory usage when changing MAX_GLYPHS should be?
19:35:48 <frogbotherer> when i had it at 4096, it wanted 300Meg
19:35:59 <frogbotherer> at 256 it's more like 30 or 40
19:36:39 <frogbotherer> the trouble is it creates a byte array for the whole texture's RGBA values for each font size
19:36:47 <blmarket> when I try 4096 with korean characters, some old-fashioned PCs needed to reduce font size to play properly
19:37:43 <blmarket> (yes, tested with my hardcode...)
19:37:49 <frogbotherer> :)
19:38:19 <kilobyte> all glyphs fit within 32x32, so 1024 of them should use 4MB.  Yet I get VSZ/RSZ 487516/95648 with 256, and 471148/144856 with 1024
19:38:37 <kilobyte> completely no idea why VSZ is going down...
19:39:42 <kilobyte> and it's consistent between runs
19:40:02 <kilobyte> 4096 glyphs is 799136/342900, indeed not really acceptable
19:40:06 -!- edlothiol has joined ##crawl-dev
19:40:43 <frogbotherer> there might be a smart way to pull the texture data back from GLES, instead of keeping the array of pixels
19:40:51 <frogbotherer> that would halve the amount of memory needed
19:41:15 <frogbotherer> we could also use fewer fonts, when more characters are required
19:41:37 <frogbotherer> (currently i think there are five permutations of font size and outline, each of which gets its own texture)
19:42:28 <blmarket> if I check with my codes that would be helpful? which informations do you need?
19:42:35 -!- eeviac_ has quit [Quit: ChatZilla 0.9.88.2 [Firefox 13.0.1/20120614114901]]
19:47:30 <blmarket> and one big character textures can't solve the problem at all... chinese have many characters
19:48:44 <kilobyte> frogbotherer's way seems to work well
19:48:54 <frogbotherer> blmarket: i put a least-recently-used cache on top of the texture, so we're now only limited by the number of unique characters on the screen
19:49:14 <blmarket> oh cool. that would be nice.
19:49:27 <blmarket> I was wondering crawl needs SDL_ttf or SDL_pango
19:49:44 <kilobyte> for 256 glyphs, we pay 12.5M
19:50:28 <frogbotherer> i looked at SDL_ttf, but it didn't seem to help very much
19:51:27 <kilobyte> I guess, interrupting the draw if we get 256 replaces in a single text region, replacing what has already been written with blanks, and retrying, would remove any limits
19:52:16 <Gretell> mudo (L27 VpAE) ASSERT(zombie_class_size(cs) == Z_NOZOMBIE || zombie_class_size(cs) == mons_zombie_size(base)) in 'mon-place.cc' at line 2053 failed. (Zig:27)
19:52:18 <frogbotherer> cor, can you do that? :)
19:53:35 <frogbotherer> the trouble is, i bet all the really good technical literature for solving this sort of problem is in Chinese, Japanese or Korean :D
19:55:42 <blmarket> sry, what I tried was "all common-used korean characters" into one big wchar_t array
19:58:49 <frogbotherer> how many are there, altogether? also, do you need the latin characters as well for loan words?
20:00:11 <blmarket> sure, latin 256 + euc-kr characters... but I can't remember its count
20:00:16 <blmarket> but it's less than 4096
20:01:33 <frogbotherer> a more difficult question: how many different (unique) characters would you typically get in a page of text? :D
20:02:21 <blmarket> commonly used characters are less than 1024, but can be more than 256... this is statistical, so I can't sure
20:03:00 <blmarket> but you can believe it's less than 1024, but... wait...
20:03:25 <blmarket> is there so many screen space which can display more than 256 characters?
20:05:10 <frogbotherer> well kilobyte managed it by putting this UTF-8 test page (https://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-demo.txt) through instead of the manual
20:05:31 <frogbotherer> i think that's a bit of an extreme case though
20:09:42 -!- edlothiol has quit [Ping timeout: 272 seconds]
20:15:19 <kilobyte> wto the contrary, it's a mild test
20:16:06 -!- aleksil has joined ##crawl-dev
20:16:20 -!- headz0ne has joined ##crawl-dev
20:19:36 -!- headzone has quit [Ping timeout: 276 seconds]
20:28:16 -!- alefury has quit []
20:28:32 <frogbotherer> ah ha!
20:28:50 <frogbotherer> kilobyte: i think i've cracked the memory problems
20:30:52 <frogbotherer> our fonts are stored as uncompressed RGBA8888, but only because that's what the code did before - we could store it in a compressed format (like RGBA4, i'll investigate), and get a load of memory back
20:31:44 <frogbotherer> also: i kept a whole copy of the pixel data in the font wrapper, because our GL wrapper wants to overwrite the whole texture each time you change it
20:32:59 <frogbotherer> it appears that there's a glTexSubImage2D in GLES1.0 that lets you redefine part of a texture
20:33:05 -!- eeviac_ has joined ##crawl-dev
20:33:30 <frogbotherer> which means we no longer need to keep the whole of the pixel data separately, if i extend the GL wrapper API a bit
20:34:37 <frogbotherer> i'll have a go at it tomorrow
20:38:46 -!- frogbotherer has quit [Remote host closed the connection]
20:48:28 -!- ainsophyao has joined ##crawl-dev
20:51:36 <eeviac_> after using grep and it lists the references, how do you get back to the command prompt?
20:52:04 <Wensley> eeviac_: are you using grep or git grep?
20:52:09 <eeviac_> git grep
20:52:14 <Wensley> eeviac_: q
20:52:23 <eeviac_> I swear I pressed that button
20:54:00 <Wensley> eeviac_: when the results of a git grep are long enough, it automatically shoves you into the "less" utility. it's the same as if you did "grep whatever | less"
20:54:31 <eeviac_> now I google what the less utility is
20:54:36 <Eronarn> 'man less'
20:56:30 <eeviac_> wikipedia is full of useful information
21:02:56 <Wensley> tbh wikipedia is more useful than man pages
21:03:48 <eeviac_> tbh I ha
21:04:02 <eeviac_> woops wrong window
21:06:10 -!- st_ has quit []
21:24:40 -!- phunktion has joined ##crawl-dev
21:27:16 -!- mikee_ has joined ##crawl-dev
22:02:31 -!- Dixie has joined ##crawl-dev
22:16:33 -!- ussdefiant has joined ##crawl-dev
22:22:25 -!- _dd has joined ##crawl-dev
22:33:15 -!- Pacra has joined ##crawl-dev
22:37:06 -!- MarvinPA has quit [Ping timeout: 265 seconds]
22:38:41 -!- Patashu has quit [Disconnected by services]
22:38:42 -!- Patashu[Zzz] has joined ##crawl-dev
23:12:31 -!- eeviac_ has quit [Quit: ChatZilla 0.9.88.2 [Firefox 13.0/20120601175215]]
23:16:42 -!- eeviac_ has joined ##crawl-dev
23:23:06 -!- dtsund has quit [Quit: reboot tiem]
23:23:13 -!- Nomi has joined ##crawl-dev
23:26:19 -!- dtsund has joined ##crawl-dev
23:37:16 -!- Wensley has quit [Ping timeout: 246 seconds]
23:59:11 -!- aleksil has quit [Ping timeout: 244 seconds]
23:59:42 -!- Grunt has quit [Quit: leaving]