00:39:03 -!- dtsund has joined ##crawl-dev
00:57:56 -!- ZChris13 has joined ##crawl-dev
01:33:05 -!- monqy has quit [Quit: hello]
01:39:56 -!- MarvinPA has quit [Read error: Connection reset by peer]
01:41:54 -!- evilmike has quit [Quit: something happened]
02:00:11 -!- wtface has quit [Quit:  HydraIRC -> http://www.hydrairc.com <- Nine out of ten l33t h4x0rz prefer it]
02:06:38 -!- absolutego has joined ##crawl-dev
02:09:46 -!- absolutego has quit [Client Quit]
02:19:29 -!- ixtli has quit [Quit: Computer has gone to sleep.]
02:45:42 -!- edlothiol has joined ##crawl-dev
03:01:26 -!- dtsund has quit [Quit: dtsund]
03:02:43 -!- iasov has quit [Remote host closed the connection]
03:03:22 -!- iasov has joined ##crawl-dev
03:21:40 -!- ZChris13 has quit [Ping timeout: 245 seconds]
03:25:46 -!- jeanjacques has joined ##crawl-dev
03:45:52 -!- alefury has joined ##crawl-dev
04:04:49 -!- edlothiol has quit [Ping timeout: 252 seconds]
04:49:46 -!- MarvinPA has joined ##crawl-dev
05:14:00 -!- Wensley has quit [Quit: Page closed]
05:14:19 <CIA-46> 03kilobyte 07stone_soup-0.10 * r5d55c714081c 10/crawl-ref/source/ (player-stats.cc player-stats.h player.h): Pass stats as int, not int8_t.
05:14:20 <CIA-46> 03evilmike 07stone_soup-0.10 * rfce85275c0f9 10/crawl-ref/docs/changelog.txt: Mention new PoG in changelog.
05:14:20 <CIA-46> 03kilobyte 07stone_soup-0.10 * recda7439579d 10/crawl-ref/source/dat/database/quotes.txt: Don't force premature wrapping in a couple of quotes.
05:14:20 <CIA-46> 03kilobyte 07stone_soup-0.10 * r708cc81d68ab 10/crawl-ref/source/mutation.cc: Fix the message for gaining bone plates as seafood/kitteh.
05:14:21 <CIA-46> 03MarvinPA 07stone_soup-0.10 * r88778a2f88d7 10/crawl-ref/source/misc.cc: Don't scramble free from drowning when unable to move
05:14:21 <CIA-46> 03kilobyte 07stone_soup-0.10 * r1683f8429f4c 10/crawl-ref/source/mon-stuff.cc: Fix servant kills of dual demon/undead holiness under Kiku and Trog.
05:14:27 <CIA-46> 03kilobyte 07stone_soup-0.10 * rf4ae63ebda35 10/crawl-ref/source/dat/database/quotes.txt: Consistently align quote attributions.
05:14:28 <CIA-46> 03kilobyte 07stone_soup-0.10 * r0ff0ad942cc0 10/crawl-ref/source/ (mutation-data.h mutation.cc): Forbid thin bones to seafood.
05:35:13 -!- Lollipop has joined ##crawl-dev
05:41:09 <phyphor> crap, Mara
05:41:13 <phyphor> I hate Mara
05:41:39 <phyphor> @?mara
05:41:39 <Gretell> Mara (13R) | Speed: 10 | HD: 18 | Health: 140 | AC/EV: 10/14 | Damage: 30 | Flags: 05demonic, evil, see invisible, !sil | Res: 06magic(144), 05fire++, 03poison, 04rot, 13neg+++, 13torm | Vul: 08holy++ | XP: 7145 | Sp: blink; mislead, b.fire (3d27), mara summon, illusion, pain (d17), teleport self.
05:41:56 <due> I <3 mara
05:42:19 <phyphor> how do I shot web?
05:42:23 <phyphor> s/shot web/kill Mara/
05:47:18 <kilobyte> forget airstrike, sticky flame and freezing cloud NOW.  Get plenty of rF.  Wield a banana and don't potion up until the clone spawns.  Then just kill him (but remember to abandon+run earlier than usually).
05:47:38 <phyphor> Oh, wait, this is -dev
05:48:53 <kilobyte> yeah, but this is even more relevant here: it'd be good to get rid of spoilers like this
06:08:29 <alefury> the clone could adjust to current player state (within reason)
06:08:53 <alefury> haste yourself ==> clone is hasted too
06:08:57 <alefury> switch weapon ==> clone does too
06:09:19 <alefury> that way its always a fair fight
06:09:55 <alefury> although there would be some new abuses too, especially with weapon switching
06:10:08 <BlastHardcheese> clone dies, you die too
06:10:37 <alefury> the clone simply being invincible (its just an illusion) would work too
06:10:42 <phyphor> If you are intentionally weak until clones then the Maras might kill you anyway
06:10:46 <alefury> only way to get rid of it is to let it time out
06:11:14 <alefury> that info leak about whos the real mara is nice. i didnt test it, but sounds like a nicely abusable bug.
06:11:27 <phyphor> info leak?
06:11:35 <phyphor> (I've just killed Mara so it doesn't help me now)
06:11:44 <alefury> target mara. then when mara makes illusions of himself, targeting still defaults to the real one.
06:12:49 <alefury> its on mantis, so not exactly a secret
06:13:42 <phyphor> oh, that's interesting
06:13:48 <due> phyphor: Hence why I thought that was balanced and not necessarily a spoiler -- it might help improve your odds, but it doesn't ensure you victory.
06:14:33 <alefury> something that might be worth fixing in general is ghosts/illusions with ranged weapons
06:14:55 <alefury> they have no ammo, so they cant shoot, and they cant switch to a real weapon either
06:15:17 <alefury> would be a pretty big buff to some ghosts, though
06:22:13 -!- edlothiol has joined ##crawl-dev
06:23:19 <kilobyte> alefury: especially, any player with throwing.
06:25:09 <kilobyte> it is tedious, but if you wanted to play optimally, against any non-ranged opponent, a turn spent doing something else than pelting it with throwables is a turn lost (assuming no choke point or similar tactical considerations)
06:32:43 -!- syllogism has joined ##crawl-dev
07:06:19 -!- Ragdoll has joined ##crawl-dev
07:34:47 -!- HangedMan has joined ##crawl-dev
07:40:50 -!- MarvinPA has quit [Ping timeout: 245 seconds]
07:47:00 <Cheibriados> [Zotdef] Unable to wield vampiric weapons (https://crawl.develz.org/mantis/view.php?id=5272) by Ragdoll
07:47:25 -!- st_ has quit [Ping timeout: 276 seconds]
08:09:14 -!- Lollipop__ has joined ##crawl-dev
08:09:19 -!- Lollipop___ has joined ##crawl-dev
08:10:40 -!- Lollipop has quit [Read error: Operation timed out]
08:13:25 -!- Lollipop_ has quit [Ping timeout: 276 seconds]
08:20:17 <CIA-46> 03dolorous * r31fb67cd0770 10/crawl-ref/source/item_use.cc: Fix tentacles reference for two-handed weapons with shields.
08:20:18 <CIA-46> 03dolorous * rc54081713378 10/crawl-ref/source/mutation.cc: Generalize the bony plates check; anyone with level-3 tentacles has no arms.
09:02:11 <CIA-46> 03kilobyte * r11ee62d43b01 10/crawl-ref/source/ (acquire.cc invent.cc item_use.cc): Fix vamp weapons being unwieldable in ZotDef.
09:07:39 <Henzell> heteroy the Thaumaturge (L12 DgFE) ASSERT(smc) in 'mon-util.cc' at line 1390 failed on turn 18334. (Lair:1)
09:37:06 -!- ZChris13 has joined ##crawl-dev
09:44:43 -!- ZorbaBeta has quit [Read error: Connection reset by peer]
09:44:57 -!- ZorbaBeta has joined ##crawl-dev
09:54:04 <Cheibriados> Monsters turned to Frenzy with needles show neutral but not berserk (https://crawl.develz.org/mantis/view.php?id=5273) by XuaXua
09:55:40 -!- ais523 has joined ##crawl-dev
10:01:35 -!- ais523 has quit [Remote host closed the connection]
10:40:01 -!- ZChris13 has quit [Ping timeout: 245 seconds]
11:00:03 -!- blabber has joined ##crawl-dev
11:05:18 -!- ais523 has joined ##crawl-dev
11:20:22 -!- Lollipop___ has quit [Remote host closed the connection]
11:20:23 -!- Lollipop__ has quit [Read error: Connection reset by peer]
11:23:25 -!- CIA-125 has joined ##crawl-dev
11:24:51 -!- CIA-46 has quit [Ping timeout: 244 seconds]
11:24:51 -!- BlastHardcheese has quit [Ping timeout: 240 seconds]
11:24:52 -!- bhaak has quit [Ping timeout: 240 seconds]
11:24:52 -!- HangedMan has quit [Ping timeout: 245 seconds]
11:25:49 -!- HangedMan has joined ##crawl-dev
11:25:57 -!- bhaak has joined ##crawl-dev
11:27:00 -!- Lollipop has joined ##crawl-dev
11:27:02 -!- Lollipop_ has joined ##crawl-dev
11:27:16 -!- BlastHardcheese has joined ##crawl-dev
12:02:09 -!- iasov has quit [Quit: Leaving]
12:07:41 -!- iasov has joined ##crawl-dev
12:10:58 -!- blabber has quit [Remote host closed the connection]
12:23:26 -!- blabber has joined ##crawl-dev
12:23:43 -!- heteroy has joined ##crawl-dev
12:24:12 <heteroy> As a Draconian something has gone amiss...
12:24:14 <heteroy> It has a very clean taste.
12:24:15 <heteroy> Your talons look more like feet. x2
12:24:17 <heteroy> Your talons dull and shrink into feet.
12:26:37 -!- squee121 has joined ##crawl-dev
12:27:55 -!- squee121 has left ##crawl-dev
12:45:42 -!- elliptic has joined ##crawl-dev
12:52:27 -!- alefury_ has joined ##crawl-dev
12:54:52 -!- petete has quit [Ping timeout: 276 seconds]
13:04:31 -!- heteroy has quit [Quit: ChatZilla 0.9.87 [Firefox 3.6.25/20111212142243]]
13:11:05 -!- monqy has joined ##crawl-dev
13:12:00 <Cheibriados> "Your giant spore resists." (https://crawl.develz.org/mantis/view.php?id=5274) by ortoslon
13:16:18 -!- metaentity has joined ##crawl-dev
13:19:05 <metaentity> last week I sent an email to crawl-ref-discuss, talking about a possible fox for the stepdown function. I've been adviced to come here and talk to a developer. Any dev here?
13:20:02 <ais523> there are quite a few devs here, not sure if any are paying attention right now though
13:20:43 <metaentity> that suggests me you're not a developer, are you?
13:21:01 <ais523> I'm not
13:21:12 <ais523> I've written a couple of lines of code for DCSS altogether
13:21:14 <ais523> not really a lot
13:21:40 <metaentity> let's say a "would-be developer" :D
13:22:01 <ghallberg> ais523 is in the same seat as me pretty much :)
13:22:41 <metaentity> how many lines have you written? 3?
13:22:43 <metaentity> ;D
13:22:50 <ais523> I thought it was 2, not sure exactly though
13:23:23 <ghallberg> Ok, I'm more like 20 but yeah...
13:23:29 <ais523> to allow Crawl to be backgrounded by a shell's job system without misinterpreting the codes telling it about that as a crash
13:24:09 <metaentity> well, it's great anyway, it doesn't really matter if it's 2, 20 or 200 lines of code
13:24:20 <ghallberg> metaentity: If you wanna explain your idea to us I think we'd be happy to hear :)
13:24:31 <ghallberg> Though you might have to say it again...
13:24:54 <metaentity> sure, no problem, as long as you are not getting bored
13:24:54 <ghallberg> I remember reading your email but... my attention is otherwhere right now.
13:25:10 <ghallberg> AS in... this time of year, right NOW I'm interested ;)
13:25:34 <metaentity> well, at least it seems somebody reads that mailing list
13:25:49 <metaentity> about the stepdown function
13:26:32 <metaentity> the stepdown function, as it is in the last version, is just an approximation of the function s*log_2(1+x/s)
13:26:57 <metaentity> but it's coded in a way it does not look like that
13:27:52 <metaentity> it's working, but even the guy who wrote it down was not fully satisfied with it, as he also commented:
13:27:52 <metaentity> // I hope this permits easier/more experimentation with value stepdowns
13:27:52 <metaentity> // in the code.  It really needs to be rewritten to accept arbitrary
13:27:52 <metaentity> // (unevenly spaced) steppings.
13:28:08 <ghallberg> mm
13:28:51 <metaentity> then my point is: why not use directly the explicit form (the logaritmic one)?
13:29:37 <metaentity> I think it would be smoother to use real-valued functions for these purposes, and maybe use gaussian distributions (or some other real-valued distribution) for probability purposes.
13:30:44 <metaentity> david ploog, a former developer, replied to my email saying:
13:30:44 <metaentity> I think it would be useful. I don't believe performance matters here, and having the  logarithm should make it easier to understand what's going on; remove  spoilers (players remembering cutoff points); and make it easier to adapt  the parameters for new uses.
13:30:57 <ghallberg> IT sounds very reasonable.
13:33:01 <elliptic> I agree that smoother stepdowns would be better in principle
13:33:02 <metaentity> actually, my whole suggestion would be to rewrite the math part using real-valued functions and distributions, in order to maintain clarity and flexibility WITHOUT weird effects like sharp change of behaviour found in the stepdown function i have analysed
13:33:30 <elliptic> changing the current stepdowns would have a lot of balance effects though, unless we carefully looked at every place where they are used
13:34:21 <elliptic> probably best would be to write some generic smooth stepdown function that approximates the current one and then gradually try to shift to it
13:34:39 <metaentity> we already have such a smooth stepdown
13:35:08 <metaentity> stepdown_value(x, s, s, +infty, -1) is equal to the function  f(x,s)=s*log_2(1+x/s)
13:35:39 <metaentity> plotting both function removes any doubt: they are very close together
13:36:11 <metaentity> obviously, they are not strictly equal
13:36:12 <elliptic> they aren't equal
13:37:00 <elliptic> in particular, if x < s then stepdown_value is just x
13:37:18 <metaentity> actually, the logarithmic one is always greater of equal the stepdown, as the stepdown is a piecewise linear interpolation of the logarithmic curve
13:38:52 <elliptic> what I was suggesting is that we could have a smooth_stepdown_value function that takes the same arguments but is a smooth approximation
13:38:52 <metaentity> I have both functions plotted down: how do I upload an image?
13:39:33 <elliptic> and then we could gradually try to change stepdown_value to smooth_stepdown_value in the code
13:39:35 <ghallberg> metaentity: Throw it on imgur or somesuch I guess.
13:40:18 <metaentity> sounds nice
13:40:26 <elliptic> I guess stepdown_value isn't called as many times as I thought, though
13:41:14 <elliptic> so maybe this isn't such a big undertaking to shift things to a smooth system without messing up balance
13:42:18 <elliptic> I'll note though that a fair amount of the time stepdown_value is used with only a couple of steps
13:43:57 <metaentity> i don't think the smooth system can mess up balance, though it does require a careful look
13:44:04 <metaentity> I have the plot http://imgur.com/ntG6N
13:44:45 <metaentity> the step in that picture is 40
13:46:10 <elliptic> not in any huge way, but it looks like the logarithmic formula is over 10% higher than the PL formula at points
13:46:41 <elliptic> changing things by 10% at dozens of places in the code requires some care :)
13:48:47 <metaentity> calculations at hand, it is 1.06 higher at most, apart from the very low values
13:49:03 <metaentity> so it is like 6%
13:49:54 <ghallberg> metaentity: You really have done your research :)
13:50:07 <elliptic> metaentity: uh, what do you mean by "very low values"
13:50:43 <metaentity> 60 or less
13:51:03 <metaentity> sorry, 40 or less
13:51:07 <elliptic> those values are very important! we can't just forget about them
13:52:08 <elliptic> an example: spell power uses a step of size 50. the logarithmic formula changes 25 to 29.2, so more than 15% increase
13:52:20 <ghallberg> It seems to me the function is often called with small values.
13:52:23 <ghallberg> "small"
13:52:37 <elliptic> 25 power is #####... a lot of early game you don't have that much
13:52:44 <metaentity> small values are values smaller than the step
13:54:10 <metaentity> as it is coded right now, from 0 until step the stepddown_value is perfectly linear with coefficient 1, but ANY smooth approximation which meets the same values at multiple of steps will be higher, as the stepdwon is convex
13:54:19 <elliptic> sure
13:55:02 <elliptic> the point is that increasing early game spell power by 5-10% is something that requires at least a little thought to be sure that it is okay
13:55:13 <metaentity> i got it, it's clear
13:55:21 <elliptic> and stepdown is used a couple dozen times
13:56:52 <elliptic> I think changing stuff to be smooth is definitely a good idea :) it just requires looking at each instance of stepdown_value in the code and making sure that this is a reasonable thing to do
13:57:13 <metaentity> when it turns out it's not, what would you do?
13:57:36 <metaentity> find another smooth alternative, or keep the existing one?
13:58:32 <elliptic> well, in some cases I think there clearly isn't any reason to worry about smoothness... for instance, pot_quantity = stepdown_value(pot_quantity, 2, 2, 6, 6);
13:59:39 <elliptic> like, we need to remember that these are all really functions on a discrete set of integers
14:00:45 <metaentity> in damage and such, are calculations done with integers or folating point numbers?
14:00:59 <metaentity> (with rounding at the end)
14:01:28 <elliptic> integers
14:02:46 <elliptic> I should say that the main motivation for smoothing out stepdowns for me is to remove invisible breakpoints
14:03:17 <metaentity> that is even for me the main motivations also
14:03:37 <elliptic> for instance, your dexterity gets stepped down starting at 24 for dodging purposes, so you might want to stop raising dex at exactly 24
14:03:51 <metaentity> and that's bad
14:03:57 <elliptic> or stop training dodging when your EV gets to exactly 30
14:04:37 <metaentity> with a smooth functions you can't exploit any exceptional value
14:05:17 <elliptic> so I'd probably just try to convert those to smooth functions and leave things that don't affect gameplay (like that pot_quantity) untouched
14:05:40 <metaentity> I don't know whether the result is worth the trouble, anyway
14:06:36 <elliptic> well, removing spoily breakpoints is worth the trouble... they exist in other parts of the code too, and we've been trying to eliminate them
14:07:07 <metaentity> ok, so if somebody has already tried to get rod of them, this looks like a nice solution
14:08:06 <metaentity> who is going to decide if this change is worth a try or not?
14:11:17 <elliptic> well, patches are always considered for inclusion ;) or one of the other devs might take an interest in it, or I might look at it in a couple months (I won't have the time until then) if nobody else has by then
14:12:17 <metaentity> unfortunately, I have little experience coding, otherwise I would be glad to help
14:12:43 <metaentity> mathematically, I could be useful though (I'm a mathematician)
14:17:06 <elliptic> well, probably it would be helpful if you wanted to put this issue on our bug/issue tracker (http://crawl.develz.org/mantis) so that we are less likely to forget about it... it isn't a bug per se, but the invisible breakpoints the non-smooth stepdowns introduce are definitely not desired
14:17:46 <elliptic> currently we are just a week or two away from releasing 0.10, which is probably part of why you haven't gotten much response
14:17:59 <Cheibriados> Animate Skeletons not arming themselves (https://crawl.develz.org/mantis/view.php?id=5275) by XuaXua
14:18:53 <metaentity> elliptic: which page is right to post about such a topic?
14:21:00 <metaentity> should I use the wiki or the bugtracker?
14:22:06 <elliptic> either using "Report Issue" on http://crawl.develz.org/mantis or somewhere in our wiki (http://crawl.develz.org/wiki/doku.php)... you'll need to register an account for either one
14:22:12 <elliptic> either will work... I'm not really that familiar with the wiki so I'm not sure what page to suggest there, but maybe someone else here can
14:27:08 <metaentity> elliptic: thank you for discussion and support
14:40:27 -!- syllogism has quit []
15:06:40 -!- metaentity has left ##crawl-dev
15:08:45 <Cheibriados> Using smooth functions and smooth random distributions (https://crawl.develz.org/mantis/view.php?id=5276) by squee
15:31:39 -!- petete has joined ##crawl-dev
15:32:43 -!- Textmode has joined ##crawl-dev
15:46:00 <Ragdoll> hangedman: that is a most interesting situation
15:46:36 <Ragdoll> are you going to forget to breathe
15:49:59 <Zannick> ?
15:52:05 -!- evilmike has joined ##crawl-dev
15:53:49 -!- blabber has quit [Ping timeout: 240 seconds]
15:54:22 -!- dtsund has joined ##crawl-dev
16:01:49 -!- cbus has quit [Ping timeout: 245 seconds]
16:03:42 -!- cbus has joined ##crawl-dev
16:06:46 -!- iasov has quit [Remote host closed the connection]
16:07:20 <Ragdoll> Your Executioner suffers a backlash!
16:07:21 <Ragdoll> haha
16:09:39 -!- Ragdoll has quit [Quit: Up, up and away!]
16:50:48 <Cheibriados> Ctrl-E spoils the map. (https://crawl.develz.org/mantis/view.php?id=5277) by XuaXua
16:53:56 * due gives kilobyte some blankets.
17:02:51 -!- MarvinPA has joined ##crawl-dev
17:07:11 <kilobyte> due: meow?  (ECONTEXT on blankets)
17:08:05 <due> kilobyte: cold snap in your part of the world!
17:09:43 <kilobyte> ah, yeah.  No snow though so it's almost not noticeable (unless you have to stay long outdoors).
17:11:40 * due nod.
17:19:49 -!- dtsund has quit [Read error: Operation timed out]
17:29:25 -!- upsy has joined ##crawl-dev
17:54:49 -!- jeanjacques has quit [Read error: Connection reset by peer]
17:58:48 -!- CIA-125 has quit [Ping timeout: 255 seconds]
18:13:15 -!- MarvinPA has quit [Read error: Connection reset by peer]
18:13:38 -!- CIA-52 has joined ##crawl-dev
18:13:39 -!- MarvinPA has joined ##crawl-dev
18:19:11 -!- ais523 has quit [Remote host closed the connection]
18:24:13 -!- ussdefiant has joined ##crawl-dev
18:27:35 -!- dtsund has joined ##crawl-dev
18:51:58 -!- st_ has joined ##crawl-dev
19:06:08 -!- alefury_ has quit [Remote host closed the connection]
19:08:38 -!- dtsund has quit [Quit: dtsund]
19:11:50 -!- Textmode has quit [Remote host closed the connection]
19:17:58 <Cheibriados> "Foo is repulsed!" does not ID wand (https://crawl.develz.org/mantis/view.php?id=5278) by minmay
19:22:07 -!- Xiberia has quit [Read error: Connection reset by peer]
19:22:57 -!- Xiberia has joined ##crawl-dev
19:38:48 -!- dtsund has joined ##crawl-dev
20:06:23 -!- bmh has joined ##crawl-dev
20:16:13 -!- Xiberia has quit [Remote host closed the connection]
20:21:58 -!- bmh has quit [Quit: Page closed]
20:38:19 -!- upsy has quit [Quit: Leaving]
20:56:54 -!- ixtli has joined ##crawl-dev
21:12:33 <CIA-52> 03kilobyte * r50fa62c58545 10/crawl-ref/source/dat/des/builder/uniques.des: Veto Ignacio's vault instead of placing it without him.
21:12:43 <CIA-52> 03kilobyte * rf211c3e44544 10/crawl-ref/source/beam.cc: Id wands of enslavement on auto-re-anger.
21:44:16 <st_> I'm thinking about changing dis_old and dis_mu a little to make them a bit more challenging and take more time to run through... considering metal gargoyles are a dis:7 only thing they should be buffed a lot I believe
21:45:53 <elliptic> make them quicksilver gargoyles
21:46:09 <elliptic> partly the old dis endings are just easy because nearly everything in them is slow
21:46:24 <elliptic> so even a very weak char can run past everything into the vault
21:46:58 <due> Make htem levitate!
21:47:27 <st_> removing the stairs from the vaults so you start further away is something I was going to do
21:47:47 <elliptic> st_: currently stuff is bugged so that you often start away from the vault anyway and it doesn't help
21:47:59 <elliptic> this is partly because the rest of the level is boring, of course
21:48:12 <st_> and then changing the Dis throne rooms so you have to go behind him instead of skirt around the edges
21:49:30 <st_> yeah the rest of the level is boring, but I think Dis and Tar have a problem that the time from entering to getting next to the lord is very short
21:49:37 <elliptic> changing the throne rooms would help, but I think the part that needs the most help is the part of the vault between the entrance and the throne room
21:49:42 <elliptic> and as you said, changing gargoyles to be more of a threat might help
21:50:18 <st_> solution: pit fiends everywhere!
21:50:24 <st_> (just kidding)
21:51:11 <st_> what do you think about the cacodemons in my dis end by the way?
21:51:20 <st_> I kind of ran out of stuff to use so I used those
21:51:54 <elliptic> I think cacodemons are reasonable there, though maybe the flavour is a bit better if they are primarily pan monsters
21:52:43 <elliptic> this reminds me though that it is really strange how makhleb is supposedly the god of the hells and yet the greater servants he grants are primarily pandemonium monsters...
21:53:53 <elliptic> also I do think that the older dis:7 maps need a fiend or two before the throne room so that they aren't completely outclassed by yours :P
21:54:30 <HangedMan> swamp_icy v original swamp vault
21:54:45 <HangedMan> no, not a big enough disparity, hmm
21:57:08 <st_> elliptic: if we change metal gargoyles we could change molten ones too, perhaps another way to make rF more valuable in geh
21:57:40 <st_> considering how rare 9s are not unjust to make them similar in threat to 1/2s
21:58:01 <elliptic> maybe, yeah... gargoyles really have tons of room for improvement
21:58:10 <Cheibriados> Brimstone Fiend (041) | Speed: 10 | HD: 18 | Health: 77-123 | AC/EV: 15/6 | Damage: 25, 15, 15 | Flags: 05demonic, evil, see invisible, fly, !sil | Res: 06magic(288), 05hellfire, 03poison, 04rot, 13neg+++, 13torm | Vul: 12cold, 08holy++ | XP: 4720 | Sp: hellfire (3d20), torment symbol, melee.

21:58:10 <elliptic> %??brimstone fiend
21:58:10 <due> uh
21:58:28 <due> just make sure you don't nuke the gargoyles in the earl D map
21:58:38 <elliptic> what about making some of that fiend's attacks AF_FIRE?
21:59:20 <Cheibriados> fire bat (05b) | Speed: 30 | HD: 18 | Health: 49-76 | AC/EV: 1/14 | Damage: 304(fire:18-35) | Flags: sense invisible, fly | Res: 06magic(24), 05hellfire | Vul: 12cold | XP: 2826.

21:59:20 <HangedMan> %??fire bat hd:18
21:59:25 <elliptic> I have no desire for ice fiend 2.0, so not all of them, but maybe one of them
21:59:35 <Cheibriados> Ice Fiend (161) | Speed: 10 | HD: 18 | Health: 77-123 | AC/EV: 15/6 | Damage: 2512(cold:18-53), 2512(cold:18-53) | Flags: 05demonic, evil, see invisible, fly, !sil | Res: 06magic(288), 12cold+++, 03poison, 04rot, 13neg+++, 13torm | Vul: 04fire, 08holy++ | XP: 3036 | Sp: b.cold (3d27), torment symbol.

21:59:35 <HangedMan> %??ice fiend
21:59:42 <HangedMan> ...huh.
21:59:47 <elliptic> though I guess AF_COLD is more damaging anyway
21:59:59 -!- HangedMan has quit []
22:02:02 <st_> no harm in doing that, that I see
22:03:21 -!- HangedMan has joined ##crawl-dev
22:11:42 -!- edlothiol has quit [Ping timeout: 272 seconds]
22:13:59 <Eronarn> so apparently if you have 0 evo you can have nothing happen with an experience card
22:14:05 <Eronarn> it should do something even if minor
22:18:17 <SamB> Hmm. It seems that we're not supposed to call raise() *OR* abort() from our (abort) signal handler. Instead, we should return from it.
22:19:03 <SamB> (Of course, we also aren't supposed to use any C++-only features in our signal handler, so I guess that's kind of moot ...)
22:58:27 -!- dtsund has quit [Read error: Operation timed out]
23:10:26 -!- HangedMan has quit [Ping timeout: 245 seconds]
23:19:57 -!- dtsund has joined ##crawl-dev
23:26:19 -!- elliptic has quit [Quit: Leaving]
23:29:19 -!- elliptic has joined ##crawl-dev