00:03:10 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.9.0-a0-42-gb071afa (32)
00:10:54 -!- valrus has quit [Remote host closed the connection]
00:11:32 -!- valrus has joined ##crawl-dev
00:11:47 -!- valrus has quit [Remote host closed the connection]
00:59:31 -!- Pseudonut has quit [Ping timeout: 246 seconds]
01:02:01 -!- Pseudonut has joined ##crawl-dev
01:31:05 <Napkin> moin
01:34:06 <due> hiii Napkin <3
01:34:25 <Napkin> hey due :)
01:34:38 <Napkin> How's it going?
01:34:51 <CIA-45> 03MarvinPA * rd07d678ad933 10/crawl-ref/source/transform.cc: Allow extending spider form while clinging over dangerous terrain
01:35:03 <CIA-45> 03MarvinPA * rf3fa798408a0 10/crawl-ref/source/player.cc: Message fix for levitation while clinging
01:35:14 <due> Not bad!
01:35:44 <Napkin> :))
01:48:01 -!- evilmike has joined ##crawl-dev
01:53:18 <Zaba> moin
01:53:25 <DrPraetor> Ah, there we are!
01:54:03 <DrPraetor> So, taking a skill from level 0 to level 3 costs a *base* of 500 skill points.  Taking a skill from level 8 to level 10 costs a *base* of 1,050 points.
01:55:11 <DrPraetor> However, evocations only costs 80% of that, so with a single level of aptitude difference, Stabbing 0->3 is the same price as Evocations 8->10.
01:58:45 <Henzell> LuckyNed the Warrior (L21 DSCr)  (D:25)
02:03:21 <DrPraetor> Now, these 500 *skill points* are not exactly the same as XP, but they *are* what matters, because they're what contributes to your calc_total_skill_points field, which determines how expensive all of your skills are.
02:04:20 <Napkin> moin Zaba
02:10:31 -!- Twilight has joined ##crawl-dev
02:10:53 <kilobyte> elliptic: even with same apts, the split at the very least doubles the cost
02:10:54 <Henzell> kilobyte: You have 3 messages. Use !messages to read them.
02:10:57 -!- Twilight is now known as Guest64988
02:12:43 -!- Twilight-1 has quit [Ping timeout: 250 seconds]
02:14:12 <kilobyte> Napkin: every single non-ASCII map in a morgue from before the Unicode overhaul is corrupted.  Those with char_set=ibm recoverably, those with char_set=unicode not.
02:14:54 <kilobyte> Napkin: the new ones don't display correctly since your apache is configured to use ISO-8859-1, even though not a single page on CDO uses that
02:15:11 <kilobyte> Mantis and WordPress override that to UTF-8
02:16:47 <kilobyte> could you just set apache's encoding to UTF-8?  Either for morgue / rc dirs, or even globally.
02:16:55 <Napkin> specifying a charset in apache is a bad idea - so that's not done at all
02:17:55 <bhaak> hehe, that's why i only output ASCII chars in text dumps :)
02:18:14 <bhaak> oob encoding is a bitch even in HTML it might go wrong :/
02:19:07 <kilobyte> Napkin: uhm, _not_ specifying the charset is a security hole if you have any user-supplied data
02:19:28 <Napkin> because?
02:19:30 <kilobyte> Napkin: this is why any semi-new version of Apache forces an encoding
02:20:41 <kilobyte> a text inside a page can otherwise set the encoding to something else and use that to sneak through data that would be otherwise rejected
02:21:37 <kilobyte> heck, a Debian security update several years ago changed that, breaking (in stable!) pages which relied only on http-equivs
02:22:04 <Napkin> yeah, I already love this conversation
02:22:08 <Napkin> talk to you later o/
02:22:32 <kilobyte> heh
02:23:27 <bhaak> hehe, sarcasm or not? we might never know :)
02:25:36 <Napkin> $> cat conf.d/charset
02:25:36 <Napkin> # Read the documentation before enabling AddDefaultCharset.
02:25:36 <Napkin> # In general, it is only a good idea if you know that all your files
02:25:36 <Napkin> # have this encoding. It will override any encoding given in the files
02:25:36 <Napkin> # in meta http-equiv or xml encoding tags.
02:25:38 <Napkin> #AddDefaultCharset UTF-8
02:27:30 <bhaak> the problem with text files is of course that you have a problem if they have different encodings
02:31:15 <Napkin> kilobyte: which morgues are you talking about?
02:31:28 <Napkin> http://crawl.develz.org/morgues/trunk/Cab/morgue-Cab-20110309-181823.txt <- ibm as char_set
02:31:58 <Napkin> http://crawl.develz.org/morgues/trunk/Butch/morgue-Butch-20101016-084559.txt <- ascii as char_set
02:32:37 <Napkin> http://crawl.develz.org/morgues/trunk/Napkin/morgue-Napkin-20110329-103611.txt <- unicode as char_set
02:33:14 <Napkin> all their maps look properly drawn to me (firefox, character encoding on auto-detect)
02:33:29 <kilobyte> as you can see, neither of these has any non-ASCII characters since walls were hacked to ASCII inside Crawl
02:34:05 <bhaak> also, there is no encoding info in the header
02:34:06 <kilobyte> but try this: http://crawl.develz.org/morgues/trunk/KiloByte/morgue-KiloByte-20110129-220313.txt
02:34:10 <bhaak> *http-header
02:34:19 <kilobyte> or, Unicode version: http://crawl.develz.org/morgues/trunk/KiloByte/morgue-KiloByte-20110406-105848.txt
02:36:07 <kilobyte> for example, this nice string: "+ADw-script+AD4" -- does it sound harmless?
02:36:49 <bhaak> he, utf-7 is stupid
02:37:15 <kilobyte> yeah, so is EBCDIC and the rest
02:37:58 <bhaak> there are only two things you could do (if you want properly displayed maps):
02:38:02 <Napkin> the first one looks fine, firefox detects cp1251
02:38:13 <bhaak> 1. write only ASCII
02:38:25 <Napkin> the unicode one looks like parts are missing
02:38:27 <bhaak> 2. convert everything to utf-8 and add a BOM at the beginning
02:38:52 * bhaak is not a big fan of BOM
02:39:18 <kilobyte> uhm, BOM is just as bad as http-equiv
02:39:45 <kilobyte> and even worse for text files, since it breaks concatenation and a good deal of Unix tools
02:39:48 <Napkin> so.. the second morgue is broken, no matter char char-encoding I choose
02:39:50 <bhaak> no, it has the advantage that it is in the document
02:39:54 -!- jld has quit [Ping timeout: 248 seconds]
02:40:03 <Napkin> and if I enforce utf8 the first one will be broken too
02:40:20 <bhaak> "no" to your first answer, not the second, ofc
02:40:30 -!- jld has joined ##crawl-dev
02:41:00 <kilobyte> Napkin: not any worse, å becomes �
02:41:12 <kilobyte> and we can detect and correct that with a simple script
02:41:50 <bhaak> hmm, are the pizza names sanitized?
02:42:53 <kilobyte> Napkin: if you feel uneasy about changing the default for the rest of the website (even though I can't fathom what could possibly be ISO-8859-1), you can add the charset just for text files, or inside the morgue and rc dirs
02:43:25 <kilobyte> bhaak: I doubt that
02:43:33 <Napkin> well, there are morgues from before your changes, kilobyte
02:44:32 <kilobyte> Napkin: a vast majority of them are pure ASCII, and we can fix the rest (at least the cset_ibm ones)
02:47:14 <Napkin> I don't see the reason yet for enforcing charset
02:47:32 <kilobyte> iso-8859-1 not cp1252 (nor 1251, but I guess you typoed)
02:48:09 <Napkin> in your first link
02:48:13 <kilobyte> all new morgues will display wrongly
02:48:15 <Napkin> what are those monsters below you?
02:48:23 <Napkin> e?
02:48:38 <Napkin> or å?
02:48:39 <kilobyte> no, no monsters
02:48:48 <kilobyte> lemme view the ttyrec
02:48:50 -!- syllogism has joined ##crawl-dev
02:49:19 <Napkin> no, no typo - firefox detected cp1251
02:49:27 <Napkin> -> e
02:49:38 <Napkin> in iso-8859-1 they are å
02:49:50 <DrPraetor> Are the timers for sewers etc. based on when you see the sewer or when you enter the level?  I remember there was some discussion about this.
02:52:27 <kilobyte> Napkin: they were stone walls (ᛥ) that for some reason were not converted
02:52:52 <kilobyte> there's no way to recover these characters from the morgue :(
02:57:19 <Napkin> I added it do <Directory /> of crawl.develz.org
02:57:34 <kilobyte> bhaak: you're right, pizzas are not sanitized in any way.  You can even even put newlines inside.
02:57:44 <Napkin> but I don't see any mentioning in lynx -mime_header still
02:59:37 <kilobyte> did you tell apache to reload the config?
03:00:01 <Napkin> please, kilobyte
03:01:23 <kilobyte> uhm sorry, "please" what?
03:02:50 <Napkin> that's the same as me asking you, if you recompiled crawl after doing some source change and not seeing an effect
03:03:35 <kilobyte> hey, I do that from time to time, a frequent error :p
03:03:51 <Napkin> ;)
03:03:51 <bhaak> you would be surprised how often you think you compiled something when you didn't :)
03:04:14 <Napkin> but it's the first thing you remember when you don't see the effect ;)
03:04:27 <kilobyte> for source, it might be pretty rare, but for DNS, I'd say I fail to bump the serial and rndc it like half the time
03:05:01 <Napkin> yeah, dns has special loopholes ;)
03:07:05 <bhaak> then you recompile, think you see the change, try to debug and half an hour later you notice that you've been compiling all the time in the wrong directory ;)
03:08:55 <kilobyte> I tried putting "AddDefaultCharset somejunk" into a <Directory> section, and it seems to work
03:09:17 <Napkin> how did you check it?
03:11:34 <kilobyte> HEAD http://w1.angband.pl (a test website), same as lynx --mime-something you mentioned but without the contents
03:12:10 <Napkin> "HEAD"?
03:13:43 <kilobyte> it comes from libwww-perl, depended upon by 958723589372589 things and thus present on most systems
03:14:27 -!- monqy has quit [Quit: hello]
03:17:07 <Napkin> hmm...
03:21:57 <Napkin> Description: Default charset parameter to be added when a response content-type is text/plain or text/html
03:22:10 <Napkin> Syntax: AddDefaultCharset On|Off|charset
03:22:22 <Napkin> Context: server config, virtual host, directory, .htaccess
03:22:36 <Napkin> And I can only get it to work when setting it in "server config"
03:22:52 <Napkin> neither virtual host not directory contexts work :-O
03:22:59 <Napkin> *nor
03:24:11 -!- Zaba has quit [Ping timeout: 240 seconds]
03:30:25 <Napkin> ok.. i don't seem to be the only one with that problem
03:33:39 -!- Zaba has joined ##crawl-dev
03:36:52 <kilobyte> I've made a script to re-encode only the morgues that are not already in UTF-8, but I'm not sure what metadata should be preserved.  mtime, obviously -- but what with rights and owner?
03:36:59 <Napkin> ok, got it
03:37:20 <Napkin> stupid proxy.conf turned it off!
03:37:28 <Napkin> once off, always off *sigh*
03:37:31 <Napkin> ?!
03:37:33 <Napkin> grr.
03:37:56 <Napkin> would be nice, but I can fix it afterwards
03:40:44 <Napkin> oh - and clever me was adding it to the https version of crawl.develz.org
03:43:24 <kilobyte> !seen paxed
03:43:25 <Henzell> I last saw paxed at Mon Apr  4 00:55:16 2011 UTC (6d 7h 48m 9s ago) quitting with message Ping timeout: 250 seconds.
03:52:02 <Napkin> hmm, seems like the default "dump_order" shown in configs is not actually the default dump_order
03:52:13 <Napkin> unless you changed your's recently, kilobyte?
03:54:35 <kilobyte> the default is commented out, in my config for some reason I forgot it isn't
03:54:45 <Napkin> it's not
03:54:59 <Napkin> http://ipxserver.dyndns.org/misc/wins/morgue-Napkin-20110406-151811.txt <- default currently
03:55:10 <Napkin> for example notes are at the end
03:55:27 <Napkin> turns_by_place and kills_by_place is not shown at all by default
03:55:46 <kilobyte> is CDO's default config different from the official one?
03:55:51 <Napkin> no
03:56:32 <kilobyte> the one in the config file might me outdated
03:56:44 <Napkin> yes
03:56:57 <Napkin> i saw that, when comparing morgues
03:57:08 <Napkin> and was curious, where those 2 nice tables came from
03:57:16 <kilobyte> new_dump_fields("header,hiscore,stats,misc,inventory,"
03:57:17 <kilobyte> "skills,spells,overview,mutations,messages,"
03:57:17 <kilobyte> "screenshot,monlist,kills,notes");
03:57:50 <Napkin> because i think turns_by_place and kills_by_place are interesting
03:58:00 <kilobyte> # dump_order  = header, hiscore, stats, misc, notes, inventory,
03:58:00 <kilobyte> # dump_order += turns_by_place, skills, spells, overview, mutations,
03:58:00 <kilobyte> # dump_order += messages, screenshot, monlist, kills_by_place, kills
03:58:02 <Napkin> yet.. i'll put them at the end, behind notes
03:58:31 <CIA-45> 03kilobyte * r29bf59374290 10/crawl-ref/source/windowmanager-sdl.cc: Non-BMP characters are not ASCII.
03:58:40 <CIA-45> 03kilobyte * r4537a1d26c20 10/crawl-ref/source/rng.cc: Use 160 rather than 128 bits of randomness.
03:58:40 <CIA-45> 03kilobyte * r589bc7780431 10/crawl-ref/source/player-equip.cc: Don't insta-kill people on vampiric weapon unmeld.
04:02:32 <Napkin> and i think "monlist" is empty for some reason
04:02:45 <Napkin> in both dumps
04:03:14 <Napkin> or is that the line with "You could see a spatial vortex (confused), a friendly iron dragon, a friendly storm dragon and a friendly quicksilver dragon."?
04:06:17 <CIA-45> 03kilobyte * r76c372e0eb0f 10/crawl-ref/source/player-equip.cc: Tell weapon/jewel artefacts it's only a meld.
04:07:09 <kilobyte> it is just that
04:07:22 <Napkin> roger
04:11:21 <Napkin> hmm.. is dump_order ignored for manual dumps?
04:13:43 <kilobyte> it appears it is used
04:14:16 <CIA-45> 03kilobyte * r9a35ab8cf4bd 10/crawl-ref/source/player-equip.cc: Don't show "it sticks to your hand" on cursed unmeld.
04:17:02 <Napkin> heh.. damn cache
04:17:20 <Napkin> sorry
04:27:41 -!- Kurper has quit [Remote host closed the connection]
04:29:05 <paxed> kilobyte: ?
04:30:17 <kilobyte> paxed: do you have a clue about newer versions of ee and virus
04:30:34 <kilobyte> ee seems to be dead upstream for ten years
04:30:56 <kilobyte> has virus been modified since getting taken from busybox?
04:31:40 <paxed> kilobyte: yes, there's a newer version of virus in busybox
04:32:19 <paxed> but i haven't bothered with it yet.
04:33:45 <kilobyte> not sure if it's good enough, too
04:35:22 <kilobyte> even the new version has broken cursor movement in some cases :(
05:02:47 -!- eith has joined ##crawl-dev
05:11:15 <Ashenzari> ugly white pixels on Leda's Liquefaction (https://crawl.develz.org/mantis/view.php?id=3785) by KiloByte
05:15:54 -!- Mu_ has joined ##crawl-dev
05:59:17 -!- eith has quit []
05:59:38 -!- eith has joined ##crawl-dev
06:04:28 -!- valrus has joined ##crawl-dev
06:24:10 -!- eith has quit []
06:24:37 -!- eith has joined ##crawl-dev
06:26:26 -!- Pseudonut has quit [Remote host closed the connection]
06:43:23 -!- eith has quit []
06:46:43 -!- valrus has quit [Remote host closed the connection]
08:12:53 -!- valrus has joined ##crawl-dev
08:17:11 -!- valrus has quit [Ping timeout: 240 seconds]
08:18:45 -!- |eith| has joined ##crawl-dev
08:25:04 -!- |eith| is now known as eith
08:31:33 -!- edlothiol has joined ##crawl-dev
09:08:15 -!- upsy has joined ##crawl-dev
09:23:42 -!- upsy has quit [Read error: Connection reset by peer]
09:26:47 -!- upsy has joined ##crawl-dev
09:40:48 <Napkin> *sigh*
09:40:57 <Napkin> the skill screen is broken
09:41:20 -!- DrPraetor has quit [Quit: KVIrc 4.0.2 Insomnia http://www.kvirc.net/]
09:42:05 <Napkin> unless, it's not supposed to show that maces&flails have crosstraining with axes, when axes skill is >1 and maces is not
10:52:45 -!- valrus has joined ##crawl-dev
11:08:54 -!- valrus has quit [*.net *.split]
11:08:55 -!- Eronarn has quit [*.net *.split]
11:08:56 -!- paxed has quit [*.net *.split]
11:08:56 -!- jlewis has quit [*.net *.split]
11:08:56 -!- stabwound has quit [*.net *.split]
11:10:55 -!- Eronarn has joined ##crawl-dev
11:10:56 -!- paxed has joined ##crawl-dev
11:10:56 -!- jlewis has joined ##crawl-dev
11:10:56 -!- stabwound has joined ##crawl-dev
11:20:46 -!- elly has quit [Ping timeout: 246 seconds]
11:21:07 -!- bf has quit [Ping timeout: 246 seconds]
11:21:48 -!- not_evilmike has joined ##crawl-dev
11:26:17 -!- bf has joined ##crawl-dev
11:26:53 -!- evilmike has quit [Ping timeout: 246 seconds]
11:27:00 -!- ivan has quit [Ping timeout: 246 seconds]
11:27:01 -!- OG17 has quit [Ping timeout: 246 seconds]
11:33:53 -!- ivan has joined ##crawl-dev
11:36:07 -!- MarvinPA_ has joined ##crawl-dev
11:36:52 -!- upsy has quit [Read error: Connection reset by peer]
11:37:00 -!- upsy has joined ##crawl-dev
11:38:41 -!- MarvinPA has quit [Ping timeout: 240 seconds]
11:43:59 -!- elly has joined ##crawl-dev
12:05:50 -!- monqy has joined ##crawl-dev
12:11:16 -!- elly has quit [Ping timeout: 246 seconds]
12:13:22 -!- MakMorn has quit [Ping timeout: 246 seconds]
12:13:23 -!- due has quit [Ping timeout: 246 seconds]
12:13:58 -!- due has joined ##crawl-dev
12:18:20 -!- MakMorn has joined ##crawl-dev
12:20:11 -!- bf has quit [Ping timeout: 240 seconds]
12:21:11 -!- upsy has quit [Read error: Connection reset by peer]
12:21:36 -!- upsy has joined ##crawl-dev
12:23:01 -!- elly has joined ##crawl-dev
12:24:51 -!- bf has joined ##crawl-dev
12:25:28 <greensnark> Napkin: Chrome's suddenly started complaining about the StartCom CA cert on cao and cdo :)
12:27:26 <kilobyte> none of the certificates on the chain have expired, so if they worked for you before, something is strange
12:28:18 <kilobyte> works for me even in Chromium
12:30:37 <CIA-45> 03kilobyte * rd6169c4135d6 10/crawl-ref/source/hints.cc: Purge an useless comment.
12:30:40 <CIA-45> 03kilobyte * r48ae4b43c04e 10/crawl-ref/source/rltiles/dc-mon/animals/jumping_spider.png: An update to jumping spider's tile.
12:30:42 <CIA-45> 03kilobyte * ra0b4e07a01b2 10/crawl-ref/source/spl-tornado.cc: Don't blame the player for trees felled by monster-cast tornado.
12:30:42 <CIA-45> 03eronarn * r8457c3a67da5 10/crawl-ref/source/delay.cc: Make recite train more often.
12:30:42 <CIA-45> 03kilobyte * rb2857e023672 10/crawl-ref/source/showsymb.cc: If an item ends up in a wall, show the wall (underlined), not the item.
12:30:43 <CIA-45> 03kilobyte * r415db94f50e0 10/crawl-ref/source/spl-tornado.cc: Don't delay mirror damage due to tornado until the turn's end.
12:30:45 <CIA-45> 03kilobyte * r4fdebf294f1b 10/crawl-ref/source/ (externs.h fineff.cc): Use mid not mindex for delayed "final" effects.
12:30:46 <Napkin> actually.. chrome is is telling me the ssl website is down :-O
12:30:46 <CIA-45> 03kilobyte * r53c4d2c47f4e 10/crawl-ref/source/rltiles/dc-mon/demons/shadow_imp.png: Denzi's flightless shadow imp tile.
12:30:49 <CIA-45> 03kilobyte * r6b978d6cd4a2 10/crawl-ref/source/hints.cc: Hints: you don't eat with '%'.
12:30:53 <CIA-45> 03kilobyte * rff9534c76587 10/crawl-ref/source/beam.cc: Prohibit flame-branded missiles from entering trees and wax.
12:31:04 <CIA-45> 03kilobyte * ra646b10eb535 10/crawl-ref/source/tags.cc: Fix a crash when reloading a game after restart_after_game.
12:31:49 <greensnark> Although it may be that a Mac update screwed things up; I believe Chrome uses the system keystore to check CA certs
12:32:07 <Napkin> it doesn't recognize the certificate chain here
12:32:18 <Napkin> but i'm a bit busy - back in 10
12:33:29 <kilobyte> greensnark: would you have a while to edumacate me about map cache files?
12:33:45 <kilobyte> (and db ones too, but this one seems straightforward)
12:35:59 <kilobyte> the thing is: especially on slow machines and network drives, rebuilding the cache takes ages.  It is likely an user interrupts that.  The next time Crawl is started, it compares the timestamps, sees the db/{dsc,lux,idx} are newer than the source, and is happy.  Which leads to fun bugs.
12:37:22 <greensnark> Yeah
12:37:28 <kilobyte> since you messed with them before, I'd want your opinion how to detect incomplete/corrupted builds
12:37:54 <greensnark> I wanted to switch to a sqlite db with a flag table to indicate the db is complete that's set at the end of the load
12:38:02 <greensnark> I'm not really happy with the binary blob deal
12:38:13 <greensnark> Actually now that we have a binary save container, we could use that for the map caches
12:38:15 <kilobyte> for the db, I'd store the timestamp of youngest source inside the db as the last operation; if it commits, we're set
12:38:30 <greensnark> Yes, something like that
12:39:44 <kilobyte> yeah, this is one of possibilities, except currently it assumes it can lock the file exclusively.  This can be changed -- trivially if we allow it to grow unchecked, can be tricky for properly reclaiming unused space.
12:40:23 <kilobyte> of course, replacing the whole file can remove any need for writing concurrently with reading
12:41:21 <greensnark> Does it do an exclusive lock even for reads?
12:41:41 <greensnark> Exclusive lock for writes sounds perfectly normal and ok to me
12:42:28 <kilobyte> the options I came up with so far: 1. writing {dsc,lux,idx} to new files, doing a fsync()+rename() dance; 2. using the save container -- to a new file, then fsync()+rename; 3. using sqlite
12:43:42 <greensnark> I'd prefer to switch to the save container or sqlite
12:43:59 <greensnark> Because having a separate index file is annoying and fragile
12:44:09 <kilobyte> I learned from a long flamewar on debian-devel recently that to be reasonably secure that your write succeeds, you need to: fsync(temp file);rename(temp file,regular name);opendir(directory);fsync(directory)
12:44:14 <greensnark> Much easier to be able to pick the map by name from the file
12:44:22 <kilobyte> the last two might be not needed if it's a single file
12:44:36 <kilobyte> or that, every map can be a separate chunk
12:44:46 <Napkin> greensnark - it could be, that chrome verifies the revocation list, when accessing a ssl-secured website
12:44:59 <kilobyte> since the save container is a filesystem-in-a-file basically
12:45:04 <greensnark> Napkin: Firefox is also acting up :)
12:45:11 <Napkin> and i think startssl.com is offline - that would also explain, why it took so long for the first access to crawl.develz.org
12:45:48 <greensnark> kilobyte: Yes, if you have the time, please move it to the save container
12:45:57 <Napkin> no, my firefox is working fine
12:46:04 <Napkin> what does yours tell you, greensnark?
12:46:30 <greensnark> You could have one chunk that stores the map selection info that remains loaded during the game, and a chunk for each map so that the game can load the map def when it wants to use the map
12:46:58 <greensnark> Napkin: sec_error_unknown_issuer
12:47:09 <Napkin> more details
12:47:20 <Napkin> continue and then check the certificates
12:47:51 <greensnark> Shouldn't I see two StartCom certificates in the chain?
12:48:01 <Napkin> exactly
12:48:02 <greensnark> I see the StartCom Class 2 Intermediate
12:48:08 <Napkin> and none above?
12:48:10 <greensnark> But not the StartCom root
12:48:37 <Napkin> check authorities in firefox's certificate manager
12:48:41 <greensnark> Unless the UI has changed so as to hide the root, I'm not seeing the root
12:49:27 <greensnark> I have the StartCom certification authority in the firefox ca list
12:49:40 <Napkin> "StartCom Ltd." -> "StartCom Certification Authority"
12:49:55 <greensnark> Yes, O StartCom Ltd, CN as you said
12:50:08 <greensnark> Expiring 9/17/36, all good
12:50:19 <greensnark> But I don't see this CA cert in the chain from CDO
12:50:44 <greensnark> Do you know some other site that uses the same intermediate CA?
12:50:51 <greensnark> Apart from CAO, which also does not work for me :P
12:52:14 <Napkin> chrome is definitely checking revocation list - and since it can't reach that, it gives "Error 7 (net::ERR_TIMED_OUT): The operation timed out." for the first time, and then the error you mentioned for every next try.
12:53:42 <greensnark> The thing that weirds me out is that I don't see three certs in the chain. Shouldn't I see three certs in the chain?
12:53:49 <greensnark> Or was I hallucinating that :D
12:53:57 <Napkin> no, you are right
12:54:09 <Napkin> my firefox has automatically downloaded them
12:54:25 <Napkin> they are listed in "Authorities"
12:54:38 <greensnark> Your intermediate CA is listed in Authorities?
12:54:46 <Napkin> not mine
12:54:47 <Napkin> theirs
12:54:59 <greensnark> Yes, StartCom Class 2 Intermediate, I mean :)
12:55:02 <Napkin> it was automatically loaded by firefox - since i didn't do that
12:55:10 <greensnark> Huh, why would it do that
12:55:29 <Napkin> because I specified it as ChainFile in apache
12:55:38 <greensnark> The intermediate CA is signed by the main StartCom root, right? I don't see the need for Firefox to import the intermediate, unless I'm horribly confused
12:55:49 <greensnark> A-ha
12:55:58 <Napkin> you have to
12:56:08 <Napkin> otherwise it won't work
12:56:23 <greensnark> Really?
12:56:36 <Napkin> firefox of course loads both certificates (intermediate CA & server certificate)
12:56:46 <Napkin> but i didn't expect it would save the intermediate CA
12:56:51 <Napkin> but it seems it does
12:56:59 <greensnark> It doesn't save it here :)
12:57:05 <greensnark> Did you import it manually? :)
12:57:16 <Napkin> yes, the server certificates are signed by the intermediate CA
12:57:19 * greensnark suspects Napkin of installing the intermediate CA cert manually.
12:57:33 <greensnark> No, did you install the intermediate CA cert in your browser, I mean
12:57:33 <Napkin> so it needs to load it, otherwise i can't verify the server certificate
12:57:41 <Napkin> no, of course not, greensnark
12:58:08 <greensnark> But you said your Firefox has the intermediate cert as a trusted root?
12:58:18 <greensnark> Or did I misunderstand
12:59:54 <Napkin> http://ipxserver.dyndns.org/misc/firefox_certs.png
12:59:57 <greensnark> Sorry if I'm making you pull your hair out :)
13:00:04 <Napkin> you are not
13:00:10 <Napkin> but, you need to believe me
13:00:16 <greensnark> Oh... wow
13:00:23 <Napkin> I didn't install them manually, so firefox installed them itself
13:00:24 <greensnark> I just have the one StartCom in my authorities list
13:00:32 <Napkin> is it a new installation?
13:00:47 <greensnark> My Firefox is newish, yes
13:00:58 <greensnark> Chrome is what I've been using for a while
13:01:12 <greensnark> One sec, I'll dig up my old laptop
13:01:30 <Napkin> i think, both are unhappy about startssl.com being offline, greensnark
13:02:33 <greensnark> Hm, sounds reasonable
13:02:59 <Napkin> i disabled "check for certificate revocation" in chrome now and will try again
13:03:08 <greensnark> Dangerous :P
13:03:20 <greensnark> But yes, worth checking to see if that's the problem
13:03:38 <Napkin> no, doesn't help
13:03:47 <Napkin> it still gets a timeout
13:04:10 -!- OG17 has joined ##crawl-dev
13:04:14 <Napkin> so - it seems it's trying to load the intermediate certificate from startssl.com and not from our servers
13:04:31 <Napkin> restart your chrome
13:04:47 <Napkin> try https://cdo and i bet you'll run into a time-out as well
13:05:01 <Napkin> every next tries will give "can't verify certificate"
13:05:26 <greensnark> I don't get the timeout, just "Can't verify cert" within 2-3s
13:05:34 <Napkin> on new start?
13:05:37 <greensnark> Yes
13:05:47 <Napkin> hmm - different behaviour then
13:06:06 <greensnark> I am on a hotel wifi that keeps poking its head in periodically, don't know if that makes it worse :)
13:06:48 <greensnark> Anyway, nothing major, beyond all the panicked hordes who will shortly be complaining to you :D
13:07:44 <Napkin> well, going to bed soon anyways ;-P
13:07:56 <greensnark> Gnight :)
13:08:07 <Napkin> not that soon!
13:08:09 <Napkin> ;)
13:08:34 <Napkin> but indeed interesting
13:08:50 <Napkin> hope they didn't get hacked!
13:08:57 <Napkin> that would be so much work ;)
13:09:48 <greensnark> Good thing you do all the cert work so I can be all airy and nonchalant :P
13:09:59 <Napkin> lol :-P
13:10:21 <Napkin> btw
13:10:31 <Napkin> i decided to just clone CDO to the new server :-$
13:10:50 <greensnark> Nice migration strategy :)
13:10:53 <Napkin> so.. I guess I need to take CDO offline for a week or so
13:10:55 <greensnark> When's the grand move
13:11:00 <greensnark> A week?!
13:11:03 <Napkin> until DNS is updated
13:11:08 <Napkin> 4 days, usually
13:11:26 <greensnark> I thought 24 hours was more than enough for propagation?
13:12:01 <Napkin> may be
13:12:06 <Napkin> check dig crawl.develz.org
13:12:41 <greensnark> What's the right answer? :P I get 80.190.48.234
13:12:57 <Napkin> the number of seconds in front of the "IN A"
13:13:13 <greensnark> 2213
13:13:19 <greensnark> But this is my local DNS server
13:13:22 <Napkin> check develz.org
13:13:23 <greensnark> Doesn't cache long
13:13:39 <greensnark> 85715
13:13:46 <Napkin> mine does 52330
13:13:58 <greensnark> Less than a day :P
13:14:23 <Napkin> damn! ;D
13:14:39 <Napkin> well, let's see how fast it goes
13:14:54 <Napkin> :)
13:15:35 <Napkin> the cloning, I mean
13:15:49 <greensnark> Don't you have a 1Gbps pipe :P
13:17:23 <greensnark> How are you cloning it, btw?
13:17:48 <Napkin> rsync or tar&ssh, I guess
13:18:02 <Napkin> do you see any other option?
13:18:27 <greensnark> No, just curious
13:18:45 <greensnark> One of my friends recently started using safekeep for backups, have you tried it?
13:18:50 <greensnark> Apparently it's built on rdiff
13:19:08 <greensnark> And has an XML config :P
13:19:16 <greensnark> Which appeals to my friend, who's a complete Java type :P
13:19:56 <Napkin> geez
13:20:07 <greensnark> Oh, forgot to ask: are you planning to make cdo dgl a separate VM from the website?
13:20:24 <Napkin> yes, one day
13:21:07 <Napkin> But I'll probably run out of time if I want to do both at once ;)
13:21:21 <Napkin> cloning and moving crawl&dgl to a vm
13:21:30 <greensnark> True
13:21:46 <greensnark> As it is it sounds like you'll lose at least a day in all the cloning and setup
13:22:41 <Napkin> I believe so
13:25:16 -!- elly has quit [Ping timeout: 246 seconds]
13:30:49 <kilobyte> Napkin: can you just reduce the TTL for the time being?
13:31:28 <kilobyte> to a day max...  heck, perhaps even 5 minutes
13:32:53 -!- elly has joined ##crawl-dev
13:34:59 <Napkin> kilobyte: good idea, done
13:35:42 <greensnark> Yep, that'll give you no excuse to not flip the switch immediately :P
13:36:00 <greensnark> So shiny new CDO tomorrow? :D
13:36:03 <Napkin> thank you guys, pleasure working with you... :>
13:36:23 <Napkin> no way - never clone on monday!
13:36:42 <greensnark> Aw, we have to wait a week? :(
13:36:43 <Napkin> nah, need to inform the other projects first
13:38:10 <Napkin> also, i'm playing a MDFi right now :-P
13:39:20 <greensnark> So you'll be free in the next 5 minuts? :)
13:41:33 <Napkin> lol
13:41:49 <Napkin> i had such a shiny nice char
13:41:53 <Napkin> and then i did slime
13:43:00 <kilobyte> ah, right
13:43:16 <Napkin> -> berserk 2, blurry vision 2, herbivore 1, deteriorating body 2, -20%mp, -2 str/dex.. oh.. and they gave me regeneration 2!
13:43:23 <kilobyte> I've learned to not do Slime before exping and gearing up in a Zig first
13:43:44 <greensnark> I've learned to not do Slime
13:44:01 <Napkin> thank god no teleportitis.. otherwise I would have quit
13:44:17 <greensnark> I hope that deteriorating body is working out well for you :P
13:44:17 <Napkin> having found only one cure mutation
13:44:28 <Napkin> that's what I got rid of! \o/
13:44:38 <greensnark> Autozerk is pretty nice too
13:44:42 <Napkin> and i can read a bit better.. and damn it, not so angry anymore ;)
13:44:47 * greensnark clings to hope.
13:45:02 <greensnark> Oh damn, you cleared all the nice mutations
13:45:22 <Napkin> i was actually expecting to get regeneration removed - but must be my lucky day!
13:45:28 <greensnark> I'll have to look elsewhere for my daily dose of schadenfreude
13:45:38 <Napkin> move along ;-P
13:47:29 <Napkin> btw, could it be, that elf conjurers were made "more fun"?
13:48:10 <Napkin> don't remember ever gotten so many times drained and put on fire by them
13:49:37 <greensnark> Those are just perks for playing a heavy armour character
13:49:50 <Napkin> ah, ok, yeah
13:50:00 <Napkin> this is the first heavy armour char for years
13:50:03 <greensnark> Vaults will be fun too, enjoy your shadow dragons
13:50:17 <Napkin> ran from him
13:50:19 <Napkin> ;)
13:50:23 <Napkin> *it
13:51:00 <Napkin> but found the ring of Polytheism (right hand) {Hunger- rF++ rN+ Dam+4} wobbling on the royal jelly
13:51:12 <Napkin> so.. will pay vault8 another visit later ;>
13:52:18 <greensnark> !lm Napkin uniq=Mennas
13:52:18 <Sequell> 2. [2011-04-09] Napkin the Executioner (L18 MDFi) killed Mennas on turn 54017. (D:20)
13:52:25 <Napkin> oh yes!
13:52:26 <greensnark> !lm Napkin uniq=Mara
13:52:27 <Sequell> 6. [2011-04-05] Napkin the Swashbuckler (L27 KoHe) killed Mara on turn 97071. (Snake:5)
13:52:32 <greensnark> There's still hope!
13:52:40 <Napkin> lol
13:52:43 <greensnark> !lg Napkin KoHe
13:52:44 <Sequell> 21. Napkin the Eviscerator (L27 KoHe), worshipper of The Shining One, escaped with the Orb and 3 runes on 2011-04-06, with 1403561 points after 158528 turns and 21:45:46.
13:52:49 <greensnark> Oh, congrats :)
13:52:55 * Napkin whispers: I stabbed Mara! o/
13:53:02 <Napkin> what a sweeeeeeet revenge :D
13:53:05 <Napkin> Thank you :)
13:53:10 <greensnark> !lm Napkin uniq=Mara -tv
13:53:11 <Sequell> 6. Napkin, XL27 KoHe, T:97071 (milestone) requested for FooTV.
13:53:26 <Napkin> second win this year!
13:53:45 <Napkin> that's like.. omfomg!
13:53:49 <greensnark> I like your fountain symbol
13:53:54 <greensnark> Is that the official symbol
13:54:02 <Napkin> stole it from Eifeltrampel ;)
13:54:08 <Napkin> no, i don't think so
13:54:23 <greensnark> Very psi
13:55:11 <greensnark> @rc Eifel
13:55:14 <greensnark> Brrh
13:55:20 <greensnark> You should add @rc to Gretell :P
13:55:36 <Napkin> yeah, like I added all the other commands ;)
13:55:46 <greensnark> Exactly :P
13:56:16 <Napkin> hehe
14:00:06 <Napkin> omg... got angry at a toadstool? grrrr...
14:17:47 <st_> greensnark: hey, I've had a bug in my game - a full level vault on D:24 that caused no pan gate to spawn (and none elsewhere.) is there anything you can do to regenerate the level for me or something? (clouded on CAO, playing erocrawl)
14:18:11 <kilobyte> dammit, reading rc files without literals is a pain in the ass
14:18:34 <kilobyte> Napkin: which editor do you prefer to use?  ee or virus?
14:18:54 <Napkin> virus
14:19:17 <kilobyte> I personally just nuke the whole file (tedious since you can't do that faster than line by line), then paste everything, so I don't care which one.
14:19:39 <kilobyte> new Crawl allows fountain:ψ
14:19:54 <Napkin> old did too ;)
14:19:56 <kilobyte> except neither ee nor virus allow you to use that on CDO
14:20:27 <kilobyte> uhm no, you could only put single-byte characters there
14:20:29 <Napkin> ah, i'm used to use the character code thingy
14:21:08 <greensnark> st_: I could theoretically blow away D:24, but can you file a bug first :P
14:21:13 <Napkin> 596 cset_unicode = item_gold: 8364
14:21:13 <Napkin> 597 cset_unicode = invis_exposed: 126
14:21:13 <Napkin> 598 cset_unicode = fountain: 968
14:21:18 <st_> I'm filing it right now
14:21:28 <Napkin> that's why i used and it worked fine, kilobyte
14:21:30 <st_> I made this char just for zig raiding!
14:21:38 <greensnark> st_: Feel free to file it against regular Crawl
14:21:56 <greensnark> Because I think the gates-on-vault thing was broken during the dungeon reorg
14:21:58 <kilobyte> greensnark: not really, it will assert if you do so
14:22:07 <greensnark> kilobyte: Oh
14:22:10 <kilobyte> because there's some silly counter of generated levels
14:22:15 <greensnark> That silly thing that tracks visited levels?
14:22:19 <greensnark> Ugh
14:22:20 <kilobyte> yeah
14:22:25 <greensnark> Well, let me look at it anyway
14:22:32 <greensnark> I should put in the save backup script for EroCrawl
14:23:24 <greensnark> Not enough hours in the day :P
14:23:32 <greensnark> Napkin: Can I borrow some hours from you
14:23:51 <Napkin> all reserved already ;-P
14:24:18 <greensnark> st_: Hang on while i add the save backup menu for EroCrawl :)
14:24:41 <st_> oh whoops, just hit report - I'll put it in the comments
14:24:54 <greensnark> You can't attach files after the fact?
14:25:14 <Ashenzari> Getting a full level vault on D:24 can cause no pan gates to generate (https://crawl.develz.org/mantis/view.php?id=3786) by st
14:25:20 -!- ais523 has joined ##crawl-dev
14:25:20 <st_> I usually just post the link
14:25:40 <greensnark> Oh right
14:30:49 <greensnark> Napkin: Btw, Chrome's happy with CDO cert again
14:30:55 <greensnark> st_: You should be able to back up the save now
14:31:18 <greensnark> You might need to log out of dgl and log back in
14:31:21 <Napkin> awesome, startssl.com is back too ;)
14:33:27 <st_> posted the save
14:33:33 <greensnark> Got it, thanks
14:34:02 <Napkin> you chose CDO instead of CAO, st_
14:34:23 <Napkin> oh, and btw, Eronarn wants to use the github tracker
14:34:31 <st_> oh whoops, that was pure habit :P
14:34:45 <greensnark> Napkin: This bug will also exist in regular Crawl, though
14:34:56 <greensnark> i.e. it's not LO or Octopode specific
14:35:00 <greensnark> I'll just add a note
14:37:10 <kilobyte> !tell Zaba Could you look at https://crawl.develz.org/mantis/view.php?id=3786 ?  Even setting the pan entry vault as transparent doesn't help -- you have more clue here.
14:37:11 <Henzell> kilobyte: OK, I'll let Zaba know.
14:37:45 <greensnark> What's Zaba's CAO id
14:37:47 <greensnark> !lg zaba cao
14:37:47 <Sequell> 90. Zaba the Ruinous (L3 SpCj), blasted by Ijyb (wand of frost {zapped: 2}) on D:2 on 2010-12-06, with 105 points after 1720 turns and 0:04:52.
14:41:39 <Napkin> oh nono, not complaining, just saying
14:42:57 <greensnark> Oh, I need to move CAO to the 0.8 release branch
14:42:59 <greensnark> Ulch
14:43:09 <greensnark> Are saves compatible? :P
14:43:48 <Napkin> not yet
14:43:56 <greensnark> No? Noooo
14:44:09 <Napkin> i mean, jpeg asked me not to put CDO and mingw builds to 0.8
14:44:19 <greensnark> Oh, why's that?
14:44:45 <Napkin> she said, she's using the reports to check what to cherry-pick
14:45:06 <Napkin> if i understood her correctly
14:45:13 <greensnark> But won't she get the bug reports anyway from 0.8 if the same bugs exist
14:45:50 <greensnark> Anyway, it saves me work for now so I won't complain :D
14:45:51 <Napkin> well, i didn't poke too deep
14:46:02 <Napkin> ;)
14:46:05 <Napkin> same here :D
14:55:41 <kilobyte> perhaps one server can run 0.8 and the other 0.9?
14:59:22 -!- Kurper has joined ##crawl-dev
15:05:26 <Ashenzari> Ashenzari identification of items on the floor sometimes doesn't happen immediately. (https://crawl.develz.org/mantis/view.php?id=3787) by elliptic
15:12:13 -!- Cryp71c has joined ##crawl-dev
15:12:19 <Cryp71c> Afternoon
15:32:56 <greensnark> st_: Can you try reloading your save, it should be fixed now
15:33:43 <st_> greensnark: seems to be working, thanks a lot!
15:33:51 <greensnark> Actually, now that I think about it, it would have been simpler to just create a pan gate by hand rather than delete the old level
15:34:00 <greensnark> I hope you didn't have anything valuable on the level :P
15:36:35 -!- syllogism has quit []
15:38:31 <st_> new D:24 has the trident of the octopus king for my octopode heh
15:38:46 <greensnark> Beware the octopus king's wrath
15:38:59 <Eronarn> i added a mention of the octopus king in the octopus dummy monster desc
15:39:15 <Eronarn> didn't give them a bonus or anything to using it though, though it's a pole weapon so they have good apts for it
15:39:15 <st_> I saw that, gave me a chuckle
15:39:46 <greensnark> We need an octopus king
15:40:03 <greensnark> Eronarn: Didn't you implement the Glaive of Prune pruneform somewhere?
15:42:03 <monqy> I half-implemented half of it, at least
15:42:44 <monqy> prune blood for glaive of prune wielders, but it was kind of ugly so if I ever pick it up again I'll have to figure out some other way to go about it
15:43:42 <Eronarn> greensnark: it was discussed that gen 2 of wizlabs could be for more of the non-caster types
15:44:07 <Eronarn> like Zhor's Cave or whatever
15:44:39 <Eronarn> there are a lot of named randarts we haven't paired up with wizlabs yet
15:52:48 <greensnark> monqy: Aha, so it was you
15:52:54 <greensnark> Pruneform would be awesome
15:53:08 <monqy> not sure what it would do
15:54:12 <greensnark> Make you enormously attractive to constipated ogres, no doubt
15:54:38 <CIA-45> 03kilobyte 07stone_soup-0.8 * rd33160bcc100 10/crawl-ref/source/player-equip.cc: Don't insta-kill people on vampiric weapon unmeld.
15:54:48 <CIA-45> 03kilobyte 07stone_soup-0.8 * r92f60d290283 10/crawl-ref/source/player-equip.cc: Don't show "it sticks to your hand" on cursed unmeld.
15:54:49 <CIA-45> 03MarvinPA 07stone_soup-0.8 * rb2adebcdd7d7 10/crawl-ref/ (docs/template/apt-tmpl.txt source/spl-cast.cc): Rename some casting/aptitude adjectives for clarity
15:54:49 <CIA-45> 03MarvinPA 07stone_soup-0.8 * r7ac1a4b0f7bc 10/crawl-ref/source/dat/des/variable/float.des: Don't autoexplore into lemuel_bear_cage
15:54:50 <CIA-45> 03MarvinPA 07stone_soup-0.8 * r78acc2d9b695 10/crawl-ref/source/mon-cast.cc: Add a message for monsters casting Trog's Hand
15:54:51 <CIA-45> 03MarvinPA 07stone_soup-0.8 * r9ff32208f01b 10/crawl-ref/source/dat/database/monspell.txt: Adjust Roxanne's casting messages, since she can't move
15:54:53 <CIA-45> 03MarvinPA 07stone_soup-0.8 * r785b02ca42ea 10/crawl-ref/source/player.cc: Message fix for levitation while clinging
15:54:54 <CIA-45> 03kilobyte 07stone_soup-0.8 * r73db83fb8fae 10/crawl-ref/source/player-equip.cc: Tell weapon/jewel artefacts it's only a meld.
15:54:56 <CIA-45> 03MarvinPA 07stone_soup-0.8 * r886192ffd486 10/crawl-ref/source/transform.cc: Allow extending spider form while clinging over dangerous terrain
15:54:59 <CIA-45> 03eronarn 07stone_soup-0.8 * r3a02f47a89bb 10/crawl-ref/source/delay.cc: Make recite train more often.
15:55:00 <CIA-45> 03kilobyte 07stone_soup-0.8 * r5484a1a500c8 10/crawl-ref/source/spl-tornado.cc: Don't delay mirror damage due to tornado until the turn's end.
15:55:01 <CIA-45> 03kilobyte 07stone_soup-0.8 * r5a98ad8cdf4a 10/crawl-ref/source/rltiles/dc-mon/demons/shadow_imp.png: Denzi's flightless shadow imp tile.
15:55:02 <CIA-45> 03kilobyte 07stone_soup-0.8 * rcdeebc26fdfa 10/crawl-ref/source/spl-tornado.cc: Don't blame the player for trees felled by monster-cast tornado.
15:55:03 <CIA-45> 03kilobyte 07stone_soup-0.8 * r94096b0e5baf 10/crawl-ref/source/rltiles/dc-mon/animals/jumping_spider.png: An update to jumping spider's tile.
15:55:03 <CIA-45> 03kilobyte 07stone_soup-0.8 * rf5cea864f61f 10/crawl-ref/source/tags.cc: Fix a crash when reloading a game after restart_after_game.
15:55:04 <CIA-45> 03kilobyte 07stone_soup-0.8 * rfaf690d40ddb 10/crawl-ref/source/hints.cc: Hints: you don't eat with '%'.
15:55:08 <CIA-45> 03MarvinPA 07stone_soup-0.8 * rbed6f2e9bb0c 10/crawl-ref/source/spl-transloc.cc: Don't avoid glow by cancelling semicontrolled blinks with enter instead of escape
15:56:27 <Eronarn> imo make pruneform not actually happen but just three levels of an only-zin-removeable prune mutation
15:57:03 <Eronarn> (and at pruneform 3, eating fruit counts as cannibalism, of course)
16:01:00 -!- barnex has joined ##crawl-dev
16:08:26 -!- jld has quit [Ping timeout: 248 seconds]
16:14:19 -!- jld has joined ##crawl-dev
16:19:21 -!- Cryp71c_ has joined ##crawl-dev
16:19:25 <Cryp71c_> join #linux
16:19:31 -!- Cryp71c_ has left ##crawl-dev
16:20:31 -!- casmith789 has joined ##crawl-dev
16:23:04 <Napkin> {Spirit, +Rage rN+ MR} <- what's that? :-O
16:23:20 <ghallberg> Bear spirit hat?
16:23:25 <Napkin> yes
16:23:34 <Napkin> spirit guardian?
16:23:41 <greensnark> Guardian spirit, yeah
16:23:45 <ghallberg> mhm
16:23:53 <greensnark> Although a hat in which you could store spirits in a little flask would also be a-ok with me right now
16:23:55 <Napkin> oops, that way
16:24:15 <CIA-45> 03greensnark * rd70fb184fe63 10/crawl-ref/source/delay.cc: [3784] Fix Lua macros not being able to long-rest (5) using process_keys.
16:30:21 <kilobyte> is there any reason to keep unidentified appearances of unrands?
16:30:52 <kilobyte> the bad ones should be random or no one spoiled will ever wear them, good ones can be recognized at sight too
16:32:13 <greensnark> Yeah, it's a spoiler
16:32:22 <Eronarn> kilobyte: i think some of the unrand descriptions are cute... what about having them appear yellow (rather than white), and autoID on pickup?
16:32:22 <greensnark> But I'd hate to lose the robe of misfortune in the cull
16:32:35 <Eronarn> that way it's also clear that it's not a normal artifact
16:32:49 <kilobyte> greensnark: good heavens!  How could you suspect me of wanting to do that!?!
16:32:49 <Eronarn> people get confused all the time over this, especially the ones like 'smoking scimitar'
16:32:55 <greensnark> Or we could make two randarts for each description :P
16:33:05 <greensnark> *unrandarts
16:33:25 <Eronarn> greensnark: imo: robe of misfortune shows up as something entirely different, with entirely different stats, and only changes into the robe all of a sudden in a high-tension situation
16:33:27 <kilobyte> I'll make it have a regular randart description
16:33:50 <greensnark> You seriously proposing to screw with the fabulously ornate robe? :P
16:34:10 <Eronarn> 'You feel a cold sweat come on. Your robe of the Eight Boll Weevles is actually the robe of Misfortune!'
16:34:26 <greensnark> I like Eronarn's idea better, or just add another decent robe unrand with the same description :P
16:34:41 <greensnark> I'm fine with getting rid of the uninspired unrand names :P
16:34:45 <Eronarn> greensnark: don't forget the issue of RC? trivializing it
16:34:47 <greensnark> But not the fabulously ornate robe
16:35:00 <greensnark> Eronarn: Well yes, but your idea requires more work :P
16:35:07 <ais523> having three unrands which are all fabulously ornate robes makes sense
16:35:14 <ais523> but spoiled players would just make sure to have a scroll of RC before putting it on
16:35:22 <Eronarn> oh, for sure - but yeah, that
16:35:34 <Eronarn> it's almost not even worth doing, it's only marginally better for the status quo
16:36:06 <greensnark> Newsflash: any sane player will have RC before trying on any randart :P
16:36:19 <greensnark> So it's not like this is an earth-shattering development
16:36:22 <Eronarn> solution: randarts never gen cursed :P
16:36:39 <Eronarn> (except for recurse ones, of course)
16:36:44 <st_> misfortune gives "braindead" status, no more RC for you
16:37:06 <ais523> but then you just couldn't remove it at all
16:37:16 <ais523> I think curses are uninteresting in Crawl mostly because it's RC or nothing
16:37:28 <ais523> they're interesting in NetHack because there's something like 15 ways to remove them, with various levels of risk
16:37:40 <ais523> and interesting in Pokémon Mystery Dungeon because of the way the item system works
16:38:01 <greensnark> That's fine, curses can be quite interesting in Crawl; they're situational
16:38:16 <greensnark> Crawl doesn't go for the stuff-the-main-game-with-minigames-because-the-main-game sucks approach :P
16:38:20 <ais523> (you can't carry more than 48 items, you can't backtrack so no making stashes, cursed items have no effect, and the only uncursing item affects all of inventory and is very rare)
16:38:43 <Eronarn> curses in crawl really almost never come up
16:39:23 <ais523> (so it's an interesting issue about whether you're going to carry good cursed items with you until you can uncurse them, and about whether you want to use your uncursing item now or save it till later)
16:39:26 <Eronarn> early on before you have rc, or if you do dumb stuff (but then you either retreat to your stash and fix it, or die in a way that sucks)
16:40:00 <greensnark> I've had some quite hilarious episodes of cursed-maceitis, and I usually don't die, but it produces some nice tense early-game play
16:40:06 <ghallberg> I had a bit of a curse issue today, but yeah it rarely shows up.
16:40:10 <greensnark> Also cursed blowgunitis
16:40:27 <greensnark> All my fault, but I still find it entertaining
16:40:40 <ais523> isn't that NetHack-style, then, where the interesting gameplay only comes from mistakes made by the player?
16:40:43 <Eronarn> early, sure, cursing can matter. later on though, pretty much never
16:40:59 <greensnark> ais523: Hardly, you're being deliberately obtuse again
16:41:01 <ais523> hmm, if you had an earlyish-game monster that cursed items, it might be itneresting
16:41:07 <Eronarn> except for the aforementioned situation where you made a mistake and then either die because of it and feel really, really dumb or get back to your stash
16:41:10 <greensnark> That's like saying optimal play is to search each square for traps in NH
16:41:19 <ais523> it isn't, earlygame at least, you run out of food
16:41:34 <ais523> and possibly even lategame, because you need to search something like 20 times to get a decent chance of finding a trap
16:41:44 <ais523> so carrying enough food to make that worthwhile would encumber you
16:41:58 <ais523> not to mention searching that much would cause a lot more monsters to spawn
16:42:08 <greensnark> I do think Crawl has too many RC scrolls
16:42:16 <greensnark> I don't think it needs fancy curse effects
16:42:21 <ais523> if they were very rare, it might be interesting PMD-style
16:42:32 <ais523> a use-now-or-save-for-later approach
16:42:39 <ais523> but they'd just be treated like potions of cure mutation
16:43:12 <Eronarn> greensnark: well, i think the basic curse effect itself could stand changing - but until it gets changed, we definitely don't need more stuff that curses or more ways to uncurse things
16:43:41 <greensnark> More a curse of the Nauglafring kind of curse? :)
16:43:45 <Napkin> not being able to uncurse your things reasonably often sounds like nethack
16:44:03 <Eronarn> (one proposal I had: you can unwield cursed items, but they stay in your inventory, and you might suddenly find yourself wielding them + unable to unwield them immediately during tense fights)
16:45:55 <Ashenzari> screenshot (in character dumps) is broken (https://crawl.develz.org/mantis/view.php?id=3788) by reid
16:48:53 -!- valrus has joined ##crawl-dev
16:50:57 <Ashenzari> magic-mapped stairs are the wrong color (https://crawl.develz.org/mantis/view.php?id=3789) by reid
17:11:04 <Ashenzari> monsters that get pushed past leave monster trails in unmapped areas (https://crawl.develz.org/mantis/view.php?id=3790) by reid
17:12:36 <CIA-45> 03kilobyte * r1514f761a7eb 10/crawl-ref/source/dat/database/monspeak.txt: Don't try to suggest certain resistances. [#367]
17:12:46 <CIA-45> 03kilobyte * rc2e0dba7fcec 10/crawl-ref/source/travel.cc: Don't auto-attack toadstools when wielding the Wrath of Trog.
17:12:47 <CIA-45> 03kilobyte * rf5508d19fbf7 10/crawl-ref/source/viewmap.cc: Fix a misbehaviour of shifted vi keys on 'X' near map borders.
17:12:48 <CIA-45> 03kilobyte * r329193e8847d 10/crawl-ref/source/godconduct.cc: Fix Okawaru not accepting kills of holy beings.
17:12:50 <CIA-45> 03kilobyte * rabfc77003b2a 10/crawl-ref/source/invent.cc: Prompt before wearing or wielding a known-cursed item.
17:12:51 <CIA-45> 03kilobyte * rd055d1247ab1 10/crawl-ref/source/dat/des/portals/wizlab.des: Mention that wizlab portals are fading away in their long description.
17:12:52 <CIA-45> 03kilobyte * r27097e33c5d8 10/crawl-ref/source/religion.cc: A message when you switch from a good god to a good god.
17:12:52 <CIA-45> 03kilobyte * re0c89c607ae6 10/crawl-ref/source/travel.cc: Auto-kill toadstools if a mummy, ghoul or in lich form.
17:25:09 -!- Pseudonut has joined ##crawl-dev
17:31:47 -!- eith has quit [Ping timeout: 264 seconds]
18:05:58 -!- edlothiol has quit [Ping timeout: 276 seconds]
18:06:34 <elliptic> hm, I managed to crash via the asserts in cgotoxy()
18:06:43 <elliptic> I wonder how I did that
18:07:36 <due> You ended up going off the screen?
18:08:41 <elliptic> yes
18:14:31 <CIA-45> 03kilobyte * r262acb623819 10/crawl-ref/source/view.cc: Fix corrupted screenshots in dump/morgue files.
18:34:58 <Gretell> Unstable branch on CDO updated to: 0.9.0-a0-70-g262acb6 (32)
18:37:31 <greensnark> Zaba: http://mywiki.wooledge.org/XyProblem <-- Last one might interest you strangely
18:41:31 <Ashenzari> Remove staff monks (https://crawl.develz.org/mantis/view.php?id=3791) by reid
18:43:26 <due> "lol"
18:44:22 <due> Surely I'm not the only one who makes the historical connection between quarter staves and martial arts?
18:44:59 <Eronarn> heck, they let you do more offhand punches than other weapons
18:45:03 <Eronarn> in crawl
18:55:24 <ais523> it looks like a flavour vs. gameplay argument there
18:56:14 <elliptic> the flavour is pretty weak, IMO
18:56:36 <Eronarn> it's perfectly flavorful, lots of RL martial arts use stick weapons
18:56:53 <Eronarn> mechanically it's not so good, sure
18:57:02 <due> Mechanics might be problematic but the flavour is rock-solid.
18:57:48 <Eronarn> can we make high priests lose rHellfire yet
18:59:11 -!- Zao has left ##crawl-dev
18:59:27 -!- Zao has joined ##crawl-dev
19:21:09 <greensnark> Eronarn: Please s/dwarf/qwarf/ in your variant
19:21:18 <greensnark> And I will update CAO asap
19:21:31 * greensnark <3 qwarves
19:24:21 <Eronarn> heh
19:24:29 <Eronarn> i should sneak that in next time i do an update
19:25:35 <Eronarn> (there isn't one for tonight, btw... probably not for several days, most of the remaining bugs are of the annoying to fix variety)
19:27:35 <ais523> since when was Xom a good god?
19:28:21 <due> I'd support qwarves.
19:35:55 -!- ZorbaTHut has joined ##crawl-dev
19:40:19 -!- MarvinPA_ has quit [Ping timeout: 258 seconds]
19:50:57 -!- MarvinPA has joined ##crawl-dev
19:56:20 -!- not_evilmike is now known as evilmike
20:06:55 <Ashenzari> Ashenzari allows switching out cursed weapons. (https://crawl.develz.org/mantis/view.php?id=3792) by elliptic
20:16:36 -!- upsy has quit [Quit: Leaving]
20:29:57 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.]
20:58:31 -!- valrus has quit [Read error: Connection reset by peer]
21:15:53 -!- DrPraetor has joined ##crawl-dev
21:16:45 <DrPraetor> Some more misc. trunk feedback.  Elephants are now too common.  They significantly outnumber blink frogs.  They're less dangerous, but take longer to kill and it's getting boring.
21:17:38 <DrPraetor> Wight packs show up at too low a depth.  Did they get added at the same depth that individual Wights used to appear?
21:18:04 <DrPraetor> Other than that, haven't noticed many other monster distribution changes but others seem fine (good to see more orange rats.)
21:20:20 <DrPraetor> Also, has the number of hydras been dropped?
21:22:29 <Adeon> selective perception
21:24:50 <Eronarn> also adding more other monsters = less hydras because it's a weighted list, afaik (i don't think elephant packs came out of something else's weight)
21:26:22 <DrPraetor> Well, the weight for elephants needs to drop significantly.  I've had multiple packs on each level of lair 5-8.  They're not so difficult with this build, but they're a chore, which is not good.
21:27:31 <st_> not sure how many times we need to complain about elephants before their frequency is reduced
21:28:04 <ais523> st_: make sure the complaints don't drive the frequency /up/
21:28:04 <DrPraetor> I've never understood who it is that commits changes to that kind of thing, to be honest.
21:28:38 <DrPraetor> The proper complaint is not "aiee! elephants are too hard!", but "ugh, elephants are *so boring*".
21:28:54 <DrPraetor> If they made Lair more difficult I could see the argument.
21:29:03 <DrPraetor> (although for that I think we just want more hydras).
21:29:09 <st_> ais523: the danger would be that they get some gimmick to make them "more interesting." thankfully elephants already have one
21:29:42 <monqy> I like trampling and elephants and trampling, but the frequency is kind of ridiculous
21:30:10 <DrPraetor> No monster gimmick I agree with monqy.  The elephant is a good monster we just need about 1/4th as many packs of them.
21:30:58 <greensnark> Actually why is the elephant frequency not filed as a bug? I keep seeing reports of it and I think due agreed that they're too common
21:31:08 <greensnark> So if it's not reported as a bug, it should be
21:31:14 <greensnark> Unless something changed there
21:31:19 <due> Elephant frequency is due to band generation.
21:31:35 <due> We need to modify frequency functions to divide weight/chances/etc by the average band number automatically.
21:31:36 <greensnark> due: Yes, didn't we agree they're too common?
21:31:41 <due> greensnark: yes, totally
21:31:47 <monqy> is ugly/slime frequency due to band generation too?
21:31:54 <due> monqy: yes
21:31:59 <monqy> and yaktaurs etc
21:32:03 <monqy> imo all of them are too common
21:32:11 <due> monqy: For instance, where it might place two or three of a single monster on a level due to the standard generation algorithms
21:32:19 <due> it instead places 5 or so as a band instead.
21:32:31 <due> which is fine if you like 3 -> 15.
21:33:00 <monqy> and from the way you described it, I'm guessing adjusting band generation wouldn't be easy
21:33:01 <greensnark> Do we have a bug tracking this
21:33:43 <st_> it's been put on some wiki page and there was a recent bug about elephants fleeing too much which I complained in that was told "go to the wiki"
21:33:46 <greensnark> We may not have time to fix mon-pick.cc to do anything sane, but we can certainly tune the frequency
21:33:54 <DrPraetor> Maybe I'm just set in my ways, but I *like* the high frequency of Slimes, Yaktaurs and Uglies.  To me, they are signature crawl monsters.
21:34:23 <greensnark> I'm interested in elephants here
21:34:37 <st_> the whole no FRs on mantis thing can be a bit frustrating
21:34:44 <greensnark> This isn't an FR, it's a BR
21:34:55 <greensnark> Ok, I'll file it
21:35:21 <DrPraetor> Back to kvetching about stabbing training.  Being forced to train stabbing to 3 gives you 500 additional *skill points*, which increases the cost all all future skills by 1/8th of a skill_cost_needed.
21:36:03 <DrPraetor> For a high level character, that is.
21:36:28 <greensnark> Ok, that I'm totally not worried about, but feel free to discuss it :)
21:36:31 <MarvinPA> i'm pretty sure the answer to this is still pretty much "tough"
21:37:05 <DrPraetor> Point is, that's roughly the same penalty to all your skill costs as taking Invocations from 8->10.  It's a big penalty.
21:37:24 <MarvinPA> i'm pretty sure the answer to this is still pretty much "tough"
21:37:34 <due> greensnark: yes, file it; I will try to get some time to actually do crawl things and fix it.
21:37:47 <due> greensnark: at the very least I can divide current elephant generation by average band size manually
21:38:33 <MarvinPA> if crawl claimed to or was supposed to have a system where you get to allocate 100% of your experience exactly where you liked then maybe it would be an issue
21:39:20 <monqy> is this "reverse victory dancing" actually optimal
21:39:32 <DrPraetor> Monqy, yes, it certainly is.
21:39:50 <greensnark> due: Yes, I think for 0.8 we should just run stats and use them to figure out how many elephants we want
21:41:57 <due> good idea
21:42:11 <due> average over a thousand generations of Lair:5?
21:42:25 <Ashenzari> Elephants are too common (https://crawl.develz.org/mantis/view.php?id=3793) by greensnark
21:42:35 <due> We could manually set the band size to 0 temporarily to determine current actual generation rates.
21:42:37 <greensnark> place-pop uses 150 iters
21:42:58 <greensnark> I'd say: just run place pop and drop the rarity until you get the absolute number of elephants you like
21:43:11 <due> good idea
21:43:25 <greensnark> I'll do a run as soon as this thing finishes rebuilding :P
21:43:39 <due> :) thanks
21:43:48 <DrPraetor> The base XP-per-skill-point (which is not the same as a skill level) starts at 1+skill_level (and goes up by 1/3rd again for every skill_level above 9).  It then increases by 1+4% for every 800 total skill points you have, roughly speaking.
21:44:01 <due> mon-pick is so horrendously broken though.
21:44:13 <DrPraetor> So that 3-stabbing is giving you a -1 penalty to your primary skill, if it's in the 20s.  Just for starters
21:44:23 <greensnark> due: Yeah, we have a wiki page on replacing it and we promise to look at it each release and each release we punt it to the next :D
21:44:33 <due> I'd almost be inclined to say "let's do monster generation in Lua" but I don't think vaults would quite cut it.
21:44:46 <greensnark> Doesn't have to involve vaults at all
21:45:09 <due> I was thinking more of a parsed text-file which lists branch/level range and weight per-monster per-branch (or generic).
21:45:12 <DrPraetor> The *real problem*, if anyone cares, is that skill_cost_level uses the total number of skill points you have to calculate cost multipliers, rather than the total XP-sunk-into-skills, which is a big penalty for generalists.
21:45:23 <due> which we could certainly do in Lua insead
21:45:57 <DrPraetor> A big penalty for generalists is okay, of course, under the assumption that all skills are useful.  But if you have any useless skills at all (i.e., stabbing, armour skill if you're wearing leather, although with AEVP maybe not that.)
21:46:22 <DrPraetor> Then there is a *sizable* incentive to avoid training the useless skills, because it really costs you a lot.
21:46:45 <due> do people actually sit and worry "oh my skills aren't training exactly how i'd like"?
21:46:56 <monqy> DrPraetor: go ashenzari
21:46:58 <MarvinPA> no, they turn off stabbing and stop caring
21:47:05 <DrPraetor> No, most people are ignoring it, but they wouldn't if they knew how big a penalty it was.
21:47:05 <monqy> they reskill their worries away
21:47:11 <elliptic> drpraetor is literally the only person I have ever heard worry about this
21:47:27 <due> DrPraetor: I have *never* heard anyone complain about this before; I have heard people complain about *everything*.
21:47:28 <monqy> maybe someone on tavern worries about it
21:47:30 <greensnark> Well, he's not the only one, but I'm not buying it anyway
21:47:39 <greensnark> There's no end of people on the tavern who obsess over this
21:47:52 <elliptic> I don't read tavern :P
21:47:54 <greensnark> Including one person who insists that he has an awesomely superior system that he can't disclose because he's under NDA
21:48:08 <MarvinPA> hahaha yeah
21:48:16 <due> my only question is: if this is a sizable impact on playability and difficulty, why aren't other people complaining?
21:48:37 <DrPraetor> Because it's hidden in the unspeakably opaque training code and no-one (myself included), *entirely* understands how it actually works?
21:48:48 <due> DrPraetor: wrong answer.
21:48:59 <DrPraetor> Also, why on earth is it split arbitrarily into skills.cc and skills2.cc
21:49:00 <DrPraetor> ?
21:49:08 <due> correct answer is: it has no discernable impact on actual gameplay and difficulty, so nobody has noticed an issue to investigate.
21:49:28 <due> DOS file size limitations or something ridiculously dumb like that, I imagine.
21:49:28 <DrPraetor> I'll allow that the effect is very hard to discern.  That doesn't mean it isn't there.
21:49:34 <greensnark> due: Your scrabble skills are failing you: "discernable" <-- :(
21:49:37 <MarvinPA> i don't think you're ever going to convince anyone that 3 levels of stabbing instead of one level of conjurations/axes/whatever is gamebreaking and makes things significantly harder
21:50:05 <DrPraetor> It's certainly not *gamebreaking*, that's true.  But it is pretty easy to fix.
21:50:08 <due> greensnark: meh, I'm tired.
21:50:12 <greensnark> skills2.cc is completely rational compared to the use of variables like 'gluggy'
21:50:21 <due> ooh we have a gluggy variable? <3
21:50:23 <greensnark> I'd really love to know just how drunk the person who invented 'gluggy' was
21:50:27 <elliptic> drpraetor: no it isn't :P your suggested fixes have been awful
21:50:33 <greensnark> due: Not any more, it met Zaba, and now it's toast
21:50:36 <due> aw :(
21:50:42 <monqy> doesn't sublimation of blood code have a variable named loopy?
21:50:44 <monqy> good stuff
21:50:50 <due> greensnark: possibly "that was really gluggy!"?
21:50:53 <DrPraetor> That stabbing only trains with the weapons that get the bigger stab bonuses?  You think that's an awful suggestion?
21:50:57 <elliptic> yes
21:51:03 <greensnark> due: That might make sense, except that this was in dungeon.cc :
21:51:10 <elliptic> if weapons are getting any bonuses from stabbing skill, stabbing skill should train
21:51:15 <elliptic> this is how skill training works
21:51:23 -!- valrus has joined ##crawl-dev
21:51:24 <greensnark> "That minivault was really gluggy"
21:51:31 <DrPraetor> So hitting someone with an enhancer staff should train invocations?
21:51:34 <due> greensnark: "I slurped it down real fast though"
21:51:43 <due> DrPraetor: that is not even a valid vomparison.
21:51:47 <due> DrPraetor: erm, comparison.
21:52:07 <due> greensnark: You're right, my spelling is somewhat horrendous this afternoon.
21:52:11 <elliptic> you aren't evoking a staff when you hit something with it
21:52:21 <monqy> it respects evocations
21:52:22 <due> You are stabbing someone when you ... stab someone?
21:52:26 <greensnark> So elephants are too common in Lair or in dungeon as well
21:52:32 <greensnark> ^^ anybody
21:52:33 <DrPraetor> Sorry, evocations, oy.
21:52:46 <DrPraetor> So hitting someone with an enhancer staff should train evocations?
21:52:50 <elliptic> no
21:52:53 <elliptic> because you aren't evoking it
21:52:58 <elliptic> you are hitting someone with a staff
21:53:14 <MarvinPA> hitting someone with a pain weapon doesn't train necromancy, etc
21:53:19 <DrPraetor> "if weapons are getting any bonuses from stabbing skill, stabbing skill should train [22:51:16] <elliptic> this is how skill training works"
21:53:38 <elliptic> I meant that if you are stabbing something, stabbing should train
21:53:50 <due> If you stab someone, you train stabbing.
21:53:56 <due> Regardless of what you stab them with.
21:54:28 <due> oh look, i have things I should be doing instead
21:54:30 * due runs away
21:54:53 * DrPraetor stabs due with a club.
21:55:02 <due> "coshing" in this instance
21:55:55 <greensnark> Is this going to be another discussion about just what weapon is ideal when you want to while away the long hours by mauling an unconscious body?
21:56:08 <elliptic> ??elliptic[3]
21:56:09 <Henzell> elliptic[3/3]: < elliptic> ... hitting an unconscious person with a great  sword sounds pretty good to me
21:56:22 -!- OG17 has quit [Ping timeout: 246 seconds]
21:56:26 <elliptic> !learn edit elliptic[3] s/  / /
21:56:26 <Henzell> elliptic[3/3]: < elliptic> ... hitting an unconscious person with a great sword sounds pretty good to me
21:58:06 <DrPraetor> Even I am now bored with all of this, but let the record show I think the lot of your are clinging to something *transparently nonsensical* for increasingly arbitrary reasons.
21:59:36 <elliptic> "transparently nonsensical" = stabbing should train stabbing
21:59:45 <greensnark> Mm, the script claims that there are ~16 elephants on average in the Lair
21:59:46 -!- OG17 has joined ##crawl-dev
22:00:00 <DrPraetor> "transparently nonsensical" = stabbing someoen with a STICK
22:00:08 <greensnark> As compared to ~14 yaks on average
22:00:24 <greensnark> I'm not sure I can trust this wretched thing, let me cross-check
22:00:27 <elliptic> greensnark: sounds plausible... of course respawning means you'll see more in actual play
22:00:34 <greensnark> Yes
22:00:42 <greensnark> The script only counts initial level pop
22:00:54 <greensnark> Good point, I hadn't thought of that
22:01:02 <monqy> ime I got more yaks in early Lairs and more elephants in later Lairs
22:01:41 <monqy> D:8 or so and hardly any elephants; D:13 and hardly any yaks
22:01:51 <greensnark> Elephants are the single largest source of experience in the Lair
22:01:55 <greensnark> Let me pastebin this
22:02:17 <elliptic> personally I don't mind the elephants
22:02:17 <greensnark> http://sprunge.us/GTWU
22:02:30 <greensnark> The count doesn't seem as high as I'd imagined
22:02:50 <greensnark> The number in the main dungeon is even less, so unless respawns are really inflating their numbers
22:02:59 <elliptic> what depth is this using for lair entrance?
22:03:12 <elliptic> that does affect things a lot as monqy said
22:03:15 <greensnark> Good question, it's a random Lair entrance
22:03:19 <greensnark> Let me try a D:13 Lair
22:03:39 <elliptic> I think people mainly complain after getting a deep lair, which probably has substantially more
22:03:53 <elliptic> and then they don't complain about the early lairs
22:04:28 <elliptic> one lair had 42 elephants, I see
22:04:36 <greensnark> Yeah, lucky :P
22:04:59 <greensnark> sigma is 8 though
22:08:00 <Eronarn> greensnark: !rc is pulling rcs from lorcs, just fyi
22:09:20 <greensnark> Eronarn: Yeah, it uses the most recently modified rc
22:13:14 <due> CountMin 3 sounds like the smallest possible band size.
22:13:38 <due> greensnark: Can you get it to ignore plants?
22:13:38 <greensnark> Yes, and note that CountMin is the smallest non-zero count
22:13:52 <greensnark> Yes
22:13:58 <greensnark> But does that help?
22:14:13 <greensnark> Oh, I guess if you want an accurate %
22:14:18 <due> Well, plants and bushes do skew the percentages.
22:14:34 <greensnark> Hm, this is weir
22:14:47 <greensnark> Either the script is lying to me or deeper Lairs have fewer elephants than shallow Lairs :P
22:14:56 <greensnark> Not much difference either way
22:15:07 <greensnark> Perhaps the depth-setting option no workee
22:15:11 <elliptic> that could be, actually... I bet they have more death yaks though :P
22:15:37 <elliptic> (if they don't have more death yaks, you should worry about the depth-setting option)
22:16:06 <elliptic> part of the issue with elephants is that there are also (death) yak packs
22:16:16 <elliptic> and elephants are the most annoying of the 3 of them
22:16:27 <greensnark> 4.65 on average at D:13 vs 4.53 on average at D:8, so let me check that the depth setting works
22:17:29 <due> I have half a mind to disable elephants bands.
22:17:33 <due> But I'm unsure that would help.
22:17:54 <elliptic> I like elephant bands :/
22:18:10 <monqy> I like them too
22:18:10 <elliptic> can be quite interesting tactically
22:18:15 <st_> the bad is the point though, they push you around so it's hard to fight them 1v1
22:18:17 <doy> yeah, elephant bands are good
22:18:20 <st_> band*
22:18:25 <DrPraetor> It's actually a much bigger deal for weapon switchers.  Having a second weapon skill at 8 (assuming 0 aptitude), raises the total XP required, for *every single other skill you have*, by about 10%, or a level or two.
22:18:57 <doy> DrPraetor: and?
22:19:00 <DrPraetor> Elephants should definitely come in clumps, just not six clumps in every lair.
22:19:14 <elliptic> drpraetor: that's not how skills work
22:19:24 <greensnark> Depth setting totally works right
22:19:30 <monqy> curious
22:19:40 <DrPraetor> elliptic: Really?  Because I have the code open in front of me, and parts of it are hard to understand but this isn't.
22:19:53 <monqy> is there just a lot of variance in elephant generation?
22:19:57 <due> I wish I had a mute button.
22:20:05 <greensnark> due: You do :P
22:20:05 <due> monqy: There's a lot of variance in all of the generation.
22:20:13 <due> greensnark: Well, yeah, but.
22:20:21 <doy> see also: mon-pick is so terrible
22:21:30 <due> doy: yup
22:21:34 <elliptic> greensnark: hm, that's strange about the death yaks then... just look at the data you listed, level-by-level... substantially more death yaks on lair:8 than any previous level
22:21:49 <elliptic> this should be more noticeable than the numbers you mentioned
22:23:12 <greensnark> True, but how well-established is it that deeper Lairs have nastier monster distributions? :)
22:23:55 <elliptic> I thought they did, but maybe not, yeah
22:24:29 <greensnark> Deeper Lairs could certainly produce nastier uniques
22:25:21 <st_> so that's means D:19 vaults doesn't mean V:8 is nastier?
22:25:44 <st_> I really seems like it does
22:25:47 <elliptic> d:19 vaults is definitely nastier
22:25:50 <greensnark> Vaults is a special case in that it uses the main dungeon generator
22:25:52 <elliptic> yeah
22:26:06 <greensnark> So V:8 at D:19 is effectively D:27 in depth
22:26:18 <greensnark> And on Vault:8 you get a generous helping of '8' monsters
22:26:37 <greensnark> *main dungeon monster picker
22:26:54 <st_> I love D:19 vaults :P
22:27:02 <greensnark> mons_standard_level is used for both D and Vault in mon-pick.cc
22:27:21 <greensnark> The other branches all appear to get depth-adjusted monster levels so in theory the depth of the Lair entrance doesn't matter
22:27:40 <greensnark> D:8 Lair: http://sprunge.us/HUPN, D:13 Lair: http://sprunge.us/KEOX
22:27:51 <greensnark> I should include absdepth in the repots
22:29:38 <elliptic> okay, that simplifies things :)
22:31:34 <greensnark> I wouldn't mind changing Lair monster gen to be more dangerous with a D:13 entrance since everyone expects it :P
22:31:35 <elliptic> I'd suggest just decreasing elephant and yak generation rates by a small amount... 20% or something
22:31:55 <greensnark> Let me make a quick tweak and see how it would really affect numbers
22:32:18 <elliptic> could also look at elephant generation in D
22:36:53 <greensnark> Lair numbers if Lair difficulty really increased with entry depth for a D:13 Lair: http://sprunge.us/cOfP
22:37:27 <elliptic> heh
22:37:47 <Jordan7hm> holy god 152 elephants
22:37:48 <greensnark> Lots of death yaks, but also super-bonus elephants :P
22:38:30 <DrPraetor> There's a bit of a counterbalancing effect, b/c once skill points cost more in all your other skills, you have less skill points total from other skills.  But I don't object to having a big penalty for trying to have two weapon skills.  It's not an ideal design decision, but it is what it is.  I *do* object to having a big, super-spoily, poorly documented penalty.  Is the fact that characters have
22:38:45 <DrPraetor> a general skill level that raises the cost of all skills mentioned *anywhere* in the docs?
22:40:14 <greensnark> due: Shall I make elephants closer to between-yaks-and-death-yaks rarity?
22:40:52 <greensnark> So around 9 elephants per Lair as opposed to ~14 currently
22:41:14 <greensnark> I'm not sure where we've positioned elephants
22:41:48 <elliptic> I'd rather keep about the same number of elephants and yaks in lair, personally
22:42:28 <greensnark> Hm, they're already pretty close in numbers: is that ok?
22:43:08 <DrPraetor> !dump
22:43:10 <Henzell> http://crawl.akrasiac.org/rawdata/DrPraetor/DrPraetor.txt
22:43:18 <DrPraetor> !!dump
22:43:26 <elliptic> yeah, that's why I was suggesting reducing their generation equally earlier... mainly I just find regular yaks really boring :P
22:43:30 <DrPraetor> ??gretell
22:43:32 <Henzell> gretell[1/3]: The counterpart to Henzell, for crawl.develz.org.  Only accepts @whereis and @? from IRC. The evil twin of Henzell. Now accepts inquiries on PM. Currently @? points at trunk due to code rot. Sorry!
22:44:49 <DrPraetor> ??dump
22:44:51 <Henzell> dump[1/3]: Press # in game, then type !dump in IRC. Macro S to #S to auto-dump when saving.
22:44:58 <DrPraetor> ??dump[2]
22:44:59 <Henzell> dump[2/3]: To see a midgame dump that you made on the CDO server, visit http://crawl.develz.org/morgues/<game>/<charname>/<charname>.txt. "Game" is one of "ancient," "hex," "stable," or "trunk." For CAO see !dump.
22:45:18 <elliptic> drpraetor: you realize you can read all this in pm, right?
22:45:49 <DrPraetor> Oh, oops.  Moved windows by mistake.
22:46:47 <DrPraetor> Anyway, 28 elephants.  12-16 elephants is okay, but 0-30, which seems to be how it works now, is a bit much.
22:47:23 <DrPraetor> And yes, I know that the current code doesn't allow 12-16 as an option.
22:47:26 <doy> i'm okay with a high variance
22:47:38 <doy> getting 12-16 elephants every single game would be boring
22:47:43 <DrPraetor> I'm okay with high variance as long as it doesn't touch 28.
22:47:47 <doy> yeah
22:47:55 <elliptic> 28 elephants doesn't sound too bad to me... that's like 5 or 6 packs
22:48:10 <doy> i guess it would depend on how rare that is
22:48:11 <DrPraetor> It took forever to hack through them all, was very boring.
22:48:41 <DrPraetor> This is the second time I've encountered >20 elephants in lair.  I've been reproducably seeing more elephants than blink frogs, which are an old favorite of mine and make lair much more... exciting.
22:48:48 <doy> "boring" is not really a very helpful term when discussing game mechanics, pretty much ever
22:49:12 <DrPraetor> Low threat level, long time commitment.
22:49:19 <elliptic> elephants are substantially more interesting than yaks
22:49:29 <DrPraetor> @??elephant
22:49:29 <Gretell> elephant (03Y) | Speed: 10 | HD: 9 | Health: 52-83 | AC/EV: 8/2 | Damage: 20, 5 | Res: 06magic(60), 12drown | XP: 430.
22:49:32 <DrPraetor> @??yak
22:49:32 <Gretell> yak (07Y) | Speed: 10 | HD: 7 | Health: 24-52 | AC/EV: 4/7 | Damage: 18 | Res: 06magic(28) | XP: 204.
22:49:33 <due> greensnark: I am happy with this.
22:50:00 <due> greensnark: Also, re: blink frogs. I seem to see one or two packs every few games on Lair:1.
22:50:04 <DrPraetor> They're marginally nastier than yaks, I suppose.
22:50:38 <DrPraetor> But mainly they just take forever to kill, which is, pardon me if this isn't helpful, boring.
22:51:17 <DrPraetor> I propose - make them show up in considerably smaller numbers, and make them speed: 12.  *That* would be a source of some excitement.
22:51:29 <due> ...
22:51:49 * due summons Eronarn for the daily faster-isn't-better lecture.
22:52:26 <DrPraetor> Faster isn't *always* better, but when you have a monster with the ability to trample you off of stairs, and otherwise push you back to surround you?
22:52:28 <greensnark> I'll have to return to the elephant thing tomorrow, bedtime :)
22:52:52 <DrPraetor> Btw, by smaller numbers I mean fewer pods, not that the individual pods should be smaller.
22:53:12 <elliptic> drpraetor: if you want to see fast tramplers, go to the vestibule
22:53:28 <due> Technically, "band", but "herds" also suits.
22:53:30 <elliptic> elephants are already quite dangerous to a lot of characters, they don't need to be speed 12
22:53:37 <doy> yeah, i think beasts getting trample is pretty wonderful
22:53:38 <due> @??elephant
22:53:39 <Gretell> elephant (03Y) | Speed: 10 | HD: 9 | Health: 52-83 | AC/EV: 8/2 | Damage: 20, 5 | Res: 06magic(60), 12drown | XP: 430.
22:53:49 <doy> geryon is actually something of a threat these days
22:53:50 <doy> (:
22:53:57 <due> elephants are actually quite scary.
22:54:01 <greensnark> Well, his beasts are anyway :P
22:54:03 <DrPraetor> Again, I have fond feelings for being savagely killed by a horde of yaktaurs, so if something else kills me in the vestibule, I will be offended.
22:54:06 <due> it's funny because they were never, ever, ever supposed to generate outside of vaults.
22:54:19 <due> but that's the way thte cookie crumbles!
22:54:23 <doy> due: "they"?
22:54:34 <CIA-45> 03greensnark * r7e89148de275 10/crawl-ref/source/scripts/place-population.lua: Show absolute depth when generating levels in place-population, filter plants, fungi and bushes by default.
22:54:49 <due> doy: elephants
22:54:54 <doy> really
22:55:13 <due> doy: yes, i created them for a one-off vault as dire elephants, then purge made some super-cool tiles so i added hellephant and plain elephants in
22:55:19 <due> doy: then someone added plain elephants to the lair starting vault.
22:55:29 <due> then someone actually made trample do stuff.
22:55:37 <doy> (me!)
22:55:38 <doy> (:
22:55:39 <due> and then somewhere along the lines elephants got added to the standard monsterset.
22:55:49 <doy> i think the first i saw of them was the elephant lair end vault
22:55:52 <due> I think I was the one who added elephants in
22:55:54 <due> yes
22:56:08 <doy> i was the one who added elephants to the generation lists
22:56:12 <due> ah
22:56:14 <doy> to test the trample stuff more
22:56:16 * due nod.
22:56:30 <greensnark> Aha, the skeletons are really emerging from the closets today :P
22:56:33 <due> but yeah, the initial intent was dire elephants, only in vaults.
22:56:37 <DrPraetor> Well, that's always the problem with overbiasing towards any given monster.  It's either monotonous to hack through them all if they're no threat, or it makes the game impossible if you're build is weak to that monster type.
22:56:41 <doy> i figured people would tweak it if it was way out of whack, since i didn't feel like trying too hard to understand mon-pick
22:56:45 <doy> (:
22:56:46 <due> i do not regret adding them in though
22:57:00 <doy> yeah, i think elephants are pretty great overall
22:57:33 <due> i mean elephants <444
22:57:53 <due> I have the massivest elephant obsession out, anyway. My room is full of abstract elephant figurines and art.
22:58:14 <greensnark> I was restraining myself from pointing out due's elephant obsession
22:58:22 <DrPraetor> Oh, lest my kvetching be misinterpreted: yes, the elephants are cool.  In more modest doses.
22:58:37 <due> greensnark: :)
22:58:50 <due> greensnark: I heard your timezone had changed again!
22:58:59 <greensnark> due: Yeah, this new timezone sucks :P
22:59:07 <due> Also, I'm not going to be coming to India this year as it stands. Finances, etc.
22:59:09 <greensnark> All the interesting people go to bed at the wrong time now
22:59:25 <due> (Are you still in India? I forget.)
22:59:38 <due> (Also, possibly invasive personal question, sorry!)
22:59:41 <greensnark> Maryland
23:00:05 <greensnark> Old slave country, or so I'm told :P
23:00:12 <due> Oh.
23:00:23 <DrPraetor> The hive should have bee mages, who come with silence fields, are silence proof, and cast lots of hexes (paralyze, etc.)
23:00:26 <due> Maryland -> maryland chicken.
23:00:44 <greensnark> Ok, off for real, early day tomorrow
23:00:45 <due> DrPraetor: That's very Raymond E. Feist, actually.
23:00:51 <due> Ciao snark which is green!
23:07:52 <Ashenzari> Lemuel's Castle variety pack (https://crawl.develz.org/mantis/view.php?id=3794) by carwin
23:08:27 <due> oooh tha sounds interesting
23:09:22 <due> Castle is such an excellent vault and concept
23:10:16 <evilmike> probably the scariest vault in the game if it appears in late D, too!
23:10:26 <due> my favourite early vault
23:10:34 <due> we need more early encompass/large vaults.
23:10:39 * due considers this.
23:10:46 <elliptic> I never ever touch that vault if it is elves or dracs
23:10:53 <evilmike> I actually think clearing lemuel_castle in late D is harder than clearing zot:5 in a lot of cases
23:11:01 <evilmike> it certainly was last time I tried it
23:11:17 <due> I totally bypass Elf castles.
23:12:27 <Gretell> chexuma (L20 SETm) ERROR: range check error (52 / 52) (Shoals:5)
23:17:53 -!- st_ has quit []
23:20:18 -!- Kurper has quit [Remote host closed the connection]
23:28:19 -!- elliptic has quit [Ping timeout: 248 seconds]
23:38:02 <Gretell> Windows development builds on CDO updated to: 0.9.0-a0-71-g7e89148
23:38:50 -!- st_ has joined ##crawl-dev
23:43:54 -!- MadCoyote has joined ##crawl-dev
23:51:04 <Gretell> Unstable branch on CDO updated to: 0.9.0-a0-71-g7e89148 (32)