00:00:26 Unstable branch on crawl.develz.org updated to: 0.9-a1-928-gea7ead2 (32) 00:00:30 Unstable branch on crawl.akrasiac.org updated to: 0.9-a1-928-gea7ead2 (32) 00:11:53 -!- valrus has joined ##crawl-dev 00:12:25 Windows builds of master branch on crawl.develz.org updated to: 0.9-a1-928-gea7ead2 00:14:45 -!- valrus has quit [Remote host closed the connection] 00:24:27 -!- Zaba has quit [Ping timeout: 246 seconds] 00:35:28 -!- Zaba has joined ##crawl-dev 01:02:11 -!- ainsophyao has quit [Remote host closed the connection] 01:02:51 -!- ainsophyao has joined ##crawl-dev 01:05:26 -!- Twilight-1 has quit [Read error: Connection reset by peer] 01:05:49 -!- Twilight-1 has joined ##crawl-dev 01:08:24 -!- ainsophyao has quit [Remote host closed the connection] 01:20:52 -!- ixtli has quit [Ping timeout: 252 seconds] 01:28:09 -!- ixtli has joined ##crawl-dev 01:53:11 -!- monqy has quit [Quit: hello] 02:01:56 -!- ixtli has quit [Ping timeout: 252 seconds] 02:35:46 -!- galehar has joined ##crawl-dev 02:36:09 hi 02:36:25 galehar: You have 1 message. Use !messages to read it. 02:36:41 !messages 02:36:41 (1/1) due said (4h 7m 42s ago): Gandelf (L4 DDBe) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed. (D:3) 02:36:52 There's been a couple of those over the past few hours. 02:37:08 ok 02:58:20 -!- ainsophyao has joined ##crawl-dev 03:06:55 do we have any more info on the circumstances of those crashes? 03:08:38 not that i know of, sorry 03:33:26 -!- syllogism has joined ##crawl-dev 03:49:37 !lm Gandelf type=crash -log 03:49:37 2. Gandelf, XL4 DDBe, T:1995 (milestone): http://crawl.develz.org/morgues/trunk/Gandelf/crash-Gandelf-20110706-032413.txt 03:51:13 ooh, this time there are cores on CDO. Should I dig in them? 03:53:24 Mmm delicious cores. 03:54:43 first one I found: Garhauk, melee_attack::player_attack ; practise ; reset_training 03:54:59 same assertion 03:56:38 -!- Pacra has joined ##crawl-dev 03:56:40 gdb mumbles something about executable possibly not matching core, but I can query variables and backtrace just fine 04:06:36 kilobyte: I'd be interested in the state of the training array at the time of the crash 04:06:44 and also the skills array 04:06:51 -!- jeanjacques has quit [Read error: Connection reset by peer] 04:08:56 hrm, not sure how to force gdb to show array of chars as not a string 04:09:02 you.skills: {mData = '\000' , "\002\002\000\000\000\000\001\000\000\000\000\005\000\000\000\000\000\000\000\000"} 04:09:34 you.training: "\016\000\000\000\035\000\000\000\000\000\000\034\000\000\003\000\004\000\000\000\000\027\000\000\000\000\000\000\000\000\000\000" 04:12:33 thanks 04:12:54 I think it happens when there is a lot of training for unknown skills 04:20:33 -!- edlothiol has joined ##crawl-dev 04:27:53 I think I'll cap the sum of all unknown skills training at 50% 04:29:00 antiskillrobin (L4 SpVM) (D:3) 04:31:13 -!- dpeg has joined ##crawl-dev 04:52:15 03dpeg * r0b190e40a97e 10/crawl-ref/source/shopping.cc: Price reform: wands. (nrook) 05:12:35 Napkin: i'm confused, what are the rules when P on CDO takes you to trunk and when it takes you to stable? :) 05:13:24 -!- Pacra has quit [Ping timeout: 255 seconds] 05:14:28 bhaak: a couple days after the start of the 0.8 tournament the default was switched to 0.8 and hasn't been updated since 05:15:37 probably no need to bother for the next 7 days, when we'll ask Napkin to set it to the 0.9 beta (for testing) and leave it at least for the duration of the official tournament 05:16:28 kilobyte: i remember the switch to 0.8, but i thougt i also remember the switch back to trunk. 05:16:37 -!- golgepapaz has joined ##crawl-dev 05:17:09 !messages 05:17:09 No messages for golgepapaz. 05:19:05 golgepapaz: just pushing your patch! 05:21:18 -!- Hehfiel has quit [Ping timeout: 276 seconds] 05:21:43 golgepapaz: it doesn't work here?! 05:22:16 golgepapaz: created a deck in wizmode, read ?id and nothing happened. 05:22:34 ah, worked with the second scroll 05:24:16 03kilobyte * rfdf5a12385d3 10/crawl-ref/source/mpr.h: Let GCC issue warnings about mprf() arguments. 05:24:26 03kilobyte * r9348c12b2254 10/crawl-ref/source/ (19 files): Fix format string warnings. 05:25:08 -!- Hehfiel has joined ##crawl-dev 05:30:12 you say it's not working? 05:30:33 maybe I am dumb 05:30:37 it does work on the second try 05:31:04 can you try again? 05:31:27 yes, I did 05:31:33 a new deck and all 05:31:45 oh ok 05:32:00 I did &0} to get some deck, used &I to forget its type. Reading ?id will identify the type of deck, but not the top card. 05:32:16 Reading another ?id will give away the top card. This could be an artefact of wizmode, though. 05:32:51 oh first id identifies the deck if it's not id ed. you can peek the card after knowing the deck first 05:32:55 Ah nice, the topmost card id.. I might start using found decks yet! :) 05:33:25 golgepapaz: May I suggest two changes: id the top card right away. And list known decks with unknown top cards as "unidentified". 05:33:39 Keskitalo: hi! I've added a billion vaults in the last week :) 05:33:44 lunch! 05:33:46 I noticed, good work! 05:34:26 dpeg: I noticed :D So many awesome valuts now! 05:36:48 -!- RichardHawk has quit [Ping timeout: 252 seconds] 05:38:08 Duplicated entry vault des files (https://crawl.develz.org/mantis/view.php?id=4223) by mumra 05:38:30 bhaak: as soon as someone complains, since i got confused and stopped using it :D 05:38:30 ah, so quite predictable 05:38:43 apart from that, it's supposed to start trunk, unless there's a tournament 05:39:22 that's what I thought, too 05:39:28 fixed :) 05:39:51 but you'll need to re-login, otherwise the config isn't re-read 05:40:35 do you mean showing it in identify inventory prompt? 05:40:52 no need ATM, i found trunk by myself :-) 05:44:42 -!- galehar_ has joined ##crawl-dev 05:45:03 -!- ainsophyao has quit [Remote host closed the connection] 05:45:33 -!- RichardHawk has joined ##crawl-dev 05:54:17 -!- Ashenzari has quit [Ping timeout: 276 seconds] 05:55:05 -!- Ashenzari has joined ##crawl-dev 05:55:05 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 06:01:32 -!- Pacra has joined ##crawl-dev 06:02:41 -!- petete has quit [Read error: Operation timed out] 06:04:52 -!- petete has joined ##crawl-dev 06:04:54 -!- RichardHawk has quit [Ping timeout: 246 seconds] 06:14:07 back 06:14:16 golgepapaz: still here? 06:20:25 yes 06:20:25 yes, I mean the inventory prompt 06:20:25 golgepapaz: ?id not giving the top card on first use (but only the deck type), is this intentional? 06:20:25 yes 06:20:25 but I'll get on working on your suggestions 06:20:25 dpeg: do we want players to use-id decks? 06:20:25 or do we want non-Nemelex worhsippers to use decks at all? 06:20:25 -!- Ashenzari has quit [Ping timeout: 246 seconds] 06:21:10 -!- Ashenzari has joined ##crawl-dev 06:21:10 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 06:23:03 dpeg: great idea for using ? for both skill descriptions and general help. 06:23:44 alright it's trivial enough. showing the unmarked card would require some work. I need to hunt down where prompt_invent_item uses OSEL_UNIDENT 06:24:38 dpeg: what's the drawback of no skill training for unknown skills? 06:36:43 galehar_: in automatic: you use them, so there is some interest in having them level, even if slowly. For example: I turn of all skills but Spc but I am still happy that Stealth, Con etc. slowly rise. 06:36:43 Here is how I see it in the new system: if you want to focus on a skill, you'll use manual mode. If you just want to train the skill you use, you activate auto. If you disable a skill in the latter, you really don't want it. 06:37:33 galehar_: I understand. Perhaps I am just too entrenched in the old mode, where it is "focus on selected skills, train deselected skills but more slowly" 06:38:18 but don't worry -- I have hope that the usage system will make everything work 06:38:26 yes. experience will tell us if this need adjustment. 06:38:43 no need to waste many words when we'll see it in practice soon enough 06:38:55 golgepapaz: I already pushed your patch. Should I retract locally? 06:39:01 s/push/commit/ 06:39:15 -!- [1]GrimmSweeper has joined ##crawl-dev 06:39:16 -!- GrimmSweeper has quit [Disconnected by services] 06:39:16 -!- [1]GrimmSweeper is now known as GrimmSweeper 06:39:40 no, I am sending a new patch implementing your suggestions in ten minutes or so 06:40:15 awesome 06:40:51 dpeg: I've been mising "train this slower" on my KoBe. Turning Short Blades on and off all the time is fiddly. 06:41:07 ghallberg: yes, we will see... 06:41:31 If worst comes to worst, we need three levels for training rate: full, slow, none. 06:41:38 Yep, it's no biggie, just wanted to chime in with my thoughts. I like it alot in other ways :) 06:41:51 There will be a solution, hopefully a better one :) 06:42:18 dpeg: or the opposite: fast, normal, none (and only one skill can be fast). 06:42:27 -!- ais523 has joined ##crawl-dev 06:48:24 denial of service from jeanjacques? 06:48:44 is there a zombie process challenge nobody told me of? ;) 06:49:02 Kill the napking's processes, kill them! :) 06:49:36 jeanjacques (L3 MuFi) (D:2) 06:49:37 jeanjacques (L4 MuAs) (D:4) 06:52:15 -!- casmith789 has joined ##crawl-dev 06:53:45 -!- edlothiol has quit [Ping timeout: 255 seconds] 06:56:16 03galehar * r8ea535ee7bab 10/crawl-ref/source/wiz-you.cc: Reduce the number of exercises done by the &g wizmode command. 06:56:19 03galehar * r669482f48a58 10/crawl-ref/source/skills.cc: Cap the total of training for unknown skills at 50%. 06:56:19 03galehar * r7186bbc06efa 10/crawl-ref/source/skill_menu.cc: Wizmode: allow exiting the skill menu with no skill selected. 06:57:46 galehar_: sending the next m help screen version. If you think the content is okay, you get a full colour version. 06:57:59 cool :) 06:58:10 we'll need to update the manual too. 06:58:28 and tutorial and hint modes. 06:58:28 Unstable branch on crawl.develz.org updated to: 0.9-a1-931-g9348c12 (32) 06:59:11 damn, I wanted to push the fix on CDO, but I started the compile before the push made it to gitorious 06:59:20 and now I have to go back to work 06:59:27 -!- galehar_ has quit [] 07:03:30 Unstable branch on crawl.develz.org updated to: 0.9-a1-934-g7186bbc (32) 07:03:35 removing unused trunks is cheap 07:07:23 -!- edlothiol has joined ##crawl-dev 07:14:03 damn. 07:14:16 I fucked it up. 07:14:37 I tried to fix this bug in a hurry during my lunch time, and I didn't test it properly. 07:15:21 I'll pastebin a patch in a few minutes (since I can't push from here) 07:17:13 away for a while 07:22:04 -!- ainsophyao has joined ##crawl-dev 07:27:36 http://pastebin.com/xSwGvcjg 07:34:18 !tell dpeg I've added a second patch file to the mantis concerning deck identification 07:34:19 golgepapaz: OK, I'll let dpeg know. 07:38:14 galehar: Air 17 at xl 9 seems damn broken 07:38:15 03kilobyte * r21b92b4539e1 10/crawl-ref/source/ (invent.cc spl-book.cc): Fix variable strings used for format. 07:38:18 03kilobyte * r06661719f19d 10/crawl-ref/source/ (player.cc player.h tags.cc): Remove an unused field. 07:38:19 03galehar * rcde2f9f57075 10/crawl-ref/source/skills.cc: Fix a bug in skill scaling. 07:39:18 kilobyte: what? You were able to train air to 17 at XL 9 with the new system? 07:41:01 kenku of course 07:41:54 -!- Pacra has quit [Ping timeout: 255 seconds] 07:42:22 other skills at 9, 6, 6 and lower too, so not an unviable char too 07:43:06 in 0.9-a1-928-gea7ead2 so you might have fixed that already, though 07:44:23 -!- Pacra has joined ##crawl-dev 07:49:28 tried in wizmode: a KeAE dumping 100% into air can real skill level 14 by XL9 07:49:35 *reach 07:50:43 I'm not sure how it could have happen. The new system only changes how and when skills are trained, not by how much. 07:50:55 pretty sure they could do that previously 07:51:07 did you try wizmoding it on the old skill system as well? 08:01:53 well, I just did and got the exact same result: L14 08:15:28 back for a moment 08:15:28 dpeg: You have 6 messages. Use !messages to read them. 08:15:36 golgepapaz: send to me, please 08:24:11 galehar: that would be a good result, if I'm not mistaken 08:25:22 Pacra: yeah, it means the new system doesn't mess with the training speed. Except with kilobyte's game. Don't know what happened there. 08:33:09 -!- ainsophyao has quit [Remote host closed the connection] 09:05:33 There’s an extra empty line in the status area in ZD. I’d guess it was added for the Zot points field and didn’t get removed along with the experience points field. 09:07:22 the XP pool was on the same line as XL, so no line was actually removed. 09:08:16 03dolorous * ra7a07758c97d 10/crawl-ref/docs/template/apt-tmpl.txt: Fix wording. 09:08:26 03dolorous * rbfc5f3c28d62 10/crawl-ref/docs/template/apt-tmpl.txt: Attempt to document how the -1/+1 values have different meanings for HP/MP. 09:08:27 03dolorous * r03ef2d27215a 10/crawl-ref/docs/template/apt-tmpl.txt: Fix inconsistent wrapping in the aptitude template. 09:09:10 Ah, i don’t remember the previous layout exactly. Just seeing an empty line in there now. :-) 09:15:25 -!- ixtli has joined ##crawl-dev 09:18:05 seems I modified it by mistake. Thanks for noticing. 09:19:53 galehar: Is there place for "Remaning Xp until level" in the status area now? 09:20:01 I fins myself missing that sometime in the early game. 09:20:03 find* 09:20:50 well, we can put it in place of the old XP pool. I'm not sure how useful it is though. 09:21:01 the place my be better used for something else 09:22:08 ghallberg: how does knowing remaining is useful? Does it affect any decision? Or are you just impatient to level up? 09:22:16 03kilobyte * rfd747ee93842 10/crawl-ref/source/dat/des/variable/altar.des: More silly names for Ashenzari's book (jeffqyzt) 09:22:23 03kilobyte * r2a4dde5e7b89 10/crawl-ref/source/main.cc: ZotDef: toadstools are not a threat to the Orb. 09:22:23 03kilobyte * r164948bb98a8 10/crawl-ref/source/output.cc: ZotDef: fix misaligned display of Zot Points. 09:23:33 <3 the new Ash book names. 09:24:10 So, now we can tell players that we do have cooking (even if it is a bit creepy) 09:25:43 galehar: that comment scares me 09:29:17 galehar: I may just be impatient I guess, I'll think about it. 10:04:34 !when 10:18:16 03dolorous * r00c388b5f81e 10/crawl-ref/source/dat/des/variable/altar.des: Comment fixes. 10:18:48 -!- MarvinPA has joined ##crawl-dev 10:30:44 what's the simplest way to print some scrollable formatted static text (help)? 10:30:50 formatted_scroller? 10:39:02 -!- ais523 has quit [Read error: Connection reset by peer] 10:40:27 -!- ais523 has joined ##crawl-dev 10:51:05 -!- ainsophyao has joined ##crawl-dev 10:58:38 -!- monqy has joined ##crawl-dev 11:08:51 -!- galehar has quit [Quit: Page closed] 11:10:48 -!- MarvinPA has quit [Ping timeout: 252 seconds] 11:10:53 -!- evilmike has joined ##crawl-dev 11:11:06 -!- MarvinPA_ has joined ##crawl-dev 11:13:12 In Kal’s ZD game on CDO, monsters seem to just spawn in random places around the level instead of coming down the stairs. Yesterday (after the exp pool removal) the game crashed when trying to play, today it works except for that issue. 11:13:41 but that's how monsters normally spawn 11:13:56 In ZD? 11:14:03 oh, wait, I missed that part 11:14:06 no idea about zotdef, sorry 11:14:36 iirc they spawn at random if the stair areas are too crowded 11:16:58 He made a backup of the save in case it’s of interest. 11:18:37 I'm wondering, is there still a plan manual skill training mode to let you choose between fast/normal/off for training skills? 11:19:14 -!- golgepapaz has quit [Read error: Connection reset by peer] 11:19:24 well yeah, the way they're placed in zotdef, it falls back to 'away from monster' if it finds to place it near the stairs 11:19:47 -!- golgepapaz has joined ##crawl-dev 11:20:30 evilmike, something like that seems to be planned, but you really should ask galehar 11:21:10 I even thought it already existed 11:21:19 * Zaba rebuilds 11:25:26 about ZotDef: the game really lasts too long. 15 hours might fit a regular game, but ZotDef is too monotonous for that. 11:25:52 back in 11:25:53 what about doubling all XP gains and cutting the wave clock by half? 11:26:01 kilobyte: +1 11:27:38 another thing is, either maps should be balanced or there should be a way to pick one. Most people immediately quit on the map with a diamond orb chamber, as you'd need to control two areas not one. 11:29:04 classical Zot has a spawn point that allows making a death row but monsters are hard to contain -- leftovers keep spilling but you can (with difficulty) manage that 11:30:55 the one with pear-shaped chamber has a fight area in a columned hallway -- easier to block but you can't have all your "turrets" shoot newcomers right as they spawn 11:31:29 but I'd say these two are roughly equivalent 11:34:39 -!- bmh has joined ##crawl-dev 11:34:45 hi 11:34:45 bmh: You have 3 messages. Use !messages to read them. 11:34:58 !messages 11:34:59 (1/3) dpeg said (17h 17m 7s ago): You sure you wanted one jelly and two sheep? 11:35:09 !messages 11:35:11 (1/2) kilobyte said (16h 38m 54s ago): uhm, is Abyss supposed to morph just next to you? Walking and getting lava or walls spawn just in your way is tedious, and forces you to watch your step... while old Abyss allowed quite rapid movement. 11:35:56 -!- RichardHawk has joined ##crawl-dev 11:36:04 kilobyte: we should really make an effort to keep ZD alive but disabling for 0.9 is an option. 11:36:36 !messages 11:36:36 (1/1) kilobyte said (16h 11m 35s ago): features turn into another every a few turns (like, a pool of lava turns into stone, then the same pool becomes rock, etc). Seems like a nasty regression, and thus I'm not merging yet. 11:36:58 ZD should be more sprint-like in pace 11:37:29 bmh: I added your vault with one sheep and two jellies. I think. Tell me if I should revert. 11:39:30 oops, messaged you by accident 11:39:58 should be 2 sheep / one jelly, after the Monty Hall problem 11:40:28 * dpeg googles Monty Hall. I played Monty on the Run as a child, in case any grizzled veterans remember that one. 11:40:37 bmh: I think I didn't get the joke then :) What's the story behind it? 11:40:43 I should script it so that when you pick a door, Xom opens a door containing a sheep 11:41:28 Monty Hall was a gameshow host. he had a game where there were three doors, two with goats, one with a car 11:41:44 after you picked a door, he opened another door to reveal a sheep 11:41:56 you then have the option to switch doors 11:42:08 the monty hall problem is great, because it angers people who don't know math 11:42:39 I'd rather have a goat than a car. 11:42:42 Oh, I treated it in my class. It is not hard to understand, not sure why the Americans make such a fuzz about it :) 11:42:49 bmh: <3 11:43:40 for starters, London traffic is horrendous, and goat cheese is delicious 11:43:59 on the internet its sort of a lesser version of the "0.999... = 1" "debate" 11:44:08 -!- galehar has joined ##crawl-dev 11:44:22 do we have a drop dead date on the release? I need to rest my wrists for a few days 11:44:34 (ie. no computer use I'm not being paid for) 11:45:01 bmh: we freeze the code the 13th 11:45:22 dpeg: you got my explanation for the decks? 11:45:22 one week to go! 11:45:22 galehar: ok 11:45:33 golgepapaz: yes, I hope it is okay :) Just pushing the commits. 11:45:53 alright 11:46:08 I tried in wizmode without &o}-ing the decks by going Nemelex. At one point I still got "Oops, counted 1 marked cards, but num_marked is 0" 11:46:17 03hsolter * r66ac17732a9d 10/crawl-ref/source/ (decks.cc decks.h spl-goditem.cc): Identify scrolls on decks reveal top card 11:46:21 03dpeg * r6993a28e3334 10/crawl-ref/source/dat/descript/items.txt: Update description for scrolls of identify. 11:46:21 03hsolter * r19645669d108 10/crawl-ref/source/ (invent.cc spl-goditem.cc): Mark the top card in any case when identified 11:48:47 yes I've tried to tell you in the personal messaging that it only happens in WIZ mode 11:49:04 !message kilobyte: I think I understand the root cause of the regression. will fix. 11:50:04 I think I did that wrong 11:50:24 -!- galehar_ has joined ##crawl-dev 11:50:25 -!- galehar has quit [Ping timeout: 260 seconds] 11:50:27 and setting num_card to other than 0 will prevent you using Nemelex's draw four and stack five abilities. 11:50:37 -!- galehar_ has quit [Client Quit] 11:50:53 -!- bmh has left ##crawl-dev 11:51:54 golgepapaz: ok 12:00:42 -!- ais523 has quit [Remote host closed the connection] 12:07:55 -!- ainsophyao has quit [Remote host closed the connection] 12:08:16 Monsters seem to spawn in random locations even in new ZD games. 12:09:38 -!- casmith789 has quit [Quit: Leaving] 12:09:41 hm, yes 12:10:05 or no. 12:10:22 hard to tell whether it tries to put them near stairs but not hard enough 12:11:25 And there’s no loot in the walled areas. :-D 12:11:46 if there's no valid place within distance 3, PROX_NEAR_STAIRS turns into PROX_AWAY_FROM_PLAYER 12:12:12 There’s nothing near the stairs. 12:12:25 kilobyte, I put in some dprf's and the if (proximity == PROX_NEAR_STAIRS) doesn't even seem to be reached 12:12:27 somehow. 12:12:30 The only things in the whole dungeon are me, the stairs and the orb. 12:12:45 Ah, there actually seem to be eight items of loot. 12:13:01 At least in this game (ion @ CDO). 12:13:10 kilobyte, indeed, they don't spawn at stairs even on an otherwise empty level 12:14:34 maybe it's because of mon-place.cc:779. 12:14:56 oh wait 12:15:15 I've just said I don't get dprfs that I put into that code 12:16:24 maybe it's because zotdef sets the monster type, so _is_random_monster returns false. 12:17:16 ok, nevermind me, that code's logic is too involved to be figured out from a quick glance 12:23:01 -!- blackpenguin has quit [Read error: Operation timed out] 12:27:19 -!- blackpenguin has joined ##crawl-dev 12:27:47 -!- galehar has joined ##crawl-dev 12:32:11 -!- upsy_ has joined ##crawl-dev 12:32:48 evilmike: yes, there's definitely a plan for a 2 speed system 12:33:03 ctrl+letter to activate fast mode in the skill menu. 12:33:16 the questions are: can you activate several skills? 12:33:21 couldn't it just toggle between fast/normal/off? 12:33:42 the skill menu already uses loads of buttons :P 12:33:44 cycle, not toggle i guess 12:33:45 well, it would make toggling off a skill more annoying 12:33:48 -!- MarvinPA_ is now known as MarvinPA 12:34:00 ah, maybe 12:34:09 I'd rather keep the manual skill system at the current level of simplicity, personally 12:34:12 ctrl+letter doesn't add a button 12:34:42 there's no reason why people need multiple speeds 12:34:50 elliptic: really? You don't think being able to select 2 skills with 2/3 - 1/3 would be helpful? 12:34:57 i'd be tempted to suggest leaving it with just on/off for a while at least, yeah 12:34:59 and see how that plays out 12:35:02 I think it would. But that can wait until .10 12:35:24 fine, let's keep it simple for now 12:35:42 galehar: nobody actually has a very good conceptual idea of what splitting xp in a 2-to-1 ratio means 12:36:14 in terms of what your actual skills will look like 12:36:26 for auto mode, I think the speed of spellcasting training should be increased. Maybe fighting too but I'm not sure. 12:37:33 spellcasting has a silly formula: one_chance_in(3) ? 1 : random2(1 + random2(spell_level)) 12:38:15 Zaba: found it, my fault. I marked the loot chambers KMASK: no_monster and made spawning near stairs obey that, but the whole zotdef vault has that setting from a tag as well. 12:38:15 elliptic: I would like my disabled skills in auto-mode to slowly train, so I don't have to toggle them off and on all the time only because I want to boost some other skill most. 12:39:03 kilobyte, oh, I see 12:39:04 -!- MarvinPA_ has joined ##crawl-dev 12:39:32 -!- MarvinPA has quit [Ping timeout: 252 seconds] 12:39:45 argh 12:39:57 difficulty of training spellcasting is the only reason why i would use manual mode 12:41:01 anyway yeah, on my ogbe fighting seems to be on quite a low percentage in auto mode, it's at 19% compared to 50% m&f (and obviously hitting things with a club is pretty much all i'm doing) 12:41:14 dpeg: well, I haven't tried auto mode (because it doesn't appeal to me at all)... training just one skill at a time in manual mode feels fine to me though 12:42:20 marvinpa_: fighting always trained much slower than melee skills in the past too, though 12:42:26 elliptic: my concern is that auto-mode should work as naturally as possible for new/uncaring players. As such, if you deselect a skill, the old system was better. But this needs time and I am fine with playing one release version with the current setup. It is much better than the old one. 12:42:35 -!- MarvinPA_ is now known as MarvinPA 12:43:26 ive been playing in auto mode on trunk, and the only real difference ive noticed (aside from no victory dancing) is that it seems like stealth trains faster 12:43:36 the old system actually got changed a bit in 0.8 (i think it was 0.8, at least), in that deactivated skills got hardly any training 12:43:44 Strange way of obtaining skills (https://crawl.develz.org/mantis/view.php?id=4224) by Stewas 12:44:04 dpeg: I'll say that my experience with people being confused about the (old) skill system on ##crawl was that players did not expect that turned off skills would still train 12:45:11 MarvinPA: fighting trains 1/3 compared to weapon skills. I think 1/2 would be more appropriate. 12:45:17 also as evilmike says, the system got changed in 0.8 so that skills that were turned off hardly trained at all after the first several levels 12:45:47 elliptic: these are players who already try to optimise. They are annoying by xp leak. I am concerned about players who are more casual than that. 12:45:49 considering that you can just go into manual mode and focus all your xp into fighting, it seems unecessary to arbitrarily train certain skills less often, yeah 12:46:08 03kilobyte * r160996b64e90 10/crawl-ref/source/dat/des/zotdef/zotdef.des: ZotDef: fix monsters spawning everywhere out of LOS. 12:46:11 if people are micro managing skills in the skill menu, i dont think they will mind when off means off 12:46:13 03kilobyte * r4b5286f2a430 10/crawl-ref/source/mon-place.cc: ZotDef: don't limit PROX_NEAR_STAIRS to a radius of 3. 12:46:13 -!- ais523 has joined ##crawl-dev 12:46:13 03kilobyte * r5ac16cc316eb 10/crawl-ref/source/ (defines.h player.cc): ZotDef: experimentally double XP gains, cut clock in half. 12:46:13 dpeg: well, I'm talking about people who have no idea how the skill system works 12:46:16 for spellcasting, I don't know about the formula, but it would help if it trained as fast as school skills at first, and a bit slower later on. 12:46:24 03kilobyte * rcf85c8c185b5 10/crawl-ref/source/dat/des/zotdef/zotdef.des: ZotDef: rework two maps, one used to be disabled. 12:47:00 dpeg: this was often in the same conversation where we explained to them what the xp pool is 12:47:27 anyway, I believe that full disabling means that more toggling is necessary than should be 12:47:46 -!- blackpenguin has quit [Quit: WeeChat 0.3.5] 12:48:19 dpeg: again, we tested near-full disabling in 0.8 and I haven't heard any negative feedback on that 12:48:46 kilobyte: Yay, thanks 12:49:06 -!- blackpenguin has joined ##crawl-dev 12:49:33 elliptic: I am your negative feedback. But as I said, it is okay for now. 12:49:58 dpeg: did you review the code from the patch or only the functionality? 12:50:05 I didn't realize you minded how things are in 0.8 though! you didn't mention that 12:51:38 galehar: only functionality. It felt a bit risque. 12:51:50 elliptic: sorry! 12:52:08 ok, someone should take a look at it. Can't tonight, I have to leave soon. 12:52:26 -!- ainsophyao has joined ##crawl-dev 12:53:10 dpeg: anyway I don't feel particularly strongly about this... I just wanted to mention my impression that players expect skills that are turned off to really be off :) 12:54:12 elliptic: I can relate... I am just saying that not really off may mean less hassle (and then the same or other players complain about xp leak, I am aware of that) 12:55:10 Unstable branch on crawl.develz.org updated to: 0.9-a1-951-g4b5286f (32) 12:55:51 galehar: will you add the m screen help? Perhaps we should give a statement that skill levels can be gained any time? (The bit that's confusing so many players right now...) 12:56:07 03galehar * ra54a4af07a80 10/crawl-ref/source/tilereg-skl.cc: Update the skill tab to the new system. 12:56:17 03galehar * ra0a87fa417a4 10/crawl-ref/source/skills2.cc: Fix dumping skills not showing disabled skills with -. 12:56:48 -!- ais523 has quit [Remote host closed the connection] 12:57:30 dpeg: yes, I'll add the help later tonight. So you have the time to add a bit about gained skills. 12:58:02 -!- cesium has joined ##crawl-dev 12:58:42 -!- galehar has quit [Quit: see you later] 12:59:37 hehe 13:05:21 Which file do I look into for wand effects? 13:05:26 -!- Twilight-1 has quit [Read error: Connection reset by peer] 13:06:54 -!- Twilight-1 has joined ##crawl-dev 13:15:12 When XOM switches weapons around, is there any writeup in Notes? 13:15:14 I can't find it.. 13:15:29 -!- Wensley has joined ##crawl-dev 13:25:44 -!- mumra has joined ##crawl-dev 13:32:12 Daek the Grave Robber (L4 DDNe) (D:3) 13:34:22 Early ZD seems considerably easier now. 13:34:32 (kilobyte: ↑) 13:54:42 Who had the idea of specifying all prices for rings and amulets and then multiplying them by 7??? 14:08:54 -!- ainsophyao has quit [Remote host closed the connection] 14:09:26 -!- Pacra has quit [Read error: Connection reset by peer] 14:09:35 -!- Pacra has joined ##crawl-dev 14:29:35 -!- elliptic_ has joined ##crawl-dev 14:31:17 !tell galehar I have updated the manual, although the skill descriptions still say "You train Armour by getting hit" and similar. Should we remove all of those? 14:31:18 dpeg: OK, I'll let galehar know. 14:33:00 -!- elliptic has quit [Ping timeout: 250 seconds] 14:33:17 -!- hoody has joined ##crawl-dev 14:44:20 -!- elliptic_ is now known as elliptic 15:07:49 andrewsmash the Covered (L4 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 2975. (D:4) 15:08:39 andrewsmash the Covered (L4 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 2773. (D:4) 15:09:26 andrewsmash the Covered (L4 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 2779. (D:4) 15:10:14 andrewsmash the Covered (L4 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 2821. (D:4) 15:29:22 -!- elliptic has quit [Quit: Leaving] 15:36:16 andrewsmash the Covered (L4 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 2816. (D:4) 15:39:39 -!- bmh has joined ##crawl-dev 15:39:57 hey 15:40:19 kilobyte: fixed the regression, but somewhere along the way I've introduced an infinite loop 15:55:52 -!- galehar has joined ##crawl-dev 15:57:34 Bonsoir! 15:57:51 hey 15:57:51 galehar: You have 1 message. Use !messages to read it. 15:57:54 !messages 15:57:54 (1/1) dpeg said (1h 26m 37s ago): I have updated the manual, although the skill descriptions still say "You train Armour by getting hit" and similar. Should we remove all of those? 15:58:30 well, getting hit still exercise the skill 15:58:48 which is different than training 15:59:08 galehar: there were a bunch of ASSERT(scaled_total <= scale_to) in 'skills.cc' lines recently. 15:59:22 what's CDO version? 15:59:32 oh right may not be updated. 15:59:35 >.< 15:59:48 Probably not even. 16:00:06 well, actually, I can tell just from the ASSERT message that it is not up to date :) 16:00:21 good :D 16:00:26 Never mind me!!! 16:01:52 I think the fix is already on CDO 16:02:08 it was fixed around 2PM today 16:02:21 GMT 16:02:25 andrewsmash the Covered (L4 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 2773. (D:4) 16:02:36 Four of those ~1 hour ago. 16:03:24 that's weird 16:04:03 I mean this specific code line is now ASSERT(scaled_total == scale); and it's at line 408 16:04:31 no wait it's another. But still. 16:06:09 -!- kilobyte has quit [Ping timeout: 264 seconds] 16:06:11 !lm * crash -log 16:06:12 1222. andrewsmash, XL4 MDFi, T:2816 (milestone): http://crawl.akrasiac.org/rawdata/andrewsmash/crash-andrewsmash-20110706-203615.txt 16:06:36 -!- dpeg_ has joined ##crawl-dev 16:06:39 ??herbivore 16:06:40 herbivore[1/1]: You get less nutrition from meat food and chunks, but more from veggie food -- including cheese. At the third level you can't eat meat at all -- if you've played a Spriggan, you're already familiar with how this works. Mutually exclusive with {carnivore} and generally regarded as less useful, since it makes eating chunks more difficult. 16:07:12 anyone knows how much herbivore X increase nutrition for plant food? 16:08:06 -!- lorimer has quit [Quit: brb, video driver manglage] 16:09:41 dpeg_: the formula isn't simple, it depends on food 16:09:47 food.cc:1738 16:10:22 I think someone made a table some time ago 16:11:08 http://crawl.chaosforge.org/index.php?title=Food#Nutritional_values_of_Food_Items 16:11:33 the wiki may be usually outdated, but this time I think it may be accurate :) 16:13:42 -!- lorimer has joined ##crawl-dev 16:15:48 dpeg_: around? I thought about your problem with training disabled skills. 16:17:54 -!- mumra has quit [Read error: Connection reset by peer] 16:18:02 ghallberg: I'm not sure how andrewsmash is playing 928 after CDO has been updated to 951. Maybe he's on CAO. 16:18:39 myeah, I have no idea. Just thougt I should mention those to you. 16:18:47 -!- mumra has joined ##crawl-dev 16:23:40 -!- nrook has joined ##crawl-dev 16:24:00 galehar: do you have any tips for finding the source of infinite loops in crawl? I think I implemented one and my GDB-foo is meek 16:24:22 galehar: here! 16:24:54 firing interface shouldn't pretend you could fire through toadstools (https://crawl.develz.org/mantis/view.php?id=4225) by doublep 16:25:07 bmh: no, sorry. I just use the debugger. 16:26:02 dpeg_: so, about skill training. Sometimes you want to focus on a skill, but still want some other to train too, right? 16:26:31 I think it's more intuitive if disabled skills are not trained. 16:26:57 I'm beginning to suspect I'm missing a stepdown function somewhere in the shop code 16:27:47 About elliptic's earlier comment, I understand his point. Optimal play is to train skill one after the other. Training several skills at the same time just mean you have XP spread over several skills while you could have one skill a level higher. 16:28:36 skills at 3.5 means you've got 0.5 worth of XP doing nothing for you. 16:28:48 nrook: I had a look at jewellery today, but let me discuss skills with galehar first 16:29:03 dpeg_: go for it 16:29:48 galehar: one after the other is risky business: you really need to know what you are doing. I don't like to do it that way, it is too complicated for me. 16:30:18 even if it's just one level 16:31:06 galehar: perhaps there should be a third mode instead of three levels for skills (no training, some training, full training)? 16:31:17 automatic, selected, relaxed :) 16:31:45 it's optimal, to do one at a time. The XP spent becomes useful more quickly. We can't do anything about that unless we make half skill level worth something, but it's not worth it. 16:32:14 and how relaxed mode work? 16:32:34 I don't care about optimal, I want easy. 16:32:58 sure, we don't have to fight it 16:33:13 I don't mean that in an angry way. 16:33:23 it's ok. 16:33:26 I am happy to play a suboptimal game as long as it is easy. 16:34:03 galehar: what was your suggestion? 16:34:05 So, after having thought about it, I still like my initial proposal I put on the wiki. Allow the player to select one skill to focus on it. 16:34:28 double the weight of this skill. Both in automatic and manual mode. 16:34:44 this funnels all xp into the skill? 16:34:48 ah, I see 16:35:30 easy to implement, easy on the interface. Just use ctrl+letter to toggle it. 16:35:52 -!- golgepapaz has quit [Quit: ChatZilla 0.9.87 [Firefox 4.0/20110318052756]] 16:35:55 I think I can explain why I am not so happy with the current setup: when I now disable a skill a relevant skill (say in order to get Spc up), I _have_ to enable it again later, or I'll be crippled. My play is way too sloppy to allow for that :) 16:36:03 galehar: sounds good. 16:36:36 for Spc, I think it needs to train faster in auto mode anyway 16:36:56 yes, but that does not solve my predicaments (probably) 16:37:01 -!- MarvinPA_ has joined ##crawl-dev 16:37:15 but I think it would be more natural for your case, to set Spc as "focus", while keeping the others activated. 16:37:21 yes, I agree 16:37:50 it is slightly less convenient than the current setup (because I have to choose on special skill) but a clear improvement, would work for me 16:37:53 rather than disabling others to have them train slower. That's rather unintuitive. 16:38:05 (I believe I am the only dev to speak up for the sloppy players out there :) 16:38:36 -!- MarvinPA has quit [Ping timeout: 252 seconds] 16:38:43 but when you're in this situation, is there several skills you want to train fast and several slow? I don't think so. 16:39:00 Most of the time, you want to train one faster, some slower, and some not at all. 16:39:25 yes, I think it captures the most common situations 16:39:36 -!- hoody_ has joined ##crawl-dev 16:39:56 galehar: instead of Ctrl-letter, could also be a third mode: Focused (just saying for interface purposes) 16:40:04 -!- hoody_ has quit [Remote host closed the connection] 16:40:13 -!- hoody has quit [Read error: Connection reset by peer] 16:40:37 -!- hoody has joined ##crawl-dev 16:40:44 and it also makes the "sloppy" play more efficient, since focusing on one skill is better. 16:41:03 okay 16:41:15 I don't understand how the focus mode would work 16:41:50 I mean you can set a skill as focus both in auto and manual mode 16:42:02 ah 16:42:21 the focus skill would be white? 16:42:35 yeah, could be 16:43:03 I use white in view transfer though 16:43:46 but since we have darkgrey for disabled and lightgrey for enabled, it would make sense to use white for the focus skill 16:44:33 yes, that's what I was thinking 16:45:00 Food reform (https://crawl.develz.org/mantis/view.php?id=4226) by dpeg 16:45:00 I'll see if I can find another colour for transfer. Or figure something out. 16:45:08 yay :) 16:45:19 does having one focus skill really solve dpeg's problem, which is that he wants some low-priority skills to grow over time? I think that "relaxed" option might be better than a "focus" option, or possibly work alongside it, where "relaxed" gave a skill 1/10 weight or something 16:46:19 huh, I'm hitting an infinite loop in random_choose_weighted 16:46:30 -!- dpeg has quit [Quit: leaving] 16:46:37 ah-ha 16:46:47 I wizmoded my skills up to level 27 and they're blowing up 16:46:59 that makes me happier than my code being broken 16:47:02 bmh: what's your version. I fixed that. 16:47:15 Wensley: galehar made it clear that disabled skills will not train. 16:47:24 oh yeah, I think you merged trunk with this bug. Update, I fixed it. 16:47:25 74abd235 16:47:29 nrook: still around? 16:47:45 dpeg_: yep 16:48:11 just typed up jewellery base prices: http://pastebin.com/esiVibgH 16:48:23 cool 16:48:25 haven't looked at artefacts yet 16:48:53 nrook: don't think there is some fancy modification for prices. Wished up a jewellery shop, and prices were as listed in shopping.cc: =MR 280 etc. 16:48:57 bmh: http://gitorious.org/crawl/crawl/commit/7c3afbb422668eab54b9cda085e35488b5374af5 16:49:13 nrook: do you see any reason for the 7* at the end? 16:49:17 dpeg_: yes, that's what I'm saying :) by using a "relaxed" training option, you could imitate the effect of turning off an often-used skill in the old skill system, thereby allowing people to migrate over more easily and without losing any functionality 16:50:00 galehar: I'm out of date 16:50:02 Wensley: galehar coded the new system, I can understand that he wants to burn the bridges to the old one 16:50:02 dpeg_: no, though I suspect it's been there for a very long time 16:50:42 Wensley: we will see how playes adapt. If there is a sizable fraction that cannot get along with fully disabled skills, we surely can react somehow. But quite possibly a focused skill is just what is needed. 16:50:44 the code would be slightly cleaner without it 16:50:49 I think the same 16:50:51 I'm all for burning bridges, but I'd like it if the new system gave me more control in all ways, rather than more control in most ways 16:51:04 (but either way the new system is great!) 16:51:25 people are free to control their skills in manual mode. If auto mode brings unatural results, it's up to us to fix it. Not to the player to fiddle by disabling skills that train too fast. 16:51:30 alright, I've fixed things up so that the abyss doesn't horribly thrash in a way that makes kilobyte sad 16:51:31 Wensley: I don't think we want the player to specify "weight for Fighting: 30%, for U&C 20%, for T&D: 5% etc" :) 16:52:03 dpeg_: well, we could just cut out the middle man and let players choose what level they want each skill to be at from the very beginning :P 16:52:13 then when it reaches that level it stops training 16:53:00 so, do you think weapon skills train too fast? Spc too slow? Let's fix that and keep the base system simple. 16:53:09 -!- MarvinPA has joined ##crawl-dev 16:53:24 Wensley: that'd be definitely too complicated for me. 16:53:29 bmh: good work! 16:53:48 I'm gonna go cut out that *7 now 16:54:45 Wensley: I prefer to throw them into the cold water. We will see after a while how easily they can adapt. 16:55:04 nrook: does it have something to do with the artefact properties? 16:55:15 dpeg_: about food reform, I think there's still some design issues, so maybe it's not the best time to call for implementation. 16:55:35 dpeg_: I think it's the other way around, artefact_value returns values meant to be multiplied by 7 16:55:50 but I think it all just stems from "rings were coded up to be multiplied by 7 at the end" 16:55:50 -!- MarvinPA_ has quit [Ping timeout: 252 seconds] 16:55:58 galehar: which ones? I spent three hours thinking about it now. 16:57:33 dpeg_: well, I have to admit I haven't think about it much. But kilobyte noted that preventing herbivore 1&2 to eat meat before very hungry and near starving could be problematic. 16:57:38 nrook: let's do the mundane items first (including rings with pluses) and randart properties afterwards 16:57:57 dpeg_: agreed 16:58:33 -!- Pacra has quit [Ping timeout: 255 seconds] 16:59:29 galehar: I added slow digestion and what they get from plant food 16:59:47 nrook: also, got to type maths over here, so may sometimes be amiss 17:00:39 dpeg_: oh, you updated it? Fine. I trust you :) And most of the original proposal was good anyway. Adjustments can come later. 17:01:07 galehar: yes, that's were the three hours went into 17:01:24 adjustments will be necessary, it is a big change, but we'll see 17:01:45 ok, that's cool. I don't have the time to look into this anyway :) 17:01:49 when I proposed Spriggan full herbivore to Erik Piper years ago, he responded by "unplayable" :) 17:01:49 -!- kilobyte has joined ##crawl-dev 17:01:51 galehar: yes, I know 17:01:58 kilobyte will complain any moment 17:02:01 kilobyte: Hi! 17:02:33 dpeg_: on the food reform page, it says that each level of the herbivore mutation reduces digestion to 4/5, should it be reduces digestion *by* 4/5, so that at herbivore 3 you'd have about 1/2 normal digestion speed? 17:02:50 Wensley: yes, of course. Would you modify. Meant as factors. 17:02:56 sure thing 17:03:00 thx 17:05:29 nrook: some prices that are obviously off (all imo). Too cheap: cTele, rMut, Faith, Rage. Too expensive: regen, invis (but that may be my mistake), Fire/Ice (they used to be a lot stronger), rN (why twice the price of rF,rC,rP?) 17:06:58 It would be nice if one could add single spells to the shopping list. It would then show the cheapest book that has the spell for each one of them. 17:07:28 <-- thinks it would be nice if pressing M over a book shop showed the normal memorization-from-ground screen for all the books in the shop 17:07:53 <-- hates memorizing the spell lists in books! 17:07:55 think about those and add them to the relevant pages :) 17:08:05 agreed with all of those, also i'd probably move warding down and conservation up a fair bit 17:08:14 MarvinPA: thank you! 17:08:39 alright, I'll move those around a bit 17:08:53 MarvinPA: I use Cons very much, but I am never sure if this is a personal bad idea. 17:09:00 -!- hoody has quit [Ping timeout: 252 seconds] 17:09:14 not at all, i'd say it's the amulet i use the most (assuming i find it) 17:09:25 I have a new version up, but all I changed was making the prices rounder 17:09:32 MarvinPA: perhaps you make the same mistake :P 17:09:36 so don't complain that I haven't done anything :p 17:09:39 heh, entirely possible! 17:10:52 -!- ixtli has quit [Quit: Textual IRC Client: http://www.textualapp.com/] 17:13:38 dpeg_: on food reform, did you intend to have spriggans have both herbivore 3 and explicit slow digestion 3? also, what's the reasoning for only allowing gourmand to affect chunks? 17:14:06 -!- Pacra has joined ##crawl-dev 17:14:26 -!- ion_ has joined ##crawl-dev 17:14:42 Wensley: you can already eat permafood up to Engorged in the proposal. 17:14:49 ah, good point 17:14:53 Wensley: and Sp should keep both Herb 3 and SD 3, yes. 17:15:14 -!- ion has quit [Remote host closed the connection] 17:15:18 -!- ion_ is now known as ion 17:15:28 "ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed." Latest local. 17:15:32 They have to play the whole game with Herb 3, so should have it as easy as possible. If a random char gets Herb, they can react in various ways. Sp can't. 17:16:22 RichardHawk: this is fixed already. Let me update the windows build. 17:18:44 new working copy: http://pastebin.com/esiVibgH 17:19:37 general question: are we fine with reducing prices for cursed items in shops _unless_ the player follows Ash, in which case they'll be more expensive? 17:19:52 cTele, rMut, Rage, Faith, Gorumand, Cons, Stasis made more expensive; Invis, Regen, rMagic, +MP, Fire, Ice, rN+, Lev, Warding made cheaper 17:20:35 I think unknown at 50 is too cheap. 17:20:48 -!- ixtli has joined ##crawl-dev 17:20:59 dpeg_: I like "prices are the same no matter your circumstances" as a general rule 17:20:59 so the shopkeeper will recognise that you follow ash and adjust prices? :p 17:21:07 Why would the shopkeepers care about your god? 17:21:18 Or even know. 17:21:25 nrook: but a clever shopkeeper would do that 17:21:47 RichardHawk: well, they can sell cursed shit^H^H^H^H gold to Ashenzarites 17:21:58 I suspect unknown jewellery was accidentally not multiplied by 7, since 350 is much closer to all the others; I'll bump it up to 250 17:22:08 You can spot an Ash follower miles away from their cursed hats. 17:22:56 what next, though? Cheap bread rations for ogres? Expensive barding for centaurs? Discount /Draining for TSOites? 17:22:58 (The gameplay point is that the Ash guy should choose between cheaper uncursed + blowing scroll and more expensive but readily cursed) 17:23:05 And then some poor bastard who has just wiped out an ossuary gets mistaken for one. 17:23:06 it's a road I don't like 17:23:30 ok, but you're the backward conservatives :P 17:24:37 guilty as charged 17:25:05 -!- upsy_ has quit [Quit: Leaving] 17:25:31 but on the other hand, I believe nethack makes food rations more expensive if you obviously need them :) 17:25:49 Well, just because you worship Ash doesn't mean -all- properties of cursed things flip upside down. You can't magically switch out your cursed equipment either! (Ignoring a shortlived commit) 17:27:30 I'd like to put a minimum price of 50 for identified things, too 17:27:44 Galehar: Much appreciated. 17:27:51 nrook: to stop people from scooping up cursed items for 1? 17:27:56 the big negatives for bad jewellery are useful for randarts, but you're still getting an ID 17:28:02 yes 17:28:08 well, actually, 50 is too high, maybe around 20 17:28:40 although people would still buy at 50 for acquirement-improving purposes, I suppose 17:29:01 nrook: how much is rCurse? 17:29:11 -!- dpeg_ has quit [Ping timeout: 240 seconds] 17:29:15 -!- dpeg has joined ##crawl-dev 17:29:50 I don't think there is an rCurse property, is there? 17:30:14 nrook: er... !rmCurse 17:30:15 er 17:30:18 ?rmCurse 17:30:28 oh 17:30:40 remove curse is 30 17:30:41 Windows builds of master branch on crawl.develz.org updated to: 0.9-a1-953-ga0a87fa 17:30:48 nrook: yes, just like with bad scrolls and potions 17:30:49 20 is good 17:31:01 (I've seen new shops and !para for 14 was good.) 17:32:38 in any case, I'm going to go eat now, I'll be back in around an hour 17:32:45 I'll think about plusses while I'm digesting :) 17:34:55 see you 17:35:09 * dpeg gives nrook a ring of slow digestion. 17:35:22 * bmh sends dpeg back to Nethack 17:37:25 :) 17:40:35 I'd be sort of fine with shops adjusting prices for temporary ailments, like upping the price of ?rc if you have a cursed item in hand, or upping the price of food if you're starving. but adjusting prices based on permanent properties should be avoided, imo 17:40:38 !tell kilobyte I fixed the annoying abyss shifts to be less bothersome. Pushed to github 17:40:38 bmh: OK, I'll let kilobyte know. 17:40:44 -!- bmh has quit [Quit: bmh] 17:41:14 -!- galehar has quit [Remote host closed the connection] 17:41:38 -!- ixtli has quit [Quit: Computer has gone to sleep.] 17:44:53 Wensley: I won't press the point further, but there is zero choice space for Nagas who get bardings made more expensive, whereas there is a potential decision for the Ashenzarite who gets cursed items made more expensive. 17:46:20 I think it would be interesting if players were forced to make a choice for proactive purposes rather than just reactive purposes. say, should I buy this potion of cure mutation now, or save my money but if I need it later it'll be more expensive? 17:47:29 Wensley: it is very similar with Ash, actually. But I guess volatile prices is a completely different chapter which should be explored on a brainstorm wiki page first. 18:35:34 -!- MarvinPA has quit [Ping timeout: 252 seconds] 18:36:45 dpeg: back 18:36:52 -!- MarvinPA has joined ##crawl-dev 18:38:16 nrook: it was delicious, I hope! 18:39:06 it was! 18:40:14 when it comes to point rings, I'm thinking we should move all of the base prices to 0 (this isn't like wands, a +0 ring of protection is worthless) and raise the per-point cost to compensate 18:41:32 this would probably make it necessary to differentiate between points on different rings; +1 AC and +1 int are quite different 18:41:39 I am not sure the formula should be linear, btw. 18:41:40 and of purchasing negatively-enchanted rings of whatever for identification? 18:41:52 monqy: about 20, as for scrolls and rings 18:42:01 there's a catchall minimum price of 20 for any jewellery 18:42:04 aha 18:42:32 we need to make sure that negative modifiers don't reduce the price too much -- this will probably be used in randarts, right? 18:43:49 nrook: the point (for non-linear) is that +2 (whatever) is much better than two +1 rings 18:44:46 we can probably put negative modifiers off until we deal with randarts, but yes 18:45:34 yes 18:45:52 something simple like "double the modifier for all points over 3" might suffice for nonlinearity 18:45:53 * dpeg proposed quadratic growth 18:46:22 nrook: I don't understand -- isn't that exponential? 18:46:49 nrook: can you see how rings of slaying 0,1 and 0,4 and 0,7 differ right now? 18:47:27 dpeg: oh, I was unclear, let me restate 18:47:36 rule of thumb: killer items (like a 0,8 slayer) can be so expensive that a single ?aquirement may not suffice to pay for it 18:47:44 I believe the difference between such rings is actually less than linear right now 18:47:54 well, no, it's linear 18:47:59 not sure about artefacts 18:48:27 03MarvinPA * r2e6381dcce7f 10/crawl-ref/source/monster.cc: Make Inner Flame expire after a while 18:48:27 dpeg: I forgot all about ?acquirement, that deals nicely with my worries about huge costs from quadratic growth 18:48:34 03MarvinPA * r3cf6573e28fb 10/crawl-ref/source/monster.cc: Make Zin's Silver Corona affect the same monsters as the silver brand 18:49:03 I was worried that +7 slaying would cost an unobtainable amount and thus be without purpose, but pretty much any amount of $ can be gained by using acquirement 18:49:21 nrook: it is one of the best uses of ?acquirement: you already know the item exists; if ?acq underperforms, you just collect money somewhere else -- this is much less a gamble that other acquirements (except in bazaars) 18:49:40 MarvinPA: does the Shroud spell work reasonably well? 18:49:40 yeah, quadratic works then 18:50:57 dpeg: i've not used it very much but it definitely seems helpful 18:50:57 say, for slaying we use 100 per base damage boost, so that 0,n goes to 100*n(n+1)/2, this would be 3600 for 0,8 -- seems reasonable? 18:51:14 MarvinPA: okay -- I love the concept (and I keep saying that) 18:51:34 * dpeg will not rest until Lugonites can desecrate altars and Repel Missiles works Shroud-style! 18:51:43 heh :D 18:52:06 I am serious about both! 18:52:11 that seems reasonable 18:52:36 What would a good target price for +8 AC be? 18:52:47 doesn't AC only go up to +6 18:52:58 repel missiles working more like shroud might be interesting, yeah 18:52:59 MarvinPA: thank you very much for your work, btw. You rock! 18:53:14 that's what I'm thinking about now; my instinct is that it's good but not a gamechanger, but I think I'm wrong 18:53:25 monqy: how would I know? I barely know what language this game is written in. 18:53:29 shroud is elliptic's work btw, just in case you were crediting me for it :P 18:53:34 MarvinPA: I know! 18:53:38 aha ok :) 18:53:52 elliptic also rocks, but he's not around :) 18:54:05 dpeg: well I remember at least one ring only goes up to +6, and there's a +8AC unrand ring, so I figured protection is probably one of them 18:54:07 * dpeg wonders if kilobyte is present, cause he also rocks 18:54:07 perhaps something around 1800; half is very simple 18:54:32 yes, Robustness and Shaolin 18:54:40 I have no clue what the relative utility of str/dex/int/ev/ac is though; simplest to make all worth the same 18:54:42 i believe evasion/protection both go up to +6 (and robustness/shaolin are +8), and slaying goes up to +0, +8 18:54:59 and stats go up to +6? 18:55:11 i think so, not certain though 18:56:23 nrook: even if they're wildly different, imo crawl should at least pretend they're worth the same, and balancing them to all be about equally worthwhile can be done later 18:56:42 crawl should definitely pretend str/dex/int are worth the same, and same with ev/ac 18:56:49 yes 18:56:56 and with to-hit and to-dam 18:57:09 but I have no reason to think stats and ev/ac have the same value 18:57:22 as in, that those two classes ("stats" and "ev/ac") are comparable 18:57:46 no, I don't think that stats, defense and slaying are the same 18:57:47 for combining +hit and +dam prices on slaying, would addition suffice, or is something more complex needed? 18:57:57 hm, good question 18:58:05 I'm thinking something more complex is probably needed, but straight multiplication is terrifying me 18:58:22 I don't think the numbers are independent -- are there things like 7,7 ? 18:58:24 +6,+6 slaying should be in the 8000-10000 range I think 18:58:30 nrook: cannot multiply :) 18:58:51 it would be 4 million 18:59:09 ??sword of jihad[2] 18:59:10 sword of jihad[2/2]: Seen in a shop for 32544 gold. 18:59:14 dpeg: ha, of course 18:59:55 i have no idea what the distribution is but i think the highest you can get is +6, +6 or +0, +8 19:00:08 yeah 19:00:11 ??slaying[2] 19:00:12 slaying[2/2]: A ring of slaying cannot boost accuracy more than +6, and it cannot boost damage more than +6 unless the accuracy boost is +0. Thus the best possible rings of slaying are (+0, +8) (probability 1.4%) and (+6,+6) (probability 0.1%). 19:00:27 if we just do n(n+1)/2 and set n to (to-hit + to-dam) that's probably alright, though it gets expensive fast 19:00:51 the asymmetry with slaying is a bit peculiar 19:01:33 if we keep the coefficient at 100, +3,+3 slaying costs 2100 19:01:36 10k is way too high even if slaying is godly 19:02:05 evilmike: I agree 19:02:11 evilmike: but 3600 for 0,8 is okay? 19:04:05 dpeg: That would be worth paying. But I'm not too opinionated about shop prices. I just say 10k is too high because I almost never have that much gold 19:04:09 how much would a -6,6 ring cost? 19:04:26 and I think shop items should at least be possible to buy 19:04:46 Wensley: this is the other interesting question, yes 19:05:01 are such rings generated? 19:05:29 on randarts I've seen something similar, iirc 19:05:59 we can chicken out and accept the fact that right now, to-hit means very little... and just associate base like 50 to it 19:06:41 -!- syllogism has quit [] 19:07:04 minuses on rings should be linear, I think 19:07:24 since they'll only matter for otherwise good rings, and in that context, -4 AC really is just twice as bad as -2 AC 19:07:45 we can put that off until we deal with randarts, though, since it is only relevant in that context 19:08:24 nrook: good point 19:10:32 if we don't weight +hit and +dam differently, one of +4,+4 and +8 will have to be unfairly priced 19:10:36 huh, looks like it's possible to get plain slaying rings with negative to-hit and positive to-dam, but only pretty rarely 19:10:46 slaying generation looks weird :P 19:11:26 nrook: yes, which is why I propose to accept the inconsistency right now and use different base prices 19:11:48 If/when to-hit and to-dam are repaired, someone will hopefully recall to address the rings. 19:11:56 *the ring prices 19:12:52 I'm not too familiar with the mechanics; exactly how much worse is to-hit than to-dam? What would x have to be for a +x,+0 slaying ring to equal a +0,+1 one? 19:13:26 in cost? or effectiveness 19:13:37 as to the latter, there isn't really an answer 19:14:15 dpeg: That would be worth paying. But I'm not too opinionated about shop prices. I just say 10k is too high because I almost never have that much gold 19:14:40 in effectiveness 19:14:42 from what I understand, part of the problem is that crawl generally favors lighter, faster weapons that can be wielded with a shield, and since lighter weapons are inherently more accurate much of the +acc goes to waste 19:14:43 by vaults:8 you could have from 3-7k gold, regularly, based on frequency of things bought 19:15:14 Pacra: yes, but I think 5k or so should be the upper limit. Dragon King can get funny prices. 19:15:24 agreed 19:15:35 having 10k gold usually means you have been unlucky with shops (nothing good, or just not many shops), or you have been acquiring gold a lot, or you have been farming 19:15:36 I would even go a bit lower 19:15:36 Wensley: accuracy in generally matters too little, afaiu 19:15:38 thats in my experience anyway 19:15:40 4k as upper limit 19:15:50 dpeg: not if, say, you're an ogre with a holy giant spiked club 19:15:58 Pacra: two ?acq can net 6k, it is okay 19:16:09 that's fair 19:16:11 Wensley: then I don't know if +4,0 is doing much good 19:16:21 although I remember gold acquirement having quite a huge spread! 19:16:39 I'm not sure of the curves myself, but having +acc definitely helps with two-handers, as little as they are used 19:16:40 (Min=220, Max=5520, Mean=1218, Median=880, Std=911) 19:16:40 my current opinion is that we should do a quadratic progression for slaying off "k(to-hit) + to-dam", where k is some number less than 1 19:16:45 it -can- net 6k 19:16:47 wait 19:16:49 550k 19:16:54 er 5.5k 19:17:03 but it rarely will, dpeg 19:17:22 gold acquirement is weird enough as it is 19:17:37 as in the formula therefor 19:17:41 Wensley: you mean old or new formula? 19:17:47 ??gold acquirement 19:17:47 gold acquirement[1/2]: Gives 10*(20+1d20+1d8*1d8*1d8) gold. (Min=220, Max=5520, Mean=1218, Median=880, Std=911) 19:17:49 10*(20+1d20+1d8*1d8*1d8) 19:17:55 wow 19:18:00 I understand that it's been improved, but still 19:18:26 Wensley: I reckon for you "improved = more gold". 19:18:46 given that ?aqc on gold is the least gamble of all (apart from food), I say high variance is a must 19:18:49 dpeg: not at all, I hate players >:D I just want it to be less silly 19:18:56 what is silly about this? 19:19:01 the formula? 19:19:10 I can explain it just fine 19:19:24 there's nothing silly about it 19:19:33 it seems unnecessarily complex 19:19:46 product of three random factors produces high variance, I add a single random term to increase the range of numbers, and I ensure a minimum value 19:19:54 Wensley: come up with something better 19:19:54 I was just replying that "two ?acq can net 6k" is fairly unlikely 19:20:02 Pacra: yes, you are right 19:20:20 I remembered the numbers to be bigger :) But still, 3.6k for a killer ring is fine. 19:20:29 dpeg: I think it would be better to just have a function that produces random numbers according to a normal distribution and then come up with a sane min and max 19:20:40 just because it has a lot of die rolls doesn't make it a bad formula, yeah :P 19:20:52 Wensley: I think min and max are fine, thankyouverymuch. 19:20:58 it seems to work pretty well in practice 19:21:04 that wasn't what I was implying :P 19:21:13 I like this current distrib because it stays with the flavour of acquirement scrolls 19:21:32 i.e. you can get something very very good (2k-5.5k gold) or something awful (220 gold etc) 19:21:50 MarvinPA: would be awesome to have statistics about what players acquired, at which levels, and what they got 19:21:57 oh, wow. 19:22:03 that would be an incredible log 19:22:05 Pacra: yes, chance for dud is crucial 19:22:07 I would read that for days 19:22:11 haha 19:22:33 -!- ion has quit [Read error: Connection reset by peer] 19:22:37 nrook: shall we gonna go on with the prices? =) 19:22:41 -!- ion has joined ##crawl-dev 19:22:43 might be reasonable to put acquirement results in the character dump 19:23:13 yes 19:23:28 they probably are, or at least you can get them there 19:24:08 in the morgue files do acquirement scroll uses and their products get logged? 19:24:21 dpeg: sure! What do you think of doing quadratic growth over the variable n = .5 * to-hit + to-dam? 19:24:25 haven't looked at one in ages 19:24:32 same 19:24:41 will remember to do so if I find an acq scroll 19:25:12 nrook: sounds good, makes 6,6 at 4500 a bit more expensive than 0,8 at 3600 but it is okay 19:25:15 would be awesome to run a script capturing all that information, separate it into different item classes, seewhat comes out 19:25:22 yes 19:25:50 so slayers are dealt with... now we need base prices for AC/EV and for Int/Dex/Str 19:25:58 how much is Int+8 worth? 19:26:05 surely more than 500? 19:26:15 +stat can only go up to +6 iirc 19:26:29 we already had that, sorry 19:26:31 (in case that affects the curve) 19:26:42 and yes, much more than 500 19:26:57 surely, but I'm not sure by how much 19:27:06 tell me a target price, I'll divide by 21 19:27:33 with base 40 we get 840 for Foo+6 19:27:52 looks okay for a start... what does it cost right now? 19:28:15 now it costs 490 19:28:39 that is a reasonable inflation rate, imo 19:28:47 which is clearly not enough, of course; around 1000 sounds reasonable 19:28:50 what is AC+6 atm? 19:29:02 exact same thing 19:29:03 base 50 makes it 1050 19:29:42 reasonable as well 19:29:52 yes, either way is fine by me 19:29:54 we might have to scale these numbers down a bit if randart prices start to be absurd 19:30:30 can randarts still have AC/EV ? I seem to recall they were nerfed away when kilobyte found out that players get insane AC on randarts alone 19:30:57 only if their base type is a ring of protection/evasion 19:31:04 not as a random property 19:31:10 thank you 19:31:23 thank you, I had no idea one way or another 19:31:28 and 840 sounds reasonable enough to me for a +6 ring, i think? 19:31:34 MarvinPA: I would guess that AC/EV is more attractive on randarts than stats, is that correct? 19:32:14 it depends 19:32:16 sure 19:32:39 does it depend on whether the stat is int? :) 19:32:50 for the most part, although +int is always nice (but that's obviously caused by int being way more desirable than str/dex) 19:32:54 heh 19:33:25 or str, for heavy armour users below the str requirement for their desired armour 19:33:39 true 19:33:41 particularly if the +str brings them to/above it 19:33:56 there is an issue that AC+4,EV+4 on a randart will be cheaper than AC+6 19:34:29 what about AC/EV base 40, and stat base price 30? 19:34:46 if we do a superlinear scale on these rings, I think we're forced into one for randarts 19:34:58 isn't the only way to get AC+4,EV+4 on a randart the brooch of shielding, or am I missing something? 19:35:01 er 19:35:03 not randart 19:35:05 artefact 19:35:26 monqy: you are right, what a relief. Thanks to kilobyte 19:35:29 right, since ac/ev can only be the base type 19:35:41 same issue with stats, but it matters less 19:35:57 so that settles all the base types? 19:36:26 it's not really that much different than, say, AC+4,+rage,rN+ being cheaper than AC+6, though, is it? 19:36:31 randarts are hard 19:36:35 but yes, I'm satisfied with base types now 19:36:46 anyways, i should be off to sleep, have fun with pricing things :) 19:36:51 nights 19:36:58 * dpeg should also sleep, 2.30 am 19:37:05 goodnight 19:37:06 -!- MarvinPA has quit [Quit: Page closed] 19:37:19 nrook: where is the place in shopping.cc that does the randarts? 19:37:55 dpeg: it's artefact_value, I believe, and then the prices coming out of there are multiplied by 7 and added to the base price 19:38:49 // This should probably be more complex... but this isn't so bad: 19:39:52 it does look a bit bad 19:39:58 ah well, we'll deal with it later 19:40:02 I'll type up a patch for jewellery 19:40:03 perhaps it was much worse before :) 19:40:07 okay 19:40:16 I am not sure if just adding them is good 19:40:18 I think I saw a comment saying there was no price modification for artefacts 19:40:29 so it was certainly much worse! 19:43:15 -!- edlothiol has quit [Ping timeout: 255 seconds] 19:45:07 we should add, and have multipliers (of order 1.1 to 1.5 for each positive randart property) and slap on those in the end 19:45:33 some of the meaningful negative properties can get multipliers <1 19:46:02 artefact_value actually looks suitable for its return value to be put into some faster-than-linear function 19:46:20 nrook: but what I propose does just this 19:47:14 (base_AC + base_rF + base_MR) * (1.1 * 1.2 * 1.05), for example 19:47:30 so combinations of good properties will be noticeably more valuable 19:47:35 oh, and (base_rF + base_rF) * (1.2 * 1.2) 19:47:41 yes 19:48:09 our ancestors tried, though: there is a special case for rC+rF (but I think it is not nearly enough) 19:48:09 -!- ixtli has joined ##crawl-dev 19:48:27 still, then, we have to come up with two numbers 19:48:52 I think it is simpler just to have a linear "goodness" number and then to say that something with 20 goodness costs, say, 4 times as much as something with 10 goodness 19:49:28 hm, you think a single function f(total base) will do? 19:49:37 I am not sure, but I can be convinced 19:49:57 I do---but, of course, it is just intuition until I find f :) 19:50:41 the thing is that rF and rC alone will cost a lot less than a good slayer but together they should be on par or even better 19:50:51 not sure how you can see that on the prices alone 19:51:07 will all these properties affect randarts? 19:51:19 Pacra: sure! Not good? 19:51:36 Your +5 helmet of rElec rF rC should be expensive, shouldn't it? 19:51:50 well, we can even put rF and rC through the same n(n+1)/2 function we're putting slaying through 19:52:05 just concerned that a decent randart may quickly become astronomically expensive with stat cost changes et al. 19:52:14 it'd need some sort of "no, 50000 is too high" stepdown function for absurd randarts, but I think it's workable 19:52:18 Pacra: we will find a way to catch the outliers 19:52:27 expensive is good yes, but not unreachable :[ 19:52:30 ok :] 19:53:06 this is also relevant for saccing purposes 19:53:08 incidentally, final base jewellery prices: http://pastebin.com/esiVibgH 19:53:24 I think you know all of these numbers already of course 19:53:38 since I haven't changed anything other than doing what we decided for variable-strength rings 19:54:05 nrook: rF+rC is a lot better than rF++ 19:54:28 I suggested 40 and 30 rather than 50 and 40 :) 19:54:37 oh right, I forgot 19:54:41 * dpeg suspects that nrook is paid by the shopowner's union 19:54:45 hah 19:55:03 the real reason I reposted it: because I knew I'd forget something :) 19:55:32 and of course I'm paid by the shopowner's union, why else would I encourage a price revision? 19:55:42 I prefer to think of it as unlimited campaign contributions 19:55:44 nrook: looks good 19:55:50 shall we have two commits, or just one? 19:56:28 two, but just because I already made one which was "take out the *7 multiplier but multiply everything by 7 manually so there's no gameplay changes" 19:56:47 MR is more expensive right now than rF :) 19:57:05 nrook: also fine to have a separate patch for base prices 19:57:34 d - the ring "Vaoghiwn" {Str+2 Dex+5 Acc+4} 855 gold 19:57:35 and STR is more expensive than slaying! 19:57:39 yes 19:57:57 h - the amulet of Eternal Night {Gourm rPois Str+4 Dam+3} 651 gold 19:58:03 Eternal Night should cost a lot more 19:58:09 and it will 19:58:33 it certainly will 19:58:43 j - the amulet "Putuj" {Cons rF+ Dex+1} 464 gold 19:59:12 -!- ixtli has quit [Ping timeout: 252 seconds] 19:59:28 nrook: okay, so you try to find a good f, I will think about multipliers 19:59:42 alright, I will 19:59:48 weren't you supposed to be sleeping? :p 20:01:16 yes, dressed and hooded 20:01:20 later! 20:01:23 -!- dpeg has quit [Quit: leaving] 20:02:21 -!- cesium has quit [Remote host closed the connection] 20:06:27 -!- ixtli has joined ##crawl-dev 20:12:20 -!- Wensley has quit [Ping timeout: 246 seconds] 20:26:13 03brendan 07abyss * r3ae088c79f71 10/crawl-ref/source/abyss.cc: Fixed abyss thrashing 21:08:49 -!- nrook has quit [Quit: leaving] 21:14:09 -!- Pacra has quit [Ping timeout: 255 seconds] 21:17:38 !tell dpeg would one message entry for silenced Singing Sword be enough, or do you want to do it per tension tier? 21:17:39 kilobyte: You have 1 message. Use !messages to read it. 21:17:39 kilobyte: OK, I'll let dpeg know. 21:25:40 -!- ainsophyao has joined ##crawl-dev 21:41:13 -!- hoody has joined ##crawl-dev 22:00:35 -!- ainsophyao has quit [Remote host closed the connection] 22:28:42 Hah. I placed teleport traps in ZD that block direct access to the zot from the outside. All the monsters stopped moving. 22:28:49 Can’t even disarm them. 22:29:15 They don’t fight back either. 22:29:20 The monsters, that is :-P 22:42:07 < Gretell> ion the Plane Slider (L27 NaWr), worshipper of Jiyva, escaped with the Orb and 15 runes, with 99999999 points after 11280 turns and 3:11:39. 22:42:27 That win should probably be deleted from the database by someone. Crawl was buggy. 22:46:46 First of all, ZD was too easy with the new changes. Secondly, when i placed teleport traps to block monsters from reaching the orb, all movement of NPCs stopped. I just had to wait for the remaining two or so runes to appear and then hack my way through the monsters standing still to win. (A.k.a. cheating :-P) 23:10:12 03dolorous * r48e82d726fc3 10/crawl-ref/source/dat/database/monspell.txt: Properly use @possessive@ in more monster speech. 23:20:11 03dolorous * r30c896eab32e 10/crawl-ref/source/mon-info.cc: Properly ignore shedu number (used for marking their mates) in monster_info. 23:36:04 LuckyNed the Fighter (L15 OgWz) (Lair:8) 23:43:08 -!- evilmike has quit [Read error: Connection reset by peer] 23:46:50 -!- ainsophyao has joined ##crawl-dev 23:47:32 moocowpong1 the Covered (L5 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 3746. (D:4) 23:48:11 -!- st__ is now known as st_ 23:48:19 moocowpong1 the Covered (L5 MDFi) ASSERT(scaled_total <= scale_to) in 'skills.cc' at line 412 failed on turn 3646. (D:4) 23:50:43 -!- moocowpong1 has joined ##crawl-dev 23:57:14 Using identify scroll doesn't change num_marked on deck (https://crawl.develz.org/mantis/view.php?id=4227) by nubinia