00:00:08 <|amethyst> I'm assuming "items" meant "wielded/worn items" 00:00:08 |amethyst: I'm counting that with spells, "magical effects" 00:01:19 …is the constant e available somewhere? 00:01:26 <|amethyst> exp(1.0) 00:01:43 ok. I'll hardcode it since it's going in a header 00:01:52 Wensley: I'm going to make rP follow this: http://en.wikipedia.org/wiki/Sigmoid_function 00:02:35 level 1 would confer 73% resistance, level 2 88%, level 3 95% 00:03:20 Unstable branch on crawl.develz.org updated to: 0.10-a0-1282-g71db0c8 (32) 00:07:58 <|amethyst> bmh: hm... it seems like that first jump is very big 00:08:15 <|amethyst> from 100% affected to 27% affected 00:08:29 -!- syraine has joined ##crawl-dev 00:08:30 |amethyst: at the moment it's 100% to 0% :) 00:08:59 -!- MarvinPA_ has joined ##crawl-dev 00:09:05 -!- dtsund has quit [Read error: Operation timed out] 00:10:42 <|amethyst> why do you need e in a header, anyway? 00:11:03 <|amethyst> presumably you're using it for e^x, in which case you just use exp() rather than pow() 00:11:21 |amethyst: I was going to allow for general sampling from logistic functions with a tunable parameter that defaults to e 00:11:30 (I have since decided against it) 00:11:44 -!- MarvinPA has quit [Ping timeout: 256 seconds] 00:14:08 -!- blabber has quit [Remote host closed the connection] 00:18:37 Windows builds of master branch on crawl.develz.org updated to: 0.10-a0-1282-g71db0c8 00:27:31 you're making a patch that adds 3-level rPois? 00:27:48 evilmike: yes 00:28:23 <|amethyst> what about monsters? 00:28:52 |amethyst: That's not my department. 00:29:02 00:29:07 <|amethyst> so monster rPois stays all-or-nothing? 00:29:14 there's a bit about rPois on the wiki here: https://crawl.develz.org/wiki/doku.php?id=dcss:brainstorm:effect:poison 00:29:24 evilmike: yeah, read it. 00:29:35 it doesn't say much, but it looks like there was a bit of a consensus that multi-level rPois for the player would be a bad thing 00:29:40 -!- valrus has quit [Remote host closed the connection] 00:31:08 evilmike: The three tiers I've implemented at 75%, 88% and 95% 00:31:55 There seems to be one camp that really doesn't want poison to be binary. 00:32:20 <|amethyst> Nagas get rP+ or rP+++? 00:33:50 Which would you prefer? 00:33:56 <|amethyst> I guess ++ would be interesting, too 00:34:04 <|amethyst> better than a ring, but can still benefit from one 00:39:40 Where is the uselessness of items determined? 00:40:05 -!- ainsophyao has joined ##crawl-dev 00:41:12 <|amethyst> bmh: itemname.cc, is_useless_item() 00:41:19 -!- dtsund has joined ##crawl-dev 00:41:31 ty 00:42:34 <|amethyst> also, I presume undead keep their immunity? 00:42:59 |amethyst: I haven't touched that code, though I haven't come across the code that immunizes mummies. 00:46:35 ah -- mutationally 00:47:49 <|amethyst> bmh: ouch.cc, check_your_resists() handles player poisoning from poison arrow, including undead immunity 00:48:08 <|amethyst> bmh: for other things it's just from their rPois it seems 00:48:34 -!- ainsophyao has quit [Remote host closed the connection] 00:53:18 -!- syraine has quit [Ping timeout: 258 seconds] 00:58:40 |amethyst: do you think lichform should still give 100% protection? 00:59:15 -!- Brannock has quit [Read error: Connection reset by peer] 01:03:39 alright. I think I have a patch. 01:04:43 Ghouls, mummies, lich form and statue remain 100% poison immune. Other forms grant +1 rP, except for Spider Form which is -rP. Naga are rP++ 01:04:44 -!- raskol has quit [Read error: Connection reset by peer] 01:04:59 -!- raskol has joined ##crawl-dev 01:05:06 -!- syraine has joined ##crawl-dev 01:08:39 -!- ainsophyao has joined ##crawl-dev 01:14:54 Should it be possible to poison a mummy (or ghoul) who has transformed into a spider? 01:15:26 <|amethyst> they can't transform 01:15:39 That settles things 01:15:41 <|amethyst> unless I'm forgetting some weirdness 01:21:09 -!- monqy has quit [Quit: hello] 01:24:22 <|amethyst> hm.. a few other things I thought about: 1. vampire rPois should probably be a sliding scale instead of just one point; does Bloodless confer immunity? 01:24:48 <|amethyst> 2. do you get meph immunity with just rP+? 01:24:54 |amethyst: good point -- it ought to and cancel "being poisoned" 01:25:09 forgot to test meph. The patch is pretty big already. 01:25:27 <|amethyst> 3. do you get half poisoning from poison arrow at rP+ or is that a sliding scale too? 01:27:03 time for me to go to sleep, I'll poke at it more once there's wider feedback. 01:27:13 -!- bmh has quit [Quit: bmh] 01:28:30 -!- Wensley has quit [Ping timeout: 260 seconds] 01:30:33 Non-binary Poison (https://crawl.develz.org/mantis/view.php?id=4742) by brendan 01:50:41 -!- a5tp_ has joined ##crawl-dev 01:53:39 -!- a5tp has quit [Remote host closed the connection] 01:54:49 -!- edlothiol has joined ##crawl-dev 02:04:30 -!- raskol has quit [Read error: Connection reset by peer] 02:05:03 -!- raskol has joined ##crawl-dev 02:09:43 -!- Wensley has joined ##crawl-dev 02:20:24 -!- Vandal has quit [] 02:31:57 -!- syraine has quit [Ping timeout: 252 seconds] 02:33:03 -!- dtsund has quit [Ping timeout: 252 seconds] 02:34:03 -!- ixtli has quit [Quit: Computer has gone to sleep.] 02:37:56 -!- edlothiol has quit [Ping timeout: 256 seconds] 02:41:21 -!- ainsophyao has quit [Remote host closed the connection] 02:45:29 Unstable branch on crawl.akrasiac.org updated to: 0.10-a0-1282-g71db0c8 (32) 02:51:00 -!- syraine has joined ##crawl-dev 03:24:45 -!- neunon has quit [Read error: Connection reset by peer] 03:26:33 -!- neunon has joined ##crawl-dev 03:51:22 -!- Keskitalo has joined ##crawl-dev 03:52:42 -!- ais523 has joined ##crawl-dev 04:06:02 -!- evilmike has quit [Quit: something happened] 04:34:00 -!- ais523 has quit [Remote host closed the connection] 05:24:48 -!- ais523 has joined ##crawl-dev 05:32:04 -!- paxed has quit [Ping timeout: 260 seconds] 05:33:58 -!- paxed has joined ##crawl-dev 05:34:08 SamB: beh... tmpfile() being broken on Vista/7 in such a ridiculous way makes me wonder why anyone even bothers using them... 05:34:34 but judging from the fact no one noticed this build failure until now, indeed no one bothers :p 05:37:07 -!- alefury has joined ##crawl-dev 05:38:11 also, interesting how Microsoft advertises their tmpfile_s() (which is broken just the same) as "security improvements" when the only difference I notice actually reduces security (to an infinitessimally small degree though) 05:38:40 how is it broken? 05:39:27 kilobyte, yeah, microsoft's _s functions serve mostly the purpose of diverging APIs and making software annoying to port ;p 05:39:48 and spreading FUD about standards 05:40:27 Zaba: tmpfile() and tmpfile_s() create the file in C:\ which you can't write to without elevation on Vista/7 05:40:55 oh fail 05:40:58 (or XP as non-admin, but no one ever does that) 05:41:14 a fix in their libc is pretty damn obvious... 05:41:20 doesn't windows have a dedicated temp directory? 05:41:48 kilobyte: the _s stuff is just all about avoiding buffer overflows 05:41:58 Zaba: one per user, or one per system, depending on settings 05:42:01 nowadays 05:42:31 (Linux typically has two dedicated temp directories, /tmp that gets cleaned out every reboot, and /var/tmp that's cleaned out manually; /tmp is far more commonly used) 05:42:41 ais523: what exactly buffer overflow is possible in tmpfile()? 05:43:20 passing it too small a buffer to hold the resulting filename 05:44:01 the best part is where they tell you to use tmpnam() or tempnam() instead -- but these two functions are something that POSIX depreciates as insecure :p 05:44:25 ais523: FILE *tmpfile(void); -- no buffer 05:44:30 ah right 05:44:35 so what argument does tmpfile_s take? 05:44:55 errno_t tmpfile(FILE** pFilePtr); 05:45:07 the correct secure versions are mkstemp and mkdtemp, incidentally 05:45:18 kilobyte: how bizarre 05:45:31 presumably the idea's to prevent a null pointer exception if the file couldn't be created and the user doesn't check for that? 05:46:33 code with tmpfile_s() will crash just the same if it doesn't check the return value 05:46:46 the idea is to spread the "standard APIs are inherently insecure somehow" FUD 05:46:59 and to make the code vendor-locked-in 05:48:00 the _s wrappers would be easy enough to implement in glibc or whatever anyway 05:48:04 if they ever did become popular 05:48:09 they're just not a very good solution to their problem 05:48:12 of course just juggling the return value and the error code slightly differently suddenly makes code more secure. Not. 05:48:52 ais523: what's wrong with tmpfile()? Seems to be just as secure as mkstemp(). 05:49:12 it writes to stdout on error (!) 05:49:46 ah heh, silly 05:50:45 also, apparently tmpfile's file permissions are a little wonky; mkstemp's are too, sometimes, but at least its man page warns of the problem 05:50:57 but actually useful in most cases -- you have one less thing to do when handling this very rare error. Of course, having control would be better still. 05:51:34 it's not much of an issue to just call perror and exit 05:52:28 err(3) and friends are good for error reporting. Unfortunately they're not standard. 05:52:48 ais523: file permissions are not relevant actually, since the file is never actually linked 05:53:23 kilobyte: /proc/whatever/fd/n 05:53:26 not doable on Windows though as it has no concept of a deleted but opened file 05:53:39 deleted but opened files are nonetheless accessible via the filesystem 05:54:11 lrwx------ 1 kilobyte kilobyte 64 Oct 14 12:53 3 -> /tmp/tmpf57cUl5 (deleted) 05:54:44 hmm, those permissions look correct 05:55:12 (deleted files show up as symlinks, but they act both like symlinks and like the original file) 05:55:21 ais523: could you tell me how? In ways other than reading the physical disk, that is. 05:55:38 kilobyte: like you just demonstrated 05:55:42 you can cat that file, for instance 05:55:45 even though it's a broken symlink 05:56:02 yeah of course, it's a still working descriptor 05:56:22 if you can control the process, you can mess with it even without /proc/ 06:00:48 -!- ais523 has quit [Ping timeout: 245 seconds] 06:01:30 -!- paxed has quit [Ping timeout: 260 seconds] 06:02:57 -!- Pacra has quit [Quit: This computer has gone to sleep] 06:03:37 -!- paxed has joined ##crawl-dev 06:04:55 -!- ais523 has joined ##crawl-dev 06:18:07 -!- jeanjacques has joined ##crawl-dev 06:31:17 -!- paxed has quit [Ping timeout: 260 seconds] 06:33:20 -!- paxed has joined ##crawl-dev 06:37:00 -!- ais523 has quit [Read error: Connection reset by peer] 06:39:10 -!- ais523 has joined ##crawl-dev 06:42:42 -!- dpeg has joined ##crawl-dev 06:42:45 cheers 06:43:12 -!- Wensley has quit [Ping timeout: 248 seconds] 06:43:21 -!- ais523_ has joined ##crawl-dev 06:43:35 -!- ais523 has quit [Disconnected by services] 06:43:38 -!- ais523_ is now known as ais523 06:44:38 -!- syraine has quit [Ping timeout: 276 seconds] 06:52:10 -!- edlothiol has joined ##crawl-dev 06:52:14 -!- ais523_ has joined ##crawl-dev 06:52:20 -!- ais523 has quit [Read error: Connection reset by peer] 06:52:26 -!- ais523_ is now known as ais523 06:58:37 -!- st_ has joined ##crawl-dev 07:01:08 -!- Torokasi has joined ##crawl-dev 07:02:48 -!- alefury has quit [Quit: ChatZilla 0.9.85 [Iceape 2.0.11/20110701115916]] 07:04:31 -!- raskol has quit [Read error: Connection reset by peer] 07:04:58 -!- raskol has joined ##crawl-dev 07:11:36 -!- Mu_ has joined ##crawl-dev 07:12:18 -!- MarvinPA_ is now known as MarvinPA 07:22:45 -!- syllogism has joined ##crawl-dev 07:24:19 -!- dtsund has joined ##crawl-dev 07:50:03 -!- casmith789 has quit [Read error: Connection reset by peer] 07:53:27 -!- alefury has joined ##crawl-dev 07:54:00 -!- ais523 has quit [Ping timeout: 260 seconds] 07:55:30 moin dpeg! 07:56:46 -!- Pingas has joined ##crawl-dev 08:03:11 -!- RichardHawk has joined ##crawl-dev 08:04:24 -!- raskol has quit [Read error: Connection reset by peer] 08:06:05 Uhh, piety drain from Ely protection your allies is brutal. 08:06:26 At least I think that's why I suddenly have no piety.. 08:07:07 Or maybe I got penance for some reason. 08:07:57 Nope, it was the protection. 08:08:38 ..or perhaps allies dying, heh. 08:13:21 Keskitalo: using Elyvilon with summons? 08:13:37 Trying it out, yeah. 08:13:45 what summons? 08:14:47 Canines. Maybe I should try if ice beasts, if they're more durable. 08:16:13 yes, the bigger the better 08:16:21 hydras will be ideal due to short duration 08:20:27 -!- raskol has joined ##crawl-dev 08:25:25 Ice beasts seem to do nicely. 08:25:54 cool 08:46:51 -!- dpeg has quit [Quit: Lost terminal] 08:52:03 -!- Keskitalo has quit [Quit: leaving] 09:03:19 -!- mikee_ has quit [Quit: HydraIRC -> http://www.hydrairc.com <- Would you like to know more?] 09:03:33 -!- mikee_ has joined ##crawl-dev 09:04:23 -!- raskol has quit [Read error: Connection reset by peer] 09:11:31 kilobyte: I must be "no one" ;-P 09:11:47 kilobyte: I *am* on XP 09:17:05 kilobyte: and evidently there's a flag you can pass to MSVCRT's open() to get it to delete the file when you're done with it... 09:17:18 -!- ainsophyao has joined ##crawl-dev 09:20:32 -!- raskol has joined ##crawl-dev 09:30:51 -!- Brannock has joined ##crawl-dev 09:42:43 SamB: Do you have access to online saves? 4741 is probably just from before the fix but I can't be sure. 09:43:44 Fyren: No, I don't know a username/password to get them :-() 09:43:54 whoops 09:43:56 :-( 09:45:21 -!- dtsund has quit [Ping timeout: 252 seconds] 09:45:21 is there a way to enable asserts but not all the debug spew? 09:46:06 I think it uses Mantis login info, but I don't know if it's automagically hooked up or what. 09:46:48 Or if being on CAO is different. 09:49:01 but, anyway, it was reported last night ... 09:49:17 with version 0.10-a0-1274-gddaaf60 09:49:41 was that not after the fix? 09:51:33 yes, it was 09:51:42 I just think I must have missed something 09:52:58 aren't asserts always enabled? 09:53:10 no :-( 09:53:43 some of them probably should not be; there are some that get triggered if I so much as resize the terminal 09:53:57 from the makefile: 09:54:03 ASSERTS -- set to enable assertion checks (implied in debug mode) 09:54:08 ah 09:54:27 I guess I should add that to debug-lite too 09:54:34 and my compile.sh 09:55:11 yeah 09:55:14 now, what should it look like when EW both increases plusses and removes a curse? 09:55:51 should it be a "silvery" red/green/whatever glow, or not? 09:58:49 * SamB tries it in 0.8, since he happens to have that around 09:58:59 currently, it's simply a green glow, isn't it? 09:59:37 tested it already, but I wasn't sure if you were asking about current behavior or if it should be changed 09:59:40 it's not working right now anyway, so that's hardly a good basis for this! 10:00:23 since it *did* work in 0.8, I think that's a better basis for determining what it should look like after the fix :-) 10:00:38 _Your cursed +0,+0 mace glows green for a moment. 10:00:40 yeah 10:01:00 that's 0.8.1 10:01:10 although it does still work in trunk, just not for every weapon type it seems? 10:01:28 I think the code as it is will only uncurse staves, heh. 10:01:32 at least it worked when I just tested it on a mace 10:01:34 I think it only works when the plusses aren't increased 10:01:44 so it will work on artifacts and staves 10:01:50 and if the "enchant" part fails 10:02:39 although my trunk is a bit out of date, maybe it was broken since then 10:02:52 how out of date? 10:03:26 51675736824b 10:03:48 0.8.1 checked for uncurse first, and if it did uncurse, didn't enchant. 10:03:56 When it uncursed, it glowed silver. 10:04:12 Fyren: oh 10:04:17 (that's october 9, not as much out of date as I thought) 10:04:23 But the new code looks like it's trying to uncurse and enchant because of the color composition in the formatting. 10:04:23 -!- raskol has quit [Read error: Connection reset by peer] 10:04:30 Fyren: that's not what I'm seeing, though! 10:04:39 -!- raskol has joined ##crawl-dev 10:04:45 I'm just looking at 0.8.1's code, I didn't recompile. 10:05:32 I have it running, and that read of "enchant weapon I" did this, too: 10:05:40 Wp: a) +1,+0 mace 10:06:25 Guess I'm misreading. 10:06:26 oh, yeah, it got changed since the commit I have 10:06:47 doesn't the do_uncurse_item call just have to be moved out of the if block? 10:06:58 Yeah. 10:07:02 That's what I'm doing. 10:07:39 I wanted to get the glow right, though, so I wasn't sure until I'd tested in an old version :-) 10:07:50 ah, ok 10:08:17 SamB: I'm pretty sure the intention is for it to be silvery green on enchant plus uncurse in the new one. 10:08:22 Even if it wasn't like that before. 10:08:23 yeah 10:08:42 Fyren: hmm. That would be complicated! 10:09:01 Let me commit this first, in any case 10:09:13 Just split success into enchanted and uncursed bools. 10:09:56 I don't mean it would be *very* complicated 10:10:26 just more complicated than moving the do_uncurse_item() call outside the if 10:12:11 also, if that was the intent, why is the code ordered like this? 10:12:57 kilobyte: what kind of glow is desired for EW when it removes curse *and* enchants? 10:13:37 SamB: Well, as we've seen, the new code is all sorts of not working for basically every case possible. 10:13:49 not QUITE 10:14:03 it worked correctly for artifact weapons ;-P 10:14:08 But if it wasn't supposed to compose the colors why would it format to "silvery %s"? 10:14:32 -!- ghallberg has left ##crawl-dev 10:14:36 oh, it's supposed to compose them alright 10:14:41 -!- ghallberg has joined ##crawl-dev 10:14:42 you're right, it will still show the "silvery green" when it uncurses and doesn't enchant 10:14:53 so it might be intended to just be "green" if it enchants 10:15:25 Just green, just silver, or silvery green for both. 10:15:31 Or red for EW2. 10:16:15 that would be best 10:16:16 The "silvery green/red" glows are related to the self-ID behaviour 10:16:32 I think 10:16:59 Don't think so, since they're inside if (wpn.cursed()) 10:17:01 it got changed in that commit, but I don't see how it would be otherwise connected 10:17:06 that is, they are supposed to provide the explanation for how you ID it as I/II/III 10:17:08 The "briefly gains a %s sheen" is self-ID. 10:17:23 well, that too 10:17:30 ah, yes 10:17:43 depending on whether or not there was a curse to remove 10:18:00 If there was a curse then I guess it still needs to be silvery green/red, though, yes. 10:18:03 For self-ID. 10:18:15 So silvery green even if it can't enchant? 10:18:21 As long as it uncurses. 10:19:44 Well, I think I'm going to leave it so it's "silvery" only when enchantment fails 10:20:24 Then from the in-character view you can't distinguish I/II. 10:20:44 I meant "silvery green/red", sorry 10:20:56 I agree with that, then. 10:21:16 -!- bmh has joined ##crawl-dev 10:21:21 -!- bmh has quit [Changing host] 10:21:21 -!- bmh has joined ##crawl-dev 10:21:29 hi 10:22:05 this way, you'll be able to tell for sure whether the enchantment part succeeded or not, even if you don't know the plusses yet 10:22:29 best would be if it was just "silver" if the enchantment doesn't succeed 10:22:39 no, no! 10:22:52 that doesn't explain the auto-ID behaviour 10:23:12 ah, ok ;) 10:23:20 could make it "reddish silver" :P 10:23:53 I don't want to have to explain to the computer how to append "ish" to color names ;-P 10:24:27 <|amethyst> silver with a touch of %s 10:24:37 <|amethyst> maybe not :) 10:24:58 silver with a %s sheen 10:25:15 hmm 10:25:17 <|amethyst> "green sheen" sounds funny though 10:25:26 that's true 10:25:34 it doesn't really matter anyway ;) 10:25:39 but we get that anyway 10:25:54 with the "nothing happened" path 10:25:56 elliptic and kilobyte do protest multi-dot rP. Anyone have input on the subject? 10:26:03 <|amethyst> oh, you're right 10:27:05 Like "your plain deck of escape {god-gift} very briefly gains a green sheen" ;-) 10:27:41 note: I typed that in manually, so it's guaranteed to be slightly off 10:28:06 <|amethyst> "lustre" maybe? 10:28:22 I think it's fine, myself 10:28:35 Bueller? 10:30:25 * SamB makes a dagger of reaching 10:35:48 okay, good 10:36:06 * SamB checked that EW doesn't uncurse, say, wielded armour 10:37:30 I wonder why Tk doesn't support using GTK and/or Qt widgets/themes 10:38:33 I mean, on Windows it looks native, why does it have to look so unique on X11? 10:38:35 03SamB * r59dd7118197c 10/crawl-ref/source/item_use.cc: Make "enchant weapon" uncurse even after successful enchantment (#4741). 10:39:26 -!- ais523 has joined ##crawl-dev 10:45:18 -!- ais523_ has joined ##crawl-dev 10:45:19 -!- ais523 has quit [Remote host closed the connection] 10:46:33 -!- ais523_ is now known as ais523 10:54:21 -!- blabber has joined ##crawl-dev 10:55:49 bmh: I agree with elliptic. You don't really need rPois right now, so the change is weird in that light. 10:56:20 Like, rF, rC, rElec reduce damage a whole lot in circumstances you will always run up against. 10:56:55 You can usually just ignore the poison or come back later or something without rPois. 10:58:25 i'm also not keen on 3-dot rPois 10:59:34 MarvinPA, Fyren: that's fine. It's really easy to change 10:59:40 one dot not providing 100% protection still seems best, although maybe it introduces some awkwardness and/or consistency issues 10:59:58 -!- blabber_ has joined ##crawl-dev 11:00:04 marvinpa: well, rElec has one dot and doesn't provide 100% protection 11:00:09 Would it be too inconsistent to base the level of protection on XL? 11:00:11 right, that'd be fine 11:00:15 bmh: definitely 11:00:20 i just mean like, if player nagas had only partial protection 11:00:22 it'd be strange 11:01:07 -!- blabber has quit [Ping timeout: 252 seconds] 11:01:10 given that they are already poisonable via poison arrow it doesn't bother me too much 11:01:14 MarvinPA: We could have two tier rP where only mummies, ghouls, and naga are in the second tier 11:01:33 nagas aren't undead :P 11:01:40 well it'd have to be three-tier, with undead on the third tier (full immunity) 11:01:41 -!- ais523 has quit [Ping timeout: 256 seconds] 11:01:59 -!- blabber_ is now known as blabber 11:02:10 and nagas on the second, but maybe that's not even necessary and just having "regular rpois/undead rpois" for players would be fine 11:02:48 perfect rpois doesn't seem necessary for player nagas 11:03:10 poisonous snakes aren't resistant to other species' venom 11:03:18 @??snake 11:03:19 snake (09S) | Speed: 13 (swim: 60%) | HD: 2 | Health: 6-16 | AC/EV: 1/15 | Damage: 508(poison) | Flags: amphibious, cold-blooded | Res: 06magic(8), 12drown | XP: 13. 11:03:31 @??black mamba 11:03:31 black mamba (02S) | Speed: 18 | HD: 7 | Health: 24-52 | AC/EV: 4/15 | Damage: 2004(medium poison) | Flags: cold-blooded | Res: 06magic(28), 03poison | Chunks: 09poisonous | XP: 456. 11:03:46 i'm relating to real life right now 11:04:21 -!- raskol has quit [Read error: Connection reset by peer] 11:04:42 -!- raskol has joined ##crawl-dev 11:05:15 -!- monqy has joined ##crawl-dev 11:05:19 elliptic, MarvinPA: Do you think that Nagas should get any benefit from a ring of rP? 11:05:23 no 11:09:12 Ok -- so how do we set the sub 100% threshhold? 11:09:25 trial and error 11:09:48 what do you mean by "sub 100% threshold" anyway? 11:10:01 If rP isn't 100% effective, how effective is it? 11:10:44 Does it normally prevent poisoning, or just reduce the damage incurred? 11:11:07 SamB: as implemented it prevents poisoning. I didn't feel like dealing with the case where someone has been poisoned and slaps on a ring of rP 11:11:26 okay, that's fine 11:11:34 If you want to do trial and error, we could generate a new rP effectiveness for each game and hide it from the player. Then we could look at the morgues and pick the poison death rate that makes us happiest :) 11:11:57 I meant, like, try 90% and see how things go ;-P 11:12:27 so how is poisonedness measured, anyway? 11:12:58 it sets a duration that counts down. If the duration is high you're "very poisoned" and take extra damage 11:13:06 It's fairly braindead. 11:13:54 so how high are these numbers? 11:14:37 The counter is capped at 40. 11:16:40 40 doesn't mean anything like 40 turns though 11:16:48 If we were going to overhaul poison more, I'd advocate for a two-axis system with duration and severity. That way if you're stung by a killer bee and then bitten by a little snake the duration would increase, but not the severity. 11:16:57 elliptic: grepping the source, one moment please. 11:18:36 ok, if the counter is above 5, you're guaranteed to get some damage. If it's above 10, you get rand2(10) + 5 11:18:54 (er, you aren't guaranteed above 5) 11:19:12 above 5, you have a 25% chance of 3 damage, and a 25% chance of 2 damage 11:19:49 it is rather rare that you get poisoned that much 11:20:49 maybe rPois should just be 50%, then 11:21:20 * bmh would set rPois as sigmoid(1 + log(XL))... 11:22:07 it shouldn't have anything to do with XL. use MHP if you want, but not XL 11:22:09 whenever the player gets poisoned, just use the randomized division rounding thingy to figure out how much to add 11:22:38 snake pit will get to join swamp and shoals as annoying and tedious branches 11:24:45 elliptic: MHP? 11:24:49 max hp 11:25:13 or use regen rate 11:25:21 those would both make much more sense than XL 11:25:28 What's the scale on regen rate? 11:26:40 ??regeneration[2] 11:26:41 regeneration[2/5]: The base regeneration rate is your maximum hitpoints / 3. If this number is over 20, then divide the result by two and add ten. For example, 100 max health results in a regen rate of 26. RR accumulates every turn and for every 100 RR you gain, you regenerate 1 hp. For 1 hp per turn, 540 health is needed. For 1 hp per 2 turns, 240. 11:26:46 mikee_: Is there something you'd prefer? "If it ain't broke, don't fix it"? 11:27:34 i never said that 11:27:47 I was asking :) 11:27:50 but i hope that's a side effect you've considered 11:28:15 hey, there is this thing called "revert" 11:28:27 we've demonstrated the ability to use it recently, I think 11:28:31 sorry, lately people have been telling me what i have said in a way that does not sound like what i said =P 11:28:58 but, see, bmh was asking :-) 11:29:10 SamB: What are question marks for? 11:29:31 bmh: those are scrolls 11:29:32 is there a problem with the comment i made 11:29:59 mikee_: no. As one of the best players around your input is very helpful 11:30:12 i am asking too 11:30:26 hmm, I think we need to give mikee_ more scrolls 11:30:43 SamB: plz. 11:30:44 so he can put them at the ends of his questions! 11:31:22 *hint hint* 11:31:39 anyway, everything else about the proposal sounds promising to me except for that 11:31:48 any luck with compilin tiles? :P 11:32:05 Mu_: I, unfortunately, did not hit your problem 11:32:23 I think I can tweak the formula such that rP at low levels won't completely neutralize poison threats while enabling beefy players to enjoy snake without tedium. 11:32:26 blah 11:32:34 I hit a moronic problem with MSVCRT's implementation of tmpfile() 11:32:51 Mu_: what version are you trying to build? 11:33:04 bmh: I don't see what's wrong with trying the simple 90% before trying exotic things 11:33:07 why does this need to scale with xl or maxhp or anything like that, anyway? 11:33:14 MarvinPA: it doesn't 11:33:25 most recent pull from master 11:33:27 marvinpa: it doesn't, I just thought anything would be better than xl 11:33:50 bmh, alright, thanks for considering my input 11:33:59 elliptic: I dislike arbitrary cutoffs 11:34:05 ???? 11:34:08 how is it a cutoff 11:34:19 it is a constant value, same as other resists 11:34:58 Mu_: does that include the most recent contribs? 11:35:16 elliptic: We have rF+,++,+++ though 11:35:31 bmh: and each one gives a constant amount of protection 11:35:34 what is your point 11:35:37 we also have rElec 11:35:40 and just one level of rElec 11:36:16 Evokable items do not train Evocations (https://crawl.develz.org/mantis/view.php?id=4743) by XuaXua 11:36:25 I think it comes down to a matter of preference. 11:36:36 my submodules seem up to date 11:36:51 -!- Torokasi has quit [Ping timeout: 255 seconds] 11:37:17 bmh: whether to have a resist that works completely differently from any other resist in the game comes down to preference? right 11:37:21 but no other resist acts anything like that, i don't really see why you'd even come up with it in the first place 11:37:22 C:\code\crawl\crawl-ref\source>make DEBUG=y TILES=y 11:37:22 which: advpng: unknown command 11:37:22 cat: util/release_ver: No such file or directory 11:37:22 * rebuilding crawl: new build flags or prefix 11:37:22 I prefer some consistency :P 11:37:22 * Need to build contribs: pcre freetype sdl-image sdl libpng zlib lua/src sqlite 11:37:29 Mu_: do you get anything like that? 11:38:23 bmh: if it can be set in the config file, there's not much point! 11:38:42 elliptic: I think that with a little added complexity we can strike a balance between neutralizing early poison threats and making snake tedious 11:39:03 bmh: don't worry about making snake tedious until it has been tested 11:39:19 elliptic: good point. I'll code up the 90% version, applied on a point-per-point basis 11:39:46 i get those two asterisk lines 11:40:16 Mu_: does the second one include all that stuff ? 11:40:44 yep 11:40:53 -!- blabber has quit [Quit: leaving] 11:40:57 and you're still getting that odd error building tilegen? 11:41:16 yep, exactly the same error 11:41:23 elliptic, MarvinPA, mikee_: Do any of you object to a rP- mutation? Or making spider form be rP-? 11:41:36 and you're sure you don't have any half-present installs of SDL or anything? 11:41:40 I don't really see the point 11:42:23 added complexity for fairly low gain 11:42:49 (you'd need to figure out things with how it stacks with rP sources, and I'm not sure what would make sense) 11:43:07 Mu_: what does your environment look like to start with? 11:43:28 Given that spiders are poison vulnerable, would it be OK to just make TRAN_SPIDER not confer rP? 11:43:29 what do you mean 11:43:46 that is, run "set > env.txt" and paste the contents of env.txt 11:43:58 -!- Pingas has quit [Read error: Connection reset by peer] 11:44:00 spider form doesn't currently give rP, does it? 11:44:06 It does 11:44:47 that's a gigantic file do you want anything particular 11:44:48 well, I'll agree that's strange :) 11:45:45 Mu_: I don't really know :-( 11:46:01 i'm pretty sure spider form doesn't give rpois 11:46:07 if there's anything involving SDL or include paths or FLAGS or anything like that, it might be relevant 11:46:27 i have no objections to that. it sounds kind of cool actually 11:46:30 player_res_poison only lists lich, ice beast, statue, and dragon transforms. 11:46:53 no references to any of that 11:46:55 Fyren: Oh, you have the advantage of sitting in front of an unmodified copy of the source :) 11:47:12 Mu_: then maybe I'd better see the whole thing 11:47:41 bmh: git diff (or gitk) can help with that ;-) 11:49:40 Since dec_poison_player only decrements the poison counter one in eight times, maybe rPois could just make that chance higher. 11:50:22 I guess that doesn't help make it be something people want to use. 11:50:32 fyren: problem is that we aren't sure whether swapping in rPois after you've been poisoned is a good thing 11:50:45 samb: http://pastebin.com/mDMEeMmP 11:50:51 elliptic: I believe that swapping it in should have no effect. 11:51:06 bmh: I generally agree, yeah 11:51:40 Could make removing it double the existing counter! 11:51:56 … !learn add bad_ideas …. 11:51:58 But if rPois goes from how it is now to 90%, people will just use it less. 11:52:36 I agree that making an already weak resistance a little worse won't accomplish much (except make snake a tiny bit more dangerous) 11:52:47 but making it non-100% is probably a good thing in the long run 11:53:59 wait did someone say spider form gives rpois? last I checked, it doesn't. 11:54:11 monqy: I was wrong. I was looking at my modified source 11:54:14 oh that had already been resolved oops 11:54:14 ahh 11:54:20 you made spider form give rpois? 11:54:30 no, I made it give rP- 11:54:33 ahh 11:54:49 I was just about to say rP- would be consistent with other bugs in crawl 11:55:15 rP- would make some sense, yes... but I still want to know how it would stack with rP 11:55:25 given that we don't have rElec- to compare with 11:55:52 like, do you get from rP- to rP+ with two rings of rP? 11:56:01 Mu_: oh, I see what you mean 11:56:10 I didn't realize you were doing this from bash! 11:56:13 that would make the most sense i think 11:56:56 it might actually be better if you ran "printenv > env.txt" and pasted *that*, then 11:57:11 I don't really want to see all these git-related shell functions ;-) 11:57:18 in a sense that would be worse than other vulnerabilities though since there are no 'rpois++' items 11:57:24 bash: printenv: command not found 11:57:24 it gets a little strange then because normally your second ring of rPois is completely useless... and now it isn't, but only very rarely 11:57:25 :P 11:57:30 oh 11:57:39 yeah 11:57:49 thanks for tryin to help me btw 11:58:04 sure, sure! 11:58:05 well realistically your second ring of rpois would still not be that useful 11:58:37 -!- Pingas has joined ##crawl-dev 11:58:41 you can probably even win with rpois- without much difficulty 11:59:26 yeah 11:59:45 Mu_: I don't suppose you have things set up so you can attempt a build from cmd.exe ? 12:00:51 nope 12:01:19 Hmm, it mostly seems to be a matter of setting the PATH appropriately 12:02:10 also, about having it as a mutation... I don't like having a mutation that simply cancels out a common item property 12:02:54 it might be okay if it were impossible to get rP+ protection if you had rP- mutation, but otherwise it doesn't seem worth it to me 12:03:44 Is you.is_undead true for wet vampires? 12:03:54 Mu_: can you just try adding your MSYS and MinGW bin directories at the front of the path before running make in cmd.exe ? 12:04:04 (in that order)? 12:04:19 -!- raskol has quit [Read error: Connection reset by peer] 12:04:39 if you tell me the directories, I could give you the command 12:04:46 -!- raskol has joined ##crawl-dev 12:05:33 elliptic, i was thinking it would create (for some people) the same mentality with mummies and rF- 12:05:56 C:\msysgit\msysgit\ & C:\msysgit\msysgit\mingw i guess? 12:07:33 mikee_: it feels different to me because of how much harder it is to get rF++++ than rP++ and also how much less important rP is 12:08:16 I'm not hugely against the idea, I just think it would be better if set up in such a way that the mutation mattered even if you happened to have rP+(+) from equipment 12:08:30 i.e. make it stronger than rP+ 12:09:18 huh, vampires currently get a bonus against poison arrow regardless of satiation. 12:09:47 really? what sort of bonus? 12:10:08 They don't get poisoned by it! 12:10:29 -!- casmith789 has joined ##crawl-dev 12:10:34 else if (!you.is_undead && doEffects) poison_player(2 + random2(3), source, kaux, true); 12:10:57 yeah, IMO bug 12:11:33 agreed. Perhaps we should encapsulate that sort of check into something like "you.is_currently_undead()"? 12:12:05 elliptic, sure, those are reasonable points 12:12:21 I'm pretty sure poison_player() won't actually poison non-poisonable undead anyway 12:12:32 so that check could probably just be removed 12:13:11 having such a function might be useful elsewhere though for sure (it should also handle lichform, of course) 12:24:19 Mu_: they don't have bin dirs? 12:24:41 Mu_: oh, I know; what does "mount" say? 12:24:50 oh i missed that part. yes, /bin in those dirs 12:25:50 -!- blabber has joined ##crawl-dev 12:26:02 set PATH=C:\msysgit\msysgit\bin;C:\msysgit\msysgit\mingw\bin;%PATH% 12:26:35 run that at the cmd.exe prompt 12:26:49 then run "make TILES=y" 12:26:58 ok.. 12:27:22 exactly the same 12:28:04 well, run "set > env.txt" there and paste *that* ... hopefully it will be a lot smaller 12:29:31 http://pastebin.com/TvY20zdu 12:35:17 Mu_: so it's probably not because of a stray environment variable... 12:36:05 -!- st__ has joined ##crawl-dev 12:36:25 -!- st__ has quit [Read error: Connection reset by peer] 12:36:39 -!- st__ has joined ##crawl-dev 12:37:34 Mu_: the only thing that I hesitate over is that QT thing 12:37:46 -!- st_ has quit [Ping timeout: 255 seconds] 12:39:15 compile faster! 12:39:21 it's a movie player, don't remember why it's in that path thing 12:39:36 probably command line video editing 12:40:08 oh, nevermind 12:40:21 that doesn't worry me at all 12:40:25 ;p 12:40:45 I thought it might be related to trolltech's toolkit ;-) 12:41:42 -!- st__ is now known as st_ 12:41:52 -!- dtsund has joined ##crawl-dev 12:53:35 -!- dtsund has quit [Ping timeout: 258 seconds] 12:56:47 -!- jeanjacques has quit [] 13:04:20 -!- raskol has quit [Read error: Connection reset by peer] 13:04:49 -!- raskol has joined ##crawl-dev 13:06:15 Should entering lich form or becoming a dry vampire completely purge poison from your system or should it cause you to suffer no damage? 13:08:05 (removing it entirely so people don't just sit around slapping 5) 13:11:01 -!- Keskitalo has joined ##crawl-dev 13:28:11 bmh: i would say: your metabolism stops entirely; the poison kicks back in as soon as you become alive again, at the same intensity 13:31:15 -!- Torokasi has joined ##crawl-dev 13:31:26 Eronarn: I will make it so. 13:31:37 * bmh adds ::are_currently_undead() to you. 13:31:50 -!- evilmike has joined ##crawl-dev 13:32:41 bmh: this might be something to look at with vamps... how they gain rP, i mean 13:33:50 maybe make them not rPois at all, and simply at some point their heart stops beating? they are the only undead this issue comes up for 13:34:37 Eronarn: poison could do hunger damage to them? 13:39:53 bmh: i like the idea of poison encouraging them to shut their body down - while they're regenerating, they are taking poison damage 13:42:06 Eronarn: if poison doesn't decay in dry vampires, they risk dying of poison when they bite feed in melee. 13:43:28 * SamB wishes he could configure git to use different URLs for push and fetch with the same remote 13:43:53 do you ever leave bloodless just from melee as a vampire...? 13:43:53 evilmike: You have 2 messages. Use !messages to read them. 13:44:02 I've never had that happen. I suppose it's possible in theory 13:45:05 -!- dtsund has joined ##crawl-dev 13:46:26 It would be an embarrassing edge case. 13:47:45 bmh: that's the point, though - they cannot be killed by the poison, but also cannot easily cure it while starving 13:47:49 because no potions 13:48:00 (or can vamps drink pots at starving? it's been a while) 13:48:29 i guess probably blood potions, but not others...? 13:49:03 They drink potions for lesser effect. 13:49:49 Fyren: Are you a bot who answers natural language queries about the source? :) 13:50:15 I dunno, I've just lurked in here for three months and know where some stuff in the source is. 13:50:22 Maybe it's just my programming to say that. 13:50:32 (Thanks!) 13:51:02 if you want to impress people, make a lot of references to fight.cc 13:51:14 &fight.cc 13:51:33 not pointers 13:51:34 references 13:51:42 fight.cc& 13:51:52 make a lot of allusions 13:51:55 like "see foo_bar() in fighter.cc" 13:51:55 :P 13:52:08 er. fight.cc 13:52:39 is it reduced effect at hungry, or at actually bloodless, too? 13:52:47 mikee_: i've read all of fight.cc you know 13:52:52 and i even basically understand how it works 13:53:01 Eronarn, i am impressed 13:53:02 Eronarn: Reduced effect under satiated. 13:53:24 as I have the world right now, bloodless vampires do not become unpoisoned, which will lead to feeding problems 13:53:28 impressing me is easy though 13:53:49 bmh: well, think of it like so: vampires would currently, if they'd been poisoned and bloodless somehow, just plain died 13:53:58 now they will not die until they choose to be non-bloodless 13:54:23 Eronarn: don't forget about auto-biting in melee 13:54:42 bmh: doesn't give enough nutrition to set you above bloodless for long 13:54:43 is auto-explore supposed to suppress so many messages you would see if you were exploring manually? 13:55:03 like, whenever you find something auto-pickupable? 13:55:03 if someone takes a tick of poison because they didn't realize they were near the limit of bloodless, is that really a big deal? 13:55:29 It is if it happens to kill them. 13:55:53 it's such a marginal case though, especially since people rarely melee while bloodless 13:57:30 -!- Twinge has quit [] 14:00:21 -!- Pacra has joined ##crawl-dev 14:01:24 SamB: Suppress and in not show? No. But it shouldn't always stop exploration. 14:01:47 Fyren: okay, because it's been doing that since at least 0.9... 14:02:51 SamB: Seems fine for me. Autoexplore picked up a bread ration and kept going (because I already had one, I guess, don't remember the rules). 14:02:55 The log shows I picked it up. 14:03:38 Or are you talking about something else? 14:03:53 Fyren: I mean, it misses the "You see a" or "You found a" or whatever line 14:04:15 -!- mikee_ has quit [Quit: HydraIRC -> http://www.hydrairc.com <- *I* use it, so it must be good!] 14:04:20 -!- raskol has quit [Read error: Connection reset by peer] 14:04:33 -!- bhaak is now known as ais524 14:04:44 -!- ais524 is now known as bhaak 14:04:53 -!- raskol has joined ##crawl-dev 14:08:12 03Raphael * rccf3b0ce4fea 10/crawl-ref/source/ (item_use.cc items.cc): Merge branch 'master' of ssh://gitorious.org/crawl/crawl 14:08:20 03Raphael * rdb317b7b52fb 10/crawl-ref/source/skill_menu.cc: Fix inverted colours for boost/penalty in the skill screen. 14:09:22 the what ? 14:10:48 Think it means he committed locally before pulling the last EW change? 14:10:52 So he had to merge. 14:10:52 ok. I think I have a working patch with 90% poison reduction. 14:12:04 * SamB thought we were supposed to use "git pull --rebase" 14:12:45 there are lots of neat tricks you can do 14:20:19 just looks like galehar lost his git configuration or something, i guess 14:20:50 and his email address too? 14:21:12 yeah 14:21:15 apparently! :o 14:22:03 oh, wait, maybe that was on Windows or something 14:22:16 actually, who knows! 14:22:32 what kind of system doesn't have a hostname for itself??? 14:22:47 (his cell phone?) 14:22:52 Maybe he's doing it with his mind. 14:23:59 I'd think his mind would be better-configured than that! 14:24:03 funny that 14:24:08 i named my laptop "mind" 14:28:07 03SamB * r7959c5028065 10/crawl-ref/source/rltiles/ (Makefile tool/tmpfile.cc): tilegen: MSVCRT's tmpfile() is broken; use our own instead. 14:28:37 * SamB wonders how to configure git to use plink instead of OpenSSH 14:38:27 * SamB wishes MSYS processes weren't so secretive about their command-line arguments 14:41:53 -!- Keskitalo has quit [Quit: leaving] 14:43:02 SamB: Looking at travel.cc, it seems intentional it doesn't print "found" messages for autopickups. 14:43:35 -!- Torokasi has quit [Ping timeout: 260 seconds] 14:43:42 It only prints them when it decides autoexplore should stop. 14:44:42 Fyren: I don't much like it :-( 14:48:23 kilobyte: why did you sneak a change to the unwield delay into my patch? 14:51:38 Eronarn, Fyren: https://crawl.develz.org/mantis/view.php?id=4742 <-- new patch. 14:54:47 -!- dtsund has quit [Quit: dtsund] 14:57:04 -!- Pacra has quit [Quit: This computer has gone to sleep] 15:02:51 -!- Torokasi has joined ##crawl-dev 15:04:18 -!- raskol has quit [Read error: Connection reset by peer] 15:12:14 SamB: Why plink? 15:12:30 ghallberg: well, it works nicely with pageant 15:12:55 mmk 15:13:10 which is a heck of a lot easier to use than ssh-agent 15:13:21 otherwise I really wouldn't care 15:13:35 You're on windows? 15:16:19 -!- Twinge has joined ##crawl-dev 15:16:20 -!- raskol has joined ##crawl-dev 15:17:38 it depends 15:17:59 mostly I do not hack there 15:18:23 but I was trying to help Mu_ out with a problem building tiles with MinGW... 15:18:58 -!- Torokasi has quit [Ping timeout: 260 seconds] 15:19:56 ok 15:20:25 unfortunately, I ran into a *different* problem 15:21:08 which is that MSVCRT's implementation of tmpfile() is totally moronic for a multi-user operating system 15:22:34 it tries to write files into the *root directory*! 15:27:17 apparantly, setting GIT_SSH=plink is enough when plink is in the PATH already 15:27:32 the git manpages did not make that at all clear 15:35:53 <|amethyst> SamB: is the race condition with tempnam() an issue on Windows? 15:36:32 |amethyst: yeah 15:36:35 er. no 15:36:43 I somehow thought you said "non-issue" 15:36:54 <|amethyst> the temp directory is guaranteed to be not writable by others? 15:37:10 not quite, but if it is writable by others that's your own fault 15:37:30 the default is that it is inside your profile directory 15:38:06 <|amethyst> and I guess on older OSes it's no worse than using \ 15:38:15 yeah 15:38:39 the *real* problem was that it just plain doesn't work for non-admins, though 15:38:48 Who is the mantis admin? 15:38:57 bmh: probably Napkin 15:39:05 kilobyte is a "manager" though 15:43:30 <|amethyst> SamB: hm... is tmpfile.o needed if you're building with mingw, or only VC++? 15:43:39 both! 15:43:52 <|amethyst> ah, mingw uses the system tmpfile? 15:44:00 <|amethyst> I wasn't expecting that for some reason 15:44:11 libmingw is a thin wrapper around MSVCRT 15:44:15 <|amethyst> aha 15:44:18 didn't you know? 15:44:40 that's why it's called "minimalistic" 15:45:14 <|amethyst> I'm not a windows dev... I had figured it was most of a libc, but I guess that niche is filled by cygwin 15:45:23 I think it mostly gets rid of underscores in some of MS's function names, maybe rearranges some of the letters now and then... 15:46:06 yes, Cygwin is the thick one 15:46:26 MSYS uses something somewhere between 15:46:44 way closer to MinGW than to Cygwin, I think, though 15:47:51 <|amethyst> I've used cygwin a little, but otherwise djgpp was the last GCC-for-microsoft-OSes I used :) 15:48:24 DJGPP has a much more complete library ;-) 15:48:39 <|amethyst> since DOS does not have a runtime :) 15:48:48 And probably a better implementation of tmpfile(), too 15:51:36 -!- bmh has quit [Quit: bmh] 15:53:47 So, really, the main advantages that MinGW has over MSVC are: can still use the ubiquitous MSVCRT.dll (even without the DDK), POSIX-compatible function names (less screwy underscores), supports GCC extensions, and doesn't have MSC's wierd bugs 15:55:40 <|amethyst> so Posix functions that aren't in MSVCRT at all also aren't in MinGW? 15:55:53 mostly not, yeah 15:56:20 <|amethyst> no mkstemp then :) 15:57:01 pretty sure not, yeah 16:04:20 -!- raskol has quit [Read error: Connection reset by peer] 16:12:26 If I switched to open()+fdopen(), I might be able to use O_EXCL... 16:12:51 unwielding a weapon takes 0.6 turns (https://crawl.develz.org/mantis/view.php?id=4744) by reid 16:15:24 -!- Pacra has joined ##crawl-dev 16:16:19 -!- ainsophyao has quit [Remote host closed the connection] 16:20:25 -!- raskol has joined ##crawl-dev 16:29:44 -!- bmh has joined ##crawl-dev 16:48:03 Enchant Weapon success does not update display (https://crawl.develz.org/mantis/view.php?id=4745) by XuaXua 16:52:20 bmh, something wrong? 16:52:47 Napkin: Not really. I just want to close some old tickets and mulch bad patches. 16:58:06 were you able to? 17:01:12 just read your email 17:01:28 what's your mantis account name, bmh? 17:01:38 Napkin: I was not able to 0000875, 0002694 -- 'brendan' 17:03:38 hmm... well, i guess we could give more permissions to updaters 17:04:17 what ability were you missing? to mark resolved/closed? 17:04:18 -!- raskol has quit [Read error: Connection reset by peer] 17:04:25 anything else? 17:04:31 delete files? 17:04:34 -!- raskol has joined ##crawl-dev 17:04:47 irgs 17:04:55 ? 17:05:43 well, deletion is always not a too nice permission to give out ;) 17:07:09 ok, resolving and closing issues is in 17:07:31 you would like closing to mark resolved issues as closed, right? 17:07:47 yes? 17:07:57 good 17:08:18 you know the difference between "resolved" and "closed", right? 17:08:27 (sorry, just making sure) 17:08:54 closed -- no further comments 17:09:21 exactly. reporters are supposed to do that, kinda like a confirmation - but many forget or don't know 17:10:01 What the heck are feedback, confirmed, and acknowledged? 17:10:22 "confirmed" is presumably for when we see that the issue is real 17:10:27 ? 17:10:36 or is that acknowledged? 17:10:45 that's so confusing! 17:10:47 "confirmed", considering the order of those stati, probably means to confirm a reported bug/issue does really exist 17:11:03 just don't use it, if you don't think it's appropriate ;) 17:11:17 we usually seem to just leave them at "new" for some reason, though 17:11:24 and please, don't ask me to remove them - it's very thightly integrated into mantis 17:12:08 i put stuff to acknowledged, when i agree to work on something.. and to assigned once i have the time to actually work on it 17:12:14 but that's just me 17:12:40 Napkin: Thanks for the superpowers. 17:13:14 i enabled the ability to set resolved/closed and "update issue status" - i hope that's what you needed 17:13:15 is bmh being hazed now? 17:13:24 still looking for attachment handling 17:14:26 https://crawl.develz.org/mantis/view.php?id=875 <-- should this behavior get changed? 17:16:07 ask in the notes - everybody who posted a note to an issue, will receive an email (kilobyte in that case) 17:19:55 hmm, there is a preview_attachments feature 17:22:47 $g_delete_attachments_threshold = UPDATER; 17:22:51 $g_allow_delete_own_attachments = ON; 17:23:00 activated those settings 17:23:10 w00t 17:23:15 https://crawl.develz.org/mantis/view.php?id=4621 <- how does it look? 17:23:49 -!- blabber has quit [Quit: leaving] 17:24:04 i can see, that I can also allow txt, diff and patch attachments to be viewer in the browser, instead of forcing a download 17:24:22 Napkin: http://i.imgur.com/DHE2h.png 17:24:39 it's ok, isn't it? 17:24:52 yes 17:24:52 i mean.. useful :) 17:26:02 https://crawl.develz.org/mantis/view.php?id=4607 <- [show content] button behind .des attachments now 17:26:30 -!- dtsund has joined ##crawl-dev 17:26:45 max size is 100kb to be shown inline 17:26:59 neat 17:27:06 agreed :D 17:27:22 and I blame you, for making me search for attachment deletion permission ;-P 17:28:04 I wouldn't need the permission if I could just write to the source tree. 17:28:26 is there a way for me to close a thing I submitted? There's a patch I uploaded there that was since applied, but I just noticed it's still marked as open 17:28:27 O_o 17:28:45 the only options I see are changing status to "feedback" and "acknowledged" 17:29:07 evilmike: I'll close it! Link me 17:29:23 -!- Torokasi has joined ##crawl-dev 17:29:23 https://crawl.develz.org/mantis/bug_view_page.php?bug_id=4691 17:29:35 I was about to say that 17:30:26 ??unborn deep dwarf 17:30:26 unborn deep dwarf[1/1]: A nasty monster in 0.8 who likes to spam haunt and sneak injury mirror on when you aren't watching. 17:30:42 Ah. I think I saw one of those kill a guy in Zig yesterday 17:30:47 ??injury mirror 17:30:48 yredelemnul[3/4]: Injury mirror (active during prayer) causes monsters to suffer the damage they deal to you. Piety cost is the square root of damage done per hit. 17:30:53 problem with allowing reporters to close their own issue, is that the stats about resolution ("done", "won't do", etc) would probably be corrupted, evilmike 17:31:03 I see 17:31:38 but I believe you help a lot, so I could also give you updater status, if you like 17:32:28 Napkin: or we could just make a submodule for vaults and give evilmike write access. :) 17:32:49 oh no, nonono :D 17:32:55 way too much work 17:33:22 ;) 17:33:33 Seems dead simple to me. We already use submodules. 17:33:41 Updater status would be convenient, at least. I also have a bad habit of tweaking vaults after uploading them, so deleting the old versions would help make things cleaner 17:33:53 what are submodules? 17:33:58 but you probably do not have modified "report issue" pages, differently for each project 17:34:31 evilmike: people can delete their own uploaded files now 17:35:02 I got "access denied" when trying that 17:35:08 try now 17:35:19 just changed it a few minutes ago 17:35:22 hi all 17:35:22 due: You have 1 message. Use !messages to read it. 17:35:30 morning due :) 17:35:35 still access denied, I'll try again in an hour 17:35:41 !messages 17:35:42 (1/1) Mu_ said (10h 22m 34s ago): sorry, only set it for 1 day :p i'll ask you next time i see you, if it's still broken 17:35:44 seriously? :-O 17:35:52 Whoa, due is like on the other side of the world from me now! :) 17:35:53 Mu_: ok 17:35:56 won't change in an hour 17:35:58 Mu_: sorry, had a crazy week 17:36:42 hmm... maybe it only works for newly uploaded attachments? but that would be weird 17:37:18 $g_allow_delete_own_attachments = ON; - if it's not working, evilmike, then it's a bug or the attachment doesn't belong to you after all? 17:37:41 well, I attempted to delete a file I uploaded 17:37:57 although the issue is currently marked as "assigned to dpeg" so maybe that has something to do with it? 17:37:58 hmm, bug! 17:38:08 wouldn't think so 17:38:09 maybe dpeg counts as owning it now 17:38:27 # allow users to delete attachments uploaded by themselves even if their access level is below delete_attachments_threshold. 17:38:37 weird 17:38:45 grrr 17:39:30 my windows Git and my own MinGW/MSYS installs aren't really working together too well... 17:39:36 did you open the issue, you attached the file to, evilmike? 17:40:01 yeah, it's this issue: https://crawl.develz.org/mantis/view.php?id=4630 17:40:09 I figured, might as well delete the first two files uploaded there 17:40:11 the only things from the windows Git that are in PATH are .cmd wrappers 17:40:17 evilmike: I can delete 'em! 17:40:57 heh, i found a report, where even though the setting is set, still only reporters of issue can delete their own attachments - but then that's not the bug we're looking for 17:41:25 mantis developer: The problem is that the author of the attachment is not stored in the DB, so there is no way to implement that kind of check. 17:42:03 ... why would you not store the author of the attachment? 17:42:07 The simple solution here would be to do so. 17:42:14 bmh: might as well, its not a big deal in any case, though 17:42:19 read the discussion here: http://www.mantisbt.org/bugs/view.php?id=8134 17:42:50 i'll keep the setting on ON, but it doesn't seem to do what we want 17:43:07 please delete them, bmh :) 17:43:16 Napkin: done. The power is going to my head. 17:43:22 hehe 17:43:31 we're all lost! 17:43:36 (MUWAHAHA) 17:43:53 Also, I am against putting vaults into a submodule -- having dpeg go over all of the vaults before they go in is really, really good 17:44:40 ok, verified: definitely our mantis mysql table for the attachments has no user_id field 17:45:49 due: I wonder if there's anything we can do to make dpeg's job easier 17:46:16 add more developers to implement his ideas 17:46:20 bmh: Well, it doesn't have to be dpeg 17:46:35 due: %s/dpeg/the vault master/ 17:46:54 -!- HangedMan has joined ##crawl-dev 17:47:15 -!- neunon has quit [Remote host closed the connection] 17:47:39 well, in theory, there are a number of well-trusted vault builders on the team, and all of the devs can get into the mindset to judge them... 17:47:54 so commenting on tickets with vaults in them might help, I suppose? 17:48:23 Allow inline preview of attachments (https://crawl.develz.org/mantis/view.php?id=4747) by Napkin 17:48:23 Give more permissions to role "Updater" (https://crawl.develz.org/mantis/view.php?id=4746) by Napkin 17:49:43 Napkin: those are some meta-issues! 17:49:59 exactly 17:50:07 all devs get those via email :) 17:52:59 done! 17:53:25 good way to document changes, I think 17:55:31 perhaps this is nit-picky, but why does open-sea in shoals trigger fireballs? 17:55:55 because shoals is too easy ;> 18:00:35 due, maybe a put the vaults into a git submodule? 18:00:49 so you could give out write access to more people? 18:01:59 didn't we just go over that ? 18:02:24 *the vault master* would just need to pick the vaults, and do an submodule update commit on the main repository - not really sure that would make things easier though 18:02:40 that would probably not make things easier 18:03:05 if you talked about that before I showed up, I'm sorry, SamB 18:03:25 guess I still need to work on the holes in my hands ;-P 18:03:33 Napkin: just go up and look at things due's said 18:04:13 oh, i thought he was talking about the mantis submodule suggestion bmh made 18:04:19 -!- raskol has quit [Read error: Connection reset by peer] 18:04:24 * Napkin shuts up then * 18:04:37 -!- raskol has joined ##crawl-dev 18:10:04 Napkin: Holes in your hands? Like water into wine? 18:10:37 kinda like that ;) 18:18:13 -!- petete has quit [Remote host closed the connection] 18:18:44 Napkin: yeah, no, I think we actually don'[t want to give out more write access to vaults 18:18:57 they are difficult to balance and can be really badly broken smoetimes 18:19:23 -!- ZorbaBeta has quit [Quit: Leaving] 18:23:16 -!- alefury has quit [Read error: Connection reset by peer] 18:26:34 What are the advantages of Mantis over other issue trackers? 18:26:56 Fyren: all of our issues are already there 18:27:08 it is already installed and more-or-less working 18:27:11 That was not the case when the choice was made, presumably. 18:27:16 well, yeah 18:27:39 but you sounded like you might be wondering why we don't change to something better ;-) 18:34:39 another project i help with used to use mantis and swapped to redmine recently 18:35:13 with the idea that it would be an improvement, i think :P 18:36:41 Fyren: free, not completely awful, not associated with a specific site (like sf/github), integrates with wiki 18:36:54 since it has all sorts of wiki and other magic things all packaged into one, mantis seems to have a bunch of nicer functionality from a mere user's perspective at least, though 18:38:45 Fyren: plus Napkin runs it. 18:39:51 the devs checked a few and found mantis to be most suitable, Fyren 18:40:01 when they ran from sourceforge ;) 18:40:24 re: vault stuff discussed earlier, I have to say I find feedback on stuff incredibly helpful... uploading new ones to mantis does make it a lot easier to get comments before it goes into the game 18:40:29 I ask because I only have experience with other tools, mainly Bugzilla (and MediaWiki). 18:41:00 bugzilla... didn't have proper permissions, if i remember correctly 18:41:02 of course, comments from real games are even better. It can be a bit frustrating sometimes when something sits on mantis for several weeks, even though it's understandable 18:41:03 -!- Brannock has quit [Ping timeout: 252 seconds] 18:41:13 I seem to recall bugzilla being frustrating 18:41:34 I think every issue tracker is frustrating, just in different ways. 18:41:35 something was missing - and greensnark considered it a real show-stopper 18:41:40 -!- Brannock has joined ##crawl-dev 18:41:57 At least with this our auth is nice and global across the site 18:42:45 and dokuwiki was chosen, because it's simple, easy to access data storage and integrated nicely with mantis 18:43:01 It could have had more effort put into it, but we've got shared logins across vBulletin, MediaWiki, and Bugzilla. 18:43:11 In another project. 18:43:23 yeah, that's nice 18:43:33 vbulletin? 18:43:42 payware forum 18:43:44 Forum software. It's not free. 18:43:47 Ah 18:43:57 supposedly good though 18:44:01 I was about to ask if that was the one that stored passwords in plaintext, but I think that might be phpBB. 18:44:21 (Or an old version of it) 18:44:21 a colleague uses it.. but they've been hacked many times already 18:44:35 seems like payware is considered a more challenging target ;) 18:45:00 hah 18:45:06 that doesn't surprise me 18:46:15 seriously.. he started to block ip ranges, because of so many successful sql injections :-O 18:47:44 -!- petete has joined ##crawl-dev 18:50:04 but maybe they just don't keep it updated properly 18:50:54 I think considering how popular vBulletin is, that is probably the case. 18:51:31 I agree, it sseems more likely. 18:51:55 I don't care what wiki we use as long as my patches make it into trunk where they can kill players. 18:52:14 well.. look who's talking - our phpbb3 needs update too *sigh* 18:52:39 i'm starting to think, that dokuwiki wasn't such a good decision.. 18:53:13 but i'm actually not using it much - and there aren't many complaints 18:53:40 but from an admin perspective, it's missing features.. and additional features are often dirty hacks 18:54:07 well, wordpress first, then the rest 18:54:10 g'night o/ 18:54:13 night! 19:01:42 -!- ixtli has joined ##crawl-dev 19:04:16 -!- raskol has quit [Read error: Connection reset by peer] 19:04:41 -!- raskol has joined ##crawl-dev 19:09:31 -!- syllogism has quit [] 19:24:20 -!- Vandal has joined ##crawl-dev 19:41:49 -!- edlothiol has quit [Ping timeout: 240 seconds] 19:46:14 !tell Napkin mumra and I would probably like to pick any new Wiki we might switch to... 19:46:14 SamB: OK, I'll let Napkin know. 20:01:05 -!- bmh has quit [Quit: bmh] 20:04:14 -!- raskol has quit [Read error: Connection reset by peer] 20:04:44 -!- raskol has joined ##crawl-dev 20:06:49 ??draining 20:06:49 draining[1/2]: A weapon brand good early on. Has a 2/3 chance of doing on average 25% extra damage to susceptible monsters, doing 1+1d3 extra damage and removing that much from max HP. It does drop XP, but the drop is not significant. If the 2/3 chance is hit, it has a further 20% chance to drop HD by 1. 20:09:13 Bizarre display bug possibly related to jiyva (https://crawl.develz.org/mantis/view.php?id=4749) by st 20:55:33 -!- ZorbaTHut has joined ##crawl-dev 20:56:26 -!- MarvinPA_ has joined ##crawl-dev 21:00:25 -!- MarvinPA has quit [Ping timeout: 260 seconds] 21:04:13 -!- raskol has quit [Read error: Connection reset by peer] 21:20:41 -!- raskol has joined ##crawl-dev 21:30:34 ??ashenzari wrath 21:30:34 wrath[1/1]: Boosts enemy AI. Better pathfinding, they're more likely to target you when you're invisible, and they never forget about you. Also -4 to all skills. 21:39:14 ??vampiric draining 21:39:18 vampiric draining[1/1]: L3 necro melee spell; does 2 + 2d9/2 + (d(pow)-1) / 7 damage. Will not deal more damage than the hp you're missing. Heals you for half dealt damage. Does not affect monsters with rneg. It never misses. 21:49:29 -!- Pingas has quit [Read error: Connection reset by peer] 22:04:16 -!- raskol has quit [Read error: Connection reset by peer] 22:04:24 -!- HangedMan has quit [Ping timeout: 265 seconds] 22:04:51 -!- raskol has joined ##crawl-dev 22:05:51 -!- Zaba has quit [Ping timeout: 248 seconds] 22:16:48 -!- Zaba has joined ##crawl-dev 22:19:50 -!- Wensley has joined ##crawl-dev 22:24:27 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 22:46:04 -!- valrus has joined ##crawl-dev 22:49:19 -!- ainsophyao has joined ##crawl-dev 23:02:54 -!- Torokasi has quit [Ping timeout: 258 seconds] 23:04:16 -!- raskol has quit [Read error: Connection reset by peer] 23:07:27 -!- ainsophyao has quit [Remote host closed the connection] 23:19:55 -!- raskol has joined ##crawl-dev 23:34:46 -!- Wensley has quit [Quit: ChatZilla 0.9.87-rdmsoft [XULRunner 1.9.0.17/2009122204]] 23:47:58 -!- ainsophyao has joined ##crawl-dev 23:50:43 -!- ussdefiant has quit [Read error: Connection reset by peer] 23:51:11 -!- ussdefiant has joined ##crawl-dev 23:55:09 -!- drwhat has joined ##crawl-dev 23:55:53 -!- drwhat is now known as doctor_ 23:56:20 -!- doctor_ has left ##crawl-dev