00:00:53 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-226-gca05e9d (34) 00:03:13 Stable (0.23) branch on CRAWL.XTAHUA.COM updated to: 0.23.1-49-g05433c6 00:21:08 The build passed. (0.23.2 - 05433c6 #11291 : advil): https://travis-ci.org/crawl/crawl/builds/513631870 00:58:25 advil: looks like you've redeemed yourself 00:58:57 I guess that's still some kind of seed bug that mysteriously only happens sometimes though? 01:00:13 -!- unhold_ is now known as unhold 01:08:10 Stable branch on crawl.develz.org updated to: 0.23.1-49-g05433c6 (34) 01:25:43 so I had this idea for making autopickup easier to deal with 01:25:54 you should be able to turn off autopickup for an item as you drop it 01:26:21 I've had it mostly done for a while but I'm nervous about submitting it 01:27:23 it also probably (definitely) needs another few hours of polish/bugfixing before I could really ask for it to be merged, but I am worried about the possibility that I do that then find out it was the wrong approach to begin with 01:28:15 long story short, would anyone be willing to go to http://73.222.250.55:4206 , log in as "foo1" or "foo2" with password "abcdef", and take a look at the drop menu 01:28:33 (that's my laptop and there are existing minotaur fighters. lol) 01:29:01 the change is: if you press \ (backslash), things you select are both dropped and autopickup-off'd 01:29:13 (but it's explained on the _ help menu, of course) 01:29:47 sorry, weird request, but if anyone feels like trying it, there it is 01:30:18 don't pwn my laptop tho 01:31:40 I *think* this is an obviously superior way to deal with autopickup in the 90% of cases where you just want to stop picking up stones, wands of random effects etc 01:37:53 Unstable branch on crawl.develz.org updated to: 0.24-a0-226-gca05e9d (34) 02:08:17 Windows builds of master branch on crawl.develz.org updated to: 0.24-a0-226-gca05e9d 02:20:39 fourba: Looking now (FTAOD, I am not a developer...) 02:23:13 fourba: I've just got a splash screen, either the game's not loading or it is v e r y s l o w l y but in the meantime; 02:24:45 I think this is a really good idea in principle. The other suggestion from Yermak (always turn off autopickup when dropping) has some edge cases where it doesn't apply (especially making space before a portal branch or zig) 02:25:31 But I can't think of any cases where I want to turn off autopickup but still keep some of the item. 02:26:23 Pinkbeast: thanks for trying it! 02:26:42 My concern is it's not quite as obvious as "I just picked this up, go into the inventory and drop it in the special turn-off-autopickup way", but that's letting perfect be the enemy of good - this idea of yours is still an obvious improvement on the situation now. 02:28:10 I left this halfway done to go think about Yermak's way of doing it for a while, but for some reason it never quite worked out 02:28:31 I think Yermak's way is unfeasible simply because it doesn't always do what you wanted. 02:28:31 there were always cases where I wasn't sure what to do 02:29:01 now that I think about it, though: maybe all of those cases are when dropping stuff? 02:29:14 wait 02:29:18 i'm an idiot 02:29:28 when picking stuff up, that's already how it works. stuff stacks 02:29:52 Yes, the change _is_ when dropping stuff - you'll always, and always want to, pick up a stacking item 02:30:35 I see someone on foo1 not on a splash screen -- did you eventually make it in? 02:30:58 I've an additional suggestion that _optionally_ 'D', "Drop item(s) picked up last", could turn off autopickup of that item. (I suggest this because I never use it normally) 02:31:31 I still see the splash screen. (I don't play tiles, but public server webtiles do work here, so it's a bit of a mystery...) 02:31:48 this was an ambitious idea of mine lol 02:31:51 If you've recently rebuilt ofc the first game start takes forrrrever on a slow machine 02:31:54 thanks for trying it 02:32:16 re: D, I think that makes sense 02:32:23 I actually use D constantly 02:32:32 but I too would never want it to not do that 02:32:36 (I think) 02:32:47 and like you said, optionally 02:33:10 I think it makes sense (but I would, wouldn't I?) but it would have to be optional or you're going to stoat up someone's preferred gameplay 02:34:21 "Turning off autopickup drops what you have" doesn't need to be optional, but there is the "maybe make it optional because inevitably _someone_ will complain" issue. :-/ 02:35:40 Still at the splash screen. If you can link a patch, I'll look it over, apply it here, and report back? 02:36:35 https://github.com/joshgelbard/crawl/tree/drop-no-autopickup 02:36:56 that's what you meant by link a patch, right? 02:37:27 That'll do, I can git fetch it into my own repo 02:42:21 (the other thing of note is the presence of misc items, at last, in the \ menu) 02:45:32 Call me confused but the last commit here is "Add 'drop and remove from autopickup' menu action" 02:45:53 yeah that's the one isn't it? 02:46:16 I use D very often, and sometimes I use it wrongly, dropping not the items I planned to drop. 02:47:04 fourba: Ahhh ha, enlightenment has dawned. This conversation has been confusing because I was confused. 02:47:26 Pinkbeast, speaking of my proposal, do you have any objections to it other than dropping staff at the stash? 02:47:47 What I _thought_ you were describing was a change when, where you press \ _for autopickup control_, you additionally drop any item you turn off autopickup for. 02:48:01 oh 02:48:02 ... which is a new idea I thought you'd just had. 02:48:15 ohhhhh 02:48:16 Yermak: No, not you, fourba. Let me try and be more clear in my replies 02:48:32 fourba: What you have implemented is also a good idea that I would like 02:48:49 you're very agreeable, pinkbeast 02:49:04 So, it could be possible to forbid turning off autopickup for dropping items iff you're standing on a waypoint. 02:49:14 Yermak: "the stash" is an awkward turn of phrase - I don't make a stash, but your idea still seems to have unwanted consequences. 02:49:39 I don't make a waypoint - ever, but especially not when I make space before a portal branch or zig 02:49:40 I can't think of any other than when you stash items. 02:50:30 Yermak: I wish I had kept better track of why exactly I felt I was getting stuck implementing your idea. I still think it's a better default 02:50:56 I'll try to remember. it may have just been internal code-related stuff 02:51:24 Besides the one above - making space - I'm at the top of a normal branch. I have some scrolls like brand weapon which I won't need for immediate death-avoidance. I'm short on space. I drop them at the top of the branch, but I still want to pick more up if I find them in the branch, to bring them out and leave them somewhere convenient. 02:52:21 Yes, I often drop items before going into portal too. And I feel like picking an item manually once in a portal is less work, than now when you have to turn autopickup off for unwanted items manually. 02:53:07 Yermak: I think that suggests your idea would at least want to be an option, since you would use it and I wouldn't. 02:53:33 If you drop ?brand_weapon in a branch to make space, you'll have to make space for another copy of it if you find it. 02:53:40 here's a question that applies to both ideas: what is to be done about branded ammunition? 02:53:49 specifically/most frequently, needles 02:54:01 does dropping a poisoned needle turn off autopickup for all needles? 02:54:09 fourba, do you have an idea how it works now? 02:54:17 Indeed - but I might not find it (I certainly might not find _all_ the maybe-later stuff) and I can cross that bridge when I come to it. 02:54:25 for all poisoned needles 02:54:49 well, right now, you have the \ menu, which has an entry only for "needles", so it's relatively clear that you're affecting all of them 02:54:58 and you choose to do that or don't 02:55:16 with either of our ideas, you'll have access to one type of needle and... maybe? affect what you do with all the rest 02:55:41 no 02:55:51 Manuals are also a problematic category 02:55:56 There is a tristate that allows to pickup all needles, always, no matter what. 02:56:09 Manuals are clearly shouldn't be affected. 02:57:44 that "no" was: "no, that's not how it ought to work", right? 02:58:08 I agree -- which is part of why I had trouble implementing it. 02:58:24 the code right now just doesn't take brand into account at all 02:58:35 like, the existing autopickup code 02:58:41 well, mostly 02:58:51 there's also pattern matching on text from your options file 02:59:24 but the stuff that shows up on the \ menu is just one big table, and many things that you would want to have individual entries... don't 02:59:46 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-226-gca05e9d 02:59:57 typing... 03:11:41 This was all getting a bit lengthy so http://www.chiark.greenend.org.uk/~damerell/games/autopickup.txt 03:11:43 In the meantime I'm building fourba's current patch to give it a whirl 03:14:55 How it works now: there is a tristate: if you don't touch autopickup for needles (and don't start as As) - this is default, you'll pick up any type of needles you have in your inventory. If you don't have some type of needles in your inventory you won't pick up them. If you turn autopickup for them on (let's call it tristate+), you pickup every type and every stack of them (except those stacks you dropped manually). If you choose the third option 03:14:55 (tristate-), you never pick up needles, even those you have in your backpack (you still pick up those you fired). 03:15:01 How it should work in my opinion: same as above for default and tristate+ except for when you drop some type of needles, you never pick them up automatically until you pick them up manually again. Tristate- shouldn't exist. 03:15:06 There is another option: get rid of generic "needles" and make an otioon for every type of them (as well as tomahawks and javelins). It would be probably preferrable if not for special brand ammo shops where you can buy lots of other combinations such as large rocks of chaos. 03:16:16 Yermak: Oh, and some of what I wrote isn't true 03:17:07 I wonder if in some of these other ideas "turn off" should be "go to initial state", not "go to tristate-" 03:18:15 thanks both I have much reading to do, sec 03:21:20 Previous system was influenced by item weight, at some point you wanted to stop picking up ammo. 03:23:00 Tristate- for ammo exists only for that reason imo. 03:24:07 Updated that document to reflect Yermak's observation 03:26:17 To think of, isn't tristate- just an atavism of versions with item weight? 03:26:41 I mean, for all types of items? 03:27:41 I can't think of a reason for it to exist nowadays. 03:27:47 Yes but scope creep 03:29:25 More seriously it's not impossible to imagine an item that behaves differently and potentially undesirably if you have more, and if one is designed it would be a right pain if tristate- had been painstakingly removed 03:30:35 I don't think it's a good idea to have something in game just in case if something needing it will be implemented. 03:31:24 Stable (0.23) branch on crawl.kelbi.org updated to: 0.23.1-49-g05433c6f1a 03:33:45 I'm not done thinking about this and am rereading what you two have said 03:34:58 I _think_ the immediate implication is that your patch wants to set AP_FORCE_NONE not AP_FORCE_OFF _but_ my understanding of how the tristate operates is a bit limited 03:35:01 but I think the fact that there's so much room for discussion here is part of why I went with the one that people can "just ignore" and that (importantly, for a noob developer) requires minimal changes to existing, working code 03:37:29 Ha, no, because there's an interaction also with "autopickup", the option, that controls whether the category is picked up at all 03:38:23 yes autopickup checks many things 03:38:34 at the very end there's that 03:40:31 Yermak: what you described for "how it should work" is definitely better, but is just harder to do code-wise 03:40:47 for me, anyway 03:41:50 So, how does your current version deal with branded ammo? 03:42:17 there's a big table of things that should be autopickupped, and it's a 2d array with a row for each base type (OBJ_WEAPONS, OBJ_MISSILES...) and a column for each subtype (RING_REGEN, MI_NEEDLES) <-- some enum names may be misremembered 03:42:40 and all i'm doing right now is modifying that table in only, and exactly, the ways that are currently possible via the \ menu 03:42:53 What I'm getting at is that if the overall category (eg missiles) is off, and I drop-and-turn-off a posioned needle, I think I wanted AP_FORCE_NONE set. If missiles are "on", I'm not sure what I want (if AP_FORCE_NONE, it doesn't actually turn off needles at all; if AP_FORCE_OFF, it stops picking up curare needles even though I have a stack in inventory) 03:42:54 so i deal with branded ammo the same way the game currently does 03:43:20 if you don't do my special drop thing, it's like you never directly touched the "needles" entry in \ 03:43:42 I see 03:43:42 and if you do it to any needle, it's like you turned it off 03:43:49 Yermak: And that's what tristate- is for - to force off an item in an overall category that's on 03:43:56 i would really like it if that table were three dimensional and included brands 03:44:13 and, in fact, wasted a bit drawing up a "sketch" of my ideal \ menu 03:44:18 i'll pastebin it, i still have it 03:44:31 there is same issue with evocables 03:45:20 Pinkbeast, how would you have an item in overall category that's on except by having it in your inventory? 03:46:01 s/an item in overall category that's on/an item on in overall category 03:46:07 https://pastebin.com/raw/2zp0KhwT <- lol 03:46:27 If the "autopickup" option lists it, that overall category is on regardless of what I have in my inventory. 03:47:13 i have zero plans to do anything like what that pastebin has, by the way 03:48:39 Ok, I got confused. What's overall category if not autopickup option? I thought needles were overall category for example and poison needles is an item that you have on when they're in your inventory. 03:49:03 I am sorry, I have used "category" in two senses. 03:49:24 I mean that "autopickup" does a per-symbol selection, like '(', Missiles. 03:49:57 And the tristate is actually "always", "never", "respect per-symbol selection", with some additional rules like not picking up useless objects 03:51:59 I don't think I understand you. 03:52:50 Yes, I'm rewriting the document because it's easier than describing it in IRC 03:52:56 ok 03:59:51 Rewritten http://www.chiark.greenend.org.uk/~damerell/games/autopickup.txt 04:01:42 fourba, your ideal \ menu will for sure scare the hell out of newbies, hah! I think current \ menu is already scary for them. I'm not sure we need something this elaborate. 04:04:32 yeah i was just wasting time 04:05:28 my actual ideal is to move as much of the useful functionality of the \ menu as possible to other places 04:06:03 Rewritten again, sigh 04:06:26 I will take a look 04:06:54 fourba: I appreciate I'm changing my mind every two seconds but I now think what you have works, even for oddball stuff like ammo, if additionally the game will always autopickup something that stacks with existing inventory. (Which it doesn't and should because encumbrance is dead) 04:06:55 Pinkbeast, looking at words "use the per-symbol selection" and "Missiles if off", I think you either understand it wrong or describe it wrong. You can't turn autopickup on for all missiles. You can do it separately for all needles, all tomahawks, etc. 04:07:15 Yermak: Yes, you can; see the "autopickup" option. 04:07:30 oh, you mean in rc 04:07:56 I don't think we should discuss rc options when talking about \ menu. 04:08:25 We have to because the middle value of the tristate is "respect the per-symbol value set in rc with the autopickup option". 04:09:30 Doesn't rc just sets the initial values of \ menu? 04:09:35 No. 04:09:46 Could you give an example? 04:10:24 what's the difference between "initial value" and "middle value" here 04:10:37 in both cases it's the thing that pressing "reset" (^) returns you to right 04:10:41 Toggle items in the \ menu; they go between "tristate+" (a bold +), "tristate-" (a bold -), and the middle value, which is a non-bold + or - (and picks up or doesn't) depending on the value of the "autopickup" option. 04:11:00 And btw, "game will always, no matter what options are set, autopickup items that stack with existing inventory" - I agree with that. 04:11:26 You (and I, no criticism of you) are unaware of this because we never fiddle with the "autopickup" option because the default is what you wanted anyway 04:12:16 I think then middle option is '+' or '-' depending on if you have the item in you inventory or not. 04:12:24 s/if/whether 04:12:28 Try it. 04:12:31 sec 04:12:46 Especially, try it after changing "autopickup" to, say, "()". 04:13:06 At start all items are in middle state, right? 04:13:18 Unless you have autopickup exceptions in rc 04:13:50 And some of them might be '+' or '-' depending on what you've set in your rc. 04:14:14 Ah! I think what is confusing you is if the per-symbol selection is "off", having some in your inventory trumps that, setting that individual item to + 04:15:01 ok, I see. I've never per-symbol rc autopickup option, this might be the case of misunderstanding. 04:15:03 (Just as if it is on, it being a useless item trumps that) 04:15:23 You still get some on by default, it's just what you wanted. 04:15:44 That said, I can imagine gnoll might turn all missles on at start. 04:15:44 *missiles 04:16:10 nvm, you're right, "middle" state is indeed dependant on what you have your rc set to. 04:18:25 Scratch it. Let's discuss autopickup lines in \ menu that don't have multiple categories in them, such as rings of fire resistance or scrolls of identify. I'm sure they don't need tristate. Correct Me, if I'm wrong. 04:22:45 I guess I better make my own document and describe all the ideas there. 04:24:01 Tristate +, always pick up scrolls of ID. Tristate -, never do so. Tristate middle, respect my decision if I inexplicably turn off ? in "autopickup" option. 04:25:03 Sounds like unnecessarily complication. 04:25:40 Maybe, but since the player _can_ turn off '?' in "autopickup", the third state is necessary to implement their decision. 04:28:55 Ok, player has set some scrolls in their rc to be explicitly picked up, and in game they want to switch between "pickup all scroll" and "pick up only chosen (in rc) scrolls". Do I get it right? 04:29:02 fourba: I think "initial value" and "middle value" are the same, except the "autopickup_exceptions" option can cause the initial value not to be the middle value 04:29:23 oh, right, there is that difference 04:29:42 Yermak: I think that's a possible case, yes. (They'd have to save and edit their rc to do it...) 04:29:54 I don't think tristate for scrolls should exist for such a players. 04:32:17 The number of such players is very small. I'd rather have them to experience problems than for most players to press additional key when switching through scroll tristate. 04:33:10 I also think the implementation becomes trickier now we know the tristate is wanted for multiple item categories some of which it seems like it would be hard to get rid of. 04:33:42 I believe we can get rid of tristate now, when item weight is gone. It just requires careful implementation. 04:34:02 discussion and implementation, rather 04:34:40 I'm not saying you're wrong, I just think this is more complex than fourba's idea, which seems to need only "always pick up items that stack into inventory" to effect an immediate and obvious improvement 04:35:15 ... and also I'm aware this tristate business was _way_ more complex than it looked at first and as such removal might be fraught with edge cases. 04:35:42 I'm perfectionist, while it is harder to change this much, at the end the code will become simpler. 04:37:49 (But I'm not against simple QoL changes) 04:42:02 -!- TAS-2012v is now known as TAS_2012v 04:52:42 Pinkbeast, thanks very much for testing my code 04:53:06 and Yermak, Pinkbeast, thanks to both of you for that discussion 04:53:19 autopickup is a fruitful topic 04:53:57 fourba: "Testing" might be a strong word, it built but in the meantime I think the discussion has thrown up a necessary additional change. 04:54:36 thanks for verifying that I can successfully serve the DCSS splash screen from my laptop, then 04:54:42 since that seemed to work 04:54:56 in fact i think foo1 was still playing when I killed the server 04:55:16 It never got past the splash screen, I regret. 04:55:54 To be specific, I didn't see it get past the splash screen. It might have done so in between me last looking in that tab and you shutting down. But it had been a very long time and it seems unlikely. 04:55:55 wait, sorry, just because I'm worried I'll miss it in all this text: which change did you mean just now 04:55:59 nvm 04:56:05 oh wait 04:56:10 yeah sorry which one 04:56:17 long night 04:56:26 I mean to always, regardless of autopickup exceptions or \ menu selections, pick up an object that stacks with existing inventory. 04:56:54 With the removal of encumbrance I think that is a good change in and of itself and avoids the complicated difficulty with the tristate explored at length above. 04:58:26 hmm 04:58:40 even explicit force_autopickup = -1 04:58:49 ? 04:58:56 Ha, clever! 04:59:11 no sorry i meant that as a question 04:59:23 No, not if autopickup is off altogether, either by Ctrl-A or rc. You got me there. 04:59:59 but if autopickup is possible at all, then there's no possible menu setting that can prevent it from applying to something you have in your inventory 05:00:10 fourba: I don't follow you. 05:00:25 still a question. i'm going to start using question marks again even though it's 2am 05:00:35 you'll notice my capital I's left 05:00:46 so, if I'm reading you right: 05:01:55 even going into the \ menu and setting wands of flame to "white minus" 05:02:20 will not prevent you from autopickupping wands of flame -- if there's one in your inventory? 05:02:24 If that's a question... yes, except this always-pickup behaviour itself has to be optional because some crazy turncount optimiser speedrunner will complain otherwise, gah. 05:02:35 Let me rererewrite that document 05:03:12 turncounts are indeed what I was thinking of. it's lucky for me that you have such a document, since I should probably be asleep 05:03:46 and that link will presumably still exist in a few hours 05:04:32 I appreciate your willingness to go into so much detail here 05:05:11 for now, good night. I'll check your doc when I'm awake again 05:05:31 I've just rewritten it again in the light of our discussion now; good night 05:36:29 lol, crazy turncount optimiser speedrunners never have auto_pickup on. 05:37:53 If they do, I'd never call them like this. 05:46:16 03NormalPerson702 07https://github.com/crawl/crawl/pull/1024 * 0.24-a0-202-g0d52ccf: Fix spurious message on moving into door with wall_jump_move (11940) 10(11 hours ago, 1 file, 8+ 10-) 13https://github.com/crawl/crawl/commit/0d52ccf8f5b0 06:21:33 -!- amalloy is now known as amalloy_ 10:46:14 03NormalPerson702 {advil} 07* 0.24-a0-227-g1a43be2: Fix spurious message on moving into door with wall_jump_move (11940) 10(16 hours ago, 1 file, 8+ 10-) 13https://github.com/crawl/crawl/commit/1a43be200669 11:02:30 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-227-g1a43be2006 (34) 11:22:18 Stable (0.23) branch on underhound.eu updated to: 0.23.1-49-g05433c6f1a 11:57:08 The build was fixed. (master - 1a43be2 #11293 : NormalPerson7): https://travis-ci.org/crawl/crawl/builds/513769179 13:01:59 Stable (0.23) branch on crawl.akrasiac.org updated to: 0.23.1-49-g05433c6 13:06:49 Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-227-g1a43be2 (34) 14:12:57 bleh, I didn't test windows carefully enough 14:13:08 gonna have to do some dodgy stuff for the sake of seed stability for these builds 15:58:52 ok, I think the release is done, whew 15:59:31 http://crawl.develz.org/wordpress/0-23-2-and-0-22-2-bugfix-releases 16:00:17 also, I posted this: http://crawl.develz.org/wordpress/behind-the-scenes-on-retrofitting-dcss-seeding 16:03:28 Is 0.22.2 just reverting the dancing weapon message bug? 16:19:57 that's the last commit in it 16:20:53 %git 0.22.2 16:20:53 07advil02 * 0.22.1-58-gf424fa9: Update changelog date for 0.22.2 tagging 10(4 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/f424fa918fe8 16:21:24 hm, not sure why that isn't numbered right 16:21:37 Mmm, but I think froggo Grand Finale was in 0.22.1 16:21:55 %git stone_soup-0.22 16:21:55 07advil02 * 0.22.1-59-ge965b3a: Update debian changelog 10(4 days ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/e965b3aba33e 16:22:11 that should be 0.22.2-1 16:22:18 but anyways it's ~57 commits 16:22:31 Oh, hang on, I'm confused. I think? 16:23:01 yeah, I don't think that was in 0.22.1 16:23:18 given that it's from mar 3 16:24:08 Unconfusedly, I mean, last 0.22 release was up to froggo Grand Finale (and "add a header") and since then there's just been that revert 16:24:35 I didn't tag it until that revert 16:24:48 that wasn't a release, I think you're just thinking of when I did a large batch of cherry-picking 16:25:08 %gti 0.22.1 16:25:12 %git 0.22.1 16:25:12 07gammafunk02 * 0.22.1: Update the debian changelog for 0.22.1 10(7 months ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/8f2143ff1bc2 16:25:48 %git 0.23.2 16:25:48 07advil02 * 0.23.1-49-g05433c6: Update debian changelog for 0.23.2 release 10(20 hours ago, 1 file, 10+ 0-) 13https://github.com/crawl/crawl/commit/05433c6f1a00 16:26:06 gammafunk: did I do something wrong in tagging 0.22.2 and 0.23.2 such that chei isn't showing them? the build scripts seemed fine with the tag 16:28:42 advil: looks like you made 0.23.2 a lightweight tag 16:28:46 instead of an annotated one 16:28:49 oh 16:28:52 I see 16:29:03 I think I interpreted that as something only for releases 16:29:04 you might want to push an annotated one 16:29:13 no, we use annotated for all 16:29:39 at least in terms of git describe, I think it wants an annotated tag 16:31:34 sadly I think that will affect the reported version for any builds you've done 16:31:39 assuming you've done those already 16:34:50 the reported versions seemed to be correct 16:35:13 hm, is there a way of doing this without a force push? 16:36:20 yeah, reported versions on mac and windows are right, so I think git describe may be ok with lightweight tags 16:37:50 oh, good 16:38:04 hrm, then I wonder why chei specifically has a problem 16:38:18 not sure re: the tag thing 16:38:54 I have now read that git describe works "better" with annotated tags, maybe what we're doing in the build system is more robust than what chei does 16:40:09 leaving those 0.22.3 and 0.23.2 tags as lightweight may not cause any longterms problems 16:40:34 hrm, I guess it does give confusing output for chei 16:41:59 hrm, I just followed some slightly mysterious internet directions that did not require a force push 16:42:05 hope whatever I did doesn't cause any issues 16:43:16 %git 0.23.2 16:43:16 07advil02 * 0.23.2: Update debian changelog for 0.23.2 release 10(20 hours ago, 1 file, 10+ 0-) 13https://github.com/crawl/crawl/commit/05433c6f1a00 16:43:19 sweet 16:43:26 %git 0.22.2 16:43:26 07advil02 * 0.22.2: Update changelog date for 0.22.2 tagging 10(4 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/f424fa918fe8 16:44:17 for the logs, in case this ever comes up again mysterious internet directions were these: https://sartak.org/2011/01/replace-a-lightweight-git-tag-with-an-annotated-tag.html 16:44:48 %git stone_soup-0.23 16:44:48 07advil02 * 0.23.2: Update debian changelog for 0.23.2 release 10(20 hours ago, 1 file, 10+ 0-) 13https://github.com/crawl/crawl/commit/05433c6f1a00 16:44:51 nice 16:44:52 %git stone_soup-0.22 16:44:52 07advil02 * 0.22.2-1-ge965b3a: Update debian changelog 10(4 days ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/e965b3aba33e 16:49:17 also, I guess I should mention that I ended up having to compile the windows builds with sse2 and with the lua contrib manually modified; I'm not sure what was going on except that I did most of my floating point testing on linux and should have tested windows more 16:53:44 as in there's still some kind of windows seed bug related to lua's numeric conversion you'll have to track down later? 16:56:07 yeah, probably multiple 17:02:49 -!- amalloy_ is now known as amalloy 17:04:50 03advil02 07* 0.24-a0-228-g05560da: Update credits 10(2 minutes ago, 1 file, 0+ 1-) 13https://github.com/crawl/crawl/commit/05560da47184 17:11:53 so what I ran was stuff like, `sudo OS=debian DIST=stable ARCH=i386 pdebuild` 17:12:22 what might have happened is that your COW was based on ubuntu 17:12:38 it's possible I forgot to document that arg for creating the COW 17:12:45 since it's been a while since I've created a new one 17:12:54 I reuse the same instance when I do builds 17:13:17 oh, will cowbuilder use env variables from pbbuilderrc? 17:13:53 no, I think you might have to use a --distribution argument 17:14:11 but it also needs to know the debian repo urls etc, not sure where it gets that 17:14:27 need to find the reference I used when setting this up 17:16:39 can you pastebin your pbbuilderrc? 17:17:34 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-228-g05560da471 (34) 17:17:52 oh it does read pbuilderrc 17:18:06 the only modifications I should have are my email/name, let me see 17:19:28 ah 17:19:55 yeah you probably do need that OS=debian bit 17:20:09 which causes MIRRORSITE to be set 17:21:06 ok yeah it seems to be building the .cow when I set that in the command line 17:22:21 advil: you probably need to set the right distribution as well, as I think it might try to use your host dist name? 17:22:25 maybe it will get that right 17:22:34 I set it as "stretch" 17:22:40 i.e. current debian stable 17:22:43 ok 17:22:45 is that what you use? 17:22:50 I do DIST=stable 17:22:57 ok, probably the same effect 17:23:32 I guess the update command might also need this 17:23:55 yeah 17:28:55 advil: I looked at the diff of the pbuilderrc between my build instance and the default, and it is just setting my name/email for the build, but there is one other difference: 17:28:58 < export DEB_BUILD_OPTIONS="parallel=5" 17:29:05 without the '< ' of course 17:29:13 which might be helpful if you have multiple cores 17:29:20 ok, thanks 17:46:11 make[1]: git: Command not found 17:46:16 I wonder if that matters? 17:47:48 don't think it does, but then again I'm not sure how it's getting the version description if not through git describe 17:47:50 oh 17:47:57 maybe that's created in the source package? 17:48:13 ah probably 17:48:19 I seem to recall it falling back to looking for the version in a file 18:20:01 advil: brave to post about seeding on april 1 18:20:16 it's still Mar 31 here 18:21:46 Unstable branch on underhound.eu updated to: 0.24-a0-228-g05560da471 (34) 18:31:29 april fools last for 48 hours though 18:43:38 or at least so stackoverflow thinks 18:44:18 “Greedy” level generation goes back to Linley’s Dungeon Crawl as far as I’m aware, and on its own terms is a fairly simple and clean approach <-- you mean lazy? 18:45:34 greedy algorithms aren't bad by default; there are plenty of use cases where they're equal to optimal 18:58:52 tbh i can't tell which work alexjurk1 thinks should have been "lazy" 19:07:53 I thought the dungeon was always generated on-demand 19:07:53 and on-demand generation is lazy 19:14:40 each level is generated when needed, but the whole level is generated at that time 19:14:57 "lazy" isn't all-or-nothing 19:16:45 well, now that i see the article alexjurk1 is referring to, i'm inclined to agree that "lazy" better describes this than "greedy" 19:31:52 maybe "incremental" would be a better word 19:40:14 keep in mind that originally linley crawl did not even support saving nor revisiting levels 23:15:38 anyone remember a few years back when Cocytus was renamed the North Pole, and Antaeus to Santaeus for a christmas holiday period? Wonder if this could happen for other holidays too 23:16:01 not as often as Google have a doodle, but more often than currently 23:20:48 there were problems with that the last time 23:21:01 it makes weird milestones due to the renames 23:21:17 I think if it were ever done again it would need to do so in a way that didn't affect milestones 23:21:28 nor logfile entries for that matter 23:32:55 yeah, just changing the runtime display somehow 23:59:55 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-228-g05560da (34)