00:13:18 Unstable branch on cbro.berotato.org updated to: 0.35-a0-41-gef146b28bf (34) 00:56:03 Monster database of master branch on crawl.develz.org updated to: 0.35-a0-41-gef146b28bf 01:11:07 Stable (0.34) branch on cbro.berotato.org updated to: 0.34-b1-41-gfa6b30bd71 04:36:36 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 05:36:28 Unstable branch on crawl.akrasiac.org updated to: 0.35-a0-41-gef146b2 (34) 10:02:11 cirn0 (L19 DrHu) ASSERT(ranged_atk) in 'beam.cc' at line 2853 failed. (Snake:4) 10:10:23 <04d​racoomega> The naga sharpshooter wields a +0 arbalest. The naga sharpshooter gestures at you while chanting. The naga sharpshooter magically throws a throwing net. The throwing net misses you. Hmm... I don't think I realized this was a thing they could do 10:10:56 <04d​racoomega> (For that matter, I don't know for sure if they could do it before my recent refactor of that, I suppose) 10:11:53 <06r​egret-⸸nde※> To be honest, we could have something else with Portal Projectile much later than Snake (another piece of extended fodder?), which could also include throwing in its theme. 10:14:22 <04d​racoomega> Portal projectile needle assassin >.> 10:41:32 <12e​bering> Demonspawn needler 10:42:00 <12e​bering> Transdimensional dart spider 10:42:08 <12e​bering> (Hell monster) 10:53:34 <04d​racoomega> Actually, I am glad for this crash, since things are more broken than I expected 10:54:09 <04d​racoomega> (Even for using portal projectile with a launcher; it just doesn't assert about the ways in which it is wrong) 11:13:02 03DracoOmega02 07* 0.35-a0-42-g52370782d7: Fix monster Portal Projectile not firing proper ranged attacks 10(2 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/52370782d71d 11:13:04 03DracoOmega02 07[stone_soup-0.34] * 0.34-b1-42-g580c134463: Fix monster Portal Projectile not firing proper ranged attacks 10(2 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/580c1344637a 11:13:32 <09g​ammafunk> Noting again that we still need a release title, so get your bad ideas in chat 11:13:47 <09g​ammafunk> I'm planning to release tomorrow and I'm not sure if I've heard any viable suggestions 11:14:46 <04d​racoomega> I still feel like 'Doomed Geometries' is probably better than any other suggestion I saw floated 11:15:03 <09g​ammafunk> will make that the working title then 11:15:21 <09g​ammafunk> !zotorbdeaths 11:15:31 <04C​erebot> Zot:5 death rates by orb flavour: 28/267x entropy [10.49%], 18/273x winter [6.59%], 13/249x fire [5.22%] 11:15:43 <09g​ammafunk> looks like it's the most appropriate title 11:16:19 <04d​racoomega> I have to admit, I've been a bit concerned seeing them leap that much to the front after the buff (after lagging in the back for a long time). But we'll see if it holds, I guess. 11:16:31 <09g​ammafunk> yeah, numbers still pretty small 11:16:45 <09g​ammafunk> after a week of tournament things will be clearer 13:15:31 <09g​ammafunk> It does occur to me that we'll call this release "Doomed Geometries" but we're possibly not increasing the Doom in the 1993 FPS sense 13:31:36 It's not too late to add a shotgun 13:58:02 <06p​leasingfungus> sadly we removed the shotgun years ago 14:07:36 Years ago I came up with DoomRL's combat shotgun mechanic, which was pretty satisfying at the time 14:08:52 <06p​leasingfungus> that was great yeah 14:08:55 <06p​leasingfungus> big fan 16:42:57 <09g​ammafunk> We were talking on cszo in the old "tileschat days" about DoomRL's The Wall special level and how that worked in terms of a novel way of releasing enemies. That was the roguelike I had just completed before returning to actually learn DCSS. I had mentioned days earlier that I wanted to try my hand at making some complex vaults for DCSS and Arrythmia (a tiles chat regular) suggested that I make something like The Wall with deep troll 16:42:58 Unstable branch on underhound.eu updated to: 0.35-a0-42-g52370782d7 (34) 16:42:58 elementalists who wake up and dig out of their prison. I agreed and thought about it a while and eventually came up with gammafunk_dig_for_victory. Ended up only referencing Nethack since it seemed more thematically appropriate, but it was actually DoomRL that was the original inspiration. 16:44:00 <09g​ammafunk> (The Wall has lots of archviles behind a thick wall of rock that you can only dig out with a rocket launcher) 17:00:30 I'm trying to find where the `name` field of an itemdef object gets set, but having some trouble. Anyone know? 17:01:01 I assume it's when the item is created on init of the game running, but that assumption hasn't yielded findings yet. 17:05:36 <09h​ellmonk> Honestly down to add this in future ranged work 17:13:45 <09h​ellmonk> what are you trying to do with item name? I can't look into it right now but I can try to find an answer for you later 17:27:49 I'm trying to figure out what step I missed when setting up a new class of item that is making that item never appear. When I try to give myself the item in wizard mode I get gold instead. The item creation goes through parse_raw_name to connect to the item, which goes through item_kind_by_name, which does some sort of lookup I don't understand on 17:27:50 item_names_cache, which seems to be a hash of names mapping to item type and subtype. That hash gets the name from this function: 17:27:50 string name = item.name(plus || item.base_type == OBJ_RUNES ? DESC_PLAIN : DESC_DBNAME, true, true); 17:27:51 And that is where the trail goes cold for me. Something must make the .name branch out using type and subtype to get to a func that gets the name, but I've not found that connection yet. I suspect that connection is missing a link for my new item `type` 17:28:39 Or... maybe that's a func not a field? 17:36:43 Hmmm, that seems to be a func that drills down into name_aux, which I do have connected to my name func... 17:46:27 Stable (0.34) branch on underhound.eu updated to: 0.34-b1-42-g580c134463 17:59:50 Okay, I think I solved the mystery of where names come from. New mystery: 17:59:51 - If I try to give myself my new item with wizard mode it gives me gold instead 17:59:51 - If I give myself my new item by putting it in the starting equipment of a job, it works. 17:59:52 Anyone have any idea why that would happen? 18:06:46 Can we see a diff? 18:14:13 aa 18:20:24 (sorry, cat-equivalent on keyboard) 18:21:25 Webtiles server stopped. 18:21:27 Webtiles server started. 18:21:40 Webtiles server stopped. 18:21:40 Webtiles server started. 19:15:42 Sorry for the delay, had to feed cats. Here's the diff: https://github.com/DHarrison998/crawlVaultHeir/commit/75c079f47866ae4128f07836e3a0242396531ab8 19:17:14 Not relevant to the problem, but typos: "Armor", "only the worth" should be "worthy" 19:17:23 (Pinkbeast, cat equivalent?!?) This CL added the new base type, coffer, and a dozen types of items, currently in a very rough and untested state. 19:17:34 Wow, that was fast. Good eye. 19:17:40 "auxiliary" 19:18:17 I just happened to look back. I think I'd better clone your repository, which will take a little while on the connection here. 19:18:40 Fixed the spelling issues there. 19:18:43 Thanks! 19:21:09 Immediately suspicious is that this doesn't touch wiz-item.cc and as such I'm a little surprised you even _could_ ask for it in wizmode, which suggests limited understanding on my part 19:22:08 wiz mode seems to divert to create_item_named in makeitem.cc pretty immediately 19:25:22 For sure I'll need to add it to wizard_create_spec_object eventually but that didn't seem necessary yet 19:25:23 At this point I would build it and start stepping through with gdb but it will be a while before I can do that 19:26:00 cat equivalent> it wasn't a cat that typed " aa" but it was equally well not something intentional 19:27:22 Given what you said I guess wildly that stepping through create_item_named() will find a problem and it's a short function 19:28:36 Thank you for being willing to help with that! I don't have debugging set up but it's looking to be something I gotta figure out how to set up soon. 19:30:46 Well, I'm curious, but I'm also unreliable so I do suggest taking this more as an indication as to how I would proceed. Especially since now I have to chew through the submodule brain damage 19:32:32 Ah! I misunderstood. Looks like it's time for me to figure out how we debug things. 19:33:07 <08o​____0> (There are two methods of acquiring items in wizmode. &o requires selecting a category and &% requires the item specifications using vault syntax. That's probably where the confusion about not altering wiz-item.cc comes from) 19:33:14 Not so much misunderstood! I am building it with the intention of taking a look, but I am unreliable and might just go to sleep 19:33:33 <12e​bering> !chat 19:33:42 <12e​bering> urf i have forgotten 19:34:05 <12e​bering> thanks 19:34:18 <12e​bering> iocaneimmunity: if you're hacking for fun, disregard my next message 19:34:54 Also IDK if perhaps now some actual devs are awake we should take this to #crawl 19:37:03 Thanks o_0 19:37:05 <12e​bering> iocaneimmunity: if you're aiming to develop something that ultimately gets merged in trunk, your background is at odds with most of the philosophy of a crawl background in a fundamental way that might be an obstacle to its merging (I am not an active dev at present so this is a drive-by). Some time in the past an "Archaeologist" background was proposed where the character started with a dusty tome and a strange box. at XL3 the tome was 19:37:06 revealed as a manual for some skill and completing the manual unlocked a (un?)randart tied to the skill (this is my hazy memory, you can search it and it still exists in some forks). It ultimately did not land in trunk because of the same objection. 19:37:59 Good to know. What objection was that? 19:38:02 <12e​bering> That objection is: crawl backgrounds are meant to be a starting kit. A few skill points, some small items, a consumable or two to get the character started. After that point the character is left to its own to unfold as the dungeon requires (or against that, as the player wishes) 19:38:12 Ha, I mentioned them in the Tavern thread (AFAIK it's just Gooncrawl but ICBW) 19:38:59 <12e​bering> to the extent that toons have features that stay with them their whole life, we use species to build that sort of mechanic (Dr color, Co mechanism) 19:41:00 Pedantically I might say Berserker is stuck with its background longer than Arch was, but less pedantically, nice to see you around again, ebering 19:41:50 <12e​bering> I am always lurking, just learning why crawl devs slowly fade out as they move along their academic career paths 19:42:09 Ah, that's where I'm ahead by being a _failed_ mathematician 19:43:09 iocaneimmunity: to be clear, you got the pile of gold by doing &% (yes?) and was exactly what? 19:43:47 <12e​bering> Be is a special case for historic reasons. The consensus the last time I was active is we don't want more Be-like backgrounds. (See the trimming of zealot starts to Be and CK) 19:45:55 Yes I got gold by doing &% small weapon vault 19:46:43 Ebering: So to be clear the main issue is that getting one of these coffer/vault items from the get go is too much of a "stuck on a set path" to be in line with "do what the dungeon throws at you? 19:47:02 <12e​bering> Yes 19:47:13 iocaneimmunity: Can you place a "small weapon vault" in a vault (I would change the name, I mean a pregenerated dungeon level) and see if it turns up? 19:47:36 Yeah the name is "coffer" everywhere else so I can change that now. 19:48:15 How do I add an item to a vault? 19:48:49 docs/develop/levels says but that's a large can of worms 19:49:02 As for the design, is there a problem with the item being so constrained or is the problem mostly that you start with a coffer at all? 19:50:12 On the plus side it is cold here and the heat coming out my lappie from compiling is welcome 19:50:28 ebering: FTAOD I appreciate the situation with Be, was not a serious argument 19:50:34 Pinkbeast, is it useful to know that if I put "small weapon vault" (or coffer after I rename it) into a job's starting equipment that I do get the item, but that the icon shows up as a ? when it should show up as the icon I set up for it? 19:50:56 I don't know, I don't really do tiles besides testing in it 19:51:01 Sorry. 19:51:31 No worries 19:51:47 ... thinking a) I think that's a soluble problem but b) is not related to the problem we started with 19:52:31 The stuff that any commit that adds an item does to tilepick.cc and rltiles/dc-*.txt is enough to figure it out, I expect 19:53:41 iocaneimmunity: Re debugging, I (on Linux) use "gdb", which is tolerably good 19:55:29 lol, tolerably good is a solid endorsement. Sadly I'm on windows. Seems like visual studio works from the documentation? 19:57:23 I expect whatever you compiled it in can in theory also debug it but I'm afraid I know nothing beyond that; when I was working with Bcadren on developing his variant in Windows the easiest way to do some testing turned out to be to build it in Windows Subsystem for Linux 20:01:29 Fair, though I recognize the number of totally unknown things I have to learn between here and debugging greatly lengthens the time it'll take me to figure out such things, but if I bounce off of VS I'll try what you say. 20:11:26 So... &% "small weapon vault" does call create_item_named() with string name = "small weapon vault". item_list ilist does end up with an item of base_type OBJ_COFFERS. 20:12:42 That's good to know. What's the subtype? 20:12:52 0 20:13:13 So that's correct 20:13:26 What if you do "large weapon vault"? Is the subtype 1? 20:13:46 I'm chasing into dgn_place_item() 20:14:22 Sounds good. Thank you again. 20:15:59 <12e​bering> The coffer at all. It is the same as arch’s crate. 20:17:24 It's become gold by if (item_made == NON_ITEM || item_made == -1) in dgn_place_item() 20:19:47 Hmmm, so something here must be off: 20:19:47 `item_made = items(spec.allow_uniques, base_type, 20:19:48 spec.sub_type, level, spec.ego, NO_AGENT, 20:19:48 _get_custom_name(spec), fixed_props);` 20:19:56 I guess so 20:21:34 The switch (item.base_type) in items() in makeitem.cc seems to do nothing with coffers 20:21:47 // that is, everything turns to gold if not enumerated above, so ... {dlb} 20:22:43 I think that's gotta be it 20:23:12 Thank you so so much. I can't tell you how many hours I think you have saved me. 20:24:07 ofc pick the debugger that's best for you but once I had downloaded it through the string connection here and built it on the slightly ancient laptop, that was easy with gdb 20:24:55 ... not that I don't sometimes find myself staring at it going, no, that is _literally impossible_ 20:26:38 (deciding a thing that happened is impossible is often a valuable step in debugging.) 20:29:18 hours> I hope the old school approach of putting debugging prints everywhere and seeing where it turns into gold would have found this faster than hours 20:31:29 Perhaps I should have asked this first: how do I debug print in this project? I tried to figure it out a few weeks back but couldn't figure it out and got distracted. 20:32:34 (I suspect that I've been too hesitant to ask dumb questions before this week for fear of bothering folks) 20:33:08 I do it with mrpf() same as the game prints everything else. There is dprf() but that's not necessary or useful for short-term stuff just debugging one problem. 20:35:04 Got it. What's the difference between them? Where do they print? 20:35:58 dprf() is used if compiled for debugging output in general (which there is a lot of). mprf() and mpr() are literally how the game tells you "Sigmund hits!" 20:37:04 As in, they (and dprf()) appear in the regular game message log 20:39:10 Makes sense. Thanks! (also, that change has it working now! Yay!!) 20:40:50 I hoped it might although often one just taps into a Matroshka doll of bugs 20:47:43 hahaha, very true. But this time the compiler smiles upon us and the item is appearing and is as functional as I would expect it to be considering I haven't fully hooked it up to be evoked yet. 23:35:39 Unstable branch on crawl.develz.org updated to: 0.35-a0-42-g52370782d7 (34)