00:11:12 <06m​umra> @dracoomega Thanks for the feedback. Certainly makes things rather simpler if the rules are relaxed on walls, though you can see the intent. Partly this was to reduce cases where they one unexpectedly appears next to you while resting and damage you (was considering having separate handling so they won't attack you immediately if they've just come into view; I'm just concerned about a player at low HP getting killing by a monster they 00:11:13 couldn't even have known was there). And partly that if they spawn deep inside rock walls there's a chance they just never wake up - I should probably still leave the check there for random spawn positions, otherwise due to how monster placement is done, lots of them can spawn deep inside rock, if it's a level with large rock regions 00:12:48 <06m​umra> I'm surprised M_SURROUND isn't working, it's pretty much the last change i made and it was working for me, maybe there's something on the level you were on. Were you in open space or a 1-tile corridor? If the latter, were they going in the walls at all ? 00:16:47 <06m​umra> The "piercing beams" comment IS intentionally worded that way, but I'll clarify it (and double check the associated code). The issue is that a beam is allowed into the wall to his the monster, but if it's a piercing beam it could now carry on out through the other side, transcending solid matter, which would clearly be hugely broken (although i guess, it actually applies for all beams because this could happen with a normal beam if 00:16:48 the monster evaded... it's just this would always happen). But for bouncing beams, the bounce is going to get handled elsewhere, so it's fine to let them continue at that point; they will effectively clip one tile inside the wall and then bounce back out 00:22:13 <06m​umra> (Side thought, not specifically related to any of your comments, but I'm wondering about removing the ASSERT from cloud creation when clouds are created in walls, and just don't create the cloud if it's in a wall. This would actually simplify code all over the place but especially with beams that also create clouds, since we have to check everywhere that it's not inside a wall, instead could just try to create them and not have to double 00:22:14 check the terrain constantly. It's also the biggest reason for having to use cell_is_solid over cell_is_invalid_target and it'd be nice to not have to worry about it, the only time you'd need to think about it would be when targetting something that is specifically creating clouds as its main effect) 00:35:04 <04C​gettys> Or have two public methods, one of which is maybe_create_cloud and the other which is must_create_cloud, something like that 02:11:27 <06m​umra> I mean sure, but it also just feels like something that dates back to a time when a) there were more targettable cloud effects available to the player and b) targetters themselves weren't so advanced so you couldnt always see in advance where a particular cloud was supposed to go 02:17:32 <06m​umra> Most of the time we just want to throw down a load of clouds in a radius within LOS, or use them as nice visual effects like dust clouds, magic smoke etc. In either case as long as the visual targetting is correct clouds only end up where they're supposed to, everything is fine. (Conceivably as well, if a monster is in a wall, you might still want to target it with a cloud effect in order to cover adjacent floor tiles, so if it moves 02:17:33 off the wall it's in the clouds) 02:34:44 <06m​umra> Cases where its critical that a cloud got created on the tile we intended are like ... Volatile blastmotes? And its impossible for a player to be inside a wall without triggering a separate assert anyway 03:33:32 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 04:49:07 <06m​umra> (Could also have clouds go to an adjacent tile if they try to place in a wall, but that could cause more unexpected stuff in some other cases) But semantically, asserts are there to trap unhandled and unexpected situations. If its handled sensibly, it's not an assert (asserting if a cloud still somehow ends up IN a wall, maybe a good idea) 05:23:07 <06m​umra> Invocations manual clearly 05:31:56 Unstable branch on crawl.akrasiac.org updated to: 0.33-a0-1169-gcefcf6c (34) 06:00:43 <06m​umra> "Circle of Glyphs" was going for a wordplay on "circle of life" but it doesn't land nor make sense thematically 😅 these all sound like good changes anyway thanks 06:14:27 Horn of Gery can't be used when silence but Gravitambourine can which opens a lot of question 06:14:52 is the sound component important for one but for the other it's just fluff and it just blasts gravity? 06:20:59 does the gaym tell you about this 06:40:13 <04d​racoomega> @mumra Partly this was to reduce cases where they one unexpectedly appears next to you while resting and damage you Well, the player should still be alerted when one comes into LoS (no matter how much wall they can walk through) and besides: -This can happen with normal monsters when you're by corners -This can still happen with wall-walkers, just inconsistently based on the terrain on the other side of the wall I do think it 06:40:13 makes sense for random spawn positions, so that they spawn in a place the player could actually encounter. (In fact, we do something somewhat similar for water monsters in Shoals, where we don't spawn them in deep water, since they might end up by the edge of the level where the player will never see them.) But for normal movement, I think these restrictions cause more 'inconsistent'-seeming behavior than they are helpful. I didn't test M_SURROUND 06:40:14 thoroughly (since I thought it might have been 'unfinished') but I was just in a perfectly normal-seeming diagonal corridor, where one lichen was next to me and another 5 were piled behind it. I think removing the assert for clouds in walls makes sense. place_cloud is already not guaranteed to place a cloud (if there's another cloud at the tile that it cannot overwrite). Simply not doing anything if the cell is solid seems fine to me. If something 06:40:14 really counts on placing a specific number of clouds, place_cloud could always return a bool for whether it actually placed one or not. ...in fact, there may be some 'inconsistent' behavior at present with uses of check_place_cloud (which automatically veto if there's a wall/cloud at the destination), which ignores the cloud-overwriting hierarchy. (ie: it will refuse to replace even a decorative cloud; it probably predates that logic altogether.) I 06:40:15 haven't yet, but it probably makes sense to review these and see if any of the uses of that cause resultant weirdness. 06:49:26 03code282802 07https://github.com/crawl/crawl/pull/4181 * 0.33-a0-599-g91c2373520: Don't assign non-draconian monsters a base_type 10(10 minutes ago, 1 file, 12+ 12-) 13https://github.com/crawl/crawl/commit/91c237352039 08:00:24 <06m​umra> Hmm, maybe there's an issue with diagonal corridors, I maybe only tested it in horizontals and verticals ... (alternatively, I'm wondering if there were no_tele flags around stopping them using some walls, and removing that restriction will certainly prevent various potential weirdness) 08:08:57 <06m​umra> Re runed vaults, I did have a way in mind to do this a slightly better way so the walls can be properly flagged as "no_monster_pass" or similar named flag, until the player breaches the door at which point flood fill a removal of said flag - but yes, simplest to not worry about this for now and maybe implement later if nicolae anyone wants to use them in such a place 08:10:53 03WizardIke02 07https://github.com/crawl/crawl/pull/4383 * 0.33-a0-1135-ged2abdb6ae: Speed up tiles rendering, especially when resting 10(13 days ago, 23 files, 1257+ 474-) 13https://github.com/crawl/crawl/commit/ed2abdb6ae2c 08:36:18 <06m​umra> !tell Master_Clown I went through the same thought process a while back re Gravitambourine, and pushed a change to make it silenceable; this was reverted ultimately as it was concluded it's not the sound that causes the affect but rather the vibration 08:36:19 <04C​erebot> docpaisley: OK, I'll let master_clown know. 08:39:35 <06m​umra> !tell Master_Clown On a gameplay level, evokers generally ignore silence, and the tambourine is in rotation with another item that definitely shouldn't be affected by silence, and we want their power levels to be in the same ballpark really. The horn's effect is rather more clearly caused by the noise, and it's a one-off guaranteed item not in a rotation so it's more reasonable for it do have additional disadvantages. 08:39:36 <04C​erebot> Maximum message length is 340 characters, but you had 354. Eschew verbosity, Gladys! 08:40:08 <06m​umra> !tell Master_Clown On a gameplay level, evokers generally ignore silence, and the tambourine is in rotation with another item that definitely shouldn't be affected by silence, and we want their power levels to be in the same ballpark really. The horn's effect is rather more clearly caused by the noise, and it's a one-off guaranteed item not in a rotation so it's more reasonable for it to be disadvantaged. 08:40:09 <04C​erebot> Maximum message length is 340 characters, but you had 354. Eschew verbosity, Gladys! 08:41:18 <06m​umra> !tell Master_Clown On a gameplay level, evokers generally ignore silence, and the tambourine is in rotation with another item that definitely shouldn't be affected by silence, and we want their power levels to be in the same ballpark. The horn's effect is more clearly caused by the noise, and it's a one-off item not in a rotation so it's more reasonable for it to be different. 08:41:19 <04C​erebot> Maximum message length is 340 characters, but you had 354. Eschew verbosity, Gladys! 08:41:34 <06m​umra> Please cerebot, it can't still be 354 chars 08:42:11 <06m​umra> !tell Master_Clown On a gameplay level, evokers generally ignore silence, and the tambourine is in rotation with another item that definitely shouldn't be affected by silence, and we want their power levels to be in the same ballpark. 08:42:11 <04C​erebot> docpaisley: OK, I'll let master_clown know. 08:42:40 <06m​umra> !tell Master_Clown The horn's effect is rather more clearly caused by the noise, and it's a one-off guaranteed item not in a rotation so it's more reasonable for it to be more disadvantaged. 08:42:41 <04C​erebot> docpaisley: OK, I'll let master_clown know. 09:46:04 03DracoOmega02 07* 0.33-a0-1170-g8bb8084e21: Warn the player when firing Mule could trigger a trap (CarefulOdds) 10(66 seconds ago, 1 file, 13+ 0-) 13https://github.com/crawl/crawl/commit/8bb8084e21a1 10:16:45 03WizardIke02 {GitHub} 07* 0.33-a0-1171-g7451b4b137: Fix closing doors with the mouse (#4276) 10(37 seconds ago, 5 files, 51+ 38-) 13https://github.com/crawl/crawl/commit/7451b4b13742 10:53:48 03gammafunk02 07* 0.33-a0-1172-gd2c03782b8: Fix a DES typo (AutarchDCSS) 10(40 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/d2c03782b806 11:10:37 04Build failed for 08master @ d2c03782 06https://github.com/crawl/crawl/actions/runs/14158524770 11:43:32 New branch created: pull/4416 (1 commit) 13https://github.com/crawl/crawl/pull/4416 11:43:33 03WizardIke02 07https://github.com/crawl/crawl/pull/4416 * 0.33-a0-1173-ge022bf94b4: Fix harmless beams prompting when hitting allies (cool 3) 10(17 minutes ago, 1 file, 3+ 0-) 13https://github.com/crawl/crawl/commit/e022bf94b49e 13:42:07 New branch created: pricingredux (10 commits) 13https://github.com/crawl/crawl/tree/pricingredux 13:42:08 03hellmonk02 07[pricingredux] * 0.33-a0-1090-g40b285ebfa: New spell: Detonation Catalyst 10(10 days ago, 18 files, 195+ 11-) 13https://github.com/crawl/crawl/commit/40b285ebfa63 13:42:08 03hellmonk02 07[pricingredux] * 0.33-a0-1091-gc4b0b63fa1: adjustments and fixes 10(7 days ago, 2 files, 24+ 9-) 13https://github.com/crawl/crawl/commit/c4b0b63fa11a 13:42:08 03hellmonk02 07[pricingredux] * 0.33-a0-1092-gc8846ca573: add missing random 10(7 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/c8846ca57390 13:42:08 03DracoOmega02 07[pricingredux] * 0.33-a0-1093-gef176482a4: Fix normal melee attacks always extending Catalyst by 1.0 turns 10(6 days ago, 2 files, 3+ 8-) 13https://github.com/crawl/crawl/commit/ef176482a4e6 13:42:08 03DracoOmega02 07[pricingredux] * 0.33-a0-1094-g06fce89e22: Make Detonation Catalyst explosions actually make noise 10(6 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/06fce89e226f 13:42:08 03DracoOmega02 07[pricingredux] * 0.33-a0-1095-g2920ae7f87: Make Detonation Catalyst actually care about which coglin weapon hit 10(6 days ago, 5 files, 38+ 28-) 13https://github.com/crawl/crawl/commit/2920ae7f8785 13:42:08 03hellmonk02 07[pricingredux] * 0.33-a0-1096-g1376d7c62f: rework damage (again) 10(6 days ago, 1 file, 7+ 4-) 13https://github.com/crawl/crawl/commit/1376d7c62f02 13:42:08 03gammafunk02 07[pricingredux] * 0.33-a0-1160-g0c3a307c66: Merge branch 'master' into theyrebombs 10(22 hours ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/0c3a307c6633 13:42:08 03hellmonk02 07[pricingredux] * 0.33-a0-1161-gddfbbbf091: damage back up 10(16 hours ago, 1 file, 4+ 3-) 13https://github.com/crawl/crawl/commit/ddfbbbf09106 13:42:08 03hellmonk02 07[pricingredux] * 0.33-a0-1162-g66d3e10174: adjust shop prices 10(3 minutes ago, 4 files, 79+ 55-) 13https://github.com/crawl/crawl/commit/66d3e101745c 13:42:50 <09h​ellmonk> hmm, fucked this branch up again, 1 sec 14:05:27 New branch created: pricingthreedux (1 commit) 13https://github.com/crawl/crawl/tree/pricingthreedux 14:05:27 03hellmonk02 07[pricingthreedux] * 0.33-a0-1173-g09ef983c2a: adjust shop pricing 10(2 minutes ago, 4 files, 78+ 54-) 13https://github.com/crawl/crawl/commit/09ef983c2ac0 14:09:41 New branch created: pull/4417 (1 commit) 13https://github.com/crawl/crawl/pull/4417 14:09:41 Branch pull/4417 updated to be equal with pricingthreedux: 13https://github.com/crawl/crawl/pull/4417 14:28:45 03WizardIke02 {GitHub} 07* 0.33-a0-1173-g17c16fb7e5: Fix harmless beams prompting when hitting allies (cool 3) (#4416) 10(23 seconds ago, 1 file, 3+ 0-) 13https://github.com/crawl/crawl/commit/17c16fb7e507 14:37:51 <09h​ellmonk> #4073 looks mergeable 14:39:22 03Hellmonk02 {GitHub} 07* 0.33-a0-1174-g57531695c7: Adjust shop pricing (#4417) 10(20 seconds ago, 4 files, 78+ 54-) 13https://github.com/crawl/crawl/commit/57531695c7e5 14:50:40 <06m​umra> @dracoomega Aha, on some testing with M_SURROUND in diagonal corridors, yes it appears that particular geometry confounds the logic (i think due to some LOS checks being blocked in odd ways). easy fix tho 15:01:47 <06m​umra> Well done for finding the buggy case straight away without even seeing the correct behaviour first 😂 15:02:51 <04d​racoomega> Haha 15:03:14 <04d​racoomega> I mean, as I said, I didn't test it so exhaustively since I just assumed it might have been part of it being unfinished, and focused on the other stuff first 15:13:15 03Heinrich Ody02 {GitHub} 07* 0.33-a0-1175-g5d77c6c51e: Optionally add target location description (#4073) 10(24 seconds ago, 2 files, 22+ 2-) 13https://github.com/crawl/crawl/commit/5d77c6c51ed7 15:21:21 <09h​ellmonk> reading the github wiki and learning about proposals to remove mindelay 15:21:28 <09h​ellmonk> is it finally time... 15:24:20 <02M​onkooky> I for one am in favor of remove mindelay no other changes 15:24:58 <02M​onkooky> let me make -10 aut dagger attacks 15:32:03 <08n​icolae> whenever i see the word "mindelay" the opening riff of this plays in my head: https://www.youtube.com/watch?v=aeJJR_uMYlQ 15:37:38 03dolorous02 07* 0.33-a0-1176-g74edd5530c: Add another adjective to artefact books. 10(2 minutes ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/74edd5530cc1 15:40:30 Unstable branch on underhound.eu updated to: 0.33-a0-1174-g57531695c7 (34) 15:43:24 <08n​icolae> a report of an almost irrelevant bug: the opening comment in necropolis.des cuts off mid-line: # Many cautious players will ignore earlier subva 15:58:38 <06m​umra> yes it's missing a 'y', it's a sandwich shop for vampires 16:44:13 <09h​ellmonk> https://github.com/crawl/crawl/wiki/Level-Feelings 16:44:28 <09h​ellmonk> wrote a thing about the level feeling stuff I've been kicking around 16:46:29 <09h​ellmonk> need implojin to come back, he's prob got notes for this 17:13:53 <02M​onkooky> riffing on your bullet 2 positive: feature gives a one-time moderately long buff- e.g. fountain of might gives you 50 turns of might 17:14:31 <09h​ellmonk> Sure, an expiring consumable 17:14:48 <02M​onkooky> so you've got incentive to find and reach the feature, but not incentive to lure back to it 19:05:09 <09h​ellmonk> should I go ahead and close 2786 19:19:21 <09h​ellmonk> would also like an opinion on 3445. It's unclear to me that the changed condition is right, but it does seem plausibly better than what we're doing now 19:19:54 <09h​ellmonk> actually, it might be best to veto troves based on skill total being < some fixed threshold rather than one value being less than another 19:23:08 03Alex Jurkiewicz02 {GitHub} 07* 0.33-a0-1177-gfe11bc45f1: feat(webserver): Add game_id to rcfile_contents (#3496) 10(59 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/fe11bc45f188 19:40:00 03Hellmonk02 {GitHub} 07https://github.com/crawl/crawl/pull/3880 * 0.33-a0-1181-g67168520ea: Merge branch 'master' into fix-coglin-unwields 10(3 minutes ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/67168520ea28 19:43:11 03Charlie Gettys02 {GitHub} 07* 0.33-a0-1178-gf145acab4e: fix: AUTOFIGHT while immotile should not wait (CrawlOdds) (#4399) 10(33 seconds ago, 1 file, 5+ 3-) 13https://github.com/crawl/crawl/commit/f145acab4e77 20:00:08 03radio-gra02 {GitHub} 07* 0.33-a0-1179-g26fe8ce49a: console: draw waypoints over other items and features (#4323) 10(57 seconds ago, 1 file, 6+ 12-) 13https://github.com/crawl/crawl/commit/26fe8ce49ab3 22:25:45 Airwolf (L26 PoFi) ASSERT(defender->alive()) in 'fight.cc' at line 411 failed. (Crypt:1) 22:28:36 Airwolf (L26 PoFi) ASSERT(defender->alive()) in 'fight.cc' at line 411 failed. (Crypt:1) 22:35:31 Unstable branch on crawl.develz.org updated to: 0.33-a0-1179-g26fe8ce49a (34) 22:47:53 <06m​umra> I don't understand the purpose of fe11bc4 (Add game_id to rcfile_contents) and the commit message is completely vauge (and hard to read due to the spacing being all messed up) 22:49:00 <06m​umra> Since the client requested the game_id across the socket, it knows the game_id, I don't see how this simplifies anything or enables any new behaviour 22:58:41 Windows builds of master branch on crawl.develz.org updated to: 0.33-a0-1179-g26fe8ce49a 23:08:47 <06m​umra> Ok I slightly see the point, but now I'm questioning why this property is added when the js client was not updated to read the property 23:10:42 <06m​umra> it'd be totally reasonable to read this property in the front end and check that the rcfile you've received is the one you're trying to edit (theoretically, if you clicked two different "edit rc" buttons very quickly in the lobby, you'd get the wrong rc briefly or even the totally wrong one if the messages arrived out of order (although i'm not convinced the latter is even possible with how the websocket setup works)) 23:12:30 <06m​umra> i feel like if we're going to add an undocumented API feature to websockets then at the very least the correct usage of it should be shown in the frontend (the commit even specifies that this is not even used "But will simplify the design of future webtiles clients" which is rather cryptic) 23:23:50 Unstable branch on cbro.berotato.org updated to: 0.33-a0-1179-g26fe8ce49a (34) 23:42:23 03mumra02 07* 0.33-a0-1180-g1e05e1a6a0: fix(webtiles): Don't flash wrong rcfile if two different "edit rc"s clicked 10(3 minutes ago, 1 file, 4+ 0-) 13https://github.com/crawl/crawl/commit/1e05e1a6a0a8 23:43:28 <06m​umra> Did consider reverting and reopening the PR to request this (and to request fixing the formatting in the original commit) 23:49:55 <06m​umra> I appreciate the desire to get PRs closed but I'm wondering whether this one was really looked at very closely 23:55:38 Monster database of master branch on crawl.develz.org updated to: 0.33-a0-1179-g26fe8ce49a