00:03:53 03CrawlOdds02 07* 0.35-a0-556-g69c1e3e389: Fix some feeling-safe checks 10(8 hours ago, 1 file, 6+ 8-) 13https://github.com/crawl/crawl/commit/69c1e3e389cd 01:08:29 <04dracoomega> Well, hopefully that's the last of the core UI. Now to test like 100 different spells >.> 01:27:23 <04dracoomega> M'thinks this isn't actually what this is: 01:27:24 <04dracoomega> https://cdn.discordapp.com/attachments/747522859361894521/1518895229829120000/image.png?ex=6a3b94ea&is=6a3a436a&hm=05a821915e27ea97d217c86c088bf18d6e316a7638a99a3e8ee715fcf106a39f& 01:27:46 <11Odds> The dreaded zombie squared 01:29:43 <04dracoomega> (So far in this testing, the significant majority of spells work correctly, and sadly we have so many of them that this still means a lot of things to try to fix ^^; ) 01:30:25 <11Odds> Is there an overview of the plan somewhere? 01:35:17 <04dracoomega> I've explained it in bits and pieces at different times, but basically I am changing a bunch of UI regarding invisible monsters to better-track known and recently-known information, and allow most of the interface to treat 'You can't see it but you definitely know it's there' monsters as present in terms of things like autofight and targeting. (And then I'm planning to also make some mechanical tweaks to the invis spell to 01:35:18 actually create ambiguity more often, while also replacing it on some monsters with other spells instead) 01:37:34 <04dracoomega> I've treatened to do this for a couple versions now, and recent conversations prompted me to actually do so 01:38:07 <11Odds> Hooray for no more manually targetting Yaras at the senses invisible thing 🙂 01:38:31 <04dracoomega> I mean, assuming you know where they are for sure ^^; 01:38:39 <04dracoomega> (Which you would if you do it immediately, of course) 01:39:22 <04dracoomega> But spell targeting is handled in a lot of different bespoke ways, which is why I am actually going through spells one by one here. Most of the major cases are already handled fine, but it's hard to anticipate which spells are slightly different about it 01:39:46 <11Odds> Yeah, there is a lot of special targetting logic 01:40:05 <11Odds> With a lot of minor bugs 🙂 01:40:12 <04dracoomega> Undoubtedly! ^^; 01:59:54 03Davidescu Victor02 {CrawlOdds} 07* 0.35-a0-557-g7e1f3d532c: Add a space to menu colour search string 10(25 hours ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/7e1f3d532ca4 02:24:55 <04dracoomega> Well, it looks like there are at least 34 spells that still have some issues I need to fix here >.> 02:25:18 <04dracoomega> (A few of these should be very simple, but I suspect some of them will be rather hard) 02:27:52 <02Monkooky> what are you changing about invis? I've got a half-planned monspell that I'm probably never going to get around to 02:49:22 <04dracoomega> A plan I wanted to try was making the Invis spell include an automatic short-range blink when the monster goes invisible, so that their position was actually fuzzed instead of usually unambiguous. 02:52:16 <02Monkooky> ok so not something fundamentally different about invis itself, but the spell specifically makes sense 02:53:03 <02Monkooky> I was thinking about a pair of spells, one which is just invis self + ally, one which is invis self + ally + swap our positions 02:55:07 <11Odds> Some PRs I'm planning to test/review/merge, in case anyone has given them thought already: - 5287: Refactor passage of golubria - 5283: Don't spend a turn when aborting channeling. - 4914: Change solo tentacles only to attack targets in our LoS - 3340: Enable a confirmation prompt for "bump" attacks with ranged weapons; per the comments probably needs a lot of merging or just redoing at this point Also, 5148 on the elven twins; for this 02:55:07 one the comments suggest we should just stop delaying the buffs, but I'm not so sure because of the messages on empowerment (and we do need to delay anyway for the case where they are frenzied and shouldn't be empowered) 03:00:04 <04dracoomega> I've had some vague thoughts about a unique making use of certain forms of mass invis (whether 'convoke + invis' or 'invisible shadow creature' like a reverse boggart). The sort of thing I think could be interesting when all the UI improvements are in place to actually track known information better. 03:04:18 <04dracoomega> 3340 certainly sounds reasonable (I am surprised it wasn't already the case), but a lot of the code involved has changed quite a bit since then, yes. (I changed a bunch of it myself ^^; ). Suspect it might actually be easier to rewrite. I'm not sure I understand your issues 5148. We're already doing part of Dowan's empowerment when out of sight (and I see nothing wrong with the player not getting a message for something they 03:04:19 didn't see happen - you do already get the 'wail of despair' regardless - when you can tell from the buffs on them upon seeing them, either way. 03:06:14 <06regret-⸸nde※> Staring at our own internal design chart for the current version and talking it out a bit with Draco, you could also possibly put up PRs for a few decent sized pieces of busywork. Don't feel any particularly obligated or pressured to get around to any of these, though. * Ripping out all the Xom mood code, making each Xom effect calling upon mood use the median mood value, and replace the current piety bar with a tension meter. 03:06:15 (Xom's amusement messages can still be kept, but in darkgray rather than yellow.) * This is a desired part of the big Xom overhaul I've planned for several versions now of making Xom only provide strategic gifts with exploration rather than randomly over time, to prevent timing exploring when Xom's in a good mood or thinking that mood management means much at all, and to kill Xomscumming while avoiding people getting overly focused on boredom as the 03:06:15 recent changes there have struggled with. * Make Pan finite, with 11 floors for now, exits between non-rune floors only inside pandemonium lord vaults, and with the exit out of Pan only on Pan:11. (This means ensuring the four rune lords on each trip, save compat code to shove people currently in Pan into Pan:1, and no exits through the Abyss.) * A lot of Hell-Pan Roulette work (+ Tomb revisions + Purgatory) here has been on new monsters and new 03:06:16 tiles, and with not wanting 0.35 to run late after 0.34 did, we're splitting some of this project between versions, which means getting some pieces of it done now and the overall roulette in 0.36. * Make examining spells display what of AC/EV/SH/Will/resists they check or half check, by going through and tagging all the spells in spl-data.h with this. * This is much more tedious than the other two, but there's tons of misconceptions for both new 03:06:16 and experienced players alike around a lot of spells and a lot of awkward spell descriptions trying to mention a few of these points, so it'd be fairly good to just have ingame displays of this. 03:06:25 <11Odds> The message I'm slightly worried about using in particular is Dowan seems to find hidden reserves of power!, because the spellbook upgrade in particular I think is not very visible 03:08:03 <04dracoomega> I wonder how much more obvious the message makes that this is what is actually happening when he hastes, anyway 03:08:21 <04dracoomega> (And if you have any familiarity at all, the presence of haste is unambiguous) 03:08:36 <11Odds> Fair, it's probably not so clear he's not just hasted 03:10:04 <11Odds> Thanks! I think I will have a crack at Pan 03:11:53 <11Odds> So the plan is that you have to do 11 floors of Pan in a row and go through all the non-rune lord vaults? I.e. there's exactly one portal per non-rune floor? 03:13:17 <04dracoomega> Yes 03:14:26 <04dracoomega> (There's a bunch of plans about randomized per-level modifiers for Pan that won't be making it into this version, but the backbone of Pan being finite makes sense to me even before the full roulette happens) 03:14:38 <11Odds> And actually the "hidden reserves of power message" doesn't even appear at the same time as the spellbook upgrade at the moment 😛 03:15:05 <04dracoomega> So many tasks I've worked on this version have turned out to be pretty painstaking and slower than expected >.> 03:15:37 <06regret-⸸nde※> We'll keep the demonic rune for now with some encompass vault plans of mine to mix together hellion island and the other guaranteed demonic rune vaults on the last floor of Pan, which will still function for randomizing the exit out of Pan once the roulette's being started. (I can handle that part.) 03:15:49 <11Odds> Aren't they always 🙂 03:34:23 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 03:42:08 03WizardIke02 07* 0.35-a0-558-g4f1e9c7b8d: Fix some out of view door info leaks 10(35 minutes ago, 11 files, 106+ 59-) 13https://github.com/crawl/crawl/commit/4f1e9c7b8d33 06:32:28 garnell (L14 TeSu) Crash caused by signal #6: Aborted (D:12) 07:35:08 03regret-index02 07* 0.35-a0-559-g19be720228: Don't create star jelly wretches (Undo) 10(43 seconds ago, 1 file, 3+ 2-) 13https://github.com/crawl/crawl/commit/19be7202286d 09:55:46 03CrawlOdds02 07* 0.35-a0-560-g1b0f8bdc6c: Empower the twins even if out of LoS 10(6 hours ago, 2 files, 21+ 27-) 13https://github.com/crawl/crawl/commit/1b0f8bdc6c19 09:55:46 03hellmonk02 {CrawlOdds} 07* 0.35-a0-561-g1b8bb01e35: Refactor passage of golubria portal selection 10(2 weeks ago, 1 file, 22+ 41-) 13https://github.com/crawl/crawl/commit/1b8bb01e3597 09:55:46 03CrawlOdds02 07* 0.35-a0-562-g2dfc88af3e: Change solo tentacles only to attack targets in our LoS 10(7 months ago, 1 file, 9+ 0-) 13https://github.com/crawl/crawl/commit/2dfc88af3e63 09:55:46 03CrawlOdds02 07* 0.35-a0-563-g4eb44f5760: Confirm bump firing with !f inscriptions 10(5 hours ago, 1 file, 6+ 0-) 13https://github.com/crawl/crawl/commit/4eb44f57603e 10:04:13 <04dracoomega> @Odds While the logic used by the tentacle commit seems largely correct, I thought I'd point out that we have a specific method for checking ally/player los for attacks, which seems likely to have been better to use there instead: monster_los_is_valid() 10:05:15 <11Odds> Aha, yes indeed 10:06:18 <11Odds> Now wondering what tentacles outside my LoS do... 10:09:12 <04dracoomega> Yeah, I noticed you didn't check for that part there (but I feel like somewhere else must have already been preventing that) 10:09:24 <04dracoomega> Since at some point in past they were stopped from attacking while out of your LoS, I'm pretty sure 10:10:15 <11Odds> Tentacles' target selection and actually hitting things is rather decoupled though 10:11:06 <11Odds> Turns out tentacles will merrily move towards foes when the player can't see the tentacle, which is probably also bad (though way less common IME) 10:11:40 <11Odds> So probably right to use monster_los_is_valid, I'll check and fix it 10:12:40 <04dracoomega> (We have so many different functions for different purposes that it can be very easy to work on the codebase for quite a while without noticing that some specifically suiting function already existed for something or another.) 10:16:59 <11Odds> (Yep, thanks for pointing them out 🙂 ) 10:25:33 03CrawlOdds02 07* 0.35-a0-564-gc910cec6f5: Fix tentacle check to use appropriate function 10(6 minutes ago, 1 file, 2+ 6-) 13https://github.com/crawl/crawl/commit/c910cec6f559 10:50:01 03CrawlOdds02 07* 0.35-a0-565-g7ec83fe5aa: Fix coglin hunters saying farewell on turn 1 10(9 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/7ec83fe5aabc 11:27:09 <11Odds> Re: https://github.com/crawl/crawl/pull/5150 (making dissipation make -swift go away faster). I tend to think we should either do this or make -swift unaffected by cancellation, yara etc - either it's a cooldown or a dispellable magical effect (Codewise the reason it winds up this way is that -swift is the same duration as +swift, and that duration is dispellable but not negative) 11:30:33 03CrawlOdds02 07* 0.35-a0-566-g9b33262cfa: Update credits 10(48 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/9b33262cfa44 11:34:22 <04dracoomega> Yeah, I was aware of the technical reason for this. Slightly weird old implementation (that probably just seemed 'good enough', I guess). I think I'd be fine with the negative part being separate and undispellable. (Inversely, I have occasionally pondered if it would be interesting if the amulet did actually make cooldowns shorter. (Or some item, anyway). But cooldowns differ so much in their scope and scale that I'm uncertain 11:34:22 about how uneven this might be. Is there stuff that it makes 'too good' in the process?) 11:39:59 <11Odds> I don't use it that much, but Death's Door seems a potentially load-bearing cooldown 11:41:18 <11Odds> Can't think of anything else I'd be especially worried about 11:44:49 <12ge0ff> having a shorter ru's exhaustion would be a bit op 12:42:49 better be op if it cost me an eye and a hand 😤 12:43:18 and also might be an uninformed take but dissipation has always felt /extremely/ meh to me 14:13:23 03CrawlOdds02 07* 0.35-a0-567-g28ce492bd5: Refactor swiftness and make -swift uncancellable 10(55 minutes ago, 10 files, 33+ 33-) 13https://github.com/crawl/crawl/commit/28ce492bd52f 15:44:25 Unstable branch on underhound.eu updated to: 0.35-a0-567-g28ce492bd5 (34) 22:36:04 Unstable branch on crawl.develz.org updated to: 0.35-a0-567-g28ce492bd5 (34) 23:00:10 Windows builds of master branch on crawl.develz.org updated to: 0.35-a0-567-g28ce492bd5 23:34:15 Unstable branch on cbro.berotato.org updated to: 0.35-a0-567-g28ce492bd5 (34) 23:56:25 Monster database of master branch on crawl.develz.org updated to: 0.35-a0-567-g28ce492bd5