00:02:56 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.22-a0-389-g8519d86 (34)
00:09:51 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-389-g8519d868e9 (34)
01:24:41 <Gretell> Unstable branch on crawl.develz.org updated to: 0.22-a0-389-g8519d86 (34)
01:57:44 <Gretell> Windows builds of master branch on crawl.develz.org updated to: 0.22-a0-389-g8519d86
02:30:02 -!- Bammboobies is now known as Bammboo
02:53:29 <Gretell> Monster database of master branch on crawl.develz.org updated to: 0.22-a0-389-g8519d86
03:04:05 <Cheibriados> New branch created: pull/750 (1 commit) 13https://github.com/crawl/crawl/pull/750
03:04:05 <Cheibriados> 03NormalPerson702 07https://github.com/crawl/crawl/pull/750 * 0.22-a0-349-g2bc5a04: Fix potential teleport closets in nooodl_elf_gasket 10(7 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/2bc5a042ef36
03:12:51 <Rotatell> Unstable branch on crawl.beRotato.org updated to: 0.22-a0-389-g8519d86 (34)
03:17:37 <gammafunk> aidanh: re: that rc dump option PR, the way that this option tries to "verify" the user's RC by providing it in the morgue doesn't make sense, since anyone can change their RC at any time, including by saving and loading a different RC at some convenient time or by changing options directly from lua console
03:18:17 <gammafunk> it's not exactly harmful to add an option to include the RC in the dump, but it's not a very good option for its stated purpose
03:19:18 <gammafunk> so I don't know if you've thought about that at all and whether there's just a better general way to fix a users options at game start, or something
03:24:26 <gammafunk> advil: you asked earlier about those webtiles menu hangs, and no, in my last 6 hour session I didn't get any, so looking good so far at least
04:34:51 <Cheibriados> Dragon scales seem intact enough to wear deep under water 13https://crawl.develz.org/mantis/view.php?id=11464 by Leszczynek
05:33:34 <Cheibriados> PoG portals drawn below items 13https://crawl.develz.org/mantis/view.php?id=11465 by Leszczynek
05:49:34 <Cheibriados> Using Shift-X can be used to reveal location within new level 13https://crawl.develz.org/mantis/view.php?id=11466 by RoGGa
07:30:09 <aidanh> gammafunk, thanks, that hadn't occurred to me; in general I'm not too familiar with the environment in which people would be verifying speedrunners' rc files
07:31:24 <aidanh> it's probably a good idea for NormalPerson7 to explicitly specify what the running environment and requirements are; i.e. a threat model
07:36:16 <aidanh> I think some state will need to be stored in the save file no matter the approach; one possibility is to store an array of all used configurations in the save file
08:10:44 <advil> re the speedrunning issue, my suggestion to them was another option that prints a milestone message when the rc file/macros change
08:10:58 <advil> there was a discussion of these issues here: https://www.reddit.com/r/dcss/comments/7vbgng/idea_for_the_real_time_speedrunning_issue/
08:16:23 <wheals> a way to dump rc would be nice anyway if that could lead to an in-game options menu
08:18:56 <advil> the dream is alive
08:29:36 <aidanh> in-game options is probably going to have to be separate from rc files
08:31:29 <aidanh> for options storage, that is; otherwise the rc file has to be modifiable by crawl, which gets complicated fast
08:35:19 <aidanh> in fact there's already a read_persist_options option; not sure what that's about
08:35:33 <aidanh> but I think the hardest part of an in-game options system would be the UI
11:19:00 <Cheibriados> 03NormalPerson702 {MarvinPA} 07* 0.22-a0-390-g8baa81b: Remove Engulf's movement penalty and make it block all god abilities 10(24 hours ago, 6 files, 30+ 22-) 13https://github.com/crawl/crawl/commit/8baa81b392f3
11:19:00 <Cheibriados> 03MarvinPA02 07* 0.22-a0-391-g927ce74: Don't give monsters a movement penalty when escaping engulf 10(3 minutes ago, 1 file, 1+ 6-) 13https://github.com/crawl/crawl/commit/927ce74a54ba
11:20:13 <MarvinPA> hmm did i get the syntax wrong for closing the pr
11:41:03 <NormalPerson7> so gammafunk, re: the dump options thing, it is obvious that it is going to be necessary to record changes to the rc-file to make it work, but given how crawl currently reads in the options file every time the game is loaded and doesn't store the options elsewhere, i don't know how that could be done
11:42:42 <NormalPerson7> (i mean i guess it would be possible to have crawl store the rc-file somewhere hidden and check the player's rc against it every time the game is loaded, but that seems a touch overkill
11:42:52 <NormalPerson7> )
11:59:38 <aidanh> NormalPerson7: you could store an array of distinct rcfile contents in the save
12:00:40 <NormalPerson7> erm yes i guess that would work
12:03:08 <aidanh> but as I said earlier, you should specify what the requirements are before trying to think up a solution
12:04:18 <aidanh> e.g. are the untrusted players running on webtiles? if so, why won't foo.org/crawl/rcfiles/version/foo.rc work?
12:07:11 <Cheibriados> New branch created: pull/751 (1 commit) 13https://github.com/crawl/crawl/pull/751
12:07:11 <Cheibriados> 03NormalPerson702 07https://github.com/crawl/crawl/pull/751 * 0.22-a0-349-gd561435: Fix dragon scales message displayed after item being destroyed (#11464) 10(7 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/d561435f36cc
12:08:23 <NormalPerson7> well the requirements are that we record the rc-file of the game *when the game was played*, and it would be good to record any changes to it already, but that doesn't need to happen before dump_options can be initially merged
12:08:38 <NormalPerson7> *record any changes to it during play
12:09:34 <NormalPerson7> foo.rc returns the current rc-file for that version; that's even more trivial to change than it is to change during the play of the game
12:09:54 <gammafunk> it's not less trivial to change
12:09:56 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-391-g927ce74a54 (34)
12:10:01 <NormalPerson7> but yes all speedrunners currently play on webtiles
12:10:02 <gammafunk> you can save, upload a new RC in seconds, resume play
12:10:13 <NormalPerson7> fine, equally trivial to change
12:10:17 <gammafunk> the only way you would know this happened is to watch the ttyrec
12:11:17 <gammafunk> instead of trying to get so-and-so patch merged as fast as possible, it's better to think of what actually useful changes you'd truly need to verify online runs
12:12:51 <NormalPerson7> sure, i mean i can do - the thing we are very worried about in the realtime community though, more than confirming WR record runs etc, is the verifiability of tournament speedruns
12:13:46 <gammafunk> tournament speedruns don't actually care about this issue
12:14:07 <gammafunk> that is to say, the tournament itself doesn't care
12:14:15 <gammafunk> we do not put conditions on users' RC file at all for scoring the tournament
12:14:56 <gammafunk> we do have a condition for "bots" not being scored, but that's not really the same thing, and we don't exactly have a rigorous definition of bot there either
12:15:30 <gammafunk> if you're saying that you particularly care about verifying the runs in your community for those tournament runs, that's fun, but it's the same issue you have for non-tournament runs
12:15:55 <gammafunk> but the tournament organizers are not trying to add a lot of special RC conditions to score games in the tournament itself
12:16:11 <gammafunk> s/that's fun/that's fine/
12:17:56 <NormalPerson7> erm well we need to define "bot" then first, don't we
12:18:33 <aidanh> that's a separate issue; however bot is defined, it's in terms of the rc file, so you need a reliable way to get the rc file from people you don't trust
12:19:27 <NormalPerson7> yep, which is why we're trying to make dumping options be a thing
12:19:53 <NormalPerson7> and ye recording updates in the same vein
12:20:13 <NormalPerson7> so i guess i will have a look at that array method you mentioned aidanh
12:20:38 <gammafunk> "erm well we need to define "bot" then first, don't we" <- we have one bot that can currently play crawl to any appreciable degree, and this is easy to identify
12:21:15 <NormalPerson7> under your definition of "bot" that is
12:21:47 <gammafunk> yes, the tournament has said "bot games don't get scored"
12:22:10 <gammafunk> so this is pretty straightforward
12:22:45 <gammafunk> you seem to think we're interested in redefining the tournament rules according to what your group of players would like to see in macros, but this isn't the case
12:23:16 <aidanh> as gammafunk pointed out earlier, the dump options PR is currently not going to accomplish much
12:23:27 <NormalPerson7> currently
12:26:27 <wheals> about that PR, maybe the message about the scales should be suppressed entirely if they fall into lava
12:26:33 <wheals> no point just telling the player they missed out
12:27:09 <NormalPerson7> both messages, or which message?
12:27:36 <aidanh> you want to spare them that sense of disappointment!?
12:28:36 <wheals> about them being intact
12:28:39 <wheals> i mean, they're not really intact if they fell into lava
12:30:02 <aidanh> oh NormalPerson7 btw that's not what I meant by using [[ and ]] for multiline string literals
12:30:22 <NormalPerson7> no? what did you mean?
12:30:59 <NormalPerson7> ok so just say "You hear a splash", I mean that would be more difficult to do, but i guess we could try
12:32:12 <aidanh> NormalPerson7: there's an example here https://www.lua.org/manual/5.1/manual.html#2.1 ; the `a = ...` code bit
12:35:29 <NormalPerson7> right so ignoring the indent, i mean i tried that, just with the indent and unsurprisingly there were many extra spaces
12:36:17 <aidanh> well just use "" then
12:53:06 <NormalPerson7> btw i realised the things i called teleport closets probably aren't teleport closets
12:53:25 <NormalPerson7> btw i realised the things i called teleport closets probably aren't teleport closets
12:53:40 <NormalPerson7> but it's still weird to see earth elementals spawning in a hole in the wall when you have antennae
12:57:19 <NormalPerson7> oh aidanh, apparently the PR was merged and i didn't realise, so nvm about that i guess
12:58:00 <aidanh> oh no worries then
13:02:26 -!- amalloy_ is now known as amalloy
13:04:22 <gammafunk> it may not be a teleport closet but there's not a good reason to have diggable rock randomly encased in stone anyhow
13:05:29 <NormalPerson7> yep (but would you like me to clarify the commit message as a result, is the question i'm getting at?)
13:07:49 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.22-a0-391-g927ce74 (34)
13:07:52 <gammafunk> it doesn't need clarification, but you did miss a rock tile
13:08:17 <gammafunk> https://github.com/NormalPerson7/crawl/blob/2bc5a042ef360a2a0da8d7efa34fb7bc13476ec6/crawl-ref/source/dat/des/branches/elf.des#L2477
13:08:28 <amalloy> MarvinPA: the new Engulf status text doesn't render correctly before you've started a game
13:08:29 <gammafunk> oh and some others below that
13:08:50 <gammafunk> I'll just fix it
13:10:16 <gammafunk> why is this vault asymmetrical
13:11:47 <ProzacElf> fears symmetry?
13:12:34 -!- MadCoyote is now known as FunkyBomb
13:13:02 <gammafunk> I'll have to look into fixing that in another commit
13:13:13 <gammafunk> a bunch of vaults have asymmetrical rock padding
13:21:02 <Cheibriados> 03gammafunk02 07* 0.22-a0-392-g5dc3fd9: Replace encased rock with stone in an Elf vault (NormalPerson7) 10(7 minutes ago, 1 file, 6+ 6-) 13https://github.com/crawl/crawl/commit/5dc3fd9af9f0
13:21:29 <gammafunk> aidanh: btw, I saw you closing a PR manually; not sure if you know about closing a PR through a commit message like I do in that commit above
13:22:20 <gammafunk> obv just in cases where you address a PR in a different commit and don't want to directly merge it, or where you only partially merge a PR's commits, but it's handy
13:28:27 <NormalPerson7> i guess i didn't look thoroughly enough, thx for that
13:30:46 <aidanh> gammafunk: with the Closes: #foo. text? nifty, thanks
13:30:53 <gammafunk> yeah
13:31:35 <MarvinPA> amalloy: i don't follow, how does it display before you've started a game?
13:32:07 <amalloy> MarvinPA: start crawl, press ?/t
13:32:12 <amalloy> from the main menu
13:32:19 <gammafunk> doh
13:32:21 <MarvinPA> oh right i forgot about that
13:32:27 <MarvinPA> but it looks like it still works fine anyway?
13:32:47 <amalloy> Engulf         {{[string "db_embedded_lua"]:4: attempt to call field 'res_drowning' (a nil value)}}
13:32:48 <NormalPerson7> oh right ye i forgot about that too
13:32:57 <amalloy> this is offline console fwiw
13:32:58 <MarvinPA> i'm seeing the standard desc as if i didn't have res drown
13:33:06 <gammafunk> amalloy: do you need to recompile?
13:33:07 <MarvinPA> also offline console, weird
13:33:09 <gammafunk> I just realize I hadn't
13:33:12 <amalloy> oh, maybe
13:33:18 <gammafunk> and yeah I see the same error
13:33:27 <MarvinPA> ah yeah the commit adds that lua hook
13:33:33 <amalloy> yeah, that makes sense
13:34:02 <gammafunk> where is our 32 core build server so we can just run crawl remotely and not recompile
13:34:06 <NormalPerson7> yep works fine for me
13:34:14 -!- Euph0ria is now known as TabMasher
13:34:38 <amalloy> yes, works fine after a rebuild. sorry for the false alarm
13:40:33 <advil> re rcfile changes, the simplest thing would be to store a hash (not the whole file at each version) and just print that it has changed...but yeah, defining exact requirements might help with this process.
13:41:04 <advil> I think having an rc option to print rcfile is independently useful, though. In fact it might be worth adding to crashlogs
13:45:04 <aidanh> I thought of using rcfile hashes, but hashes are opaque to users and a pain to verify, especially with things like line-ending differences
13:45:41 <advil> well I don't mean printing the hashes
13:46:29 <advil> just printing "the rcfile changed" in the milestone list
13:47:11 <aidanh> oh, duh, yeah that'd be sufficient
13:49:13 <gammafunk> advil: it's not that simple even then, because of lua console
13:49:25 <gammafunk> you can change options on-the-fly within crawl
13:49:38 <advil> add an option to disable that
13:50:01 <gammafunk> it'd only enabled if someone has a keybinding for it I think
13:50:51 <advil> ah...I think part of the idea would be to also print keybindings / changes to keybindings
13:51:17 <gammafunk> oh, right, in-game macro changes
13:51:23 <gammafunk> so it's not even restricted to RC
13:51:30 <amalloy> well, in-game macro changes are visible to anyone watching the ttyrec
13:51:31 <advil> yeah
13:51:41 <gammafunk> amalloy: this is trying to avoid watching the ttyrec
13:51:47 <advil> I guess these speedrunners are just very untrustworthy people or something
13:52:02 <gammafunk> the whole premise of this idea is that you would "know for sure" that someone has a given RC and set of keybinding/macros the entire game
13:52:30 <gammafunk> so if it's not actually ensuring that, it's not better than just watching the ttyrec
13:53:17 <gammafunk> maybe advil's milestone idea would be sufficient if it can detect changes to RC file, keybinding, but yeah part of this feels like it's not actually well-defined
13:53:17 <amalloy> i thought as far as speedrunners were concerned, watching a ttyrec to verify a speedrun isn't *that* bad, the problem is that you currently can't tell even from watching what they secretly set up pre-game in an rcfile
13:53:17 <gammafunk> do you want to prevent users from changing their macros from RC-defined ones?
13:53:29 <gammafunk> amalloy: yes, and I agree, but these people are trying to add options to help
13:53:45 <gammafunk> which I don't think is necessarilly the right approach, but we do have some fidelity that other communities don't
13:54:09 <gammafunk> since we're running the crawl binary on the servers, some verification can be made that could theoretically prevent the need to vet ttyrecs/video
13:55:08 <NormalPerson7> gammafunk (i assume you're asking me with that question): not necessarily
13:55:29 <gammafunk> I'm not asking anyone in particular, just speaking in general
13:55:30 <amalloy> aside: i'd be interested to see the sequence of keys someone pressed for a whole game of crawl, sorta the inverse of a ttyrec. compare demise's 5-paragraph essays with my copy of War & Peace
13:55:52 <advil> that would also be handy as training data
13:56:17 <amalloy> training what?
13:56:30 <advil> anything!
13:56:35 <advil> RNNs though probably
13:56:57 <gammafunk> amalloy: probably we could save a subset of the json from webtiles
13:57:18 <NormalPerson7> ah right gamma yes i guess that thing about in-game macro changes being visible in the ttyrec means dumping the macros is less of a priority
13:57:50 <gammafunk> well, see this is why this exercise is not making a lot of sense
13:58:06 <gammafunk> if you're just going to have to watch the ttyrec to verify the run in the first place, or watch a submitted video
13:58:37 <gammafunk> you don't have a reason to go through lots of effort to detect RC changes, macro changes, etc
13:58:50 <NormalPerson7> in which case we should dump rcfile/macro changes to make that not necessary
13:59:11 <gammafunk> yes, if you could reliably detect any "bad change", there would not be a need, but what is a "bad change"
13:59:46 <gammafunk> what conditions do you need to have set at the beginning of the run so that if they remain constant throughout the run, the run is valid
14:00:16 <advil> one minimalist solution to the rcfile issue would be (in combination with the current PR) an rcfile option that disables S
14:00:39 <NormalPerson7> i mean there are pretty massive issues with that
14:01:05 <gammafunk> advil: I don't think this will work since you probably can induce a crash and that might not be easy to detect?
14:01:15 <gammafunk> not sure, but sometimes crawl will force a save
14:01:20 <NormalPerson7> yea - what happens if you close the tab?
14:01:33 <gammafunk> yeah, of course that's true as well
14:01:34 <advil> that's true
14:01:53 <advil> so don't save at all :)
14:02:16 <advil> but I think it might help to have some kind of document that lays out what the speedrun community wants here
14:02:38 <advil> I think these discussions keep going around many of the same points that were discussed in that reddit thread already
14:02:40 <NormalPerson7> yes i have let other speedrunners know that this would be helpful
14:02:46 <gammafunk> right, I don't think this is fully well-defined, and I'm not sure said community is actually interested in vetting these runs
14:02:58 <advil> well, from that discussion, many of them definitely are
14:03:21 <gammafunk> many of them have talked about it, but it seems they don't want to verify runs like other communities of speedrunners do
14:04:47 <gammafunk> if it can be made well-defined in a way that is addressable with RC options and milestones where the definition isn't going to change month to month, then having a set of changes to vet runs will make sense
14:09:53 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-392-g5dc3fd9af9 (34)
14:23:17 -!- crate_ is now known as crate
14:27:25 -!- GiantOwl is now known as Kalir
14:37:54 -!- thurin is now known as uprime
14:41:44 <aidanh> any way to get stats on options use?
14:43:07 <gammafunk> not unless you want to download all the RCs and analyze them
14:43:44 <gammafunk> would be pretty hard to even parse them reliably, but you could probably do something basic
14:45:05 <amalloy> well, to be fair we do have a program that is excellent at parsing crawl rc files
14:51:08 <gammafunk> sorry, but this is a no-crawlcode zone
14:58:01 <aidanh> that sounds like more work than just continuing to support the option (easy_exit_menu in this case)
14:58:13 <aidanh> of course, I could just r e m o v e it
15:01:03 <advil> it is a problem with options that people use them
15:03:26 <advil> that one sounds like once upon a time crawl had the other behavior and so it was changed gradually via option to prevent people freaking out
15:04:27 <advil> %git 53e26a9dce81
15:04:28 <Cheibriados> 07galehar02 {jpeg} * 0.8.0-a1-75-g53e26a9: Default easy_exit_menu to false. 10(7 years ago, 3 files, 3+ 3-) 13https://github.com/crawl/crawl/commit/53e26a9dce81
15:05:44 <gammafunk> galehar of the past gives aidanh his blessing for removal
15:06:03 <advil> galehar *and* jpeg!
15:07:15 <advil> apparently some CDO rcs are indexed by google
15:07:59 <aidanh> well well, I guess devs from ancient times aren't all bad after all
15:08:17 <aidanh> who'd have thunk it
16:09:15 -!- jim_ is now known as Guest30747
16:22:35 <Cheibriados> 03NormalPerson702 07https://github.com/crawl/crawl/pull/751 * 0.22-a0-349-gc4999fe: Dont tell player of dragon scales creation when to be destroyed (#11464) 10(4 hours ago, 1 file, 5+ 2-) 13https://github.com/crawl/crawl/commit/c4999fe891ae
17:13:39 -!- Krakhan|2 is now known as Krakhan
17:30:56 <gammafunk> ^status
17:30:56 <Rotatell> 44 Crawlers. CBRO disk usage=99% (135GB) | RAM usage=51% (4GB)| uptime/CPU= 17:30:55 up 113 days, 21:06,  2 users,  load average: 1.28, 1.01, 0.96 (4 Cores) http://status.berotato.org
18:18:52 <Lantell> Unstable branch on underhound.eu updated to: 0.22-a0-392-g5dc3fd9af9 (34)
19:02:47 <gammafunk> ^status
19:02:47 <Rotatell> 43 Crawlers. CBRO disk usage=99% (135GB) | RAM usage=47% (4GB)| uptime/CPU= 19:02:31 up 113 days, 22:38,  4 users,  load average: 0.76, 1.02, 0.97 (4 Cores) http://status.berotato.org
19:02:47 <gammafunk> johnstein: letting you know that cbro at 99% disk again
19:18:21 <johnstein> that was fast
19:18:24 -!- eb_ is now known as eb
19:26:12 <gammafunk> yeah, I was surprised, but didn't have any idea what tends to cause disk usage increases like that
19:26:50 <gammafunk> MarvinPA: A question for you, it's annoying to use static discharge and chain lit around firewood (plants, fungi) because it bounces to those, so you have to lure monsters away from that to use those spells
19:26:59 <gammafunk> Do you see a problem with having the arcs not bounce to firewood?
19:28:15 <wheals> imagine how the players who rely on this tech for nikola will respond
19:29:28 <gammafunk> !lg wheals cikiller=nikola
19:29:28 <Sequell> 1. wheals the Shatterer (L15 GrGl of Makhleb), blasted by Nikola (chain lightning) on D:14 on 2013-11-06 15:22:40, with 92814 points after 29840 turns and 1:15:44.
19:29:35 <gammafunk> tsk tsk tsk, not using the secret tech
19:30:45 <gammafunk> that game was a long time ago, wheals has learned since then
19:31:04 <wheals> i wonder if gargoyles were missing relec at that time
19:31:12 <gammafunk> ooh, was that a grotesk?
19:31:21 <wheals> not but that got reverted back and forth a few times
19:31:34 <wheals> iirc
19:31:49 <gammafunk> did it really
19:32:16 <wheals> %git 7f6b743bca1
19:32:16 <Cheibriados> 07kilobyte02 * 0.14-a0-904-g7f6b743: Revert "Revert "Revert "Don't give Gargoyles innate rElec""" 10(4 years, 5 months ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/7f6b743bca1a
19:32:23 <gammafunk> ah, the relec, sure
19:32:33 <gammafunk> I thought you meant the name for some reason
19:32:42 <wheals> oh, no
19:32:44 <gammafunk> yeah I could see that going back and forth
19:33:07 <wheals> also that commit was 10 days after my death, so it seems gargoyles did indeed lack rElec at the time
19:33:12 <wheals> which explains how i managed to die
19:35:25 <gammafunk> yeah, the wheals of 2013 was basically unkillable otherwise
19:36:11 <wheals> !lg . s=start(year) o=start(year) / won -graph
19:36:12 <Sequell> Unknown function: start
19:36:31 <wheals> !lg . s=year(start) o=year(start) / won -graph
19:36:32 <Sequell> 29/3672 games for wheals: https://shalott.org/graphs/3693d858b45eef3ba9a0da659dbfea24115e0924.html
19:37:36 <gammafunk> wonder what happened in 2014
19:37:49 <gammafunk> !lg . s=year(start) o=year(start) / won -graph
19:37:54 <Sequell> 89/799 games for gammafunk: https://shalott.org/graphs/f0ac7a2e810f7a1cb8e95075963ec499357e07ed.html
19:38:04 <wheals> are these graphs working for you?
19:38:31 <gammafunk> yeah
19:38:33 <gammafunk> don't for you?
19:38:35 <gammafunk> I'm on chrome
19:38:42 <gammafunk> I've heard others in firefox complain about them
19:38:54 <wheals> yeah, switching to chrome fixes it
19:39:00 <wheals> they definitely used to work in firefox
19:41:00 <wheals> what happened in 2014 is that i figured out how to actually play the game correctly
19:41:05 <wheals> well, what passes for correctly for me
20:17:29 <amalloy> wheals: it's just an svg file. dunno how that would stop working in firefox. maybe see if there's any help in the javascript console?
20:18:33 <gammafunk> that page also loads some css iirc
20:19:35 <gammafunk> I guess it's generating the image from javascript
20:20:37 <gammafunk> it's loading some google jsapi
20:20:57 <gammafunk> and calling some google visualization function to generate the image
20:21:03 <gammafunk> what I'm saying is that this is probably amalloy's fault
20:22:49 <amalloy> okay, i guess you're right that it's not just an svg. i looked at it and saw it was an svg once it was rendered, but i guess it is rendered client-side with javascript
20:24:17 <gammafunk> yeah it is using some google javascript library to help make that svg I guess
20:24:28 <gammafunk> perhaps there's something chrome-specific about that library?
20:53:35 <advil> wheals: yeah, they stopped consistently working for me in firefox/safari (ios and mac) maybe around a month ago
20:54:05 <advil> used to work before that
20:56:23 <gammafunk> obviously a google plot to ruin firefox's marketshare
20:57:07 <gammafunk> heh, "google plot"
20:57:53 <advil> it's specifically s=time graphs, I think
21:12:38 -!- amalloy is now known as amalloy_