IRC logs for #openttd on OFTC at 2020-11-29
⏴ go to previous day
02:49:46 *** m1cr0man has joined #openttd
04:45:54 *** tokai|noir has joined #openttd
04:45:54 *** ChanServ sets mode: +v tokai|noir
05:42:31 *** Gustavo6046 has joined #openttd
08:30:33 *** longtomjr has joined #openttd
08:37:28 *** andythenorth has joined #openttd
08:46:51 *** HerzogDeXtEr has joined #openttd
09:58:41 <TrueBrain> I didn't want to leave you hanging this long
10:01:15 <andythenorth> do I even exist?
10:02:11 <TrueBrain> we still have no confirmation of that
10:02:31 <TrueBrain> but no worries; if not, you are the worlds most advanced AI, so there is that
10:07:28 <andythenorth> I type therefore I am
10:22:07 <TrueBrain> wow .... I deployed my new DorpsGek code .. and ... it just works :o :o
10:22:13 *** jottyfan has joined #openttd
10:29:49 *** jottyfan has joined #openttd
10:32:18 *** sla_ro|master has joined #openttd
10:52:33 *** Progman has joined #openttd
11:01:15 * andythenorth assumes there an adhoc way to filter stdout when running make?
11:02:35 <andythenorth> I have 2068 nml warnings every time I make
11:02:45 <andythenorth> they all contain the same string "returns a constant, optimising"
11:04:23 * andythenorth has tried google :P
11:05:24 <andythenorth> I thought maybe "grep -v 'foo'" but I can't make that work
11:06:35 <andythenorth> hmm, seems I need to learn how nmlc connects to stdout and stderr
11:06:44 *** ChanServ sets mode: +v tokai
11:07:54 <TrueBrain> make 2>&1 | grep -v "returns a constant"
11:07:58 <TrueBrain> possibly that helps
11:08:11 <TrueBrain> as I assume these errors go to stderr, so you first have to pipe stderr to stdout before you can filter
11:18:26 <TrueBrain> okay, during deployment, DorpsGek can lose 60 minutes of Seen database .. not really a lot I can do about that honestly
11:19:25 <dwfreed> before it is replaced, force it to flush the database to disk?
11:19:52 <andythenorth> ha now I found an actual thing to fix
11:19:57 <andythenorth> which was hidden
11:20:29 <TrueBrain> dwfreed: order of execution doesn't allow that :) As it goes with modern services .. the new instance runs before the old one is shut down :)
11:20:34 <TrueBrain> andythenorth: gratz :)
11:20:52 <TrueBrain> okay, I guess I can change the flush interval to 60 seconds or something silly ..
11:21:46 <dwfreed> executing the new one before the old one is dead, in this instance
11:22:10 <TrueBrain> the problem here is running software from 2005 :P
11:23:03 <TrueBrain> it is funny to me, that ECS simply doesn't allow you a strategy where you are not aiming for 100% uptime :)
11:23:24 <TrueBrain> I wonder, does k8s still have a strategy where that is allowed ..
11:23:59 <TrueBrain> k8s still has the ability for that
11:24:07 <TrueBrain> sounds as bad as it is :P
11:31:07 <TrueBrain> okay, I get 50 KB/s .. the data is 1MB .. so that would mean that ever 2 minutes would consume 1/6th of the throughput allocated
11:34:04 *** supermop_Home_ has quit IRC
11:41:40 <andythenorth> now the Iron Horse release dance
11:41:53 <andythenorth> make, bananas, grf.farm, AWS, forums
11:47:43 <TrueBrain> okay, I think I found a way to migrate DorpsGek without lose of data .. always the challenge :)
11:48:54 <TrueBrain> AWS DataSync is pretty nice :)
11:50:49 <andythenorth> pls send me newsletter
11:51:01 <TrueBrain> you are receiving it .. live ....
11:51:09 <andythenorth> I am literally dragging-dropping files to AWS in browser right now
11:51:18 <andythenorth> and waiting for completion so I can invalidate cloudfront
11:51:24 <TrueBrain> AWS DataSync is to sync between AWS data storages
11:51:30 <TrueBrain> well, an on-prem, but you need to do shit for that
11:51:39 <TrueBrain> so it is not for you :)
11:52:06 <andythenorth> I could run the agent as a VM on my laptop? :P
11:52:59 <TrueBrain> could, dunno :P Still, that service is not for your goal :)
11:53:25 <andythenorth> I should probably get GH actions to build tags from my docs repo
11:53:37 <andythenorth> then push them and invalidate cloudfront
11:54:13 <andythenorth> hmm upload to S3 is very slow today
11:54:46 <TrueBrain> hmmmmmmm .. the seen data was not stored .... how to debug this in a sane way ..
11:56:56 <andythenorth> hmm docs repo tagged
11:57:00 <andythenorth> cloudfront invalidated
11:57:04 * andythenorth uploading banans
11:57:44 <andythenorth> now just need to convert the changelog from MD to BBCode
11:57:57 <andythenorth> bump first forum thread
11:58:01 <andythenorth> make some screenshots
11:58:07 <andythenorth> releasing is silly
12:08:54 <TrueBrain> owh supybot .. you never stop surprising me ... "flush config" is also "flush everything else" ... but the documentation doesn't mention that ...
12:15:18 *** frosch123 has joined #openttd
12:20:11 <TrueBrain> well, it is a good thing I was validating this .. otherwise it would have been a while before we noticed the seen-database wasn't being updated :P
12:21:56 <frosch123> but i like "world.flush"
12:22:56 <DorpsGek_III> [OpenTTD/DorpsGek] TrueBrain dismissed a review for pull request #40: Flush more often and deploy on staging https://git.io/Jkxqw
12:23:56 <frosch123> no idea what changed :)
12:24:34 <TrueBrain> "TrueBrain force-pushed the TrueBrain:supybot branch from 3742c32 to f7bac7e 1 minute ago" <- "forced-pushed" is a link, that shows you the diff
12:24:38 <TrueBrain> it is a bit hidden, but very useful
12:25:47 <frosch123> wow, yeah, "bit hidden"
12:26:00 <TrueBrain> that happens if you don't color links :)
13:01:35 *** WormnestAndroid has quit IRC
13:01:48 *** WormnestAndroid has joined #openttd
13:15:04 <FLHerne> is an interesting comparison
13:15:30 <FLHerne> Class 66 (which is pretty much a UK-gauge SD70) next to the full-size edition
14:37:53 <andythenorth> when shipping docs for distribution...
14:37:56 <andythenorth> they're markdown source
14:38:08 <andythenorth> ship as markdown, or render to .txt?
14:38:55 * andythenorth has a high level of disinterest in this :D
14:44:06 <frosch123> didn't nielsm add some formatting for md?
14:44:48 <andythenorth> I'll just dump them to text
14:44:52 *** Progman has joined #openttd
14:47:02 <andythenorth> oof wat, python markdown lib has no text output option
14:47:06 <andythenorth> what a mess this all is
14:47:31 * andythenorth resets everything, it's nonsense
14:47:45 * andythenorth starts a newsletter about format disasters
14:49:39 <frosch123> you can find the md->txt conversion in os.rename
14:50:28 <andythenorth> md is not text :( It's text with unreadable junk in it :(
14:50:31 <frosch123> ottd readme does the same
14:50:34 * andythenorth hates all these formats
14:52:17 <TrueBrain> you do know .txt is not a format at all, right? :)
14:52:49 <TrueBrain> as in that reasoning, no "text" format is without "unreadable" junk
14:52:56 <TrueBrain> take all these " I wrote in the last sentence
14:53:03 <TrueBrain> so .. yeah .. now what?
14:53:22 <TrueBrain> in other words: people who consider "cp README.md README.txt" silly, are silly, by definition ;)
14:54:26 <TrueBrain> rendering .md to .txt .. what a silly concept you invented there :)
14:55:07 <andythenorth> I don't know why it triggers my irrational response
14:55:19 <andythenorth> makes me unspeakably angry, how silly
14:55:29 <andythenorth> I just yelled at my kids because I was cross about it :P
14:56:06 <TrueBrain> well, now you know what caused it, you can deal with it :)
14:56:20 <andythenorth> but there must be a filter to parse out the junk?
14:56:28 <TrueBrain> what "junk" are you on about
14:56:32 <TrueBrain> you are creating a problem where there is none
14:56:39 <andythenorth> stack overflow recommends rendering to html, then using beautiful soup to pick out the content
14:56:50 <TrueBrain> it is mad because it is pointless :)
14:57:00 <andythenorth> I just want to lose the control characters
14:57:13 <TrueBrain> mostly because really nobody reads a .txt file
14:57:40 <TrueBrain> and those few that do, really survive the # and *
14:57:41 <andythenorth> I buy your idea for 1 dollar
14:57:59 <TrueBrain> people are amazing in creating problems .. really, they are. But this one is so pointless ..
14:58:32 <andythenorth> everything about this GPL compliance fucks me off :)
14:58:35 <TrueBrain> either you create a .html or .pdf, or you just leave the .md :)
14:58:45 <andythenorth> this all relates to needing to move license to repo root
14:58:54 <andythenorth> and have it not as a chameleon page template
14:58:56 <FLHerne> Markdown is text anyway :p
14:59:08 <TrueBrain> FLHerne: so is XML, not sure that carries a point ;)
14:59:41 <FLHerne> Unlike XML, Markdown doesn't look like an unreadable mess when read as plain text
14:59:44 <andythenorth> I can't be arsed with dealing with a future drama about it
14:59:48 <TrueBrain> FLHerne: I agree :)
14:59:58 <andythenorth> 'non-compliant' / hard to find /s
15:00:06 <TrueBrain> andythenorth: where in the license does it say the location the license has to have?
15:00:10 <andythenorth> and .pt isn't a recognised GPL file format
15:00:24 <TrueBrain> wtf is "GPL file format"?
15:00:30 <TrueBrain> did you just invent some drama?
15:00:34 <andythenorth> COPYING.text or LICENSE.txt
15:00:46 <TrueBrain> license.txt not found in that document
15:03:27 <andythenorth> nothing in the v2 FAQs about format
15:03:50 <TrueBrain> it would be very weird, if a license also defined the format of the license :D
15:03:58 <TrueBrain> "This license is only valid in Comic Sans"
15:04:16 <TrueBrain> "Any other font-rendering is explicit not allowed by a penalty of 100 dollar per day it is available"
15:05:50 <andythenorth> is this an acceptable way to read the license in GitHub?
15:06:03 <andythenorth> I do need to move the location, someone said they couldn't find the licese
15:06:12 <andythenorth> but maybe the format is fine
15:06:48 <TrueBrain> I like line 280 and on
15:06:55 <TrueBrain> pretty sure you were not suppose to copy that :P
15:08:37 <andythenorth> I tried to find out the facts on that previously
15:08:46 <TrueBrain> and always remember: the way GitHub renders anything is completely irrelevant for if the license applies :)
15:08:46 <andythenorth> the implication is you have to copy the whole thing, including that
15:09:04 <andythenorth> but it seems weird copying that Ty Coon stuff
15:10:03 <TrueBrain> lol, never noticed it was part of the official license
15:10:21 <TrueBrain> we distribute it with OpenTTD related projects too
15:10:28 <TrueBrain> shows how often I looked at the license file .... lolz
15:10:47 <andythenorth> you haven't read every clause, and asked advice from your attorney? :o
15:11:10 <TrueBrain> wait, you are my attorney, aren't you?
15:12:16 <TrueBrain> so ... when am I going to deploy DorpsGek_iv to production .. mmm
15:13:27 <frosch123> do it before someone requests an attorney plugin
15:14:44 <FLHerne> ^ would resolve a lot of andythenorth's questions
15:16:24 <TrueBrain> 6037 names seen by DorpsGek over 13 years
15:16:45 <TrueBrain> well, people who spoke, ofc
15:16:58 <FLHerne> Might need adjustment for _ etc.
15:17:09 <TrueBrain> so many spambots :P
15:17:30 <FLHerne> Huh, I don't actually remember ever seeing spam in here
15:17:43 <TrueBrain> be happy :) Few years ago it was pretty common
15:17:48 <TrueBrain> but OFTC did good dealing with it, honestly
15:18:20 <TrueBrain> bots are easy to spot in the logs, as they write more than normal chars :P
15:19:20 <TrueBrain> for i in `seq 2007 2020`; do for j in `seq 01 12`; do g=$(printf "%d/%02d" $i $j); h=$(printf "%d%02d" $i $j); if [ -n "$(ls *.$h*.log 2>/dev/null)" ]; then mkdir -p $g; mv *.$h*.log $g/; fi; done; done
15:19:23 <TrueBrain> do I win a price now?
15:21:13 <andythenorth> this is so lolz, because Bananas will remove license.txt anyway
15:21:36 <andythenorth> this is slightly less fun than playing tanks on a bad RNG day :P
15:26:35 <TrueBrain> okay, what I need to do: deploy new DorpsGek (this kills DorpsGek_III), scale it to 0, copy all files onto the EFS, scale to 1, validate that everything is working as we expect, scale it to 0, copy the logs/seen database from DorpsGek, kill DorpSGek, scale to 1 ... that doesn't sound complicated at all!
15:27:38 <TrueBrain> guess I do that tomorrow, when it is nice and quiet in here :)
15:29:24 <Eddi|zuHause> is that a challenge? :p
15:31:19 <TrueBrain> (the fun part is, I most likely won't have time tomorrow :P)
15:32:16 * andythenorth grinding out small to-do list items
15:32:25 <andythenorth> I have these huge to-do lists for each newgrf project :P
15:32:34 <andythenorth> it's a great tactic for actually focussing
15:32:37 <andythenorth> until they get too big :(
15:34:56 <frosch123> i use a jpeg todo list
15:35:09 <frosch123> sometimes it loses items in the compression
15:36:51 <TrueBrain> well, my big todolist is almost done :D
15:37:10 <DorpsGek_III> [OpenTTD/OpenTTD] Foxar commented on issue #6598: Possible for client to be in a nonexistent company on a server; but then the client crashes. https://git.io/fhn0d
15:37:22 * andythenorth needs some Horse feedback :P
15:37:23 <TrueBrain> haha, I changed the Docker Hub password, and now Azure is failing to build images .. no shit :P Let me fix that ..
15:37:30 <andythenorth> should I do a poll?
15:37:37 <andythenorth> is it done [yes] [no]
15:37:44 <dwfreed> why does azure need the password
15:37:52 <dwfreed> something something api keys?
15:38:10 <TrueBrain> you do know we are talking about Docker Hub here, right? :P
15:38:25 <TrueBrain> those who refused to do anything modern for .. well .. up to the beginning of this year or something?
15:38:55 <dwfreed> I have never used it, as I don't use docker
15:39:06 <TrueBrain> Docker is nice, Docker Hub ... is .................
15:41:34 <TrueBrain> but fear not! It is now a token!
15:42:01 <TrueBrain> if our lord hurries up with migrating to GitHub Actions, we can just simply remove this all, but okay :)
16:12:20 <andythenorth> frosch123 recolouring with composed sprites....
16:12:50 <andythenorth> cb 2D will only work on layer 0?
16:13:01 <andythenorth> I store recolour sprites in register to recolour other layers?
16:16:09 * andythenorth has the answer nvm
16:39:43 <andythenorth> hmm randomising both 1cc and 2cc, it's only 256 recolour sprites? o_O
17:07:15 <DorpsGek_III> [OpenTTD/OpenTTD] Foxar opened pull request #8346: Fix #6598: Added company id check for connect console command. https://git.io/JkpLU
17:39:22 *** gelignite has joined #openttd
17:44:54 <nielsm> 14:44:04 - <frosch123> didn't nielsm add some formatting for md? <-- I think it's still in an open PR, something about a help and manuals window
17:45:44 <nielsm> iirc I did separate the extended text viewer formatting out into its own commits so you can skip the help and manuals window part if you want
17:46:16 <nielsm> (there's also some optimisations in the rendering of long files and with automatic linebreaking enabled)
17:46:42 *** Wormnest has joined #openttd
18:02:44 <_dp_> btw, last few lines in the story book sometimes disappear currently
18:03:07 <_dp_> couldn't quite get a repro though, seems to be happening at low resolutions
18:34:26 <DorpsGek_III> [OpenTTD/nml] javelinnl opened issue #172: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpcn
19:06:54 *** sla_ro|master has joined #openttd
19:10:22 <TrueBrain> there are more of those
19:11:13 <frosch123> so they found a talk page? :p
19:11:21 <TrueBrain> with himself? Sure!
19:24:27 <DorpsGek_III> [OpenTTD/nml] glx22 commented on issue #172: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpcn
19:24:48 <DorpsGek_III> [OpenTTD/nml] frosch123 commented on issue #172: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpcn
19:24:55 <frosch123> damn, few seconds too lsow
19:25:34 <LordAro> i wonder how difficult it would be to add a "commit message" to the wiki, that could then be added to the actual commit message
19:25:45 <frosch123> it's some ttdp bullshit again
19:26:05 <frosch123> there are separate code points for aircraft and helicopters, and ttdp only patched one
19:26:26 <TrueBrain> LordAro: very trivial; just what is the point? Given the ~15 years of history, it was not used in any sensible way for most of the commits :P
19:27:37 <DorpsGek_III> [OpenTTD/nml] TrueBrain commented on issue #172: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpcn
19:30:13 <DorpsGek_III> [OpenTTD/nml] frosch123 commented on issue #172: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpcn
19:31:03 <TrueBrain> glx: I thought I try something new, making silly comments to provoke a more .. declarative answer :) Which I got, so I am happy :)
19:31:38 <TrueBrain> your comment was very .. "you are wrong"-vibe :) And frosch123's comment didn't really help me in understanding if this is a bug or not :P His latest comment makes sense :)
19:32:30 <glx> well it's a mix of not-a-bug, and incomplete nml doc
19:32:40 <TrueBrain> the user doesn't care :)
19:33:06 <TrueBrain> it would be awesome, if helicopters would make take-off sounds
19:33:12 <TrueBrain> that would put a smile on my face :)
19:33:34 <TrueBrain> couldn't NML warn btw, if a user tries to set that property on helicopters?
19:34:18 <TrueBrain> or is there this endless stream of shit you would need to be warned about? :D
19:34:48 <frosch123> newgrf is a bit like mediawiki :)
19:34:55 <TrueBrain> I was afraid of that :P
19:35:05 <frosch123> it does stuff because it does it that way
19:35:15 <frosch123> changing things is hard, because everyone assumed the stupid behaviour
19:35:20 <TrueBrain> the wiki describes the implementation, gotcha :P
19:36:36 <TrueBrain> still amazed people can do so much with NewGRF :)
19:37:15 <frosch123> cs did some copy&paste, ttdp only patched the aircraft occasion
19:37:24 <frosch123> then people documented it does not work for helicopters
19:37:38 <TrueBrain> and now people depend on it being broken, I guess?
19:37:42 <frosch123> and when noone cared to patch the second location, that became the specified behavior
19:37:55 *** otetede has joined #openttd
19:38:03 <frosch123> yeah, you need some compatibility/versioning thing to enable the new behavior
19:38:30 <frosch123> but at the same time, there is already a more advanced/complicated newgrf feature that can do it
19:38:57 <frosch123> harder to use for a newbie, but does everything better than the property
19:39:11 <TrueBrain> and NML does not do such translations, I assume?
19:39:49 <frosch123> not in general, but i wouldn't be surprised if it does in some special cases :)
19:39:57 <andythenorth> glx could I help add an nml flag to suppress nml optimisation warnings?
19:40:08 <TrueBrain> depending if someone took care of that specific case, I can understand :)
19:41:01 <frosch123> i do not follow aircraft sets in particular, this may be the first time in 15 years someone wants to provide a custom sound for a helicopter
19:41:34 <frosch123> even andy does not have an aircraft set
19:41:47 <andythenorth> pikka made AV9, aircraft are solved
19:41:52 <andythenorth> probably the most perfect grf
19:42:05 <TrueBrain> makes sense frosch123 :)
19:43:31 <glx> andythenorth: I didn't try to "rewrite" warnings to enable filters yet
19:45:39 <andythenorth> hmm, timber vs. wood. vs. logs
19:46:00 <TrueBrain> I rather have wood in my pants than logs .. honestly ...
19:46:33 <andythenorth> what about timber?
19:46:41 <TrueBrain> never tried, no opinion
19:46:46 <andythenorth> ok so it's a scale
19:47:02 <glx> andythenorth: but you can comment them out, they are all in switch.py
19:47:02 <andythenorth> and timber is schrodinger
19:47:21 <andythenorth> TrueBrain this is great, but it doesn't help name a wagon for carrying wood
19:47:42 <TrueBrain> nobody ever accused me of being useful, so not sure what you expect here
19:47:51 <andythenorth> the translation is en-gb, so the wagon name should be 'timber wagons'
19:47:59 <andythenorth> but timber wagons carry logs
19:48:05 <andythenorth> but the openttd cargo is wood :P
19:48:12 <TrueBrain> to bring to the wood factory!
19:48:15 <andythenorth> english sense makes
19:48:22 <TrueBrain> sounds very sensible to me!
19:48:30 <TrueBrain> unlike the helicopter sound, btw
19:48:35 <TrueBrain> that makes no sound at all!
19:48:58 <andythenorth> must have sounds
19:49:20 <TrueBrain> of me making train noises?
19:49:27 <TrueBrain> pretty sure that keeps your kids up at night
19:49:35 <andythenorth> for helicopter, great noise
19:51:12 <TrueBrain> well, as long as the sound is to scale, it should be fine
19:51:26 <TrueBrain> we can't have the 'choo choo' of a helicopter being in a different scale than the 'choo choo' of a train, ofc!
19:51:56 <andythenorth> do we have to change frequency when scaling sounds?
19:52:11 <andythenorth> hmm we have stereo mixer I think also?
19:52:26 <frosch123> does daylength affect the playback speed?
19:52:29 <andythenorth> trains on right of screen go 'choo' in right speaker?
19:52:43 <andythenorth> frosch123 it should
19:52:51 <andythenorth> but I was thinking of doppler also for travelling vehicles
20:31:30 * andythenorth adds ANOTHER_TRAIN
20:32:25 * andythenorth continues to avoid drawing car transporting wagons
20:32:51 <frosch123> i checked, the "crashing" sound is only used in this place
20:33:02 <frosch123> so it definitely is no "crashing" sound, just a toyland plane
20:37:23 <TrueBrain> I feel a PR coming up :)
20:39:16 <andythenorth> you can probably get medicine for that
20:40:36 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 opened pull request #8347: Change: [NewGRF] Use aircraft property 12 also for helicopters. https://git.io/JkpVx
20:40:45 <frosch123> yes, but not the one you were hoping for
20:51:27 <DorpsGek_III> [OpenTTD/OpenTTD] javelinnl opened issue #8348: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpre
20:53:09 <frosch123> i don't feel like calling this a fix for the issue in the commit message
20:53:59 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 commented on issue #8348: sound_effect property doesn't seem to work with helicopters https://git.io/Jkpre
21:59:41 <andythenorth> nice CDN url eh :P
21:59:49 <andythenorth> quite hard to draw :P
22:38:51 <andythenorth> looks annoyingly phallic currently :P
23:42:06 *** luaduck has joined #openttd
continue to next day ⏵
