IRC logs for #openttd on OFTC at 2024-02-19
⏴ go to previous day
00:44:35 <reldred> xrdp-egfx is impressive. Actually makes RDP fast enough to use on Linux servers.
00:44:48 <reldred> Even just using software rendering with no GPU.
03:09:52 *** gnu_jj_ has joined #openttd
03:13:00 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:30:20 *** Wormnest has quit IRC (Quit: Leaving)
04:04:01 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:12:39 *** D-HUND is now known as debdog
04:46:39 <merni> I will probably write a PR for that soon... just opening the issue so I don't forget about it again
04:46:50 <merni> Discord is not an issue tracker :)
04:58:56 <ajmiles> _glados_: Not awful for an evening's work 😆
05:02:39 *** HerzogDeXtEr1 has joined #openttd
05:08:06 *** HerzogDeXtEr has quit IRC (Ping timeout: 480 seconds)
05:48:00 *** HerzogDeXtEr1 has quit IRC (Read error: Connection reset by peer)
07:13:05 <DorpsGek> - Add: summary for week 07 of 2024 (by OpenTTD Survey)
07:23:06 <wensimehrp> survey-summary[bot]v: 🤔 Seems like quite a few people uses Canadian dollars
07:36:20 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
07:42:55 *** Ox7C5_ has quit IRC (Ping timeout: 480 seconds)
08:25:11 <_glados_> ajmiles: Does this also off load the whole rendering work to the gpu?
08:28:17 *** sweduke has joined #openttd
08:28:17 <sweduke> I see that there is an option to have multiple industries per city, which makes the freight traffic much more interesting. But I'd like to have to have only multiple base industries per city, I don't need multiple fabrics per city. Any chance to get this as an additional option?
08:28:54 <reldred> sweduke: Highly likely you'll only see this in either an industries grf or a gamescript
08:29:31 <reldred> (and it's something I'd like as well, I like to play maps with lots of primary industries and very few secondary/tertiary)
08:30:36 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
08:31:01 <reldred> One thing I have done which if you don't mind using RVG for your game script I've found works well is to set industries to funding only, set the RVG setting to prospect additional primary industries, and then build all secondary/teritiary industries by hand. Or you can combine the two; set the normal build amount to be 'minimal' so you get one of every industry, and then use RVG's feature to boost
08:42:25 <peter1138[d]> Hmm, 12000 towns and 20000 industries.
08:56:25 <reldred> so they announced a buyout of the business I work at today, should work out pretty well for us, we're keeping our engineering team intact and based down here, keeping all our products we've developed, but we're getting to offload so much crap to their existing 24/7 helpdesk.
08:57:02 <reldred> and they already run the ERP that we were looking at moving to, which saves us having to chew that off
08:57:26 <reldred> and they already run the RMM we were going to swap to end of march lmao
08:59:19 <peter1138[d]> Takeovers usually end up not as you expect.
08:59:47 <reldred> I've been through one before that set off alarm bells, this one so-far has been promising, and I'm not normally an optimistic person.
09:00:06 <reldred> met the director of technical stuff today, he was panicking about how we were going to lift and shift all our infra since the state government are kicking us out of our building in a few months
09:00:22 <reldred> was like, lol, we already moved it to our DC two weeks ago
09:02:18 <LordAro> reldred: you should've said that it'd require many hours of paid overtime for you
09:02:25 *** Ox7C5_ has quit IRC (Ping timeout: 480 seconds)
09:02:38 <reldred> this one is at least promising because we were already planning to change all our tooling end of next month anyway, so we're already basically mentally geared for that change, and the business acquiring us already does most of what we do (though with different products) interstate, the last buyout I was involved in was from a business that had very very little to do with our industry so they were
09:02:38 <reldred> just Bright Idea (tm) monkeys with zero understanding of how our industry actually works.
09:02:54 <peter1138[d]> Heh, I remember my boss, not panicking, but trying to plan moving everything out of an office...
09:03:08 <reldred> LordAro: nah I've already told them I need to do a major reshuffle of our DC infra soon.
09:03:54 <peter1138[d]> I had already moved everything to the DC during the last move, so the only thing in the office was unimportant.
09:04:55 <peter1138[d]> Currently thinking is "why do we need the DC, can't we use the cloud"
09:04:57 <peter1138[d]> Which is... meh.
09:06:13 <peter1138[d]> I added code to speed up station catchment recalculation... but it only really has an effect with Xarick's crazy saves :S
09:07:16 <reldred> gave them the good news about a bunch of our hardware being up for renewal, question is whether they deem it not worth maintaining a third DC in another state or consolidate it to one of their two DC's interstate. I'll be trying to make the case that we have ~1ms approx latencies to most of our customer base, moving interstate would jump that to 20-30ms
09:07:59 <reldred> but as for the rest of cost of operations we're a super tight focussed team of 3.2 people (the other senior engineer is semi retired and only works one day a week), so our overheads otherwise are super low.
09:08:32 <reldred> but we've got a lot of active contracts on services running out of that DC so we'll see
09:08:33 <LordAro> peter1138[d]: we've thought about that from time to time too. cloud is very expensive for the sort of compute that we'd generally need
09:08:40 <peter1138[d]> Hmm, probably a good argument for going cloud here, our hardware is quite old.
09:08:43 <LordAro> though most of the web services could probably be offloaded
09:08:58 <reldred> cloud always ends up stupidly expensive
09:09:22 <reldred> Most businesses go that way I see because they're scared of what it would cost/take to actually build a rack in a DC
09:13:36 <peter1138[d]> It would, modern servers are ridiculously expensive.
09:14:59 <reldred> Yeah we were discussing that today. Seems to be either you guy with the big vendors and pay exorbitant fees, or you go the way some of the new players are doing it and go for whitebox hardware and spec it all custom.
09:15:27 <reldred> We're currently saddled up with Dell at the moment and it feels bad man
09:15:45 <reldred> but hey, they come out at midnight to swap a mobo if they have to so
09:16:42 <peter1138[d]> In the early 2000s we used to self-build 4U servers. Of course not as good as a proper server but good enough at the time.
09:16:44 <LordAro> they do seem to have to swap an awful lot of MBs
09:16:56 <LordAro> (we use them for workstations)
09:19:20 <reldred> Yeah we had to replace one so-far
09:19:55 <reldred> and we've got a dead interface on one of the servers 4x10gbe mezzanine cards
09:20:17 <reldred> (which I want replaced with sfp mezz cards anyway)
09:29:30 <xarick> ow snap, GetHangarTile needs a bit more care
09:31:50 <peter1138[d]> Pretty sure it could be improved further but would need more special-casing.
09:35:51 <peter1138[d]> Threading NOT_REACHED... huh
09:36:15 <peter1138[d]> Oh, uncaught exception.
09:36:25 <j_n> j_n: in fact, it could make the game more interesting for me
09:37:27 <j_n> like, I can play with three other AI companies
09:37:27 <j_n> I transport cargo, while they transport pax/mail
09:37:27 <j_n> I make enough money to take over the AI companies, thus becoming a passenger transport magnate as well
09:39:01 <j_n> oh nice, 14.0 is already in RC
09:39:16 <xarick> GetOrderDestination for hangars only returns the hangar with index 0
09:43:04 <j_n> hrm, when starting OTTD 14.0 on my Win8.1 laptop the game starts with some strange thin arial font
09:43:18 <j_n> when switching off TTF fonts and switching them back on, the game crahses
09:43:43 <j_n> not sure if the devs even care about supporting systems this old at this point
09:43:57 <peter1138[d]> Is that a fresh install?
09:44:11 <j_n> although I installed graphics set separately (OpenGFX2)
09:44:39 <j_n> let me check if it happens with the worse OpenGFX
09:44:49 <xarick> the hangar index is harder to deal than expected 😐
09:44:59 <peter1138[d]> Please report it with all the crash dumps & logs etc 🙂
09:45:13 <j_n> and the password to GH acc is at home
09:45:50 <j_n> also I'm concerned about safety running on such an old system, should probably switch to linux at some point
09:46:17 <peter1138[d]> A crash in OpenTTD is not offtopic here 🙂
09:47:03 <peter1138[d]> Not sure if we dropped Windows 8 support yet, might have.
09:47:33 <reldred> we've still got folks developing grf's on win7
09:47:34 <peter1138[d]> I would have expected it to not run though, but I dunno.
09:47:43 <reldred> though he is a bit of a statistical outlier
09:48:00 <j_n> peter1138[d]: could I put the files into a zip and send them here? Or only GitHub issues are supported?
09:48:15 <j_n> peter1138[d]: well, download says `Windows 7 / 8 / 10 / 11`
09:48:29 <peter1138[d]> That's clear then.
09:48:34 <j_n> it makes sense considering how like, old this game is
09:49:34 <LordAro> we officially dropped Vista for 14.0, but it already didn't work
09:50:15 <peter1138[d]> Hmm, do I still have my old Windows 8 tablet... I think not.
09:50:22 <LordAro> j_n: Discord/irc is very much adhoc, GH would be better
09:50:59 <peter1138[d]> The battery in it failed in a bulging scarily way...
09:51:22 <peter1138[d]> You can send it here now, and then at least it's here when you forget later :p
09:51:40 <LordAro> remove battery and power it directly from the cable?
09:52:03 <j_n> peter1138[d]: I have an idea
09:52:03 <j_n> j_n does pinging myself work?
09:52:21 <j_n> I'll set the message where I ping myself as unread and then I'll have an extremely annoying ping icon on discord
09:52:21 <peter1138[d]> No, its firmware refused to boot if the battery was less than 2% charged.
09:52:29 <locosage> powering from cable didn't work with my old phone either
09:52:37 <j_n> j_n: I always check those so it's foolproof
09:52:47 <locosage> had to frankenstein some drone battery to it
09:53:25 <peter1138[d]> Mine was an Atom with 1GB RAM, it wasn't worth saving...
09:53:39 <peter1138[d]> (Windows 8 was horrible on it)
09:53:50 <j_n> I have a win7 laptop, the graphics card seems to be corrupted and the system also won't boot after the "Windows was not shut down properly" screen
09:54:02 <locosage> I only did that to get some data out
09:54:04 <j_n> the recovery menu works but the display is too garbled to use
09:54:12 <locosage> alredy had a new phone by that point
09:54:19 <j_n> I'll probably gut it for electronic parts at one point
09:54:46 <j_n> since no one will buy something that old and nonfunctional
09:55:09 <peter1138[d]> Hmm, wonder if this code improves one specific case but impacts normal performance...
09:55:57 <xarick> found a maybe a bug, adding order
09:56:15 <j_n> j_n: or use it as a server of some sort, since although the display is very fucked, it's not completely unreadable
09:56:18 <j_n> that was too off topic, sorry
09:56:20 <xarick> i have aircraft 2, and i want to copy orders from aircraft 1
09:56:37 <xarick> go to, left click on aircraft 1, and it doesn't copy the orders 😦
09:57:29 <xarick> instead, it opens up the aircraft 1
09:59:59 *** Tirili has quit IRC (Quit: Leaving)
10:00:34 <LordAro> j_n: it's very rare that something is actually off topic here :)
10:00:43 <LordAro> peter1138[d] & I often talk about bikes
10:00:49 <LordAro> peter1138[d]: how are your legs doing, btw?
10:10:59 <xarick> I don't understand this, I can't reproduce the bug in master
10:11:44 <peter1138[d]> Still, I did it properly... had to catch up with the group after the first control as they didn't stop, the buggers...
10:14:49 <LordAro> just an info control, or did they actually miss it?
10:15:02 <xarick> ah, ok I can reproduce the issue
10:15:23 <xarick> it's related to quick go to
10:16:18 <peter1138[d]> Just an info control, but most of them were not interested in doing it properly.
10:37:37 <peter1138[d]> After that we did stop at the controls, and two others did end up completing it... though they had to ask me for the answers, cheeky gits!
10:50:10 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
11:19:55 <xarick> gotta update everything from 14 to 15?
11:21:21 <xarick> how do i manually change a number from 14 to 15 in the compat files...
11:21:50 *** Ox7C5_ has quit IRC (Ping timeout: 480 seconds)
11:29:35 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
12:03:44 <_glx_> Read my comments and you won't need compat files
12:04:30 <ajmiles> _glados_: Not yet, I'm just trying to stand up a 12 backend first
12:18:10 <xarick> this is what's eating gigabytes of memory during load
12:18:31 <xarick> loading that 4k savegame with 14 AAAHogs
12:19:48 <xarick> how did it manage to save so much data so quickly
12:26:53 <peter1138[d]> telk5093: Probably best to report a 'bug' if the format you need is not supported.
12:28:40 <xarick> cargo packets approaching the 1 million
12:29:32 <xarick> seems AAAHogEx first limitation will be group pool, lol
12:31:08 <xarick> strange that he got more groups than order lists
12:31:37 <xarick> either emtpy groups, or vehicles with no orders?
12:40:28 <_glx_> xarick: Subgroups most likely
12:51:21 <peter1138[d]> Oops, crash in StationCargoList::Append.
13:02:40 <andythenorth> [in the other chat]
13:02:44 <andythenorth> "state machine"?
13:07:25 <peter1138[d]> Need a faster CPU.
13:08:25 <peter1138[d]> Also I'd like to know where 33.80ms went.
13:10:32 <LordAro> is it on the framerate window? :p
13:12:17 <peter1138[d]> ```Stop reason: signal SIGSEGV: address not mapped to object (fault address: 0x10)
13:12:17 <peter1138[d]> warning: `this' is not accessible (substituting 0). Couldn't load 'this' because its value couldn't be evaluated```
13:14:15 <_glx_> And no, using more cores will not help as they run in sync with gameloop
13:15:59 <peter1138[d]> (EffectVehicle *) 0x00007fff97ab3270```
13:17:40 <peter1138[d]> Oops, crashed the debugger :S
13:26:22 <_glx_> hmm #12120 is in the annoying to debug group, as the crash happens inside the thread
13:26:23 <peter1138[d]> Autosave 1 minute, see if I can get a save closer to the problem...
13:26:36 <xarick> the warning is complaining
13:27:01 <peter1138[d]> _glx_: Outputting the exception instead just doing NOT_REACHED would be helpful.
13:28:40 <LordAro> could be a debug vs release thing?
13:29:07 <LordAro> as in, NOT_REACHED in release, output in debug
13:30:35 <xarick> what is this warning pointing at?
13:32:07 <_glx_> xarick: missing () somewhere
13:33:25 <peter1138[d]> Meh, autosave didn't.
13:35:18 <peter1138[d]> (const Owner) 128
13:39:41 <peter1138[d]> Oops, broke the debugger again.
13:39:47 <peter1138[d]> And it didn't autosave.
13:43:45 <_glx_> xarick: the warning means only the part between | and != is used for the comparison
13:46:02 <xarick> I removed the != 0 and it stopped complaining
13:46:10 <_glx_> hmm the code looks fine actually, it's just a weird construct
13:47:14 <peter1138[d]> It's suggesting that you are comparing a flag incorrectlky.
13:47:44 <peter1138[d]> Oh, I missed the second screenshot 🙂
13:48:11 <peter1138[d]> Yes, that code terrible 🙂
13:48:33 <xarick> not my code, but i can fix it while I'm here
13:48:45 <_glx_> just ignore intellisense
13:48:51 <peter1138[d]> If it's not yours, don't touch it.
13:51:11 <xarick> `OrderDepotTypeFlags odtf = (OrderDepotTypeFlags)(ODTFB_PART_OF_ORDERS | ((order_flags & OF_SERVICE_IF_NEEDED) ? ODTFB_SERVICE : 0));`
13:51:11 <xarick> somewhere else in script_order.cpp it is done this way
13:51:28 <xarick> the != 0 could be removed
13:52:34 <_glx_> but code style says it should not 🙂
14:00:32 <_glx_> hmm might be my cfg too here
14:00:49 <_glx_> need to disable more folders
14:02:47 <peter1138[d]> Oh, I see, `i->stations_near` does not match `st->industries_near`.
14:03:34 <peter1138[d]> I incorrectly assumed it does.
14:03:37 <_glx_> they should intersect, but not be indentical
14:04:51 <_glx_> at least i and st should be in the other one _near
14:05:51 <peter1138[d]> No, `industries_near` uses cargo acceptance, and not all tiles accept (enough) cargo.
14:06:31 <_glx_> the stupid not all tiles are accepting
14:06:35 <peter1138[d]> This just means this little optimisation I was working with is wrong.
14:07:16 <peter1138[d]> I don't consider it stupid.
14:09:04 <xarick> @glx small issue where the arrow is pointing
14:20:20 <peter1138[d]> Maybe I should add another list, just of any industries near.
14:21:29 <peter1138[d]> Or: not worry too much about this as performance with normal saves is fine.
14:24:01 <LordAro> i'd probably want to go with the latter
14:28:12 <peter1138[d]> It turns out iterating 20,000 industries when updating station catchment is sub-optimal—only 2 ms per go, but it adds up.
14:29:32 <peter1138[d]> But that many industries is extreme.
14:34:18 <peter1138[d]> Probably the strategy of clearing the lists and then adding back is a bit naïve anyway.
15:08:36 <Eddi|zuHause> didn't someone run tests back then whether it's faster to iterate the industries or the map?
15:13:20 <xarick> I'm impressed by your dedication
15:16:45 <j_n> j_n: I forgot to activate the protocol
15:18:20 <peter1138[d]> Eddi|zuHause: It's PR#7235, all sorts of performance tests were done, but I don't remember precise details.
15:18:39 <j_n> do replies, like, show up on the IRC side? Just wondering
15:19:13 <Eddi|zuHause> probably? only thing i know doesn't arrive is edits
15:19:51 <peter1138[d]> The reply shows up, the quoted part of the reply does not get repeated though.
15:24:09 <j_n> j_n: hrm, do I file it as a bug report or a crash
15:24:35 <_glx_> if it's a crash then report a crash 🙂
15:24:58 <j_n> well it's improper behaviour followed by a consistent crash
15:30:41 <peter1138[d]> Bug vs Crash is about severity.
15:30:56 <peter1138[d]> Crash is always more severe 🙂
15:31:58 <wensimehrp> telk5093: In most of the times players won't even focus on the last four digits.
15:31:58 <wensimehrp> A-Tourism chose to ignore the last four digits by setting the baisc currency unit to *ichiman yen* thus it reduces reading time.
15:31:58 <wensimehrp> I wonder if there could be a setting to auto-ignore the last four digits when the number expands to five digits.
15:31:58 <wensimehrp> Another problem might be that the CJK format requires a unit at the end. $100万0000 isn't proper format and could be changed to 100万0000美元. That could be done by custom currency though.
15:33:58 <Eddi|zuHause> we can't support all currency/language combinations, that gets very... explosive
15:34:41 <wensimehrp> Sounds like Alice in a rabbit hole
15:39:51 <wensimehrp> I'd say that the current format for zhcn is fine, although there are some places normally we won't except it to use the CJK format e.g. sprite align window or factory production window.
15:39:51 <wensimehrp> Traditionally the CJK format is only used on financial related subjects.
15:40:48 <wensimehrp> Might sound depressing... Sorry if this might create any extra work
15:43:12 <j_n> does anyone know what the purpose is for light PBS signal alternate sprites?
15:43:50 <j_n> I don't see them used there either
15:44:30 <wensimehrp> Hmm you might want to download a newgrf that procides the graphics for alternate signal sprites
15:44:50 <j_n> well it's not too important for me
15:45:37 <wensimehrp> That is multi-aspect signals
15:52:29 <Eddi|zuHause> i think there are different signals depending on driving side?
16:07:47 <peter1138[d]> Hmm. Possibly worth it.
16:14:51 *** Wormnest has joined #openttd
16:16:19 <xarick> oh, nevermind, it's okay...
16:17:45 <xarick> that const Order could be moved to right before the switch
16:31:40 <xarick> CmdModifyOrder has a very ambiguous `uint16_t data`
16:34:37 <xarick> i'd rename it to order_flags
16:36:19 <_glx_> it can be OrderStopLocation
16:36:21 <xarick> oh, I see, there's other uses
16:36:38 <_glx_> it's just a generic value
16:38:39 <xarick> (OrderConditionVariable)data
16:39:06 <xarick> some places cast what it is, some don't
16:39:31 <_glx_> they all cast where needed
16:41:41 <xarick> but someone in the future could change the enum...
16:41:51 <xarick> and then it wouldn't work
16:42:26 <xarick> order_flags & OF_NON_STOP_FLAGS
16:42:26 <peter1138[d]> Eddi, thanks for the tip, btw.
16:43:01 <peter1138[d]> "If the number doesn't match" is what Xarick is thinking.
16:43:08 <_glx_> no need to cast to uint_16
16:46:27 <peter1138[d]> Hmm, I keep losing keyboard input.
16:46:43 <_glx_> and ScriptOrderFlags is far from being bigger than 16 bits
16:47:16 <xarick> ```uint data = DA_ALWAYS_GO;
16:47:16 <xarick> if (order_flags & OF_SERVICE_IF_NEEDED) data = DA_SERVICE;
16:47:16 <xarick> if (order_flags & OF_STOP_IN_DEPOT) data = DA_STOP;```
16:47:16 <xarick> it is assuming flag combinations don't exist.
16:47:47 <peter1138[d]> Yeah, no SDL_KEYDOWN events 😮
16:47:57 <_glx_> DA_ stuff don't combine
16:48:47 <xarick> that's the problem... it's not consistent with what the flags can do
16:49:14 <_glx_> but DepotAction is not a flag
16:50:38 <xarick> AreValidOrderFlags needs fixing then
16:53:58 <_glx_> current version is correct, you can't have stop and service at the same time
16:55:58 <_glx_> of course if you add unbunching then you need 2 more tests
17:30:37 *** Flygon has quit IRC (Remote host closed the connection)
17:32:07 *** HerzogDeXtEr has joined #openttd
17:36:03 <xarick> This flag combination is possible: `ODATFB_NEAREST_DEPOT | ODATFB_HALT`
17:40:35 <frosch123> ah, i only tested non-black ellipsis
17:40:56 *** virtualrandomnumber has joined #openttd
17:41:20 *** virtualrandomnumber has quit IRC ()
18:20:07 <xarick> it was bugged since...
18:20:42 <xarick> when was service at nearest depot introduced?
18:21:21 <xarick> I mean, go to nearest depot
18:35:02 <Eddi|zuHause> probably 5 years?
18:35:35 <DorpsGek> - Update: Translations from eints (by translators)
18:36:09 <Eddi|zuHause> (svn r12661) -Add: ability to send a vehicle (using default orders) to the nearest depot.
18:36:15 <Eddi|zuHause> so muuuch longer ago
18:42:54 <debdog> How do I make changes to "Transparency Options" to be permanent? If I enable stations to be visable when transparency mode is on this setting only stays until I toggle the mode off and on again.
18:43:43 <debdog> have not found anything in Settings but I might just have missed it there
18:45:17 <michi_cc[d]> So, which socials are still missing for RC1? Anybody alredy done reddit or Twitter?
18:45:59 <michi_cc[d]> Hmm, no message prepared either?
18:47:03 <xarick> oh, wait, this is possibly a Codefix?
18:47:31 <xarick> cus I have no idea who calls GetNextDecisionNode
18:48:37 <peter1138[d]> I've not a Mastodon post either.
18:48:55 <peter1138[d]> Mainly because I've been busy and/or lazy.
18:49:59 <michi_cc[d]> xarick: If you'd actually write what your PR is supposed to fix, you would probably know the answer.
18:50:54 *** kamnet has quit IRC (Quit: User went offline on Discord a while ago)
18:58:51 <michi_cc[d]> Okay, I just noticed there wasn't even a Discord announcement for RC1 😦
18:59:45 <xarick> I read the function description, it wants to know the next station it can go and return that. If it can't go to any station, it return nullptr. Stopping at depot qualifies as not going to a station
19:00:45 <xarick> what the code was doing was... getting the next station it can stop at, even though when the vehicle processes the depot order, it would stop
19:01:15 <xarick> so that's how I decided this is bugged
19:01:25 <xarick> didn't match the behaviour
19:16:21 <michi_cc[d]> debdog: Ctrl+Click on a the transparency buttons will lock them.
19:19:46 <debdog> michi_cc[d]: thanks a lot!
19:37:57 <frosch123> funny you mention the number format, we already got two "bug" reports about it 🙂
19:38:23 <LordAro> should 'asian' be capitalised?
19:38:54 <frosch123> (both in the category: insufficient specification)
19:39:18 <truebrain> it gets the ball rolling at least 😉
19:39:26 <truebrain> did someone happen to make a Steam image for RC1?
19:39:53 <michi_cc[d]> LordAro: You tell me, you'te supposed to be English as far as I've heard 🙂
19:40:21 <michi_cc[d]> truebrain: Haven't seen one yet.
19:43:04 <truebrain> too bad you can't template PRs 🙂 As we really should have all this part of a release post 😄 Avoids feeling the "pressure" of getting this done after the release is made
19:44:59 <truebrain> owh, sorry, I should have been more clear in what I meant
19:45:04 <truebrain> you can't make a PR template with choices
19:45:15 <truebrain> like: you are about to make a PR for a release, did you remember to do A, B, C
19:45:24 <truebrain> which is different from a PR for a blog
19:45:29 <truebrain> that is what I meant 😉
19:45:32 <michi_cc[d]> Okay, yeah, not every website PR is for a new release 🙂
19:45:46 <truebrain> that said, it doesn't hurt to make a generic template
19:45:53 <LordAro> michi_cc[d]: yes, it is
19:45:59 <LordAro> having googled to check :)
19:47:06 <michi_cc[d]> That reminds be, I'll PR my dev diary in a bit because it's not reading as well as I'd like it to, but I kinda out of ideas how to fix it. Hoping for some improvement ideas.
19:47:13 <michi_cc[d]> But release social stuff first.
19:47:57 <truebrain> I am going to use your text 🙂
19:49:43 <xarick> is kuhnovic around? Look what I found!
19:51:51 <xarick> is it safe to assume OT_LEAVING still has the destination of OT_GOTO_STATION?
19:52:47 <michi_cc[d]> Discord, reddit, Xitter done.
19:53:37 <truebrain> right, not the best solution, but it might help 🙂
19:53:49 <michi_cc[d]> truebrain: You can butcher my dev diary soon in return 😛
19:54:30 <truebrain> _zephyris: nice, tnx 🙂
19:55:47 <_zephyris> Oh wow, I knew it was higher resolution - didn't check how much higher!
19:56:12 <_zephyris> That's 4k screens plus 200% interface scaling in action
19:57:29 <frosch123> this text truncation business is super weird
19:57:50 <frosch123> it uses the font of the first run, but the colour of the last run
19:58:02 <frosch123> independent of where the truncation happens
20:01:52 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
20:02:11 <andythenorth> thanks for doing all this 🙂
20:07:15 <truebrain> Rubidium: I tried adding "savegame size" to the survey, but my approach sucked balls, and I can't find a better one 😛
20:07:21 <truebrain> lot of bookkeeping, which is easily to go wrong
20:07:40 <truebrain> if you have any ideas, I am all ears (as I would love to have that stat)
20:07:41 <michi_cc[d]> Hmm, can I push directly to the website repo, or do I need to make a clone first?
20:07:49 <peter1138[d]> Yeah, text truncation was always a bit weird.
20:07:51 <truebrain> michi_cc[d]: to the repo is fine
20:14:55 <michi_cc[d]> Do your worst with this rambling blog post 🙂
20:18:41 <FLHerne> truebrain: maybe it's a daft question, but what does "at the end of every game" mean for surveys?
20:18:43 <xarick> wow, I made it to the list!
20:19:38 <FLHerne> I presume it means when leaving a session, but in isolation the term seems more applicable to the highscore screen in 2050 or whatever
20:20:26 <_zephyris> michi_cc[d]: Looks great! Very happy to proof-read, but I like the content.
20:23:49 <j_n> Eddi|zuHause: (sorry for necroing) that seems to be exclusive to semaphores
20:43:53 <xarick> that's not how you (or I) should use flags
20:48:43 <_glx_> xarick: looks similar enough to be added to your PR
20:49:39 <_glx_> they are both "incorrect check of flags"
20:50:26 <xarick> `(v->current_order.GetDepotActionType() & ODATFB_NEAREST_DEPOT) == 0`
20:50:26 <xarick> `(v->current_order.GetDepotActionType() & ~ODATFB_NEAREST_DEPOT) != 0`
20:50:26 <xarick> `v->current_order.GetDepotActionType() & ~ODATFB_NEAREST_DEPOT`
20:51:56 <peter1138[d]> Yeah, I can't reproduce JGR's comment about #12129.
20:55:55 <_jgr_> I'd misremembered how it worked, the change seems fine, sorry about that
21:03:45 <truebrain> michi_cc[d]: nice read; my only feedback would be that in some places the transitions are a bit stiff, but that can easily be fixed with some fillers 🙂
21:05:19 <michi_cc[d]> Suggestions accepted 🙂
21:08:58 <_glx_> xarick: `(v->current_order.GetDepotActionType() & ODATFB_NEAREST_DEPOT) == 0`
21:12:29 <truebrain> owh right, my 2 posts also need a Steam image
21:12:48 <truebrain> maybe I can look at _zephyris really cute, maybe they can fix me up something nice 😄
21:14:05 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:14:11 <xarick> I got a feeling fixing the HALT flag is gonna make the aircraft part misbehave
21:14:48 <xarick> and then fixing the NEAREST_DEPOT part is gonna misbehave a bit more
21:15:27 <_glx_> both tests are actually broken if you have both flags set
21:15:49 <xarick> it was relying on both being broken 🙂
21:16:00 <xarick> now if both get fixed, something... might happen
21:20:51 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
21:37:13 <truebrain> sometimes, I just shoulnd't be trying to understand a blob of our code 🙂
21:37:21 <truebrain> it just makes you go: HUH?! time after time 😛
21:49:12 <peter1138[d]> left-right is not exactly complex
21:49:23 <peter1138[d]> It's just not the left-width that you might expect.
21:50:06 *** Wormnest has joined #openttd
21:51:15 <peter1138[d]> Airport goto-hangar orders are bugging me right now...
21:51:33 <xarick> surprise! the aircraft actually goes to the correct hangar, it works with both fixes applied
21:52:14 <peter1138[d]> Do aircraft go to a specific hangar tile or just any hangar at the airport?
21:52:38 <truebrain> You are the developer, you should know! 😛
21:53:28 <xarick> you need to change the hangar orders to (stop)
21:53:28 *** felix has quit IRC (Read error: Connection reset by peer)
21:53:29 *** j0anjosep has joined #openttd
21:53:29 <j0anjosep> peter1138[d]: hi! they go to any hangar tile
21:55:48 <peter1138[d]> I'm looking at RemoveOrderFromAllVehicles(). Aircraft have... extra handling.
21:56:26 <xarick> Oh.. i thought you were looking at my stuff
21:57:46 <xarick> that is because airport stations can have goto hangar and / or goto station
21:58:58 <xarick> well, that is actually my stuff as well, lol
21:59:27 <peter1138[d]> Hmm, damn, looks like I can't easily optimise this, as v->current_order is separate from shared orders.
22:02:30 <frosch123> peter1138[d]: any hanger of the airport. otherwise you will get in big trouble when upgrading airports
22:02:52 <peter1138[d]> Yeah, I got it 🙂
22:02:53 <frosch123> it's the same as "any platform of station" or "any depot in joined depots"
22:03:19 <frosch123> do we have joined depots? or is that an openPR?
22:03:24 <peter1138[d]> It's an open PR.
22:03:34 <peter1138[d]> It's a bit of a huge open PR.
22:04:22 <peter1138[d]> So I think the only way to know if any vehicle's current_order points to the destination is to iterate the vehicle pool. That makes me sad.
22:04:53 <xarick> can I add a PR in a fix?
22:05:11 <xarick> Fix 7104, 7104 being a PR?
22:05:25 <frosch123> i would assume we also have a orderpool
22:05:31 <frosch123> or even orderlistpool
22:05:53 <peter1138[d]> Yes, orderlist pool is the optimisation, but current_order may be outside of that.
22:06:02 <peter1138[d]> At least, I believe.
22:06:02 <kuhnovic> michiccviaGitHub: Great post, I think this will be a fun read for both devs and non-devs. And the ship pathfinder reference is flattering.
22:06:16 <frosch123> oh, true, automatic service orders
22:06:23 <xarick> Kuhnovic was summoned!
22:06:29 <frosch123> if an aircraft has no depot order, it may head to the nearest airport
22:06:53 <peter1138[d]> The aircraft thing was a red-herring, this applies to all vehicle types.
22:07:07 <peter1138[d]> The condition is a bit confusing 🙂
22:07:20 <kuhnovic> xarick: And now my bed is summoning me 😆
22:07:21 <peter1138[d]> (To me at least)
22:11:23 <kuhnovic> xarick: I think that could very well be the case
22:12:20 <kuhnovic> But the real question is why. Most likely there's a reason for it. It might even be a good reason.
22:13:09 <xarick> OT_GOTO_STATION > OT_LOADING > OT_LEAVING, the dest remains untouched
22:13:56 <xarick> whatever, I always forget the real names
22:17:02 <kuhnovic> Tbh, I did notice that the ship lost situation quickly resolves itself as soon as the ship "wiggles itself loose" by choosing a random direction. So it's a bug, but not that severe.
22:21:07 <_glx_> ship lost when leaving dock is not new, IIRC it was continuing forward, bump the dock and reverse, then find the path
22:40:46 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
22:54:36 <_zephyris> truebrain: Any idea what dev diary number the survey/social posts will be?
22:55:18 <truebrain> Ship is next, think the social went after that
22:56:05 <truebrain> Michi after, survey next, daylength last? Who knows!
22:56:29 <_zephyris> I'll make some .xcfs 🙂
22:58:38 <michi_cc[d]> I'm open to any order, but maybe my post in last place as it isn't actually describing any feature.
22:59:25 <michi_cc[d]> Not sure it should be stuffed somewhere in the middle.
23:00:13 <_glx_> oh I think I see what's wrong in #12125
23:00:31 <talltyler> Last place for your very nice sausage-making post makes sense to me
23:00:41 <truebrain> michi_cc[d]: Depends on Tyler and when he finishes daylength I guess
23:02:12 <peter1138[d]> Something in the OpenTTD font that Windows 8 doesn't support?
23:02:18 <peter1138[d]> (But that would be odd)
23:08:46 <michi_cc[d]> _glx_: Eh, doesn't even say if GetLastError() returns something useful in the error case.
23:11:00 <frosch123> why are the translators so spammy today? did someone advertise translations again?
23:12:09 <michi_cc[d]> No idea, who wrote the RC1 news post?
23:12:16 <_glx_> there's a mention it in RC1 post
23:15:47 <xarick> NonStopType on waypoints?
23:18:45 <xarick> nonstop means "go and don't stop at any intermediate stations if you happen to pass through them"
23:19:52 <xarick> go and don't stop at any intermediate stations if you happen to pass through them via this waypoint
23:20:17 *** Wormnest has joined #openttd
23:22:05 *** keikoz has quit IRC (Ping timeout: 480 seconds)
23:23:12 <xarick> ships can be given non-stop orders from AIs
23:24:08 <xarick> maybe instead of passing the flags like that, remove the flags before calling the Cmd?
23:25:05 <_glx_> hmm from stackoverflow it could be a non TTF font
23:38:59 <xarick> is OSL_PLATFORM_MIDDLE = 1, ///< Stop at the middle of the platform the new default?
23:39:05 <peter1138[d]> Oh of course, most logical explanation.
23:39:25 <xarick> should that default be applied to AIs?
23:41:23 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:44:23 *** ChanServ sets mode: +v tokai
23:45:02 <_zephyris> _glx_: I assume this is about the openttd font crashing Win 8?
23:45:18 <_glx_> yes, but it doesn't really makes sense
23:45:29 <_zephyris> What do you mean, like a OTF with a TTF extension or something?
23:45:43 <peter1138[d]> Is it actually the OpenTTD font, or some other font?
23:46:16 <_zephyris> Hah, I'm happier, everyone else is sadder
23:47:35 <_glx_> so it shouldn't even try to load openttd font
23:48:02 <_zephyris> In other news... This should sort the Serbian issue
23:50:43 *** Tirili has quit IRC (Quit: Leaving)
23:51:11 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
23:55:20 <_glx_> ``` so clicking on the button should not even switch to sprite font
continue to next day ⏵
