IRC logs for #openttd on OFTC at 2024-03-12
⏴ go to previous day
00:12:53 *** Leopold has joined #openttd
00:18:44 *** Leopold has quit IRC (Remote host closed the connection)
00:25:24 *** Leopold_ has joined #openttd
00:43:43 *** Leopold_ has quit IRC (Remote host closed the connection)
00:48:38 *** Leopold has joined #openttd
00:54:54 *** Leopold has quit IRC (Remote host closed the connection)
01:03:04 *** Leopold has joined #openttd
01:10:04 *** Leopold has quit IRC (Remote host closed the connection)
01:18:32 *** Leopold_ has joined #openttd
01:33:26 *** Wormnest has quit IRC (Quit: Leaving)
02:06:36 *** Artea has quit IRC (Ping timeout: 480 seconds)
02:22:56 *** Leopold_ has quit IRC (Remote host closed the connection)
02:31:44 *** Leopold has joined #openttd
02:41:35 *** Qhrille has joined #openttd
03:40:11 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
03:44:37 *** gnu_jj_ has quit IRC (Ping timeout: 480 seconds)
04:26:51 *** Leopold has quit IRC (Ping timeout: 480 seconds)
04:33:51 *** Leopold_ has joined #openttd
04:41:27 <DorpsGek> - Update: Translations from eints (by translators)
04:48:33 *** ahyangyi has quit IRC (Quit: User went offline on Discord a while ago)
06:37:45 *** pickpacket9 has joined #openttd
06:39:41 *** pickpacket has quit IRC (Ping timeout: 480 seconds)
06:39:41 *** pickpacket9 is now known as pickpacket
06:46:20 *** Leopold_ has quit IRC (Ping timeout: 480 seconds)
06:48:32 *** Leopold_ has joined #openttd
07:16:30 <johnfranklin> Is Microsoft Surface worthwhile?
07:17:04 <truebrain> openttd-deployments[: So that works 😄 and it is a vcpkg fontconfig failure .. so that is annoying 🙂
07:29:50 <andythenorth> I should probably get a newer build
07:30:01 <andythenorth> finding bugs in RC1 isn't helping 😛
07:30:06 <andythenorth> they're all fixed
07:44:14 <andythenorth> I think RV signalling would be quite interesting 😛
08:14:48 <LordAro> there was a traffic light patch somewhere, iirc
08:18:10 <kamnet> It was in Chillcore's Patch Pack and Spring 2013 Patch Pack. I quite liked it.
08:18:51 <andythenorth> 'wait in depot until a slot is free' would do it eh
08:19:04 <andythenorth> we should do **more** things depending on depot hacks
08:21:14 <peter1138> What do traffics achieve...
08:21:27 <peter1138> What do traffic lights achieve...
08:22:56 <emperorjake> I remember that traffic light patch, it was an extreme performance hog
08:25:02 <kamnet> peter1138: Eyecandy/"realism". Road traffic will stop at intersections with lights and take turns going through, and several lights in a row could be synced to each other. Players could build them and towns could randomly place them.
08:29:24 <andythenorth> traffic lights would be used to hold vehicles ahead of clusters of connected road stops, until a bay is free
08:29:31 <andythenorth> I have no other use for them 😛
08:29:58 <andythenorth> it's taking a pathfinder difficulty and converting it to a player mechanic 😛
08:55:08 <peter1138> andythenorth: That implies they don't just switch state on a timer.
08:56:43 <peter1138> Mmm, such speaker destruction.
08:58:18 <peter1138> With proper resampler... less destruction.
09:42:54 *** ahyangyi has joined #openttd
09:43:07 <ahyangyi> I don't think I can imagine the sound from this graph
09:46:47 *** Artea has quit IRC (Ping timeout: 480 seconds)
09:47:00 <xarick> alright, there is a bug in GetRandomFollowUpTrackdir, another "edge case"
09:47:28 <xarick> it removes tracks which would make the ship make a 90 degrees turn
09:47:51 <xarick> however, in this case, this is the only available track
09:48:58 <xarick> poor ship really is forbidden to make 90 degree turns
09:49:05 <xarick> even if it's the only choice
09:51:42 <peter1138> Hmm, where are library dependencies defined now..
09:59:44 <kuhnovic> GetRandomFollowUpTrackdir has an option to allow/forbid 90 degree turns?
10:00:20 <kuhnovic> So that is by design, I don't know why you think it's a bug
10:01:07 <peter1138> What's the rationale for limiting 90 degree turns?
10:19:26 <reldred> Prevent trains taking the Forbidden Turn that Ruins Immersion in compact junctions.
10:19:49 <reldred> I generally avoid designing networks where that's possible in the first place but you know how players are
10:20:03 <reldred> Dunno about ships tbh
10:20:17 <peter1138> This function is specifically ship-only.
10:20:18 <reldred> are they two seperate settings?
10:20:26 <reldred> Well NFI on that one.
10:20:26 <peter1138> And nothing to do with that setting.
10:34:02 <xarick> the ship pointing towards north can only turn 90 degrees to the left
10:34:15 <xarick> but the track is removed
10:39:10 <peter1138> Damn, it's nowhere near lunchtime 😦
10:45:05 <xarick> is it just me or CreateRandomPath isn't working as expected? Wondering if it's a me problem
10:45:27 <xarick> seems to be pushing 8 of the same trackdirs to path_cache
10:47:08 <xarick> gonna test master instead
10:48:57 <kuhnovic> xarick: An alternative would be to discourage 90 degree turns instead of preventing them entirely. Take a random non-90 direction if available, and a 90 as a last resort. And INVALID_TRACKDIR if nothing is available.
10:50:36 <kuhnovic> peter1138: The rationale was that when a ship can't find a path, it starts to aimlessly wonder around the map. And having it randomly take 90 degree turns looked silly, because the ship had to slow down all of a sudden.
10:51:36 <kuhnovic> But I guess it would be better to return a 90 deg turn if that's the only option available.
10:55:28 <xarick> it's not updating tile nor dir for each time it advances
10:56:32 <xarick> also, going north should alternate between TRACKDIR_LEFT_N and TRACKDIR_RIGHT_N
11:07:03 <peter1138> kuhnovic: Well I don't need to suggest this now 😄
11:07:10 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick opened issue #12279: [Bug]: CreateRandomPath not updating tile nor dir for each step it advances, GetRandomFollowUpTrackdir forbids 90 degree turns even when it's the only choice https://github.com/OpenTTD/OpenTTD/issues/12279
11:07:53 <xarick> erm... they are related, hmm ok sorry
11:08:13 <kuhnovic> Can you add the savegame. preferably not a 4K 70k ships one 😛
11:08:29 <xarick> it's the 180 degree savegame
11:11:07 <xarick> actually... the savegame isn't proper
11:11:52 <xarick> nevermind, I'll work on reproducing it more concisely
11:28:26 <xarick> hmm not sure how to reproduce 12280 with master alone
11:32:04 <xarick> sounds more like a me problem
11:36:04 <xarick> I need CreateRandomPath on ChooseShipTrack to be called, meaning I need high_level_path to be empty
11:36:25 <xarick> but ChooseShipTrack must be reached
11:36:36 <xarick> and in master that doesn't happen, only on my branch
11:46:33 <xarick> alright, got this reproduceable in master
11:47:41 <xarick> took me a bit to get the ship in the right conditions
11:47:49 <xarick> but eventually got there
11:54:02 <rau117> Idk if this is a new bug or a bugfix
11:54:02 <rau117> Previously, if you built non-water structures at the very corner of the map at level 0, the water would not spread to such tiles. Now the water spreads anyway.
11:54:02 <rau117> For example, _the same_ prozone 2013 for 5k trains - now has a flooded trains on one side. It turns out that the compatibility is not 100% with old saves.
11:55:34 <ahyangyi> I think there's a setting that influences this behavior
11:58:32 <LordAro> one could argue an unintended side effect
12:09:16 <emperorjake> NetTrans had that same issue of edge flooding
12:13:01 <Eddi|zuHause> rau117: i think that change was ages ago
12:13:47 <LordAro> we could just remove edge flooding entirely
12:15:00 <andythenorth> "there is a goblin here"
12:15:58 <peter1138> Wasn't that a Xarick-change?
12:22:09 <Eddi|zuHause> rau117: but savegame compatibility was never this exact, like we changed acceleration model at some point, "breaking" all kinds of people's savegames by making trains slower than previously
12:27:58 <rau117> Eddi|zuHause: hmm... were there any changes that led to the destruction of structures?
12:27:58 <rau117> slow trains are certainly unpleasant, but solvable
12:27:58 <rau117> fixing/relocating sunken technologies is a much more difficult task
12:48:26 <talltyler> Working on the posts to follow 🙂
12:55:23 <peter1138> I have... a sausage roll 😄
13:03:36 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
13:16:12 *** SigHunter has joined #openttd
13:42:11 *** Artea has quit IRC (Ping timeout: 480 seconds)
14:07:38 <xarick> rau117: flooding was only for tiles of type Clear
14:08:59 <xarick> I removed the special case 😐
14:11:04 <xarick> and gave void tiles the power to flood adjacent tiles regardless of type
14:11:41 *** nielsm has quit IRC (Ping timeout: 480 seconds)
14:11:55 <peter1138> You ruined the game 😉
14:13:11 <xarick> I mean... they're on height 0...
14:15:32 <peter1138> I thought this was already in 13.x tbh.
14:17:05 <peter1138> Jan 26, 2023, maybe, maybe not.
14:19:47 <xarick> not even tree tiles were being flooded
14:20:03 *** pickpacket0 has joined #openttd
14:20:23 <xarick> it felt very... random
14:24:27 <peter1138> Oh well. I did think about posting it to social media to get more input but I missed the deadline on that.
14:24:46 *** pickpacket has quit IRC (Ping timeout: 480 seconds)
14:24:46 *** pickpacket0 is now known as pickpacket
14:36:27 <talltyler> Anywhere we should announce voting besides Discord channel #announcements and TT-Forums? (both of which I have done)
14:39:36 <LordAro> xarick: i don't see it as a problem
14:39:55 <LordAro> sure, it affects a couple of weird edgecases
14:44:38 <peter1138> talltyler: the filenames within the zips could be better...
14:45:43 <peter1138> ErorrTNT's is 512x512 which is a bit much for the title game.
14:48:42 <peter1138> accipiter2000's is just trains and towns, yikes.
14:49:22 <peter1138> They all only have one company.
14:50:21 <merni> peter1138: All maglevs too
14:50:54 <xarick> omg my save is unreadable why..
14:52:36 <peter1138> Really only 4 candidates? There are all pretty horrible.
14:53:42 <merni> Plus C's is okay imo but rather sparse
14:54:08 <peter1138> Plus C's is the only one in the race, but it's still just somebody's game that they've been playing.
14:54:13 <talltyler> You should have entered, then?
14:54:23 <peter1138> Too busy developing.
14:54:24 <talltyler> For what it's worth I spent a lot of time on my entry 🙂
14:55:12 <peter1138> Yours does at least use new features (infinite money and stopped time)
14:55:20 <peter1138> Gamelog reveals all...
14:55:20 <talltyler> And depot unbunching
14:55:36 <talltyler> Money and time are just because those are my default newgame settings 😛
14:56:04 <peter1138> accipiter2000's had trams in it at one point, then they were removed.
14:56:38 <peter1138> Not good for a title ggame.
14:56:51 <talltyler> Lovely, there was a title game last year that had GRFs in it
14:56:57 <peter1138> (Why is gui.fast_forward_speed_limit logged?)
14:57:11 <talltyler> Fast forward is cheating? 😛
14:57:31 <peter1138> But only if you change how fast fast forward is?
14:57:46 <peter1138> I'd say scrap that one because of the NewGRFs
14:57:49 <talltyler> It's for speedruns obviously 😄
14:58:57 <peter1138> ErorrTNT is a huge map with mostly nothing on it, but because it's a huge map it's a large file.
14:59:21 <merni> peter1138: why's that a problem :p
14:59:29 <merni> I mean, sure, we have a honking red warning
14:59:46 <merni> but does it cause actual issues in these simple cases
14:59:56 <peter1138> Because title games can't have NewGRFs, and including a title game which has had them added and removed is... urgh.
15:00:12 <merni> could you expand on the "urgh" bit :p
15:00:23 <peter1138> We tell people not to do it. At all.
15:00:33 <peter1138> So including a savegame with it is a no-no for me.
15:00:39 <merni> peter1138: sure, but nobody listens
15:00:58 <peter1138> Yes, and including it would give them more reason to not listen.
15:01:07 <merni> I don't think anyone would even know
15:01:27 <talltyler> On the flip side, if we throw out entries, we will have even fewer than we do now...
15:01:37 <merni> but frankly the title competition is not going to be a motivator for anyone to change their behaviours in usual gameplay
15:01:58 <peter1138> On the other flip side, it's a bad title game anyway. Just ugly super-long maglevs and dense town builds.
15:02:22 <merni> adding newgrf in a running game is a quite useful feature, and lecturing people about it is not going to help anything. If they do something really bad like messing with industry sets they'll find out anyway :p
15:02:28 <talltyler> Anything wrong with Plus C's entry? That has my money as the player choice. 🙂
15:02:39 <merni> peter1138: Yeah, let its poor quality eliminate it naturally
15:02:58 <peter1138> It's literally in the entry rules. Do not use NewGRFs.
15:03:32 <talltyler> I am disappointed that eratonysiad didn't submit a Toyland entry. It might have had a chance this year. 😦
15:03:46 <merni> peter1138: The submitted game doesn't :p
15:04:02 <talltyler> I voted for your Toyland entry every year 😄
15:04:48 *** Leopold__ has joined #openttd
15:04:59 <talltyler> talltyler: Early voting seems to agree, 14 votes with the second-place only having 4 🙂
15:05:25 <peter1138> I should just "Mark as answer" ;p
15:06:27 <talltyler> If you're volunteering to be in charge of the title screen competition, I won't object 😛
15:06:45 *** Leopold_ has quit IRC (Ping timeout: 480 seconds)
15:07:00 <peter1138> Plus.C is a bit too busy with too many vehicles, but nothing really bad.
15:07:21 <talltyler> Odd signalling and track layout in a few spots, but nothing wrong with that
15:07:40 <peter1138> Also they start paused, not sure if that needs to be fixed 🙂
15:08:07 <peter1138> Oh, and the Dune reference might be an issue, but probably not.
15:10:04 <merni> peter1138: The guidelines specify a "busy, exciting scene" lmao
15:10:42 <talltyler> Starting paused is your client setting, no? It works fine as `opntitle.dat` 🙂
15:10:50 <talltyler> What is the Dune reference?
15:11:11 <peter1138> Can be achieved by designing a title game instead of just sending in your normal game you played 😉
15:11:41 <peter1138> Starting paused only for new games, otherwise it's saved in the game.
15:12:00 <_glx_> and intro game as nothing removing the pause
15:12:27 <talltyler> When you have multiple camera positions, it's often better to build a regular game that you can capture later
15:12:59 <talltyler> Especially since people have different resolutions and sometimes see larger areas than you intend
15:13:20 <peter1138> Did you even look at these savegames/ :p
15:13:28 <merni> I mean... the company name is not even visible in the intro game
15:13:42 <peter1138> It doesn't matter at all, I just found it amusing.
15:13:47 <talltyler> I opened them as intro games, why would I look at the rest of the game?
15:14:09 <peter1138> To check things like "is it huge" and "what's in the gamelog". I dunno 🙂
15:15:45 <peter1138> Hmm, I was waiting for CI to finish, but got sidetracked. Now I can't find that tab...
15:15:47 <talltyler> I've never looked at the gamelog 😄
15:16:06 <talltyler> I didn't even know how until I saw your screenshot, figured it was a console command, and tried `gamelog`
16:14:26 *** Wormnest has joined #openttd
16:18:16 <merni> TallTylerviaGitHub: Why should a dropdown with unrelated options like this have a default?
16:18:41 <merni> I don't think any of those actions are so much more frequent than others that it would justify being made a default
16:21:58 <peter1138> It does for the normal vehicle list, but to be honest I don't think it should.
16:22:16 <peter1138> It does not select that default action if you cancel the menu.
16:22:31 <xarick> what the heck happened to league table menu?
16:25:56 <peter1138> xarick: Dunno but it's interesting that it even works 😄
16:26:40 <xarick> it was supposed to list the title, not its contents
16:30:48 <xarick> wondering if I can reproduce it
16:33:26 <xarick> okay I think I know what's happening
16:33:45 <xarick> original savegame was 13.4, the menu was bugged back then
16:34:16 <xarick> so loading from the bugged save, this happens
16:35:26 <xarick> I don't think it's the saved data I generate
16:35:47 <xarick> it's fine if save and load versions are 14
16:38:14 <xarick> this is how it's supposed to look
17:38:46 <frosch123> hmm, every titlegame compeition i remember we wanted to change the intro menu
17:39:06 <frosch123> so it does not block the view on the vehicle in the center of the screen, which it is following
17:39:53 <_jgr_> That seems like getting the priorities back to front?
17:40:14 <_jgr_> The title save is background decor for the main menu
17:41:20 <frosch123> some consider it a tutorial
17:55:45 *** HerzogDeXtEr has joined #openttd
18:35:03 <peter1138> Hmm, something is hogging my CPU 😮
18:56:28 <peter1138> And also not possible, because virtual groups don't exist.
19:15:15 *** Leopold__ has quit IRC (Ping timeout: 480 seconds)
19:22:28 *** Leopold has joined #openttd
19:34:27 *** Artea has quit IRC (Ping timeout: 480 seconds)
19:43:28 <rutoks> LordAro: could you please take another look at my console autocompletion PR?
20:03:26 <xarick> i hate pointers and references, i never get this right
20:24:00 <_jgr_> andythenorth: It's fixed already
20:24:15 <andythenorth> I have just downloaded nightly build instead
20:24:25 <andythenorth> the crash was RC1, and the nightlies were also broken 😛
20:24:46 <andythenorth> I keep finding RC1 issues, that are already known 😛
20:24:46 <rutoks> Does github checks `Textbuf* a` automatically?
20:25:47 <andythenorth> is there a way to increase autosave frequency?
20:25:58 <_jgr_> No, generally PR reviewers look for syntax stuff like that
20:26:03 <andythenorth> I have lost progress, 10 mins it's pretty much unworkable 🙂
20:32:25 <xarick> sorry I don't know how to use & stuff
20:32:33 <xarick> going with the std::pair solution
20:40:37 <_glx_> andythenorth: use the console for less than 10 values
20:41:45 <_glx_> `set autosave_interval`
20:42:06 <xarick> let us chose 1 minute in gui plz
20:45:05 <_glx_> it's a timer else it would never autosave during pause even with build while pause
21:35:05 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:37:39 <truebrain> more detail isn't always better 😄
21:39:11 <truebrain> really need to load this data in a database, so I can make 0% another point in time for each entry .. currently 0% is the first measurement I ever did 😄
21:39:50 <truebrain> anyway, dates memory was impacted: 20240108, 20240114, 20240203
21:40:27 <truebrain> and for CPU the first two dates
21:40:43 <truebrain> later this week I will test what happens when the sprite-font is used, to see if that is what is happening on the 14th
21:43:03 <truebrain> the memory jump in 20240203 is unbunching, which adds 20 bytes per BaseConsist
21:43:19 <truebrain> (I don't have an opinion about whether those jumps are right or wrong btw; just want to make sure we know what makes them jump 😄 )
21:44:22 <truebrain> so that means the biggest unknown is why the CPU jumped this much with the improved ShipPF
21:44:25 <kuhnovic> xarick: Ironically you were already doing it with the &tile function parameter
21:44:30 <truebrain> that feels a bit odd; especially as most maps don't have ships
21:46:26 <truebrain> maybe something kuhnovic knows 🙂 (no rush, but mostly wondering why all games have an increased CPU usage since improved ship PF)
21:46:32 <_glx_> invalidation of water regions (it's just setting a flag, but happens everytime a tile is affected)
21:47:28 <truebrain> sounds like a bad trade, to have that consume 10% more CPU 🙂
21:47:43 <truebrain> (worst case, btw 😛 )
21:48:06 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:48:12 <_glx_> but it's just a guess, since you have no ships that's the only thing I can see coming from improved PF
21:48:26 <truebrain> you made it sound like you actually investigated it 😉
21:49:27 <truebrain> anyway, nothing else odd jumps out in the last few months; so that is good 🙂
21:51:05 <_glx_> there's a strange temporary huge increase in memory usage between 0111 and 0127
21:51:23 <truebrain> yeah, but something that got resolved
21:51:27 <truebrain> 0128 is the date of resolving
21:52:04 <truebrain> I also removed a line where for 1 game the CPU spiked to 200% for a few weeks
21:52:08 <truebrain> but that too was fixed 😛
21:52:43 <truebrain> (I am mostly focused on 14.0 performance, that we didn't regress 🙂 )
21:55:06 <kuhnovic> _glx_: This is indeed what could be the reason, although setting a boolean should be very lightweight. It might have something to do with cache misses though. All water regions store their validation state as a boolean, and since regions are relatively large objects in memory there's a good change the data is no longer in a cpu cache.
21:55:48 <kuhnovic> The recent PR from JGR might helpt with this. But it's just asumptions at this point.
21:55:48 <peter1138> Sounds like a job for a bitmap 😉
21:56:06 <kuhnovic> peter1138: Yup, I was thinking the same
21:56:06 <truebrain> it would be appreciated if someone can just benchmark it 🙂 It should be really visible 🙂
21:56:23 <_jgr_> If it was that expensive it ought to show up in a profile, I can't say that I've seen anything like that in any of the profiles that I've done
21:56:42 <truebrain> openttd 1.0 title game was the one with the biggest impact
21:57:41 <kuhnovic> If anyone has a benchmark then I'd you to see it 😁
22:00:37 <peter1138> Vehicle specialization is a bit out of order.
22:01:33 <peter1138> DisasterVehicle and EffectVehicle are also SpecializedVehicle, which inherits from Vehicle. So they have BaseConsist information too.
22:02:17 <_jgr_> The OpenTTD 1.0 title game doesn't seem to have very much in it, that could make small changes look relatively larger
22:04:12 <truebrain> _jgr_: Yeah, so this should be easier visible there 🙂
22:04:18 <kuhnovic> If it had many ships you would have seen an improvement 😉
22:05:04 <_glx_> it would have many buoys in this case 🙂
22:05:21 <_glx_> 1.0 was using OPF so buoys were mandatory
22:05:42 <_jgr_> The ships are not exactly going very far
22:05:49 <_glx_> oh maybe it had NPF already
22:06:52 <peter1138> 1.0? Pretty sure we have YAPF by 1.0.
22:07:10 <kuhnovic> Anyway, making things more cpu cache optimal is not a bad idea. As long as it doesn't make the code an incomprehensible mess, but I don't expect that.
22:07:41 <kuhnovic> Goes on my mental stuff-to-do-when-I-have-time-lisr
22:08:02 <peter1138> Did we use the old old title game for 1.0 as well, or is it a different save?
22:09:52 <peter1138> In the unit test. Joy.
22:10:31 <truebrain> And I have some ideas how we can have constant performance measuring ... something to test out later this week 🙂
22:11:06 *** salut3585 has quit IRC (Quit: User went offline on Discord a while ago)
22:11:42 <xarick> I believe it's town growth
22:11:54 <xarick> houses taking sea space
22:12:06 <xarick> trigger the region update thing
22:12:28 <_glx_> water regions can be full land too 😉
22:12:58 <xarick> oh, so that means all towns?
22:13:27 <_glx_> but it's just setting a bool, the real update happens when pathfinding if needed
22:13:47 <xarick> oh, right, forgot that little detail
22:14:24 <truebrain> Does it also set a bool when trees grow? (Just curious)
22:14:59 <_glx_> IIRC it's when a tile is cleared
22:15:19 <truebrain> That doesn't happen all that often
22:17:24 <xarick> so it requires actual pathfinding to do the region calculations
22:18:27 <xarick> let me investigate something
22:19:08 <xarick> `Debug(map, 3, "Updating water region ({},{})", GetWaterRegionX(this->tile_area.tile), GetWaterRegionY(this->tile_area.tile));`
22:19:16 <xarick> see if that gets triggering
22:20:18 <_jgr_> I'm having hard time finding anything to do with water regions at all in a rather noddy profile of the 1.0 game I've just done
22:25:13 <xarick> was that during the time water regions was doing a full map Force Update?
22:26:00 <xarick> yes, but on those dates
22:26:37 <kuhnovic> Not sure what you mean by that
22:28:20 <xarick> I'm not sure what TrueBrain is testing specifically, is it the intro game?
22:28:32 <kuhnovic> Ah right. That might actually be a factor. Although it's a one time thing, and on a small map it shouldn't have too much overhead
22:28:57 <kuhnovic> (bedtime now, I'll read this tomorrow)
22:32:26 <truebrain> _jgr_: I guess we need to compare the profile of before and after, to see where the 10% is 🙂
22:33:53 <xarick> how do I download that? need to test
22:33:57 <truebrain> Most games btw show a jump, but I was hoping that a small game made it easier to spot 🙂
22:39:31 <xarick> I always forget where they can be downloaded from
22:49:51 <xarick> there's a lot of invalidations, may be the act of invalidating itself?
23:12:17 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
23:18:08 <xarick> by the way, with NPF, the ship gets stuck trying to reverse
23:18:28 <xarick> with that 1.0 savegame
23:19:17 <xarick> this leads to an increase in cpu time
23:19:23 <xarick> could it be the cause?
23:21:10 <xarick> but it's unrelated to WaterRegion...
23:24:51 *** Flygon has quit IRC (Remote host closed the connection)
23:37:33 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
continue to next day ⏵
