IRC logs for #openttd on OFTC at 2023-09-04
⏴ go to previous day
00:33:16 *** wallabra has joined #openttd
02:08:43 *** Wormnest has quit IRC (Quit: Leaving)
02:17:20 *** Flygon has quit IRC (Remote host closed the connection)
03:02:05 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:10:16 *** NGC3982 has quit IRC (Ping timeout: 480 seconds)
03:13:26 *** NGC3982 has joined #openttd
03:19:30 *** D-HUND is now known as debdog
03:47:31 *** Deep3D has quit IRC (Ping timeout: 480 seconds)
04:18:55 *** ChanServ sets mode: +v tokai
04:25:53 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
05:43:38 *** bryjen has quit IRC (Quit: Leaving)
05:58:35 *** keikoz has quit IRC (Ping timeout: 480 seconds)
06:30:22 <truebrain> At least there was room to expand 😄
07:11:50 <truebrain> hmmm ... I am getting a bit annoyed by the performance of Python in our backend lately .. can Cloudflare accept random TCP connections already? No .. darn .. 😛
07:13:18 <truebrain> also, my statistics of the backend is funny .. we have a negative amount of active turn connections, it says. For some reason, connections can drop without them going via disconnect functions ..
07:14:03 <truebrain> peter1139: can I nitpick about sentences not ending with dots, or shall I just leave that? 😄
07:14:10 <truebrain> (it is such a nitpick ....)
07:15:47 <peter1139> (Can you be more specific, sorry...)
07:16:15 <truebrain> sure; I just never know if that is worth complaining about, honestly 😄
07:17:24 <truebrain> performance wise, this array is terrible 😛 All operations are O(n) .. guess as long as it is small enough, it is not worth the overhead of a more complex structure. Also not for this PR btw.
07:18:19 <truebrain> the key seems to be (grfid, stringid), and the value (textholder, def_string)
07:18:44 <truebrain> definitially not for this PR 😄
07:21:31 <peter1139> Yes, I suppose so. One detail is that the index is looked up and passed around, so using some other keyed container must also support indexed access.
07:22:01 <truebrain> so another lookup from index to that pair .. yeah, it gets complicated really quick 😄
07:24:35 <peter1139> We could additionally store a map, per NewGRF, of stringid to index. Nearly every place (but not all) that uses GetGRFStringID also has access to GRFFile *.
07:25:35 <truebrain> I guess that is only worth it if profiling shows this is a performance issue; it was more the pedanticness in me that went: wait .. a `find_if` on a vector to find a pair .. that smells like a `std::map` 🙂
07:25:38 <peter1139> Or, they are allocated sequentially, presumably one NewGRF at a time. So it may be possible to limit the search by remembering the first and last IDs used by a NewGRF.
07:26:09 <truebrain> but I doubt this show up in any profile; NewGRF does a lot more stuff that takes more time 😄
07:28:04 <truebrain> I assume you tested it; was too lazy to validate 😄
07:29:01 <peter1139> It wouldn't be the first time!
07:29:14 <peter1139> (Although I do tend to say if I've not tested it...)
07:30:14 <peter1139> It took me a little parsing to understand what the returned widget counters were being used for.
07:30:59 <truebrain> I just went: I am sure he thought about if this was the right thing to return 😛
07:35:22 <peter1139> My car tax increased from £30 to £35. Definitely paying for all the damage it does there... :/
07:35:50 <truebrain> percentage wise, that is a pretty big increase
07:37:27 <peter1139> So that 20MB static array doesn't appear to show up in RES in top.
07:37:57 <truebrain> as frosch mentioned, most likely only when a page is being used, it is mapped in
07:43:03 <peter1139> Ah yes, comparing a blank savegame with FIRS loaded, shows about 20MB less used.
07:43:51 <truebrain> tomorrow I will run my performance benchmark, see how many games changes in memory-usage because of it 😄
07:46:18 <peter1139> Weird though, I can't remember what I was looking at when I spotted that and the array size jumped out that me.
07:56:08 <truebrain> now the question .. do I keep trying to fix the backend applications, to get more performance / less issues out of them
07:56:21 <andythenorth> peter1139: don't you have to complain about cyclists not paying for roads or something?
07:56:29 <andythenorth> isn't that expected tradition?
07:56:30 <truebrain> or do I accept that sooner or later things like Cloudflare allow WASM on custom TCP ports, and already go build it in WASM ..
07:57:07 <peter1139> andythenorth, obviously I'm too busy jumping red lights and breaking the speed limit to do that.
08:54:02 <andythenorth> does this make the game snappier?
09:19:51 <peter1139> No, it is unlikely to affect performance at all, but does prevent 20MB being allocated for no reason.
09:28:09 <peter1139> Apparently I didn't press the Enable auto-merge button hard enough.
10:05:09 *** virtualrandomnumber has joined #openttd
10:05:37 *** virtualrandomnumber has quit IRC ()
10:13:36 <truebrain> lol, whether a setting is its default value or not is a lot more tricky then I would like 😄
10:13:44 <truebrain> for the GUI it is simple, as it only deals with integer-based settings
10:13:53 <truebrain> but internally, we also have more complicated settings, like strings and lists 😄
10:20:06 <truebrain> that is one weird setting ...
10:21:19 <truebrain> `autosave_interval` ... didn't we remove that?
10:21:43 <truebrain> ah ... he forgot something essential .. 😄
10:22:09 <peter1139> No, it was just changed from a list to a value.
10:22:33 <truebrain> no, it was "fixed" to not be in the list
10:22:36 <truebrain> but that commit has a mistake
10:23:57 <truebrain> so that makes it a fix on a fix, we are so good in this software development stuff 😛
10:24:37 <peter1139> Always blame the reviewer ;-)
10:25:04 *** gelignite has joined #openttd
10:25:10 <truebrain> and the reviewer not actually testing what he is approving .. and the commiter not actually testing .. and owh owh owh, so many people to blame here 😄
10:25:23 <peter1139> I have to say, I find the manual `interface->Add()` lines a bit of a weird way to do it :-)
10:25:47 <truebrain> but I aint touching that!
10:25:57 <peter1139> Indeed. Definitely not in this PR. :)
10:26:09 <truebrain> already had to look into how the GUI actually shows what settings are non-default .. which is some weird code, now we made the settings into a variant
10:26:30 <truebrain> I also got annoyed that it is `small_font` and `medium_font`, instead of `font.small` and `font.medium` ..
10:26:44 <truebrain> but ... I aint touching that either! 😛
10:28:19 <truebrain> `"pf.forbid_90_deg": "false",` ... wait, why do I have this setting at false ...
10:28:25 <truebrain> I guess I was testing shit when we talked about it 😄
10:28:47 <truebrain> my configuration is such a mess of "things I wanted to test", making it nearly impossible to start a proper game without going: what the actual fuck
10:30:28 <truebrain> there was a thing with that setting a while ago
10:30:35 <truebrain> so I was flipping it around to test the change for that
10:30:43 <truebrain> clearly ... it was left in an unwanted state 😄
10:31:26 <_glx_> I once had `select_goods` in the wrong state
10:36:43 <truebrain> okay, in #11232 I took no effort to see if a list was different from default .. turns out, only resolution is using a list in modern games
10:36:58 <truebrain> also fine to not put any effort in it 🙂
10:37:26 <truebrain> owh, I am wrong .. `music.custom_1` and friend are also lists
11:16:28 <andythenorth> why is coffee? (see also)
11:17:37 <peter1139> I've gone from "hmm, probably need to eat a little bit more" to "oh shit, I definitely need to eat less" pretty quickly :(
11:19:27 <peter1139> Although to be fair yesterday was the club BBQ so bound to be temporarily a bit higher after that...
12:57:35 <truebrain> that is annoying ... lol
13:05:40 *** gelignite has quit IRC (Quit: Stay safe!)
13:12:57 <truebrain> also interesting: `micro 1708 3.0 0.4 62915453608 67924 pts/1 Sl+ 14:59 0:00 python test.py`
13:13:05 <truebrain> yes, you read that right .. 62 TiB of VSZ ..
13:21:03 <truebrain> never have I ever had an application where I had to worry about VSZ .. but this is reaching a limit 😛
13:21:43 <truebrain> very curious what they are doing to cause that ....
13:22:28 <truebrain> it is a Rust Python module .. let's try in native Rust, see if that shows the same issue ..
13:30:15 <truebrain> in native Rust it does exactly the same .. oh-oh .. that isn't good 🙂
13:35:01 <peter1139> Rust: ensuring it won't corrupt memory by never reusing it...
13:36:53 <truebrain> it is the application I am working with ..
13:38:02 <truebrain> after drilling down .. the WASM application is requesting memory, and the engine I am testing happily does so ... but for some reason a bit more than one would assume 😛
13:38:11 <truebrain> in WASM you allocate memory in pages of 64 KiB
13:38:29 <truebrain> so I request 1 page ... and ... it allocates like .... much much much more 😛
13:39:01 <truebrain> 6GB, instead of 64 KiB ..
13:39:44 <peter1139> Sorry, I guess I was confused by you mentioning rust three times in two lines.
13:40:01 <truebrain> yeah, I wasn't really clear in what I was doing .. I was just flabbergasted 😄
13:58:19 <truebrain> okay, it is this wasm engine that is the issue ... lol. What a weird issue ..
14:19:59 <truebrain> haha, each instance of the WASM I was spinning up creates a virtual allocation of 4GB
14:22:25 *** gelignite has joined #openttd
14:22:45 <peter1139> Urgh, everything on Linux seems to have its own different looking widget toolkit :p
14:23:26 <peter1139> Some of them have (fairly prominent) settings to make them look like other widget toolkits, but don't actually do that, they just look like something else again.
14:51:34 <alfagamma7> Well changing ISP is a hassle that I don't like but nvm
14:56:37 *** urdh has quit IRC (Quit: Boom!)
15:07:02 *** Kitrana has joined #openttd
15:12:17 *** Kitrana2 has quit IRC (Ping timeout: 480 seconds)
16:00:31 *** HerzogDeXtEr has joined #openttd
16:28:48 <andythenorth> it was, I finished it
16:37:39 *** Smedles has joined #openttd
17:51:55 <Eddi|zuHause> doesn't that cause cancer?
17:52:33 <kamnet> If you drink 100 cans of it a day, maybe.
17:54:05 <kamnet> Or, if you're a lab rat. But lots of human food you give to lab rats gives them cancer. That's why they're good for testing stuff, scientists are able to accurately predict stuff with them.
17:56:54 <Eddi|zuHause> there probably are people that drink that much :p
17:59:05 *** DorpsGek_vi[1] has joined #openttd
17:59:59 *** kamnet[d] has joined #openttd
17:59:59 <kamnet[d]> That's at least 1200 oz of fluid a day. Most people are absolutely sick of drinking any liquid when they hit about 200 oz in a day. Most people only consume 32-64 oz a day of any liquid, including liquid in your solid foods.
18:00:40 <peter1139> Spoken like someone who isn't addicted to Coke Zero / Pepsi Max...
18:01:40 <kamnet[d]> Oh I definitely was for a time. I could polish off 4 liters of soda in a day at one point in my life.
18:03:27 <kamnet[d]> 200 oz would be around 6 liters. 1200 oz is seventeen 2-liter bottles.
18:06:01 *** ketsuban[d] has quit IRC (Ping timeout: 480 seconds)
18:06:01 *** _glx_ has quit IRC (Ping timeout: 480 seconds)
18:06:01 *** michi_cc[d] has quit IRC (Ping timeout: 480 seconds)
18:06:01 *** _zephyris has quit IRC (Ping timeout: 480 seconds)
18:06:03 *** frosch123 has quit IRC (Ping timeout: 480 seconds)
18:06:04 *** emperorjake has quit IRC (Ping timeout: 480 seconds)
18:06:04 *** johnfranklin has quit IRC (Ping timeout: 480 seconds)
18:06:04 *** talltyler has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** ahyangyi has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** brickblock19280 has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** belajalilija has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** _jgr_ has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** alfagamma7 has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** andythenorth has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** truebrain has quit IRC (Ping timeout: 480 seconds)
18:06:05 *** DorpsGek_vi has quit IRC (Ping timeout: 480 seconds)
18:07:30 <kamnet[d]> Hm, just to check my math on that, human body can retain between 8 and 15 lbs of excess fluid before you start hitting serious health complications and organ shutdowns, which converts to around 120-240 oz of fluid. Your kidneys would be in overdrive at that level and probably at risk of entering renal failure.
18:09:37 <kamnet[d]> So basically if you tried taking on more than seven 2-liters of soda in a day, your organs would fail before the cancer even gets a chance to set in.
18:13:09 <Eddi|zuHause> so 2 liters is 6 cans, 6*7 is 42?
18:14:05 *** kamnet has quit IRC (Ping timeout: 480 seconds)
18:20:47 *** andythenorth has joined #openttd
18:20:47 <andythenorth> I knew someone who had kidney issues due to mainlining Coke all day
18:20:55 <andythenorth> probably around 2 litres
18:21:30 *** alfagamma7 has joined #openttd
18:38:16 <DorpsGek> - Update: Translations from eints (by translators)
19:14:35 *** talltyler has joined #openttd
19:15:28 <andythenorth> why don't we have GSCheat?
19:15:43 <andythenorth> and specifically GSCheat.SetDate() ?
19:15:54 <andythenorth> I could solve daylength in one simple move 🙂
19:20:41 <talltyler> Or just wait until Christmas when I do it properly 🙂
19:21:23 <talltyler> My mouth might be writing checks that my ass can’t cash, but eh
19:22:41 <talltyler> The goal is clearly in sight, there’s just a few more steps to get there 🙂
19:30:33 <peter1139> I have had cheese. No wonder I'm putting on weight too quickly :/
19:31:52 <talltyler> Huh, haven’t seen that British spelling before
19:34:44 *** ketsuban has joined #openttd
19:34:44 <ketsuban> "Deque" is **d**ouble-**e**nded **que**ue.
19:35:23 *** ketsuban is now known as Guest1755
19:35:24 *** Guest1755 is now known as ketsuban[d]
19:36:34 <ketsuban[d]> The homophony with "deck" is mainly just to be cute. It's not how we spell "deck".
19:40:58 <peter1139> ... What are you talking about?
19:41:32 <peter1139> Nobody pronounces deque as 'deck'
19:42:17 <ketsuban[d]> How else would you pronounce it?
19:44:28 <andythenorth> can we discuss how pronounce mySQL?
19:44:33 <andythenorth> or is that over?
19:45:08 <peter1139> It's pronounced "owned by Oracle"
19:50:07 *** truebrain has joined #openttd
19:50:07 <truebrain> Owh, not "garbage"? Must have been misprouncing it ...
19:50:18 <andythenorth> lol we made a Flash game for Oracle once
19:50:29 <andythenorth> we had an existing tech stack for high scores
19:50:48 <andythenorth> which used a not-Oracle DB
19:51:07 <andythenorth> the contract required us to rewrite the high score table to use Oracle
19:51:11 <andythenorth> we got paid extra
19:51:55 <andythenorth> cool story eh? 🙂
19:58:25 <peter1139> "GPL Roadtype Collection, Pt. 1/2/3/4/4/5/6". What?
20:00:55 *** brickblock19280 has joined #openttd
20:02:52 <andythenorth> how is reddit today though?
20:10:33 *** wallaby2 has joined #openttd
20:13:02 *** asymptotically2 has quit IRC (Remote host closed the connection)
20:13:03 *** jlx__ has quit IRC (Remote host closed the connection)
20:13:03 *** pickpacket has quit IRC (Read error: Connection reset by peer)
20:13:03 *** Hobbyboy|BNC has joined #openttd
20:13:03 *** wallabra has quit IRC (Read error: Connection reset by peer)
20:13:04 *** Hobbyboy has quit IRC (Read error: No route to host)
20:13:05 *** pickpacket has joined #openttd
20:13:07 *** asymptotically2 has joined #openttd
20:56:48 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:44:25 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:06:58 <_glx_> isn't there already a filter for that ?
22:25:29 <talltyler> Yes, there’s a drop-down for cargos. I intentionally did not look at cargo types in the text filter because of that. 🙂
22:26:03 <talltyler> I will write a comment accordingly in the PR later
22:26:48 <andythenorth> if we add enough text searches, the results will become unusable :
22:28:31 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
continue to next day ⏵
