IRC logs for #openttd on OFTC at 2024-04-03
⏴ go to previous day
00:05:35 *** bertvvvs has joined #openttd
00:09:34 *** bertvvvs has quit IRC (Quit: You have a nice day.)
00:17:23 *** bertvvvs has joined #openttd
00:21:15 *** bertvvvs has joined #openttd
01:47:02 *** Wormnest_ has quit IRC (Quit: Leaving)
01:57:38 <johnfranklin> Renewed City Growth is obsolete
01:57:46 <johnfranklin> Try Renewed Village Growth
02:09:05 *** herms has quit IRC (Quit: bye)
02:10:36 *** gnu_jj_ has quit IRC (Ping timeout: 480 seconds)
02:18:19 <kamnet> klote: This was written when FIRS 2 was current, so try switching to that version instead. Or, as noted, switch it up to Renewed Village Growth.
02:33:18 *** wormnest has quit IRC (Quit: User went offline on Discord a while ago)
03:01:59 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:31:23 *** bertvvvs has quit IRC (Quit: You have a nice day.)
04:40:38 <DorpsGek> - Update: Translations from eints (by translators)
05:07:01 *** keikoz has quit IRC (Ping timeout: 480 seconds)
06:02:30 <andythenorth> FIRS detector GS lib?
06:58:48 <andythenorth> probably true in the Pilbara
07:01:48 <andythenorth> can't imagine this doesn't depend heavily on sshd
07:36:47 *** marchnex has joined #openttd
07:47:01 *** marchnex has quit IRC (Quit: marchnex)
07:51:13 *** Smedles has joined #openttd
08:03:50 *** HerzogDeXtEr has joined #openttd
09:45:43 *** D-HUND is now known as debdog
10:03:42 *** matusguy has joined #openttd
10:58:31 *** gelignite has joined #openttd
11:01:12 <pickpacket> kamnet: good morning :) How's life?
11:02:00 <kamnet> pickpacket: It trudges on, my friend. Survived yesterday's tornadoes and winds.
11:02:23 <xarick> earthquake in Taiwan, inb4 rising prices
11:04:15 <pickpacket> kamnet: oh my! I'm glad. Tornadoes seem pretty scary
12:28:34 *** gelignite has quit IRC (Read error: Connection reset by peer)
13:28:57 *** keikoz has quit IRC (Ping timeout: 480 seconds)
13:38:58 <xarick> lower profits for the same save
13:39:05 <xarick> on different versions of openttd
13:39:51 <xarick> gonna disable breakdowns, just to take that variable out
13:40:07 <_glx_> something changed in these IIRC
13:42:15 <xarick> gonna restart test without breakdowns
13:45:41 <xarick> yeah, it's definitely less in 14
13:46:42 <_glx_> what is the route length ?
13:47:30 <xarick> 274 realfake tiles, not sure about it in manhattan
13:47:58 <xarick> let me get coordinates
13:48:57 <xarick> 655 x 553 to 882 x 411
13:52:58 <peter1138> Cargo payment graph will be identical, it doesn't show the longer times.
13:53:31 <peter1138> 369 isn't that long for aircraft though.
13:55:23 <peter1138> Hmm, not the same aircraft.
13:58:37 <peter1138> > ssh always allowed remote code execution.
14:02:02 <_glx_> yeah of course, but the idea was to filter who could run the code
14:03:44 <peter1138> Urgh, can't find my rim tape :S
14:04:12 <xarick> 13.4 makes 25% more profits, these numbers seem quite round
14:04:24 <xarick> or 14 makes 80% of the profits of 13.4
14:06:19 <peter1138> `git diff 13.4 master`
14:06:43 <peter1138> > 1400 files changed, 162400 insertions(+), 86742 deletions(-)
14:06:51 <peter1138> That's quite some bloat.
14:06:55 <_glx_> changelog has a summary
14:07:55 <peter1138> > src/3rdparty/nlohmann/json.hpp | 24765
14:08:22 <peter1138> Okay, so catch2 and json are quite a chunk together.
14:08:34 <peter1138> fmt too, but that was there in 13.4 too, although a smaller version.
14:10:01 <peter1138> widget.cpp has 121 less lines, nice.
14:10:36 <_glx_> moved to other files most likely
14:11:20 <peter1138> media/baseset/orig_extra.grf | Bin 329463 -> 445331 bytes
14:11:26 <peter1138> hiding those exploits
14:12:53 <_glx_> hey we could add a hidden action for real asm 🙂
14:14:24 <LordAro> buffer overflow + stack bust is probably a thing somewhere
14:22:39 <xarick> wondering if it's this
14:24:09 <_glx_> most likely an effect of this one yes
14:28:07 <peter1138> That is the change we were refering to, yeah 🙂
14:39:37 <xarick> int periods_over_max = MIN_TIME_FACTOR - MAX_TIME_FACTOR; why isn't this also a const
14:42:03 <LordAro> it's unlikely the compiler doesn't know that
14:42:56 <xarick> periods_over_max is still being changed aftewards
14:45:30 <xarick> doesn't seem to be this that's affecting income 😦
14:57:16 *** keoz has quit IRC (Ping timeout: 480 seconds)
14:57:37 <LordAro> box_name_re = re.compile(fr"{target_version}-(.*).box")
14:57:40 <LordAro> gotta love some french strings
15:07:55 <xarick> confirmed, it's not this... something else is affecting payments
15:10:37 <xarick> my AI can't survive anymore
15:24:34 <peter1138> Probably not much effect over 369 tiles.
15:24:36 <xarick> nope, it's just an airport station
15:24:47 <xarick> no spreading with other stations
15:32:26 <xarick> let's go back to 19-09-2023, see if it changed anything
15:44:11 *** Wormnest has joined #openttd
15:49:32 *** keikoz has quit IRC (Ping timeout: 480 seconds)
16:16:02 *** bertvvvs has joined #openttd
16:21:06 <silent_tempest> I don't have a mac but my laptop is a high dpi display. I saw this issue:
16:21:06 <silent_tempest> I think I have a repro. That is a as far as I can zoom the minimap on a 64x128 map. What should it look like if the zoom actually works like it's supposed to?
16:40:06 <xarick> is v->tile safe to use on an aircraft?
16:40:48 <xarick> i guess it's fine, it's not flying
16:47:44 <peter1138> silent_tempest: Minimap extra zoom in is not implemented, so technically it's working as it's supposed to.
16:48:18 <silent_tempest> It's pretty un-readable though
16:48:50 <silent_tempest> But fixing it is "adding extra features" to enhance the zoom levels?
16:52:09 *** misterbungus3 has joined #openttd
16:55:54 *** misterbungus has quit IRC (resistance.oftc.net larich.oftc.net)
16:55:54 *** greeter has quit IRC (resistance.oftc.net larich.oftc.net)
16:55:55 *** misterbungus3 is now known as misterbungus
16:58:05 *** greeter has joined #openttd
17:10:03 <xarick> okay, I see the manifestation of the problem
17:10:09 <xarick> current_tile is 0 for mail
17:11:16 <peter1138> Is it meant to go by the head vehicle?
17:12:28 <xarick> v->tile is 0 somewhere...
17:15:17 <peter1138> `uint loaded = ge->cargo.Load(cap_left, &v->cargo, next_station, v->type == VEH_AIRCRAFT ? front->tile : v->tile);`
17:15:24 <peter1138> Might be other places.
17:15:49 <peter1138> PrepareUnload() uses v->tile too.
17:16:09 <peter1138> And ReturnCargoAction::operator()
17:16:18 <silent_tempest> _glx_: Is it a zoom level issue or a scaling issue?
17:16:26 <peter1138> And FinalizeRefitAction::operator()
17:16:38 <silent_tempest> The inteface scaling setting has no effect on the size of the map rendered in that window
17:17:14 <peter1138> It's a "nobody has implemented it" issue, as I said.
17:17:42 <peter1138> (Well, JGRPP has an implementation, long ago)
17:18:10 <peter1138> There's no particular reason for minimap zoom to be affected by interface scale.
17:18:16 <peter1138> So it just needs more zoom levels adding.
17:18:39 <peter1138> (And it's nothing to do with "extra zoom", as it's not drawn as sprites, of course.)
17:19:18 <silent_tempest> I mean the end goal is make the map readable right? Then scaling it by the gui scale factor would do that right?
17:19:46 <_glx_> gui scaling is applied, it's just there's no zoom in levels
17:20:38 <xarick> seems the asserts are only checking for INVALID_TILE, but not 0
17:21:46 <silent_tempest> maybe this would make more sense to me if I find the minimap code.
17:22:08 <silent_tempest> And that is? map.cpp seems to code for the actual world map?
17:23:54 <_glx_> it's in smallmap_gui.cpp
17:24:46 <silent_tempest> okay I had just found it before I saw your message. Thanks for confirming
17:28:19 <_jgr_> peter1138: It's really dP's code, from citimania, if I remember rightly
17:32:21 <xarick> okay, all the v->tiles are in economy.cpp
17:33:54 <xarick> I don't know how to reach those reserve stuff
17:37:53 <xarick> oh, there's more changes that happened afterwards... 😦
17:41:36 <peter1138> You could report it as a bug.
17:41:41 <peter1138> And let someone else fix it 🙂
17:42:13 <_glx_> Check 9c49a61 diff for economy.cpp and replace `v->tile` with `v->type == VEH_AIRCRAFT ? front->tile : v->tile);`
17:42:50 <peter1138> virtual TileIndex GetCargoTile() const { return v->tile; }
17:43:10 <peter1138> TileIndex Aircraft::GetCargoTile() const override { return v->front->tile; }
17:43:41 <_glx_> But a bug report should be enough
17:44:01 *** Exec has quit IRC (Quit: File:///)
17:44:25 <_glx_> I may look at it after dinner
17:45:18 <_glx_> You already did the time consuming task of finding the breaking commit
18:32:40 <xarick> I happened to be testing aircraft with maintenance costs on
18:32:56 <xarick> and was surprised my AI was struggling
18:36:53 *** As1 has quit IRC (Remote host closed the connection)
18:45:51 *** gelignite has joined #openttd
18:52:35 <truebrain> So the waiting game continues
18:54:16 <andythenorth> for a some years at work, some upstreams were so often down that we had our own cache
18:54:30 <andythenorth> as it was the only way to reliably meet SLAs for disaster recovery
18:54:38 <andythenorth> and everything was pinned versions any way
18:54:44 <andythenorth> but it was a PITA
18:55:45 <truebrain> I am more curious what their security team is analyzing 😛
18:59:13 <frosch123> maybe they deployed sshd on their own servers
18:59:18 <frosch123> would give them plenty to check
19:01:32 <truebrain> I doubt that would keep the xz repo locked and vcpkg orders to not change mirror 😄
19:01:58 <frosch123> a former employer of mine got hit by ransomware 7 weeks ago, they are still busy with recovery
19:03:04 <truebrain> "Did you test your backup strategy?" - "Yes, we have backups" ... not the question ... not the question ....
19:07:00 <xarick> changed everywhere to v->First()->tile
19:07:10 <frosch123> if gh was hosted in germany, i would suspect that authorities confiscated the harddrive storing the repo
19:07:32 <truebrain> cute to think it is a single harddrive 🙂
19:08:41 <frosch123> well, there are stories i can't tell in public 🙂
19:10:02 <_glx_> oh here they could have blocked the IP (via ISP DNS)
19:10:23 <xarick> I think mail was getting a payment of 0
19:10:38 <frosch123> oh, true, they could have blocked all gh ips. we may be lucky then
19:11:50 <klote> frosch123: Yikes, still everything on local storage?
19:12:23 <klote> I pushed Sharepoint and Onedrive usage within the company
19:12:38 <_glx_> ok let's look at fixing 12415 (I already know where)
19:12:44 <klote> has a retention of several months
19:13:17 <klote> havent had a ransomeware issue yet
19:13:58 <truebrain> you don't till you do. So here too, the question is: do you have a Disaster Recovery Plan and did you execute it at any point in time? 😄
19:14:44 <_glx_> the most important issue for ransomware is human
19:14:44 <klote> is part of our ISO27001 certification
19:14:56 <truebrain> lolz; 27k1 wants you to have it on paper
19:15:15 <truebrain> the amount of companies I know that have 27k1, but do not execute a DRP on any regular base is far longer than the ones that do 😛
19:15:25 <frosch123> oh, don't worry, they have 27k1 certificates for years
19:15:33 <truebrain> good; keep it that way 🙂
19:15:46 <truebrain> frosch123: I noticed that getting 27k1 certificate is just paperwork. It has no actual meaning 😦
19:16:01 <klote> truebrain: then your auditor is shit
19:16:10 <truebrain> haha, if only that was true 🙂
19:16:24 <frosch123> yes, it's the same as enterprise scam linux. you do not have to fix any bugs, you just have to document them
19:16:32 <klote> i need to show proof of restores being succesfullk
19:16:39 <truebrain> yup; which is trivial to do
19:16:48 <truebrain> doesn't mean you actually restored anything useful
19:16:50 <truebrain> on any useful system
19:17:01 <truebrain> so most people do it on their own laptop with a single file, or something shitty like that
19:17:35 <klote> i mean i do restores of entire cold storage envoirenment so...
19:17:55 <truebrain> repeat it every N months, and you are good 🙂
19:18:08 <truebrain> when disaster strikes, you want to have 0 doubt what to do to recover
19:18:23 <truebrain> as you will be in a panic, and you will feel the pressure of getting everyone "back to work"
19:18:38 <klote> had that happen with the kaseya hack X-X
19:19:11 <klote> all systems in field potentially hit
19:20:05 <truebrain> lolz, so 7z SDK can be downloaded only in 7z format 🙂
19:20:12 <truebrain> now that is some "eat your own dogfood"
19:35:19 *** gelignite has quit IRC (Quit: Stay safe!)
19:42:32 *** bertvvvs has quit IRC (Quit: You have a nice day.)
19:45:15 *** ChanServ sets mode: +v DorpsGek
19:45:15 *** ChanServ sets mode: +v blathijs
19:45:15 *** ChanServ sets mode: +v orudge
20:09:06 <_glx_> xarick: you have something to test 🙂
20:13:47 <_glx_> I just looked at 9c49a61
20:14:14 <xarick> there was an addendum later on 19th sept
20:18:24 <xarick> I didn't test the other vehicle types..
20:18:37 <xarick> wondering what happens with trains
20:19:10 <truebrain> okay, if we so wish, we could switch from xz to read our savegames to libarchive; that seems to be reading the savegames just fine
20:20:30 <_glx_> I think other articulated vehicles set v->tile correctly when they enter a tile
20:27:31 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:34:54 <xarick> it's correct for trains, it actually points to the location of the wagon
20:35:07 <xarick> but articulated, how do I test this?
20:39:31 *** maxshpuntozavr has quit IRC (Quit: User went offline on Discord a while ago)
20:40:24 <_glx_> trams, and you could also try some road hog vehicles
20:51:59 <xarick> articulated trams work with v->tile fine
20:52:14 <xarick> at least this function I'm looking at
20:56:43 <xarick> 1801 isn't entirely correct, I think
20:58:21 <xarick> ah no, my bad, it's correct
21:00:00 <xarick> no reason to think articulated trains wouldn't work
21:01:34 *** j_n has quit IRC (Quit: User went offline on Discord a while ago)
21:04:48 <xarick> everything works! ship it!
21:14:05 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:19:11 <truebrain> just to be prepared
21:22:13 <truebrain> oof, supporting emscripten with libarchive will be a bit annoying
21:22:15 <truebrain> but also not a priority
21:25:33 <truebrain> _glx_: that sadly enough is the same "progress" as it was a few days ago
21:25:41 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:25:42 <truebrain> the page uses caching
21:25:48 <truebrain> so it doesn't show till you hard reload 😛
21:25:57 <truebrain> still, little progress 😉
21:28:29 <truebrain> lolz, why are the Windows runners installing libiconv .. which requires MSYS2 ..
21:33:58 <truebrain> lolz, via vcpkg it depends on liblzma
21:34:01 <truebrain> okay, that was pointless 😛
21:35:02 <_jgr_> Effectively blocking it entirely in vcpkg is a bit awkward, but I can see why they've done it. All the Linux distros seem to have just bumped it back to 5.4.x
21:35:17 <truebrain> they are taking a risks, ofc
21:35:21 <truebrain> as nobody knows how far back this really goes
21:36:06 <_jgr_> Presumably Lasse Collin does 😛
21:37:30 <dwfreed> but the reality is that I guarantee you that xz-utils is getting very through code audits done right now
21:37:34 <_jgr_> Given that he's tidying up the mess and has been reinstated on Github
21:37:58 <reldred> Sounds like it's just him
21:37:59 <dwfreed> so if there's anything else wrong with it, it'll come out within the next week or two
21:38:02 <truebrain> _jgr_: he too has to check each and every commit done by mister J
21:38:42 <truebrain> and the reaction of GitHub (and by extension vcpkg), feels more like an abundance of caution, which I can appreciate 🙂
21:39:25 <dwfreed> the reality is not everybody else can just ditch xz-utils
21:39:26 <truebrain> dwfreed: more the point that we currently don't know if this is really it. We all assume as much, but I can understand people needing their time to confirm that.
21:41:43 <peter1138> truebrain: Including commits attributed to other people.
21:42:01 <truebrain> I really hope he is getting help, for example, by people from GitHub, to help him out
21:42:58 <peter1138> Probably too much, and not knowing which to trust...
21:44:47 <truebrain> either way, my attempt to have an alternative totally failed; and 7zip LZMA SDK is just .. well. It is. Not touching that 😛 So back to the waiting game 😄
21:45:05 <_glx_> vcpkg reverted to 5.4.4, and now wait on github for access to the downloads
21:45:50 <_jgr_> I think it's reasonable to assume that the relevant security lists/groups will have been in contact with him
21:46:37 <_jgr_> Unlike before, everything in that project is going to have a whole array of microscopes focused on it for some time
21:47:39 <_glx_> one good thing from all this is openssh now natively supporting systemd mechanism without external patch nor link to systemd
21:47:44 <truebrain> in other news, the vcpkg crash the other day, turns out it is because they changed their build-farm to use `-static` .. and glibc and static builds are just a bad idea 😛
21:48:48 <peter1138> _glx_: So you don't need a patched version to be affected any more 🙂
21:50:03 <dwfreed> no, you just can't be affected
21:50:47 <_jgr_> The notification is a fairly trivial unix domain socket datagram, you don't need to link the whole systemd lib for that
21:51:17 <truebrain> seems the sarcasm sign was missed 😛
21:52:16 <truebrain> so even you learnt a new trick because of this incident!
21:53:04 <dwfreed> I knew it was pretty simple, just never bothered with it; the particular environment that repo is used in was getting annoyingly complicated and fragile without it, though
21:53:35 <dwfreed> so I just said to hell with it and added it
21:53:38 <truebrain> meh; still no incoming TCP support in Cloudflare Workers .. sad.
21:54:12 <peter1138> How's ipv6 support in various places now?
21:54:49 <truebrain> been using IPv6 for months now; has never been better 🙂
21:55:04 <peter1138> (in our infrastructure, I'm referring to)
21:55:19 <truebrain> since 2007 we support IPv6 for all our services 🙂
21:55:36 <peter1138> ISTR there was some proxying for something that doesn't do it.
21:55:49 <_glx_> and I was one of the rare with native IPv6 for a long time 🙂
21:55:53 <truebrain> github.com, discord.com, etc don't support IPv6; so we proxy those requests
21:56:47 <peter1138> Stupid GitHub, stupid Discord.
21:57:16 <_glx_> stupid country reserving so many IPv4 ranges
21:57:20 <truebrain> discord especially I don't really get, as they are behind Cloudflare .. which supports IPv6 😛
21:58:28 <truebrain> but so anyway, most of our infra is running on IPv6-only on AWS (those services that need state)
21:58:55 <truebrain> so for GitHub, Discord, etc, we call a proxy running on Cloudflare, which translates it to an IPv4 request
22:00:32 <truebrain> the only IPv4 we still run ourselves, are for our incoming TCP connections, like game-coordinator, bananas-server, etc. Because Cloudflare doesn't support TCP (only HTTP, which is also TCP, I know, so I mean TCP minus HTTP, ofc)
22:00:53 <_glx_> and with my ISP IPv4 are shared to 4 users by default unless you request a full one
22:00:55 <truebrain> flipped the world inside-out: IPv6 first, IPv4 second 😛
22:01:24 <peter1138> I have /48 of ipv6, so no worries about sharing 🙂
22:01:50 <peter1138> Mind I also have /29+1 of ipv4, because nerd.
22:02:20 <truebrain> owh, Sentry also runs IPv4 only
22:02:27 <truebrain> I knew I was missing one
22:06:22 <truebrain> peter1138: yes, /29 is greedy 😛
22:11:50 <peter1138> Well, we're getting some good extra fixes in before the release with this 'delay'
22:12:02 <_glx_> I only have /60, limited to /61 using the ISP provided box, and in actual usage it's a /64 because I only use 1 of the 8 prefix delegations
22:22:12 *** matusguy has quit IRC (Remote host closed the connection)
22:42:18 <peter1138> truebrain: Also makes it hard to switch ISP....
22:42:33 <truebrain> If you actually use them, yes 🙂
continue to next day ⏵
