IRC logs for #openttd on OFTC at 2012-01-06
⏴ go to previous day
00:00:08 <swissfan91> but jvassie, in the swiss set tracking table, found it to be 426
00:00:19 <swissfan91> and that is usually pretty accurate.
00:01:12 <MinchinWeb> well, it could be different sources or generations or even different gearings of the same engine... no way to know without looking at the sources
00:01:57 <swissfan91> i don't know how the site gave 529, but you worked it out to be 526..
00:03:33 <MinchinWeb> depends on how exact a convesion you use... I used 2 digit, the site uses 16 or more
00:05:23 <swissfan91> I think I may have gone over the top in this tracking table!
00:08:02 <swissfan91> 52 swiss narrow gauge EMUs, averaging 3 or 4 liveries each..
00:21:29 <Elukka> you can use google for unit conversions
00:27:17 <Elukka> i've made it do the most esoteric unit conversions with regards to space stuff :P
00:31:41 <swissfan91> I am all number crunched out I think.
00:40:13 <swissfan91> we should hopefully have some NG trains for your rails, Snail!
00:49:48 <Snail_> I'm planning to release the tracks publicly, and the trains only to playtesters (I'm sure my set is still full of bugs :p )
00:50:40 <swissfan91> well if you need a hand with any of it, let me know!
00:52:24 <Snail_> yep I'll create a new thread when it's ready
00:52:41 <Snail_> just a matter of days hopefully, I'm just ironing out the last known bugs and adding rackrail
00:54:01 <Elukka> heh, ran into a RhB version of that while researching standard gauge versions
00:54:14 <swissfan91> ah, nice. I have drawn up a sizeable tracking table full of Swiss NG EMUs. I may even give Yoshi a hand with the drawing if it takes long enough that I have finished all my work.
00:54:16 <michi_cc> swissfan91: One important thing to note is that electric engines often have a significantly higher short time power than what they can do continuously. It's often not clear which power is meant in descriptions.
00:54:29 <swissfan91> not that I want to ruin his sprites with mine.
00:55:13 <Snail_> michi_cc: I sent a first complete set of base sprites to make custom tunnels to planetmaker, to do some initial tests
00:55:45 <Snail_> I can send them to you too if you think you could help with that ;) I included the grass and the "bowl" with a new shading for the 4 directions
00:55:53 <Snail_> plus 3 custom portals of different sizes
00:55:56 <swissfan91> michi_cc: yeah, I have found quite a few different versions of power.
00:56:21 <michi_cc> Snail_: If you send me them I'll see what I can whip up.
00:56:53 <swissfan91> how on earth do I go about finding out the cost of trains for a set?
00:57:12 <Elukka> real world costs don't really work with regards to game balancing, so you have to come up with something
00:57:36 <Snail_> I included instructions as to how the sprites need to be "mounted" together (lego-like)
00:58:06 <Snail_> swissfan91: I guess costs need to be set up with trial and error until the game is balanced...
00:58:23 <Snail_> I myself am unsure about my trains' costs and the playtesting phase will be useful to set those up, too
00:58:27 <swissfan91> Elukka: What set are you collecting standard gauge trains for?
00:59:16 <swissfan91> oh right, does that have a tracking table anywhere?
00:59:20 <Elukka> real length wagons and more sprite angles \o/
00:59:26 <Elukka> hell of a lot more stuff to draw for it
01:01:59 <swissfan91> CETS is 2CC isn't it?
01:02:51 <Elukka> for now it's real colors
01:05:49 <swissfan91> I've just been on the thread.
01:06:08 <Elukka> i didn't even know there's a thread :D
01:06:44 <Elukka> that... isn't related but has the same name
01:07:22 <Elukka> that'll result in confusion if that's still active...
01:07:28 <swissfan91> that's a bit awkward..
01:07:51 <swissfan91> I did wonder why you had SBB trains in your tracking table, but it wasn't mentioned in that thread.
01:10:51 <Elukka> a CETS template for long coaches looks like that
01:11:23 <Elukka> dunno how they made it work in the game but it does, and they're not bendy
01:13:22 <Snail_> how do they work in curves just before (or after) slopes?
01:13:50 <Elukka> probably slightly glitchily
01:15:58 <Elukka> doesn't look too bad to me
01:16:29 <Elukka> nothing quite that long in game yet but the 4 axle coaches are reasonably long
01:16:37 <Snail_> how about drawing uphill sprites? so that the wagons can be seen actually climbing slopes
01:17:06 <Elukka> whether that's possible is beyond me
01:17:15 <Elukka> oh yeah, that grf needs some recent nightly of openttd
01:32:31 <michi_cc> Snail_: The custom portals that face to the top (bottom row) won't work that way. The track base needs to be separate from the portal otherwise either the train is hidden by the tracks or the portal by the train. I'll do a quick copy-paste just for testting though.
01:40:24 <Snail_> michi_cc: you mean the portals facing away from the player? like the NE and NW ones?
01:41:08 <Snail_> in that case, then we can just separate the track from the portals themselves... like, cut them in 2 pieces.. these graphics should still work, I guess?
01:43:35 <Qinc> hey, any chance I can get some quick help on a subject?
01:43:56 <Zuu> The chance gets higher if you actually ask your question.
01:44:07 <DorpsGek> Zuu: Don't ask to ask, just ask
01:44:21 <Qinc> Configging a dedi server for ttd,
01:44:32 <Qinc> server_admin_chat, an admin chat? how do I use it?
01:45:18 <Qinc> have looked on the wiki, nothing useful there
01:45:34 <Zuu> there is a document in the doc directory about multiplayer.
01:45:43 <Zuu> In your OpenTTD installatino directory.
01:47:05 <Qinc> Yeah, looked there aswell, nothing on that specifically, unless its only accessible from another application..
01:47:23 <Zuu> there is something called the admin port.
01:48:20 <Zuu> And there are 2-3 libraries that people have started as a base for writing applications that work with it. Yet I don't know if there really is any public available program.
01:49:01 <Zuu> There is also the autopilot which is older than the admin port. It hooks up to stdout/stdin of a OpenTTD dedicated server.
01:50:28 <Qinc> Yeah.. thanks.. was considering it
01:50:37 <Qinc> thankyou for your help ;)
02:25:34 <michi_cc> NE and NW need some work I guess :) Personally I'd remove a bit from the top grass sprite to get a bit more working room for the portal.
02:26:20 <Snail_> so, make the "bowl" a little bit shorter?
02:27:04 <michi_cc> Yes. Or maybe not directly shorter, but change the shape.
02:28:22 <Snail_> yeah... the shape is problematic, coz it needs to fit other styles too
02:28:59 <Snail_> although the bowl ca be at least partly overwritten by the portal
02:30:51 <michi_cc> Now you have some thinking to do and I some sleeping :)
02:32:07 <Snail_> yep, the portals themselves will need a lot of work. I was focusing on the landscape now
02:32:23 <Snail_> I'll try to play a little bit with the northbound views and split the portal sprites as wel
04:36:45 <MinchinWeb> with the Ship Pathfinder (in game), how close does a ship have to get to a buoy before it considers the ship for have reached it?
04:37:16 <Eddi|zuHause> the center of the boat must be on the buoy tile
04:37:48 <MinchinWeb> ... it seems to me when I watch them, they skip to the next destination before that
04:38:49 <Eddi|zuHause> [citation needed]
04:42:56 <MinchinWeb> in this case, it appears to be a Manhattan Distance of 3...
04:45:28 <Eddi|zuHause> well, make a bug report about it, if you think this is disturbing your operations
04:45:58 <Eddi|zuHause> but it could also be that the sprite is misaligned
04:46:23 <Eddi|zuHause> could press Ctrl+B to confirm (needs developer tools enabled)
04:51:10 <MinchinWeb> the ship is from FIRS
04:51:29 <MinchinWeb> I'm actaully writing an AI and trying to figure out how close is close enough with buoys
04:51:47 <MinchinWeb> (I don't exactally want to spam the map with them, if it can be helped)
05:21:52 *** fjb|tab is now known as Guest22813
05:21:53 *** fjb|tab has joined #openttd
05:34:06 *** fjb|tab is now known as Guest22815
05:34:07 *** fjb|tab has joined #openttd
05:56:23 *** Eddi|zuHause has joined #openttd
06:10:08 <MinchinWeb> is there a way to download just the NoAI documentation? (rather than the full source documentation)
06:28:54 *** JVassie has joined #openttd
06:33:10 *** MagisterQuis1 has joined #openttd
06:42:29 *** MagisterQuis has joined #openttd
07:08:36 *** DayDreamer has joined #openttd
07:11:24 *** DayDreamer has left #openttd
07:18:04 *** Twofish has joined #openttd
07:38:21 *** jacob88 has joined #openttd
07:53:45 *** |Jeroen| has joined #openttd
08:24:14 *** glevans2 has joined #openttd
08:28:38 *** Progman has joined #openttd
08:34:10 *** Brianetta has joined #openttd
09:17:23 *** fonsinchen has joined #openttd
09:38:11 *** Twofish has joined #openttd
09:49:05 *** kleinerdrache has joined #openttd
09:59:20 *** sla_ro|master has joined #openttd
10:03:34 *** Devroush has joined #openttd
10:04:41 *** mahmoud has joined #openttd
10:07:08 *** TGYoshi has joined #openttd
10:10:41 *** sla_ro|master has joined #openttd
10:37:43 *** sla_ro|master has joined #openttd
11:18:27 *** Adambean has joined #openttd
11:31:26 *** fonsinchen has joined #openttd
11:51:32 *** tokai|mdlx has joined #openttd
11:51:56 *** Biolunar has joined #openttd
12:11:00 <peter1138> what's some good change control (not version control) software?
12:12:49 *** welshdragon has joined #openttd
12:28:49 *** frosch123 has joined #openttd
12:31:10 <Wolf01> peter1138, do you mean a software like diff used between revisions?
12:40:32 *** fonsinchen has joined #openttd
12:42:42 <peter1138> no, general purpose change control
12:42:54 <peter1138> not version control o_O
12:43:31 <frosch123> ${EDITOR} changelog.txt
12:45:31 <planetmaker> peter1138: in order to create a meaningful changelog.txt (if that's what you mean), it usually is a good idea to resort to the commit messages of your VCS
12:45:31 <frosch123> yay, 0 bit graphics
12:45:41 <planetmaker> otherwise... diff it
12:45:59 <peter1138> i'm talking change control
12:46:08 <peter1138> not change logs, or version control, etc :p
12:46:35 <planetmaker> where differs "change control" from "version control"?
12:46:46 <Wolf01> then I can't understand what change control is :P
12:48:09 <Wolf01> I red it, but I didn't understand it :(
12:48:22 <planetmaker> :-) I should have wiki'ed it myself, I guess
12:48:50 <Wolf01> at least not the practical use
12:49:45 <Wolf01> it looks like flyspray to me
12:54:44 <Eddi|zuHause> why does xkcd remind me of Rubidium? :)
12:54:59 <Wolf01> omg, xkcd... I forgot it
13:09:54 <Rubidium> because it's a lovely red today?
13:11:10 *** Illegal_Alien has joined #openttd
13:13:21 *** sla_ro|vista has joined #openttd
13:22:08 *** FLHerne has joined #openttd
13:29:51 *** Prof_Frink has joined #openttd
13:44:31 <xQR> i have created a bug tracker entry about a problem that could be exploited to block the games on servers - is there a way to hide it so only devs see it?
13:44:56 <xQR> i don't think it would be good to give players the idea, though i don't think that very many players actually read the bug entries :P
13:45:07 <appe_> im just trying out replacing trains for the first time
13:45:19 <appe_> i cant seem to replace to the very same engine type
13:45:31 <appe_> even though the engine is available in the depot.
13:46:42 <xQR> the screenshot doesn't show any problem actually
13:46:55 <xQR> the "Start Replacing Vehicles" button is active and waiting for your click
13:46:58 <xQR> that's what the screenshot shows
13:47:30 <peter1138> appe_, that's called "autorenew"
13:48:52 <appe_> ah, i see. though, this is a setting, not a feature, afaik
13:49:02 <appe_> and im playing on a server with that setting - not enabled.
13:49:14 <xQR> no, it's a local client setting
13:49:18 <xQR> you can enable it in your own options
13:50:39 <appe_> though, i love that replacing feature
13:51:50 <FLHerne> The problem, though, is when you need to replace short wagons with longer ones, and your trains are already the longest practical...
13:52:27 <FLHerne> A 'replace wagons of type x with the same length of type y' would be nice
13:53:16 <appe_> it would be a 0.5 longer+
13:53:28 <xQR> yeah but that is covered by the wagon removal option
13:54:01 <appe_> btw, how much does passenger traffic to oil rigs stimulate the production? none?
13:54:39 <FLHerne> I don't think I noticed that button...
13:54:45 <appe_> that autorenew option didnt come cheap on a 200 train+ map..
13:55:45 <xQR> i don't think that moving passengers there has any direct effect onto the oil production
13:56:11 <xQR> well appe_ that's why it is an option i guess :P
13:56:20 <xQR> if you want to have the full control over it, you gotta do it manually ;)
13:58:36 <FLHerne> Just tried it - the Wagon Removal button does exactly what I wanted...how've I managed to miss it for 2 years?
14:01:16 <appe_> i didnt even know openttd had autorenew, nor autoreplace
14:05:27 *** TomyLobo has joined #openttd
14:06:57 *** sla_ro|master has joined #openttd
14:26:46 <appe_> i have this big network
14:27:02 <appe_> electrified railway, of what i wich to change to maglev.
14:27:20 <appe_> can i change the trains in the depots, even though the depots are of the electrified railway type?
14:27:51 <Eddi|zuHause> but you can use the universal railtype
14:49:45 *** sla_ro|master has joined #openttd
14:56:21 *** FLHerne has joined #openttd
14:57:38 <Eddi|zuHause> we _really_ need a better way of testing availability timelines...
14:58:06 <Remi_Woler> I vote for unit testing \o/
14:59:56 <planetmaker> what / how do you suggest to test or display that, Eddi|zuHause?
15:00:08 <planetmaker> and what do you want to see considered?
15:00:44 <planetmaker> also an action6 which changes availability depending on openttd rev?
15:00:56 <Eddi|zuHause> planetmaker: the easiest way i can imagine would be: i start a game in 2030 with vehicles never expire, and when clicking through the purchase list, it shows the earliest and latest availability for each vehicle
15:01:26 <Eddi|zuHause> (the randomized values)
15:03:01 <Eddi|zuHause> currently i'm flying mostly blind here, because it's too tedious to test whether vehicle X will be unavailable before or after Y is available
15:05:05 <Eddi|zuHause> scripts/processing.py:127:values["retire_early"]=values["vehicle_life"]-4 <-- i need to test whether this magic -4 is the right value
15:06:00 <Eddi|zuHause> and using the date cheat is not working
15:06:19 <Eddi|zuHause> it doesn't make vehicles unavailable properly
15:07:34 <Eddi|zuHause> unrelated: is there a RandomActionD?
15:08:11 <Eddi|zuHause> so we could have the default for which company is played set to "random", and it'll choose a different one per game
15:18:31 <Eddi|zuHause> so things like the random seed are not accessible?
15:23:44 <frosch123> hmm, after using svn for some time, it always confuses me that hg commit is instant :p
15:24:28 *** Fosland has joined #openttd
15:24:50 <Eddi|zuHause> anyone figured out a good pricing formula yet?
15:25:10 <Fosland> Can anyone give me a hand with downloading openttd for my mac? :)
15:25:55 <Eddi|zuHause> no, we cannot reach out for your mouse to click on the download link
15:26:58 <Fosland> Well, when i download the nightly, do i then need to place the gfx, sfx and msx in the file?
15:27:24 <Eddi|zuHause> have you checked the readme yet?
15:29:06 <frosch123> you only need to succeed in installing the gfx
15:29:14 <frosch123> you can get the rest easier in game
15:29:18 <Fosland> I cant find a readme inside the file
15:32:03 <frosch123> section 4.2 in the former, section 3.2 in the latter
15:33:35 <planetmaker> Eddi|zuHause: but one gets already for each vehicle the introduction date and the lifetime, right?
15:34:02 <Eddi|zuHause> planetmaker: i'm talking about the model life, not the vehicle life
15:35:00 <Fosland> When downloading the gfx from the site (openttd.org) Do i download all of em? Because i get a list with three options
15:35:14 <Fosland> Binary files, sources and another sources
15:35:50 <frosch123> Eddi|zuHause: actually i have a tool for that in progress
15:36:01 <frosch123> just was too lazy to create a project on devzone yet :p
15:36:23 <frosch123> Fosland: only the binary
15:37:22 <Fosland> Then i make a new folder inside the nightly build named "baseset" and extract it?
15:37:29 <Eddi|zuHause> i think the vehicle purchase and running costs are the only major thing left before we can have some "real world" playtesting
15:38:10 <Eddi|zuHause> there's a bit of minor stuff like reliability
15:38:25 <Eddi|zuHause> but i have no idea how to reasonably test this
15:38:30 <Eddi|zuHause> i never play with breakdowns
15:39:29 <planetmaker> let's try the banana approach with reliability: choose some values which seem about right and let users complain
15:43:11 <Eddi|zuHause> there are a few fixed hardcoded values left that need to be differentiated for the vehicles. air drag, loading speed, cargo decay
15:45:24 *** fonsinchen has joined #openttd
15:48:47 <Eddi|zuHause> and the length refit for the DMU/EMU vehicles is not implemented yet
15:50:40 <Fosland> ive managed to install and get openttd to work, but i cant join any server? I even have the nightly build?
15:51:09 <glx> server need to have the exact same version
15:51:18 <planetmaker> Fosland: you need the exact same version. And I'm slow
15:51:34 <Fosland> And how can i get the exact same version?
15:51:48 <planetmaker> Same as you got the current one
15:52:13 <planetmaker> most servers run 1.2.0-beta1 or 1.1.4 currently
15:52:33 <planetmaker> i.e. the testing or the stable release version
15:52:46 <Fosland> Thanks, i just need to download the stable
15:53:18 <Fosland> When downloaded do i need to add grx to that file?
15:53:55 <planetmaker> depends where you placed it
15:54:12 <planetmaker> if you placed it in ~/Documents/OpenTTD/data you need do nothing else
15:54:29 <planetmaker> I'd recommend to use the beta1 though ;-)
15:54:48 <Fosland> But i wanna play on Luukland, and they run the 1.1.4
15:54:59 <Fosland> What cool features are in the beta?
15:55:16 <planetmaker> many ;-) For example all the features which obsolete Luukland ;-)
15:55:38 <planetmaker> like: goal scripts
15:56:03 <Fosland> I can have to games? One for Luukland and one for beta?
15:56:04 <planetmaker> etc pp. See the changelog
15:56:22 <planetmaker> of course. You can have as many openttds as you want (and have space for)
15:56:49 <planetmaker> just make sure to place your graphics files not local to the individual dir. But into ~/Documents/OpenTTD
15:57:00 <Fosland> Thanks for the help guys :)
15:59:54 *** sla_ro|master has joined #openttd
16:01:45 <Fosland> planetmaker, i have downloaded the 1.1.4 and put grx into the baseset folder, why is it still not working :/
16:02:08 <planetmaker> 1.1.4 requires it to be in the data folder.
16:02:21 <planetmaker> I'm afraid that's one of the changes between 1.1.x and 1.2.x
16:03:10 <Fosland> There we go, perfect :)
16:14:41 <Zuu> <planetmaker> let's try the banana approach with reliability: choose some values which seem about right and let users complain <--- Sure, and as noone has replyed to my last relaese of "Neighbours are important" they all think it is very good :-)
16:15:23 <planetmaker> Zuu: a reasonable assumption, though ;-)
16:15:38 <Zuu> However, I was happy to get a possitive feedback for PAXLink recently.
16:15:59 <Zuu> Made me think about possible integrate feeders into CluelessPlus.
16:18:15 <Remi_Woler> Zuu: I think it is almost good. Seriously. Played it twice now, and I think the balance is off for the worse. It quickly becomes near impossible, if not impossible, to satisfy the goals
16:19:01 <Remi_Woler> I haven't analyzed the algorythm enough to suggest something better, hence why I didn't speak up yet
16:19:26 <Zuu> Did you use 1.0 as scale factor or did you scale down the goals?
16:20:17 <Remi_Woler> I used it as-is, from Bananas
16:20:30 <Zuu> There is a scale parameter in the game script settings now.
16:21:05 <Zuu> You can also turn off the thing that makes towns with small neighbours hard to grow.
16:22:07 <Remi_Woler> I don't think my version has settings. So I may have a too old version then. Either way, it was the problem that up till 600~800 was doable, but then the goals suddenly becoming insance, like being multiplied by 70, instead of a linear-y increase
16:24:03 <Zuu> It could be the neighbour thing that kicks in. A factor is computed that compares the size of the town itself and the 2-3 or so closest neighbours. For towns with small neighbours this factors become > 1. For relativly small towns the factor is < 1.
16:24:21 <Zuu> As you grow your town larger this factor becomes larger if you didn't do anything to its neighbours.
16:24:46 <Zuu> So not only does the goals grow because of your town growing but also because it become larger relative to its neighbours.
16:24:49 <planetmaker> Well, might make sense for CluelessPlus :-)
16:26:19 <Remi_Woler> Zuu: that explains it. It quickly became an issue that there were not enough passengers in the map being "produced" to satisfy any of the goals, after I cultivated all towns to 600~800. (512x512 map, with "very low" towns)
16:26:20 <Zuu> Sometimes when it upgrades airports, the new one get worse coverage due to noise restrictions or larger airport size, and in that case a feeder service would be nice.
16:27:08 <Zuu> There is also a distance limit on what is considered a neighbour. Possible som 80-100 tiles or so.
16:27:24 <Remi_Woler> I think I saw 100 in the code
16:27:37 <Zuu> But again only the 2-3 closest or so counts (could be 5 too, I don't really remember)
16:28:29 <Remi_Woler> I believe 3, but it's been a week since I looked at the source. Haven't played it since, since my human opponent thought they were too hard, and instead started practicing RoRo's and not care about any of the goals at all
16:30:12 <Zuu> If you want to use the new one, download it from bananas and create a new game. (GameScripts can't be upgraded in a running game). The new one allows scaling the goals from 1 to 1000%. That setting can be changed in game (at least for single-player and non-dedicated servers) so you can fine tune it in the game if it becomes too hard.
16:31:26 <Remi_Woler> we're playing the TestTest one now, which is really easy, so she can become used to the goals, and probably then use the new Neigbours-Needs script. I was planning on adjusting the algorithm in the existing script, but your solution is much easier! :D
16:31:34 <Zuu> But I'm also open to make changes to the default values so that there are sane defaults too.
16:32:26 <Eddi|zuHause> i should probably finalize my delanuay-library
16:32:37 <Remi_Woler> I won't be able to suggest anything till I found something that works better. That's why I hadn't spoken up yet
16:32:37 <Eddi|zuHause> could be used to find the appropriate neighbours
16:33:02 <Eddi|zuHause> i was also thinking about splitting the map into regions with this
16:33:04 <Zuu> The new one also have a new (optional) feature of reducing/stopping town growth if towns get too congested. This limit can also be scaled in-game through a setting or completely be turned off.
16:33:39 <Zuu> Remi_Woler: Fine, I'm looking forward to what you come up with. But no hurry.
16:33:44 <Remi_Woler> largest I've seen so far is 70k (non-gamescript), so that hasn't become an option yet
16:34:30 <Eddi|zuHause> roughly: pick the biggest N towns, make them capitals. calculate the delaunay triangulation/voronoi partition to determine which towns get which capital
16:34:36 <Remi_Woler> Well, it mostly depends on when she has enough time to play. I can play on my own, but that won't reflect real life. Especially since there's no AI yet that understands goals (for as far as I am aware)
16:34:51 <Eddi|zuHause> possibly stack this multiple times, so you get a hierarchy
16:35:27 *** sla_ro|vista has joined #openttd
16:35:51 <Remi_Woler> Eddi|zuHause: if you can, also look at world features. Like if a town is on a real island, it should not be in the same region as one on the mainland. Probably bigger rivers as dividers too.
16:37:50 <Remi_Woler> Eddi|zuHause: and I'd probably call them counties :)
16:38:41 <Eddi|zuHause> i don't think "map features" are going to be sensibly included
16:39:03 <Remi_Woler> would probably too hard, either programming or cost wise
16:39:08 <Zuu> Remi_Woler: If you want to make it easier, especially conceptually, I think it is probably a good idea to turn off the neigbours feature. (there is a setting for that too) I've played with the script at this configuration (at the time when TrueBrain wrote it), and it provide some challenge but is not too difficult.
16:39:50 <Remi_Woler> Zuu: TestTest == yours-without-neighbours, correct?
16:40:16 <Zuu> Could be that yes. Do you have a link?
16:40:36 <Zuu> f it is the one that TrueBrain wrote originally, then yes.
16:41:41 <Remi_Woler> I guess the URL gives it away :$
16:42:33 <Eddi|zuHause> Remi_Woler: conceptually you'd give the world a non-linear distance function, which makes rivers and water "wider" than land tiles
16:42:56 <Eddi|zuHause> but that makes all the currently used algebra "wrong"...
16:44:01 <Remi_Woler> Eddi|zuHause: why? I was thinking just finding out what tiles are in between them, and if no route can be found over land, they would not qualify as neighbours? Probably too costly, especially in a gamescript, but that was how I was thinking
16:44:18 <Remi_Woler> Pretty bruteforcing though, now that I write it out, so probably not the best option
16:44:28 <Zuu> Sounds like a such graph that you can make in a GIS or macro model software over how far you get in X minutes. :-)
16:45:16 <Eddi|zuHause> Remi_Woler: that's nothing that can be reasonably integrated into an algebraic solution like delanuay
16:45:35 <Remi_Woler> delanuay doesn't ring a bell
16:47:07 <Remi_Woler> but in all honesty: I learned programming first, and algebra much much later. So I have quite a hard time expressing something in algebra, even though I could code it in C without a sweat. Okay, maybe a couple curses to whoever thought up pointers
16:47:39 <Remi_Woler> delaunay doesn't ring a bell either
16:48:26 <Zuu> I too learned to program before (advanced) math.
16:48:37 <Zuu> The best way to make a math person mad is to reuse variables :-)
16:49:14 <Remi_Woler> and the best way to make a developer mad: use another character than i for the iterator in a for-loop? :D
16:49:39 <Zuu> or leaving out identation..
16:51:56 <Zuu> Got to bike. See you later.
16:57:05 *** TheMask96 has joined #openttd
17:04:37 <planetmaker> I'd not show the random stuff but the raw dates w/o randomization
17:04:52 <planetmaker> it would show those dates when newgrf developer tools are active
17:05:44 <Eddi|zuHause> but the randomization is the interesting part... for various randomizations, do vehicle X and Y overlap?
17:06:54 <Eddi|zuHause> and your "refrigerator wagon" is buggy :)
17:07:27 *** fjb|tab is now known as Guest22852
17:07:28 *** fjb|tab has joined #openttd
17:19:37 <Eddi|zuHause> running cost: (standing | running) * vehicle age * inflation balance?
17:21:17 <Eddi|zuHause> well "*" not exactly
17:21:27 <Eddi|zuHause> but older vehicles should cost more
17:21:35 <frosch123> you should add the breakdown chance to the running cost, if breakdowns are disabled :p
17:22:03 *** sla_ro|master has joined #openttd
17:22:34 <frosch123> depot visits would then reduce the running cost
17:22:39 <planetmaker> trunk((vehicle age in years) / 10) or so. But yes
17:22:50 <Eddi|zuHause> inflation balance would be to have the ratio purchase/running cost roughly the same if you played for 100 years or started 100 years later
17:23:02 <planetmaker> I like that idea really, frosch123
17:23:16 <Eddi|zuHause> frosch123: that sounds evil
17:24:02 <planetmaker> * (5 - breakdown_chance / 20) or so
17:24:20 <planetmaker> with breakdown_chance = 0...100
17:24:32 <Eddi|zuHause> is there a var for current reliability?
17:24:33 * Remi_Woler spots a division by zero
17:24:56 <Eddi|zuHause> Remi_Woler: for particularly small values of 20?
17:25:02 <planetmaker> Remi_Woler: mind order of operation
17:25:41 <Remi_Woler> Eddi|zuHause: for the smallest value of breakdown_chance. The way I learned math, division takes priority over subtraction
17:26:01 <Eddi|zuHause> Remi_Woler: is your monitor upside down?
17:26:03 <planetmaker> and? it's the nominator
17:27:08 <planetmaker> anyway... (1 + age / 10) * (6 - breakdown / 20) * (running | stopped) * factor
17:27:14 <Remi_Woler> breakdown_chance can be 0. You divide breakdown_chance by 20, then subtract that outcome from 5, then multiply that outcome with whatever is before the line
17:27:34 <planetmaker> Remi_Woler: and now write it down and see where you divide by 0
17:28:20 <Remi_Woler> Okay, maybe not the most correct term, but still the same issue
17:28:35 <Eddi|zuHause> so something like taking var 4E as a linear factor
17:28:36 <Yexo> Remi_Woler: 0/x is valid, x/0 is not
17:29:15 <Remi_Woler> but does it do what you want?
17:29:21 <Eddi|zuHause> so if reliability is 50% you pay twice, 25% you pay quadruple
17:30:17 <Eddi|zuHause> we only have a range 0..255
17:30:39 <planetmaker> given the rounding of int, I guess it'd be 4x and 5x with what I gave. But...
17:30:39 <Yexo> divide by 51 instead of 20
17:31:20 <Eddi|zuHause> var 4E is 0..FFFF (0%..100%)
17:33:23 <Eddi|zuHause> frosch123: but that "breaks" the idea of having older-looking vehicles drive around, which you didn't send to depot for a while ;)
17:33:45 <Eddi|zuHause> var 80+4E, of course
17:33:53 <frosch123> vehicles also get older inside depots
17:34:12 <frosch123> oh, i forgot the 80 as well :p
17:34:17 <Eddi|zuHause> frosch123: but the graphics are decided for last depot visit
17:34:47 <frosch123> well, visualising the running cost via the vehicle sprite is not too bad, is it?
17:35:04 <Eddi|zuHause> no, i mean livery repainting
17:35:11 <Eddi|zuHause> not making the vehicle look old
17:35:32 <frosch123> i don't get your point
17:35:35 <planetmaker> is the livery updated by depot visits?
17:35:43 <frosch123> do you want to never send vehicle to depot?
17:36:04 <frosch123> then i think that is a bad idea :p
17:36:07 <Eddi|zuHause> frosch123: that's how i play usually :)
17:36:30 <frosch123> all my vehicles regulary visit depots
17:36:48 <frosch123> not only because i play with breakdowns, but also because i play with overflow depots
17:36:48 *** sla_ro|master has joined #openttd
17:36:50 <Eddi|zuHause> frosch123: to sensibly do that, i need engine rotation :)
17:37:13 <Eddi|zuHause> so the train moves with another engine, while the first engine goes to depot
17:37:42 *** Alberth has joined #openttd
17:37:42 *** ChanServ sets mode: +o Alberth
17:37:43 <Eddi|zuHause> so i have N trains and N+1 engines, and always one engine goes to depot
17:38:19 <Alberth> it must feel very lonely, all alone in the depot :(
17:38:55 <Eddi|zuHause> it feels even more happy when it comes out again :)
17:40:40 <Eddi|zuHause> but i wanted to introduce a parameter anyway: update livery on: {current year (immediate), last service date, build date}
17:42:47 *** Chris_Booth has joined #openttd
17:52:13 <planetmaker> Eddi|zuHause: build_date means 'never change livery'?
17:52:20 <planetmaker> of existing vehicles?
17:57:45 *** welshdragon has joined #openttd
18:07:42 <Eddi|zuHause> may make some later liveries unavailable
18:10:17 <planetmaker> why would that make them unavailable?
18:10:38 <planetmaker> ah, you mean when existing vehicles are re-painted past the model's availability?
18:11:20 *** fonsinchen has joined #openttd
18:19:38 *** epicNick has joined #openttd
18:20:04 <epicNick> hello can sombay helpme what is the player limit for open transport tycoon?
18:20:24 <Eddi|zuHause> 255 players in 16 companies
18:21:17 *** FLHerne has joined #openttd
18:22:47 <Terkhen> was that the record of "leaving after question is answered"? :P
18:23:27 <welshdragon> why is it I can't select 'large' towns when I click 'Found Town' button?
18:24:33 <planetmaker> you're not in the scenario editor probably
18:25:19 * planetmaker knows no game "open transport tycoon". I know OpenTTD, though
18:26:26 <Eddi|zuHause> we still need "dual headed" flag for articulated vehicles :)
18:33:42 <Yexo> TrueBrain: there is an issue with mail forwarded by openttdcoop.org to openttd.org addresses
18:34:34 <Yexo> one example: approximately an hour ago I send an email to nml-team@openttdcoop.org, which is forwarded to (amongst others) my openttd.org address
18:41:56 <Ammler> what michi_cc said, the question is will/can/should we implement SRS on our side :-)
18:43:11 *** Markavian has joined #openttd
18:45:22 *** TGYoshi has joined #openttd
18:46:44 <CIA-6> OpenTTD: translators * r23761 /trunk/src/lang/ (16 files): (log message trimmed)
18:46:44 <CIA-6> OpenTTD: -Update from WebTranslator v3.0:
18:46:44 <CIA-6> OpenTTD: belarusian - 1 changes by Wowanxm
18:46:44 <CIA-6> OpenTTD: bulgarian - 50 changes by kokobongo
18:46:44 <CIA-6> OpenTTD: english_AU - 6 changes by tomas4g
18:46:46 <CIA-6> OpenTTD: english_US - 1 changes by Rubidium
18:46:46 <CIA-6> OpenTTD: finnish - 1 changes by jpx_
18:48:27 <peter1138> people still using SPF?
18:49:08 <^Spike^> or atleast ottds MX seems so :)
18:53:54 <TrueBrain> Yexo: like said: SPF :)
18:54:36 <Yexo> TrueBrain: it does raise the question: how many (valid) messages are blocked by that filter?
18:55:05 <TrueBrain> just remember that any mail you send from @openttd.org has to be send FROM secure.openttd.org
18:55:56 <Yexo> I meant it more like: how many addresses are there with wrongly configured servers (like openttdcoop apparently)?
18:56:17 <^Spike^> wrongly configured server.....
18:56:22 <^Spike^> ty for that compliment of hard work :)
18:56:38 <Yexo> ^Spike^: sorry it sounded like that
18:56:50 <^Spike^> i never take it personally :)
18:57:00 <Yexo> good, it wasn't meant as such
18:57:09 <^Spike^> i'll just cry myself to sleep tonight ;)
18:57:10 <TrueBrain> Yexo: the DSN btw got rejected by openttdcoop :)
18:57:27 <TrueBrain> so openttdcoop doesn't like you relaying messages like that :)
18:57:39 <TrueBrain> it should, by RFC, rewrite the address of course
18:57:52 <TrueBrain> but that is something that went out of the room as soon as the RFC was implemented, years ago :)
18:58:02 <^Spike^> who follows a damn rfc? :)
18:58:13 <^Spike^> tinymailserver from what was that guys name again
18:58:15 <TrueBrain> Yexo: I guess you set up an alias on openttdcoop, which in fact is more a maillist?
18:58:31 <planetmaker> that's about it, yes
18:58:32 <Yexo> TrueBrain: not me, but it works like that yes
18:58:38 <TrueBrain> then it should fail yes :)
18:58:47 <TrueBrain> I understand it mostly is the easiest solution
18:58:58 <TrueBrain> but with all the SPAM these days, it is an invalid solution (by RFC, again)
18:59:06 <TrueBrain> in that respect, we live in a said world :(
18:59:19 <^Spike^> gresylisting does seem to work here :)
18:59:22 <Yexo> I understand why it fails and that is should fail, however since I don't see anything about the failure on yexo@openttd.org, I'm wondering if I've missed more messages this way
18:59:37 <TrueBrain> Yexo: like said, the DSN got rejected by openttdcoop :)
18:59:47 <TrueBrain> as there openttdcoop does follow RFC :)
18:59:55 <TrueBrain> well, not really, but partial :)
18:59:57 <Yexo> you've got to spell that out for me
19:00:09 <TrueBrain> okay: you send mail from openttd.org to openttdcoop
19:00:14 <TrueBrain> openttdcoop relays it to openttd.org
19:00:16 <Yexo> I thought openttd.org refused the messages because of the wrong "From:" header?
19:00:22 <TrueBrain> openttd.org mail filter sees @openttd.org, but from wrong IP
19:00:36 <TrueBrain> it sends a DSN to the sending server, to let the user know that it failed
19:00:43 <TrueBrain> this travels back to openttdcoop
19:00:49 <TrueBrain> which rejects the DSN because it is to @openttd.org
19:00:53 <TrueBrain> which is not in his allowed relay list
19:01:14 <TrueBrain> so it is a double issue, and both are the reason not to use aliases for maillists :)
19:01:42 <TrueBrain> of course the mailfilter could deliver the DSN to the real MX, but by RFC you are obligated to do to the MX that tried to give you the email :)
19:01:48 <TrueBrain> (again, to avoid spam :))
19:02:03 <TrueBrain> we all received "failure to deliver" emails from places we didnt know we tried to send to :P
19:02:21 <Eddi|zuHause> not that i know of :p
19:02:29 <TrueBrain> sadly, there still isn't a light-weighted maillist software
19:02:50 <TrueBrain> easiest solution is sending from @openttdcoop account
19:03:50 <TrueBrain> and if you missed more email like this; you shouldn't ,only in these scenarios where you bounce an email from @openttd.org over another server, which is not configured to relay emails in a RFC fashion
19:04:00 <Ammler> TrueBrain: does openttd mailserver do SRS?
19:04:21 <Yexo> TrueBrain: I understand now why the bounce message also didn't come back
19:04:52 <TrueBrain> Ammler: how is that relevant?
19:04:55 <Ammler> I don't think it is against RFC, else postfix would support it
19:04:57 <TrueBrain> the issue is more that openttdcoop doesn't
19:05:21 *** HerzogDeXtEr has joined #openttd
19:07:36 <TrueBrain> Yexo: one of the downsides (next to the many upsides) of using a SPF with -all :)
19:08:14 *** JVassie has joined #openttd
19:08:16 <TrueBrain> I also dont really have a solution for you :(
19:08:38 <^Spike^> why you using mailfiltering.nl in the first place? other then spam protection... :)
19:08:46 <TrueBrain> it _is_ spam protection
19:08:59 <Ammler> well, also if we would change our mailserver, you still have no clue how many other mails you reject because of that
19:09:30 <TrueBrain> Ammler: because in general I don't care about MX that do not follow RFC
19:09:34 <TrueBrain> as 90% of those hosts are spammers
19:09:47 <Ammler> TrueBrain: please show me the RFC about that
19:09:52 <Yexo> Ammler: but in general if a message would be rejectd the sender would get a bouncer mail
19:10:08 <TrueBrain> Ammler: by RFC a MX is obligated to accept a DSN from the MX he tries to send an email to
19:10:26 <TrueBrain> if the sending MX has no route back for an address he is sending from, he has no business doing that
19:10:51 <Ammler> moment, do we speak about the same issue?
19:11:09 <Ammler> what does DSN have to do with that?
19:11:35 <TrueBrain> Delivery Status Notifications
19:11:54 <Ammler> well, that we could fix
19:12:31 <TrueBrain> in this setup, I doubt you want to fix it
19:13:06 <Ammler> I thought the issue is related that we don't rewrite the sender
19:13:26 <TrueBrain> the problem is you are sending an email from an adress you dn't have a return path for
19:13:36 <TrueBrain> (which is because you use aliases to use as maillist)
19:13:45 <TrueBrain> which can be solved by sender rewriting, yes
19:13:45 <Yexo> Ammler: the mail is rejected because you don't rewrite the sender. openttd.org sends DSN which you can't send back to the sender
19:14:20 <Ammler> TrueBrain: ok, so you just used 1000 other words ot desribe the issue we already were aware of :-P
19:14:36 <TrueBrain> I have no clue what you are aware of or not
19:14:45 <TrueBrain> I just try to keep describing it in different ways till you catch on somewhere
19:15:26 <Ammler> TrueBrain: still, where is that RFC?
19:15:34 <TrueBrain> ugh, you want all of them?
19:15:50 <TrueBrain> guess 1891 is most relevant to this case
19:15:51 <Ammler> no, just the one you talk here
19:16:03 <Ammler> I wonder, how ISP works then
19:16:33 *** TGYoshi has joined #openttd
19:16:43 <Ammler> I think, the issue here is defined in your spf
19:16:53 <TrueBrain> ISPs, and gmail, dont blindly relay emails
19:17:02 <TrueBrain> they only send emails from addresses they have a return path for
19:18:54 *** andythenorth has joined #openttd
19:19:23 <Yexo> Ammler: I can send mail with gmail from my @openttd.org address, but it sends mail via openttd.org
19:19:43 <Yexo> gmail connects to openttd.org, which resends the message to the actual target
19:20:27 <Ammler> Yexo: yes, but that it has to be from openttd.org is not a RFC rule, that is defined in the spf
19:21:27 <SpComb> I know plenty of ISP smtp servers that relay mail from addresses they don't accept delivery/forwarding for
19:22:15 <SpComb> and DSNs typically get sent back to the given source address, not returned to the sending mail server, which is what lead to the backscatter problem?
19:22:39 <Ammler> yes, which you would break with SRS
19:25:28 <TrueBrain> SpComb: many of such MXes find themself on blacklists these days :)
19:25:39 <TrueBrain> for a few months it was the easiest way to send spam
19:25:42 <Ammler> bascially you should not alias to addresses you don't have
19:25:45 <TrueBrain> just say you are DSN, and all doors were open :)
19:26:39 <TrueBrain> when writing the SMTP stuff, people had so much faith in the human race: "nobody would ever say they are DSN when they are not, would they? That is only for failure delivers!"
19:26:51 <SpComb> well, that's how DSNs work
19:27:00 <TrueBrain> no, DSN should travel back in the path
19:27:07 <TrueBrain> not blindly transmit to what ever was MAIL FROM
19:27:38 <SpComb> besides, if openttdcoop tries to relay a mail from @openttd.org to @openttd.org, which fails SPF, then the recieving MX should reject it outright, not accept it and generate a DSN..
19:28:03 <SpComb> then the openttdcoop MX could generate the DSN and deliver it... although delivery of that DSN would probably also fail..
19:28:19 <SpComb> although, I guess DSN doesn't keep the origional from addr..
19:28:27 <^Spike^> SpComb it would not probably fail... it does fail
19:29:46 <Ammler> that little bit I read about, you would need content filter, which is silly :-)
19:30:04 <TrueBrain> SpComb: that is a valid question. openttdcoop should generate the DSN. I don't know why this got delegated to mailfiltering in this case, but it is generated from the session from openttdcoop; I can only guess it is because postfix decided it is the legal way .. never read into that :)
19:30:12 <SpComb> it sounds sensible, but I've never heard of an MX that routed generated DSNs back to the MX that connected in
19:30:25 <TrueBrain> SpComb: it happens more often than you think :)
19:30:33 <SpComb> rather than the envelop from addr
19:30:50 <Ammler> that is like your postman opens your mail and changes the sender to his address
19:31:11 * SpComb is responsible for public smtp servers as well
19:31:17 <TrueBrain> Ammler: nothing of this has anything to do with what is inside the email :)
19:32:33 <Ammler> ok, the postman, could put the mail in a new covert and send it that way right
19:33:10 <Ammler> hmm, why doesn't postfix do that?
19:33:28 <TrueBrain> SpComb: in this case openttdcoop explicitly asked to generate a DSN on failure :)
19:34:37 <TrueBrain> and specially for Ammler: that is defined in RFC 3461
19:35:37 <TrueBrain> so in ESMTP, it is up to the MX to deligate DSN generation
19:36:26 <SpComb> buh, too much to think about on a friday evening
19:36:31 <SpComb> someone should just fix email :)
19:36:38 <TrueBrain> SpComb: you started this part yourself :P
19:36:48 <Ammler> TrueBrain: you have any stats, how many mails you reject with that issue?
19:37:29 <TrueBrain> Ammler: which issue are you talking about?
19:37:37 <TrueBrain> failure in SPF, or failure in accepting DSN?
19:37:44 <TrueBrain> (when explictly requested :D)
19:38:00 <TrueBrain> both are different issues
19:38:09 <Ammler> well, our issue is the spf
19:38:13 <TrueBrain> SPF failure rate is unrelated to unable to route back DSN :)
19:38:24 <TrueBrain> "your" issue is more than one issue, which are all a result of eachother
19:41:15 <TrueBrain> random stats: every 1000 emails that arrive for openttd.org (assuming it gets that far in MX communucation
19:41:33 <TrueBrain> 350 of them due to SPF failure
19:41:59 <TrueBrain> mostly due to people who try to email from shit like website@ openttd.org
19:42:28 <TrueBrain> or "username"@ that tries to email to "username"@ ..
19:42:43 <CIA-6> OpenTTD: alberth * r23762 /trunk/src/lang/english.txt: -Fix [FS#4954]: Removed double words in three strings, and one change 'cost' to 'income' in english.txt (thanks Tranzistors)
19:42:51 <TrueBrain> no clue if that were stats you were looking for Ammler, but your request was a bit vague to me :(
19:45:21 <Yexo> that were the stats I was looking for earlier :)
19:45:38 <TrueBrain> sadly, they tell you nothing in relation to your problem :(
19:46:33 * Rubidium never had any problems with sending mail to the openttd mail server
19:46:37 <TrueBrain> SpComb: owh, I stand correct: if no NOTIFY was send, by ESMTP, you MUST generate a DSN :(
19:47:00 <TrueBrain> Rubidium: there is nothing wrong with openttd mail server :) Neither with openttdcoop for that matter.
19:47:27 <Ammler> TrueBrain: that's ok, thanks :-)
19:48:12 <Ammler> a generic rule might be that you should not setup alias to 3rdparty domains
19:48:51 <TrueBrain> someone should reinvent SMTP :)
19:49:03 <TrueBrain> and not write it from the naive point of view that nobody wants to do harm :(
19:49:50 <CIA-6> OpenTTD: rubidium * r23763 /trunk/src/object_cmd.cpp: -Fix [FS#4956]: check whether a water tile is really empty when overbuilding it with an object
19:50:27 *** andythenorth has joined #openttd
19:50:40 <Alberth> TrueBrain: wouldn't that simply end in not accepting any mail at all?
19:50:44 <TrueBrain> reading through the RFC again (just because I am bored I guess):
19:50:49 <TrueBrain> is forwarded it will have a new envelope return address. Any DSNs
19:50:51 <TrueBrain> which result from delivery failure of the forwarded message will not
19:50:52 <TrueBrain> be returned to the original sender of the message and thus not expose
19:50:54 <TrueBrain> the recipient's forwarding address.
19:51:02 <TrueBrain> one of the funny things that are often forgotten :P
19:51:11 <TrueBrain> Alberth: hehe; I guess :)
19:51:18 <TrueBrain> Alberth: forcing SPF for every domain would help a lot tbh :P
19:52:44 *** Chris_Booth has joined #openttd
19:55:46 <Ammler> yes, the issue is that spf is not well accepted, postfix without native srs support proves that :-)
19:56:12 <TrueBrain> SPF is well accepted, relaying with SPF isn't ;)
19:56:19 <TrueBrain> has been an issue since the day SPF was invented, sadly
19:56:42 <TrueBrain> reason most people still use ~all I guess :)
19:57:07 <Ammler> TrueBrain: you could set it to that and watch for a month or so
19:57:22 <Ammler> and if the spam really rises, you can still revert it
19:57:23 <TrueBrain> Ammler: been there, done that ;)
19:57:34 <TrueBrain> -all was not my first choice :)
19:58:14 <Ammler> maybe as you had that, you didn't have greylist
19:58:40 <TrueBrain> and sadly greylisting is less and less efficient
19:59:52 <^Spike^> different experience that i have with it... and use it even on a hosting server...
20:00:12 <TrueBrain> well, 2 years ago, greylisting was very efficient; it catched 80% or so
20:00:18 <TrueBrain> nowedays that number is more like 40%
20:00:40 <^Spike^> here it still seems to catch alot
20:00:49 <TrueBrain> it is based on the idea that spammers are using simple perl scripts to send emails, and don't listen to the reply of the MX. Just send as much as you can as often as you can, and see where it gets you
20:00:59 <TrueBrain> we already have a few scripts that in fact listen to the reply
20:01:07 <TrueBrain> and just requeue an email 5 minutes later
20:01:16 <TrueBrain> very very nasty scripts :(
20:01:37 * ^Spike^ is checking exim logs somewhere else...
20:01:50 <^Spike^> and still not seeing a bot resend
20:02:13 <michi_cc> Yeah, greylisting doesn't help at all if the spam comes from a totally genuine mail server (hacked most likely) complete with TLS encrypted connections and all :(
20:02:16 <TrueBrain> ^Spike^: to give you an idea of the volume I talk about, we filter email for thousands of domains; millions of emails flow through these filter systems a day
20:02:28 <TrueBrain> michi_cc: I havent seen TLS yet :P
20:02:58 <TrueBrain> but I have seen the scripts go from a 10 liner to just binary blobs :(
20:03:20 <TrueBrain> I guess the main issue is with the readers. For some reason peopple STILL clikc on these spam messages
20:03:54 <michi_cc> The only thing left to do at that point is content filtering, but that's always a big problem with false positives.
20:04:05 <TrueBrain> we stopped doing content filtering years ago
20:04:24 <TrueBrain> so many problems that come from it :(
20:04:35 <TrueBrain> a pharmacy company was most enjoyable
20:05:12 <michi_cc> Sure, but what else to do against spam from valid addresses with valid SPF, valid DomainKeys signature and basically valid everything?
20:05:25 <TrueBrain> I hope SMTP is reinvented by then :P
20:05:41 <TrueBrain> for now they say valid SPF is easy: they either shut down the domain, or blacklist it
20:05:47 <TrueBrain> easier to do than IPs
20:06:28 <TrueBrain> I guess if that happens, the rules of registering domains get much stricter, prices go up up up
20:06:30 <michi_cc> The problem isn't SMTP, but simply that there are ISP that professionalize in providing these services to spammers.
20:06:43 <TrueBrain> and people clicking on them :D
20:06:44 <SpComb> the wikipedia SRS article has nice background, but... doesn't actually say what SRS is
20:07:00 <TrueBrain> SpComb: SRS is a heavily debated 'solution' for SPF to allow forwarding
20:07:07 <michi_cc> SpComb: Read the german one :)
20:07:22 <^Spike^> for neither is proper implementation for alot of MTA's
20:07:46 <TrueBrain> SRS is in conflict with existing systems that store information in the envelope sender mailbox name, such as VERPs.
20:07:47 <TrueBrain> SRS is a system that, after even just two levels of forwarding, causes envelope sender mailboxes to become so long that they run the risk of hitting mailbox name length limits in mail softwares.
20:07:51 <TrueBrain> SRS creates the possibility of attackers forging "bounce" messages, reintroducing one of the very things that SPF is touted to (but, ironically, doesn't actually in any case) prevent.
20:08:00 <TrueBrain> written by a dude in 2004
20:08:31 <michi_cc> Second one is wrong though, at least in the currently accepted SRS.
20:08:43 <TrueBrain> well, that is good ;)
20:09:01 <TrueBrain> I am sure either VERP or SRS died nowedays to :P
20:09:28 <Ammler> google adds this: Return-path: <openttd+bncCPKV0ZbuFRCNo534BBoEopw64g@gmuer.org>
20:10:19 <SpComb> but upshot: if you enable SPF for your domain, you break forwarding of mail from your domain by the receipient
20:10:37 <michi_cc> Ammler: Not relevant to SRS though, as Return-path is content, and not part of the FROM and RCPT SMTP envelope.
20:11:01 <Ammler> michi_cc: well, looks better as srs
20:11:40 <TrueBrain> SpComb: basic rule of thumb: if you enable SPF with -all, don't use non-rewriting aliases on external hosts
20:11:46 <michi_cc> Ammler: It's a totally different thing though, that has *nothing* to do with the SMTP envelope MX's deal with.
20:12:44 <Ammler> but we might be to get something like that to work, then we don't need SRS
20:12:46 <michi_cc> Ammler: From the POV of mail transport, everything inside a mail (including all 'From:', 'To;' or whatever) is completely irrelevant.
20:13:15 <Ammler> for me, SRS looks very hackish
20:13:23 <TrueBrain> it doesn't only look hackish
20:13:34 <TrueBrain> it is a solution to a problem created by SPF
20:13:42 <TrueBrain> an "unwated" side-effect
20:14:05 <TrueBrain> personally, I dont see it as such. I dno't want to email to a maillist which forwards my email blindly. I want it to rewrite it, so replies go back to the maillist :)
20:14:14 <TrueBrain> "reply-all" button is often not used by so many people .. ugh .. those days :(
20:14:16 <Ammler> yes, just put the mail in a new envelope and you get your lovely DSN to work :-P
20:14:40 <TrueBrain> dont know if you ever had that, that you email a maillist, and you get a personal reply back, because the dude forgot to hit Reply All
20:14:45 <TrueBrain> so you forward it to the maillist
20:14:50 <TrueBrain> and the Re and Fwd stack up
20:15:15 <Ammler> gmail web doesn't support that
20:15:30 <Ammler> if you hit reply there, you do not send to the maillist
20:15:32 <TrueBrain> nowedays maillists rewrite the from to a single address: the list
20:15:57 *** sla_ro|master has joined #openttd
20:16:29 <TrueBrain> or, my personal favorite: have mail read-only, and a webinterface to reply :)
20:16:48 <TrueBrain> like ... our bug tracker :D
20:16:53 <SpComb> TrueBrain: but SPF happens on the sender address, and the alias/forwarding happens on the recipient address :)
20:17:10 <SpComb> unrelated policy actions
20:17:10 <TrueBrain> SpComb: that causes the unwanted side-effects :)
20:17:18 <TrueBrain> it basically killed all simple 'aliases' maillists
20:17:35 <SpComb> dunno, we do forwarding
20:17:41 <TrueBrain> that is forwarding :D
20:18:20 <SpComb> i.e. user@ourdomain.fi -> <their current email address that they actually read>
20:18:24 <SpComb> typical .forward setup (but in LDAP)
20:18:31 <TrueBrain> that is called an alias ;)
20:18:46 <SpComb> but then also 'alias' maillists and other stuff internally, but SPF shouldn't affect anything there
20:19:10 <TrueBrain> and that is were many people consider SPF 'broken'
20:19:10 <SpComb> and yeah, I guess those user@ forwards break with SPF on the sender domain
20:19:37 <TrueBrain> truth is, that SMTP was never intended to work like that; we only all got used to abusing it that way. And that causes a lot of discussion
20:19:41 <TrueBrain> and then you get solutions like SRS :P
20:19:44 <SpComb> but how else could they be done :/
20:19:56 <SpComb> I mean, the alias-forwards, in our case
20:20:23 <TrueBrain> basically, if I email from host A to host B, and it forwards it to host C, host C should, upon reply, go via host B (ideal world)
20:20:30 <TrueBrain> that would allow internal domains for mail delivery
20:20:39 <TrueBrain> this never got mainstream
20:20:43 <TrueBrain> and is kinda impossible these days
20:21:06 <TrueBrain> SPF is build on this idea, to put it in easy terms
20:21:28 <TrueBrain> what we commonly do, is mail from host A to host B, which is forwarded to host C, enters host C as if it came from host A
20:21:31 <TrueBrain> host C doesnt see host B
20:21:35 <TrueBrain> this is an SPF failure
20:21:38 <TrueBrain> and has no real solution
20:22:18 <TrueBrain> many call it abuse of forwarding
20:22:36 <TrueBrain> others call it "but it always used to work"
20:22:47 <SpComb> host B should tell host A to send the mail to host C instead :)
20:22:50 <TrueBrain> now .. 8 years later
20:22:52 <TrueBrain> still no solution :(
20:23:00 <TrueBrain> SpComb: that would be one valid solution
20:23:26 <SpComb> but I'm sure that's also vulnerable to abuse in some way
20:23:50 <TrueBrain> every solution will be abused by some people :)
20:23:56 <TrueBrain> either by spammers or legit users :P
20:24:39 <TrueBrain> so basically, use maillists, not aliases :P
20:24:59 <TrueBrain> (where a maillist, if you follow the RFC definitions, is one that rewrites the Mail From, and an alias doesnt)
20:25:50 <TrueBrain> (forward is ambigious, hence the clarification in those 2 tersm :))
20:40:34 *** mahmoud has joined #openttd
20:40:43 <Eddi|zuHause> cool, you can see a large ring around the moon
20:41:11 *** Chris_Booth has joined #openttd
20:43:06 <Eddi|zuHause> more like a rainbow :)
20:43:18 <Eddi|zuHause> (only not coloured)
20:58:04 *** Chris_Booth_ has joined #openttd
21:00:17 *** Chris_Booth_ is now known as Chris_Booth
21:06:14 *** valhallasw has joined #openttd
21:19:53 <Eddi|zuHause> ERROR: unable to write data: [Errno 28] No space left on device
21:20:55 * Alberth though space had infinite size
21:22:13 <Eddi|zuHause> the forum looks weird...
21:22:27 <Eddi|zuHause> it has "summer" text colours, but "winter" background...
21:23:57 <welshdragon> bugs my eyes to s***
21:27:17 <Alberth> clear your browser cache :)
21:32:21 *** TomyLobo2 has joined #openttd
21:37:47 <andythenorth> switch your forum theme :P
21:38:12 *** TomyLobo2 is now known as TomyLobo
21:38:56 *** Chris_Booth_ has joined #openttd
21:39:06 <andythenorth> or is it plastic pixels :P
21:40:23 <Rubidium> that can't be OpenTTD. It's 32bpp and unrealistically scaled!
21:42:30 *** fjb|tab has joined #openttd
21:44:55 *** Chris_Booth_ is now known as Chris_Booth
21:49:07 <CIA-6> OpenTTD: rubidium * r23764 /trunk/src/ (6 files in 4 dirs): -Fix [FS#4955]: make default timeouts for certain states lower and configurable
21:49:30 <appe_> stop ruining a perfect game with 3d implications.
22:10:48 *** andythenorth is now known as Guest22884
22:10:50 *** andythenorth has joined #openttd
22:19:28 <CIA-6> OpenTTD: translators * r23765 /trunk/src/lang/ (8 files in 2 dirs): (log message trimmed)
22:19:28 <CIA-6> OpenTTD: -Update from WebTranslator v3.0:
22:19:28 <CIA-6> OpenTTD: dutch - 7 changes by Yexo
22:19:28 <CIA-6> OpenTTD: english_US - 5 changes by Rubidium
22:19:28 <CIA-6> OpenTTD: french - 8 changes by glx
22:19:30 <CIA-6> OpenTTD: german - 2 changes by planetmaker
22:19:30 <CIA-6> OpenTTD: italian - 1 changes by lorenzodv
22:22:19 <CIA-6> OpenTTD: rubidium * r23766 /trunk/ (6 files in 4 dirs): -Update: some documentation
22:25:57 <CIA-6> OpenTTD: rubidium * r23767 /tags/1.2.0-beta2/: -Release: 1.2.0-beta2
22:28:42 <Eddi|zuHause> hm, i don't suppose nml supports the ** operator?
22:29:44 <Eddi|zuHause> not what i'm looking for :)
22:30:32 <Eddi|zuHause> basically i want "inflation_offset = (1.04/1.03)**(start_year-1920)
22:32:04 <Rubidium> without floating point math that looks pretty difficult
22:32:38 <Eddi|zuHause> i see... i'll end up making a lookup table i believe
22:32:53 <planetmaker> Eddi|zuHause: I don't think that inflation should be priced into the running costs
22:33:33 <Elukka> what if you play with inflation of?
22:33:35 <Eddi|zuHause> planetmaker: it makes a big difference if you start in 1990 or start in 1880 and play until 1990
22:34:19 <Eddi|zuHause> Elukka: then the factor will be 1
22:34:59 <Eddi|zuHause> assuming i can actually read that flag
22:35:54 <Eddi|zuHause> we say "zwei Fliegen mit einer Klappe"
22:36:30 <Rubidium> andythenorth: you have to evaluate your bed and only if that returns nothing you start with code
22:39:17 <planetmaker> sleep well, andythenorth :-)
22:40:52 *** andythenorth is now known as Guest22887
22:40:52 *** andythenorth has joined #openttd
22:42:21 <appe_> what does primary make new industries form on a map?
22:44:46 <Yexo> <Eddi|zuHause> basically i want "inflation_offset = (1.04/1.03)**(start_year-1920) <- that's impossible to implement in nml
22:45:05 <Yexo> except when you limit both values to constants, in which case it would be useless
22:45:32 <andythenorth> why not fix ottd?
22:45:35 <andythenorth> inflation is broken
22:45:58 <andythenorth> why another newgrf hack :P
22:48:14 *** welshdragon has joined #openttd
22:51:05 * andythenorth boards the sleep train :)
23:00:49 <Eddi|zuHause> is that the swiss embassy?
23:01:24 <__ln__> no idea actually, though would seem logical.
23:05:09 <Eddi|zuHause> google maps says yes
23:08:48 <Eddi|zuHause> "Building was constructed 1871, became property of switzerland in 1919 and used continuously as swiss consulate, survived the reconstruction of central berlin under the nazis and WWII"
23:10:08 <Eddi|zuHause> "that makes it the closest embassy to the federal power center (Reichstag and Kanzleramt)
23:12:14 <__ln__> a big swiss flag being the first sight of berlin was a bit surprising
23:16:25 *** Brianetta has joined #openttd
23:24:53 <BUTTMUNCH> love what you guys are doign with openttd
23:28:17 <welshdragon> Where's an admin when you need one to ban someone?
23:29:02 <BUTTMUNCH> wow, ban me because i pull a little joke?
23:29:21 <welshdragon> it's inapproriate - we're family friendly in here
23:29:25 <planetmaker> no, because of a yelling nick name and poor spelling ;-)
23:29:26 <__ln__> if i were an op, i would ban you for your nick, not for the joke.
23:29:30 <DorpsGek> planetmaker: English only
23:29:55 <BUTTMUNCH> sorry, it’s my secondary nickname for when i d/c
23:29:59 *** BUTTMUNCH is now known as iddqd
23:30:14 <Eddi|zuHause> how's that an explanation?
23:30:37 <iddqd> how is it not? you fill in a first nickname, and a 2nd nickname in mirc
23:30:46 <Elukka> but it's a great nickname
23:31:05 <iddqd> but some people don’t seem to like it :(
23:31:43 <__ln__> i guess hookers could be added using a newgrf, but not sure how exactly would they fit in the game.
23:32:18 <iddqd> maybe like an extra service during train passenger transportation
23:32:23 <iddqd> you get extra $$ for tickets
23:34:01 <Elukka> but surely they would be independent operators rather than part of your transportation company
23:34:50 <iddqd> but they pay you to be on your train
23:36:22 <Eddi|zuHause> MB once said he implemented a "red light district" :)
23:36:37 <__ln__> i declare your suggestion to be out of scope of the game.
23:38:34 <Eddi|zuHause> <iddqd> maybe like an extra service during train passenger transportation <-- and which country offers that service?!?
23:42:28 *** swissfan91 has joined #openttd
23:42:50 <swissfan91> on the devzone - how do I delete an old version of a tracking table I have uploaded?
23:43:34 <planetmaker> that's a bit unspecific
23:44:20 <Elukka> "Montreal-Mirabel International Airport was the largest airport ever envisioned, with a planned surface area of 397 square kilometers. That's larger than Montreal, the city that it served."
23:45:01 <Rubidium> imagine how big it'd be when it'd be an intercontinental airport
23:45:05 <swissfan91> sorry I'm new to this whole devzone malarky.
23:45:55 <planetmaker> if you have the proper permissions, you find a waste bin icon next to the filename
23:46:27 <swissfan91> ah, that'll be it then!
23:46:31 <planetmaker> on the other hand... the date of upload is listed right next to the filename. Thus it's always clear what is newest
23:46:44 <swissfan91> I just thought I was being simple :)
23:48:46 <planetmaker> swissfan91: wrt tracking table: did you consider to use a google docs spreadsheet?
23:49:25 <planetmaker> it avoids imho all kind of issues as all people who may can edit it and always have an up-to-date version of the document
23:49:27 <swissfan91> I only copied the standard gauge table Yoshi created. Perhaps I'll ask him.
23:49:35 <swissfan91> it does sound like a good idea, yes.
23:49:49 <swissfan91> does google docs tell you who has been changing it, and where?
23:49:57 <swissfan91> I could envisage it becoming a little annoying.
23:50:14 <planetmaker> only those people who were given write access can
23:50:56 <Snail_> I was wondering if you had had the chance to have a look at the custom tunnel sprites I sent you?
23:51:40 <Elukka> Eddi|zuHause: there's an entirely unrelated set called CETS that stands for the same thing that was active in.. late 2010/early 2011 or so
23:51:59 <planetmaker> I started somewhat. But there's nothing which works so far
23:52:22 <Eddi|zuHause> Elukka: we discussed that half a year ago already
23:52:30 <Eddi|zuHause> Elukka: and decided to just go ahead
23:52:53 <Eddi|zuHause> Elukka: and that was more like a southeastern europe set
23:54:56 <planetmaker> Eddi|zuHause: what does the "core core core core" mean for wagons?
23:55:10 <planetmaker> why the multiple string?
23:55:22 <Eddi|zuHause> planetmaker: it's triplets: (bav, core, 1), (sax, core, 1) etc.
23:55:24 <Snail_> I talked a little bit with michi_cc yesterday, he suggested me to shorten the northbound (NE and bowls a bit
23:56:26 <Eddi|zuHause> planetmaker: e.g. the ET 171 (S-Bahn Hamburg) has (DRG, ext, 2) and (DB, core, 3)
23:56:53 <Eddi|zuHause> planetmaker: as it doesn't really make a lot of sense appearing for DR
23:58:24 <planetmaker> he, so you coded it already? :-)
23:59:19 <michi_cc> Changing OTTD was trivial, finding the right offsets for the graphics not so much :)
23:59:26 <swissfan91> they look a little... large?
23:59:40 <planetmaker> and yes, I have to agree that the Northbound grass cover for the tubes should stretch more
23:59:51 <planetmaker> it bends quite different from the South-bound
continue to next day ⏵
