IRC logs for #openttd on OFTC at 2012-09-08
⏴ go to previous day
00:01:46 *** Rhamphoryncus has joined #openttd
02:02:40 *** tokai|noir has joined #openttd
02:02:40 *** ChanServ sets mode: +v tokai|noir
02:40:02 *** Supercheese has joined #openttd
03:03:35 <aditsu> hi, I joined a multiplayer game earlier, but now I get "network-game connection lost" every time I try to reconnect, what can I do about it?
03:26:03 <aditsu> oh, it worked eventually
04:56:41 *** Eddi|zuHause has joined #openttd
05:16:41 *** andythenorth has joined #openttd
05:24:01 *** Markavian` has joined #openttd
05:52:28 *** sla_ro|master has joined #openttd
05:58:05 *** KnogleAFK has joined #openttd
06:00:11 *** Supercheese has joined #openttd
06:37:22 *** andythenorth has joined #openttd
06:53:54 *** KnogleAFK is now known as Knogle
07:15:08 * andythenorth ponders NoConomy
07:15:19 <andythenorth> where is alberth? :)
07:18:19 <Terkhen> enjoying the weekend, I hope
07:19:59 *** valhallasw has joined #openttd
07:24:55 <andythenorth> no frosch either :)
07:26:31 <andythenorth> newgrf has multiple ways to exert control over production
07:26:49 <andythenorth> if newgrf lost ability to set prod_level, that would be cleaner
07:27:19 <andythenorth> then ottd could set prod_level to [whatever] NoConomy decides
07:27:42 <andythenorth> then GS could tell ottd NoConomy methods what prod_level to set
07:27:48 <andythenorth> and then there are no conflicts
07:28:01 <andythenorth> newgrf authors may have to write a little more code than previously, but it's not hard
07:28:28 <andythenorth> NoConomy provides a base prod_level multiplier
07:28:43 <andythenorth> each industry instance can then modify that according to any rules newgrf chooses
07:29:00 <planetmaker> sound simply like just another global variable
07:29:31 <andythenorth> in this case it involves removing some of the newgrf API
07:29:42 <andythenorth> but no functionality is lost, because there are other ways to get same result
07:30:01 <andythenorth> as Yexo proved last night with his FIRS patch, which was done without using prod_level
07:30:29 <andythenorth> because of the way it was patched it also works seamlessly with industry production cheat, which also happens to set prod_level
07:59:31 *** Alberth has joined #openttd
07:59:31 *** ChanServ sets mode: +o Alberth
08:03:19 <Alberth> 2 other devs awake already :)
08:07:12 *** valhallasw has joined #openttd
08:08:20 <Terkhen> half awake actually :P
08:08:44 <andythenorth> 2 other devs, and one andythenorth trying to delete some of the newgrf spec
08:09:20 <Alberth> good, good, please proceed mr the north.
08:10:17 <andythenorth> this lacks context
08:12:00 <Alberth> I would not mention noconomy at this time
08:13:14 <andythenorth> it's my shorthand for 'openttd should handle more economy stuff'
08:13:22 <andythenorth> and 'newgrfs should stop trying to handle it badly'
08:14:20 <Alberth> you should bring the change in a positive way, more co-operation with player actions, or better handling of probabilites or so
08:15:00 <Alberth> euhm co-ordination with player actions
08:15:27 *** frosch123 has joined #openttd
08:15:33 <Alberth> probability handling could be quite convincing imho
08:15:46 <andythenorth> you're proposing marketing? :O
08:16:39 <andythenorth> can't I just grumble about everything that is broken and should be rm-ed :P
08:16:43 <andythenorth> who needs benefits :P
08:18:09 <frosch123> if you have a method to rm -rf /tt-forums/off-topic, then i am all in
08:18:58 <andythenorth> is that a place I should go?
08:19:03 <andythenorth> I went there once by accident
08:19:07 <andythenorth> it was very annoying
08:19:18 <andythenorth> I nearly flamed everyone
08:19:58 <frosch123> actually it is fine as long as they do not show up on the rest of the forums :p
08:20:27 <andythenorth> frosch123: so I have a proposal. To avoid confusion by talking about implementation instead of result....
08:20:40 <andythenorth> ...the proposal is basically 'make prod_level read only for newgrf'
08:21:26 <andythenorth> and the 'benefit' is cleaner interface
08:21:43 <andythenorth> and possibility for more control to move back to openttd (and in future, GS)
08:23:58 <Eddi|zuHause> andythenorth: what's preventing you from simply not using it?
08:24:12 <andythenorth> this is about a cleaner interface
08:24:31 <frosch123> write a tool which ranks newgrfs as andy-clean
08:24:56 <Eddi|zuHause> then rewrite all existing GRFs to this method
08:24:58 <frosch123> anyway, users can already change prod_level using cheats
08:25:04 <frosch123> so there is no problem with gs changing it
08:25:16 <frosch123> if the newgrf conflicts, it's the user's fault
08:25:31 <andythenorth> so who wins cb29 / 35?
08:25:36 <andythenorth> you'd have to run an auction
08:25:56 <andythenorth> you now have two bidders for the result instead of one
08:26:04 <frosch123> andythenorth: just like for town growth
08:26:14 <frosch123> gs can set town growth to fixed rate, or to automatic
08:26:40 <frosch123> just always give priority to the gs
08:26:54 <andythenorth> so that's an auction
08:26:57 <andythenorth> and gs always wins
08:28:26 <andythenorth> ok, so that's a proposal to change spec of cb29 / 35?
08:28:48 <andythenorth> instead of: newgrf may set prod_level
08:29:02 <andythenorth> now: newgrf may attempt to set prod_level, but openttd may over-rule it
08:29:12 <Eddi|zuHause> in how far is there a "change"?
08:30:34 <Rubidium> Alberth: but... I was awake as well... replacing the tires on my bike
08:30:41 <Rubidium> (without the proper tools)
08:31:00 <andythenorth> currently newgrf has absolute control over prod_level
08:31:10 <Alberth> oh, and at work already. Welcome then as well Rubidium
08:31:19 <andythenorth> frosch123 proposes that newgrf does not have absolute control over prod_level
08:32:29 <frosch123> andythenorth: changing production should be no problem
08:32:34 <frosch123> overruling closure might be
08:40:27 <andythenorth> ok so on this proposal newgrf can only suggest prod_level
08:41:17 <andythenorth> now a GS can trivially break something like (current) FIRS supplies
08:41:27 <andythenorth> so we get whining from newgrf authors
08:42:01 <Eddi|zuHause> i still don't see the problem
08:42:04 *** HerzogDeXtEr has joined #openttd
08:42:23 <andythenorth> it's a change to newgrf spec
08:42:54 <andythenorth> seems to be following case
08:43:01 <andythenorth> - I propose change to newgrf spec
08:43:06 <andythenorth> - reply: don't change newgrf spec
08:43:13 <andythenorth> - instead do xyz
08:43:18 <andythenorth> - xyz changes newgrf spec
08:44:06 <Eddi|zuHause> no, what i meant was: "i do not see the situation that causes the need for any change"
08:45:17 <andythenorth> so fundamental issue: do you think openttd / GS should be able to control production level of primary industries?
08:46:20 <andythenorth> control / influence /s
08:46:56 <Eddi|zuHause> and then there comes someone along muttering "production cheat" and stuff
08:47:07 <andythenorth> so for 'no', there is no problem :)
08:47:50 <Eddi|zuHause> what happens inside an industry is clearly a newgrf task
08:48:31 <andythenorth> I see gameplay benefits in being able to influence amount of produced cargo via openttd or GS
08:48:50 <andythenorth> e.g. recessions, economic changes etc
08:49:00 <andythenorth> or unlocking stages of a challenge
08:49:24 <andythenorth> "You have completed stage 1, coal production increases across the map"
08:50:57 <Eddi|zuHause> ok, but then again, what is the problem with that?
08:51:35 <andythenorth> with doing that?
08:51:40 <andythenorth> can't be done currently
08:52:03 <Eddi|zuHause> the newgrf specs never said that cbXY was the only way that values could change
08:52:31 <Eddi|zuHause> and it also never said that the values would be the same the next time the cb is run
08:52:33 <andythenorth> that's one interpretation yes
08:52:39 <andythenorth> I don't read it that way
08:52:43 <andythenorth> I don't think other authors would either
08:52:56 <Eddi|zuHause> so whoever sets the value last, "wins"
08:53:05 <andythenorth> I don't have a strong opinion on it, but I think it's a spec change
08:53:22 <andythenorth> it breaks pretty nearly every known industry set afaik
08:53:55 <andythenorth> based on current test patch, FIRS will probably be changed to handle it nicely though
08:54:03 <andythenorth> which is nice for me and FIRS authors :P
08:54:13 <Eddi|zuHause> what kind of "break"?
08:54:13 <andythenorth> less so for others?
08:54:38 <Eddi|zuHause> what currently happens during a recession?
08:54:45 <andythenorth> depends on newgrf
08:54:53 * andythenorth wonders which other sets actually would break
08:54:58 <andythenorth> PBI I'd have to read src
08:55:10 <andythenorth> maybe it would be fine
08:55:34 <andythenorth> maybe just add a note to spec and all is well
08:56:59 <andythenorth> just add "OpenTTD may over-ride the value you return, for more precise control use the production callback to generate produced cargo"
09:02:13 <Eddi|zuHause> that is probably worded badly
09:03:03 <Eddi|zuHause> recessions and stuff are an original feature, and the production cheat is also very old
09:04:37 <andythenorth> recessions don't touch the newgrf
09:04:50 <andythenorth> unless it explicitly allows them
09:05:01 <andythenorth> newgrf has absolute control except for production cheat
09:05:49 <andythenorth> * recessions don't touch the newgrf *if* it handles cb29 / cb35 *and* it doesn't return 04
09:05:56 <andythenorth> it's not a simple situation ;)
09:08:29 <andythenorth> suggestion: allow openttd / GS to over-ride prod_level, see what breaks over time, deal with it later?
09:09:33 <Terkhen> I'm against anything that includes "deal with it later" :P
09:09:59 <andythenorth> don't solve things that aren't proven to be a problem
09:10:47 <andythenorth> if TB had stopped to think through all problems GS might cause, we wouldn't have it :)
09:11:04 <Terkhen> problems were thought :P
09:11:06 <Eddi|zuHause> andythenorth: other approach: GS influencing production will run the random callback, or some new callback?
09:11:45 <Eddi|zuHause> andythenorth: then newgrf keeps its control, but GS gets some
09:12:13 <andythenorth> that's my proposal
09:12:33 <andythenorth> openttd can reject result newgrf returns to cb 29 / 35
09:12:40 <andythenorth> GS can influence openttfd
09:12:46 <andythenorth> it's a clean set of separations
09:12:50 <Eddi|zuHause> no, that is something very different
09:13:01 <Eddi|zuHause> than what i just suggested
09:13:11 <andythenorth> oh you trigger cb29
09:13:19 <andythenorth> and newgrf handles that
09:14:21 <andythenorth> sounds reasonable
09:14:33 <Eddi|zuHause> var10 (or something) contains what the GS wants to do (double, halve, set to minimum, close, whatever)
09:14:36 <andythenorth> why does newgrf get to keep control?
09:15:11 <Eddi|zuHause> and newgrf can then agree with it, or not
09:15:17 <andythenorth> or rather, why does newgrf need that?
09:15:36 <Eddi|zuHause> _someone_ needs to have the final word
09:16:05 <Eddi|zuHause> and i think this way around is cleaner/needs fewer changes
09:17:34 <andythenorth> newgrf already has complete control anyway via production cb
09:17:43 <andythenorth> currently it has too many ways to do production imho
09:17:47 <andythenorth> more than is necessary
09:18:01 <andythenorth> it makes everything harder
09:18:16 <Eddi|zuHause> but you cannot remove any controls
09:18:35 <Eddi|zuHause> except by moving to GRFv9 and invalidating all old GRFs
09:20:29 <Eddi|zuHause> that means ALL grfs, no matter whether they provide industries or not
09:23:19 <andythenorth> I had missed that implication
09:23:29 <andythenorth> when we moved to v8, old grfs continued working
09:26:10 <andythenorth> so maybe we should not allow GS to control industry production at all
09:26:17 <Eddi|zuHause> yes, because we kept everything in place for grf<8
09:26:55 <andythenorth> it's quite frustrating being bound to this broken idea that industry newgrfs can provide most of the economy
09:27:00 <Eddi|zuHause> some callbacks were removed in favour of cb36, but all these callbacks were kept as legacy code
09:28:00 * andythenorth tries to figure out a route :|
09:28:52 <andythenorth> player advanced setting? "Industry production changes: newgrf | GS"
09:29:16 <Eddi|zuHause> now you're getting silly :)
09:29:22 <andythenorth> a new action 0 prop for newgrf? "Permit prod_level to be change"
09:29:47 <andythenorth> trying to think of simple, nice ways around this :)
09:45:17 *** Progman has joined #openttd
10:05:03 <planetmaker> to boldly go where no man has gone before ;-)
10:09:16 <Rubidium> but it isn't 'there' yet, right?
10:15:28 <planetmaker> dunno. As I understand it, it's in the transition region. Neither here nor there.
10:16:08 <Rubidium> so if Voyager I is the farthest, what probe is second farthest
10:17:20 <frosch123> in any case, there has still been no man :)
10:17:51 <frosch123> and i am happy for that one
10:35:58 <planetmaker> Rubidium, the 2nd furthest alive spacecraft is... surprise, Voyager2 ;-)
11:03:06 *** argoneus has joined #openttd
11:29:44 *** Devroush has joined #openttd
11:30:30 *** KenjiE20 has joined #openttd
11:30:43 <Rubidium> planetmaker: then when did it pass Pioneer 10
11:31:54 <Rubidium> As of July 25, 2012, Voyager 2 is traveling at 15.447 km/s relative to the Sun, and currently at a distance of about 99.13 astronomical units
11:32:17 <Rubidium> october 2009: Projections indicate that Pioneer 10 reached 100 AU
11:33:36 <Rubidium> Feb 8 2012: Its distance from the Sun is about 105.197 (Pioneer 10)
11:34:36 *** Supercheese has joined #openttd
11:36:19 <Rubidium> Feb 8 2012 Voyager 1: 119.70479 astronomical units from the Sun
11:56:25 <planetmaker> Rubidium, communication with Pioneer 10 was lost 10 years ago.
11:56:34 <planetmaker> "The final, very weak signal from Pioneer 10 was received on January 23, 2003 when it was 12 billion-kilometers (80 AU) from Earth"
11:57:14 <TrueBrain> clearly no Duracell batteries
12:07:34 <Rubidium> planetmaker: so what, it's remains man made
12:08:50 *** valhallasw has joined #openttd
12:14:03 <CIA-2> OpenTTD: frosch * r24513 /trunk/src/ (8 files in 4 dirs): -Add: [Script] ScriptIndustryType::IsProcessingIndustry.
12:15:59 *** LordPixaII has joined #openttd
12:18:08 <Elukka> one of these days we're gonna have to start launching proper nuclear reactors
12:18:13 <Elukka> if we ever want to do anything significant in space
12:20:11 <frosch123> "significant in space" is an interesting term
12:20:29 <frosch123> is a supernove something significant?
12:22:21 <Rubidium> no, those are generally only tiny and insignificant for us
12:34:52 *** FLHerne has joined #openttd
12:40:04 *** andythenorth has joined #openttd
12:49:42 *** valhallasw has joined #openttd
13:05:16 <Elukka> significant as in significant human activity
14:08:46 *** andythenorth has joined #openttd
15:02:04 <Zuu> Hmm, so currently there are company specific goals and global goals. They are stored in the same structure with company_id set to a special value for global goals. In this storage structure its non-trivial to allow global goals to store progress for all 14 companies, while the company goals only need to store one progres text.
15:02:31 <Zuu> Furthermore, I start to wonder if it makes sense to have a global goal with company specific progress trackning.
15:02:54 <Zuu> Isn't global goals, goals that all companies should cooperate on?
15:03:53 <frosch123> for nocargoal it's a goal which every company tries to achieve
15:03:57 <Zuu> If they compeete on the goal, just duplicate the goal for all companies as specific goals with the same text. Then you get capability to track progress and completeness state for all companies.
15:05:05 <frosch123> what worried me about nocargoal was that the target date is the same for all companies, indedendent on when they start
15:05:34 <frosch123> so, yeah, maybe most goals should be company specific
15:05:41 <frosch123> and only cooperative goals should be global
15:05:43 <Zuu> I just think that we might overengineer things if global goals get a special storage type with an array for storing progress/completeness and the other goals are stored as individual goal (IDs) per each company.
15:06:03 <frosch123> yeah, you might be right :)
15:06:41 <Zuu> One reason why NoCarGoal use the global goals currently is to allow spectators to see the goal cargoes.
15:07:30 <Zuu> If all players/spectators could see all goals, the global goals are not so important anymore.
15:13:35 <Alberth> Zuu: where to find 'AILibrary' needed by (game script) Superlib-25 ?
15:14:03 <Zuu> Why would a game script need AILibrary?
15:14:11 <Zuu> You can get SuperLib for NoGo from bananas.
15:14:34 <Alberth> I have superlib-25, but no AILibrary
15:15:21 <Zuu> it will need RPF 4 for nogo, but what is the error message?
15:15:26 <frosch123> i guess it should say game script lbrary
15:16:14 <Alberth> frosch123: it should omho, otherwise Superlib is not findable
15:16:59 <Alberth> Zuu: I looked for "Superlib" version 25, and I ended up with an AI one, apparently
15:17:12 <frosch123> you need superlib for nogo
15:18:20 <Alberth> yeah, I figured as much, but thanks :)
15:18:52 <Alberth> Zuu: imho it wouldl REALLY help if you give different a different name
15:18:54 <frosch123> i cannot find the error message in the source you are refering to
15:20:31 <Zuu> Alberth: What should I give a different name?
15:20:38 <frosch123> oh, you installed an ai library into the gs lib directory
15:21:01 <frosch123> and ottd fails to distinguish them :)
15:21:35 <Zuu> Or you refer to the fact that AI and NoGo version use the same short name?
15:21:40 <frosch123> the library derives its main class from AIlibrary, which is not present when ottd loads a gs lib
15:22:46 <Zuu> I'll take a look in my conversion script and see if it needs an adjustment there. Though I never had any problem with that. But it could be that OpenTTD is just kind and accept it even if its not correct.
15:23:08 <frosch123> Zuu: it's no problem with your lib
15:23:28 <frosch123> just ottd tries to load the library.nut which uses AILibrary instead of GSLibrary
15:23:35 <Alberth> how am i to know i need FOO_for_NoGo ??
15:23:47 <frosch123> ottd just has no way to distinguish ai from gs libs, except by printing that error
15:24:17 <frosch123> Alberth: AIlibrary is no external library
15:24:39 <frosch123> just that ottd does not provide this class for gs
15:25:00 <Zuu> Alberth: By downloading via bananas or investigate in what dependencies a gs/ai have.
15:25:40 <Zuu> I try to include the dependency info in text for my AIs/GSs for those who try to download manually but generally recommend people to use bananas as that removes all that hassle.
15:25:41 <Alberth> util.superlib seems clear enough to me
15:25:58 <Alberth> Zuu: yep, except my bananas is down :p
15:26:22 <Alberth> now if you use "whateverlib", I would have found the right thing
15:26:31 <Zuu> In your case you can also untar the AI/GS and look in main.nut and within the 50 first lines or so see which depnedencies there are.
15:26:49 <Zuu> However, that is not something I would suggest to the average player :-p
15:27:16 <Alberth> import("util.superlib", "SuperLib", 25); <-- that?
15:27:34 <Alberth> it still says "Superlib" not SuperLib_for_nogo
15:28:35 <Alberth> so I am to know that by magic or so?
15:28:52 <Zuu> I guess it could be more clear. But from my point of view, I found it self explaining that a GS use the GS version of a library and not the AI version.
15:29:40 <frosch123> rename the classic superlib to superlib for noai :)
15:29:48 <Zuu> I generally don't suggest that a user need to look into main.nut to find this out.
15:30:09 <Alberth> it's even useless, the info is not even there
15:31:36 <Zuu> I see your point. However, I think the way I though was that the "for NoGo" part is not really part of the name but something that explains the target/platform for the library.
15:35:09 <Alberth> is there a pathfinder.road for nogo too?
15:36:04 <frosch123> it's listed on bananas at least
15:36:10 <frosch123> generally no ai lib fors gor gs
15:36:26 <frosch123> *no ai lib works for gs
15:37:55 <Zuu> Indeed, as all calls to the API are prefixed by AI for AIs and GS for GSs, so releasing a universal library is non-easy.
15:38:37 <Alberth> it has no _for_nogo extension, it seems
15:39:45 <Alberth> which is going to be fun to have together :)
15:39:57 <Zuu> Neither does any of the other ports that came after my NoGo port.
15:41:03 <Zuu> So if you want consistency it is rather that the SuperLib port should be renamed which makse more sense now given that it now have almost all functionallity of the AI version.
15:41:23 <Zuu> At the beginning the NoGo version was limited and did only support a subset of what the AI version could do.
15:42:31 <Zuu> However when it was released it was an experimental test to see how much part of it that was possible to transfer into the new GS area.
15:43:06 <Alberth> oh, that's fine, my only problem is that naming conventions are currently confusing
16:14:05 <Zuu> Alberth: If "SuperLib for NoGo" would have been named just "SuperLib" on bananas, would you have found it easier then?
16:19:45 <Zuu> frosch123: Is your proposal still to add a new top-level button for Goals? Or to have a "goal" item in the company legue menu? Or to ... ?
16:20:30 <frosch123> what number of nogo windows are on the list to add?
16:20:42 <frosch123> i remember replacnig the league and score window
16:21:21 <frosch123> and if the score window would remain with tabs, the goals should as well maybe
16:21:23 <Zuu> We had a discussion a week ago or so to allow GSs to create additional detail windows in case the goal list is not enough.
16:23:13 <Zuu> But I also see your point that the GUI generally have these list at top level menus with all companies there.
16:23:57 <Zuu> And one way is to do like that and then on the company specific goal window in future add a "details" button or similar if we add a such capability.
16:24:44 <frosch123> anyway, width is not really a problem
16:24:59 <frosch123> and there were also buttons added in the past
16:25:31 <Zuu> Ok, I had the suspection that it might be a problem, but was not sure. That is why I asked you :-)
16:27:03 <Zuu> But then your proposal to have a new goal button that open up a window for the current company (with drop down to see the window for other companies) with the goals may be good.
16:27:19 <Zuu> Any details would then be opened from there.
16:27:39 <Zuu> We get a clear goal button then in the GUI that isn't hidden away from players.
16:28:05 <Zuu> If the button is invisible if there are no goals, that is even more a clear indication that the game have or have not goals
16:29:13 <frosch123> maybe: currently there is the league menu, with league and score
16:29:28 <frosch123> if we remove both of them if a gs is present
16:29:39 <frosch123> we can add a gs league, and one item per company for the goals
16:29:45 <frosch123> then it would work like the company button
16:30:02 <frosch123> (one item per company, plus separate client list and spectate)
16:30:58 <Zuu> For GSs that provide game changes/goals it make sense to replace the default company league table with a rating from the GS.
16:32:07 <Zuu> But maybe it is easiest to code it so that if any GS is loaded, the default company league is disabled and leave it up to GSs to decide if they want to implement a replacement or not. (but if they don't there is no league at all)
16:35:20 <Zuu> So to update that one, the fist point for Goal Window would change to "Add a new main toolbar button which opens up company specific goal windows in the same way as eg. the economy window works"
16:47:54 *** Ryton- is now known as Ryton
16:48:41 <Ryton> 120 users in the channel, wow!
16:48:53 <Zuu> 120 irc clients connected
16:50:20 <Ryton> hehe. sounds like a duth politic figure: i can still count: 74
16:50:57 <Ryton> +32 is 116. (or the like;-)
16:52:39 <Alberth> there is no "game library" in the list above
16:53:40 <Alberth> imho different libraries should be clearly distinguishable by name
16:54:16 <frosch123> let's blame TrueBrain for both :p
16:55:38 <Alberth> that may help in a few of the above cases :)
16:56:41 <Zuu> Alberth: But if it is the same library, just with all "\bAI" replaced by "GS", does that varant a different name of the library?
16:58:16 <Alberth> I have to use the right one otherwise it fails. That makes it a different library in my book.
16:58:37 <Alberth> but GameSuperLib and AISuperLib would be fine imho
16:58:38 <frosch123> it's different like 64bit and 32bit
16:58:46 <frosch123> problem is just bananas
16:59:32 <Zuu> As frosch123 said, I see it more like eg. a linux or windows build of the same library.
17:00:19 <Zuu> For Pathfinder.Road all code in the repository is for the AI target. Then the Makefile contains a sed command to produce the GS version of the library.
17:00:44 <Alberth> Zuu: pathfinder is even more fun; the filename is exactly the same
17:00:47 <frosch123> you have to put all 32bit libraries into /usr/lib, and all 64bit into /usr/lib64
17:00:56 <Alberth> from the filename, there is no way to decide what it actually is
17:01:07 <frosch123> if you switch them up, stuff fails to load and prints weird errors
17:01:10 <frosch123> like, invalid elf format
17:01:28 <frosch123> though you don't like elves
17:01:30 <Zuu> Alberth: same with Queue.BinaryHeap, Graph.AyStar and Queue.FibonacciHeep
17:01:49 <Zuu> which were ported before Pathfinder.Road :-)
17:01:50 <Alberth> sure, but they are labeled .x64_86 or ..x68 or so
17:02:01 <frosch123> Alberth: if bananas would show them on one page, with a "GS" and an "AI" download link next to each other, all would be fine :)
17:02:17 <frosch123> Alberth: the library itself isn't
17:02:25 <Alberth> Zuu: without automatic download, nobody will get the right version, ever
17:02:39 <Alberth> frosch123: here the package is also the same
17:02:52 <Zuu> Which is why I tend to suggest to people to use bananas
17:03:27 <Alberth> and that's good, but does that mean it has to be impossible to do it manually?
17:03:32 <Zuu> Another reason is that there is no way to get old versions of a library by downloading from the web frontend.
17:03:54 <Zuu> If I would release version 26 now, you would be unable to get version 25 from the bananas website.
17:04:22 <Alberth> I think it was already established the bananas site needs to be changed
17:06:37 <Zuu> Let say like this. From AI/GS author point of view, it makes sense that the import statement is exactly the same for both AI and GS.
17:07:12 <Zuu> If that were not the case, the NoGo translator for SuperLib would need to find the import statement for Pathfinder.Road and change that one when producing the NoGo version of SuperLib.
17:07:26 <Zuu> So there are reasons why it is like it is.
17:07:55 <Zuu> What you say is that there are also reasons from user perspective to make it more complicated (for AI/GS authors), to make downloading the right lib easier.
17:08:42 <Alberth> how does import xgs versus import xai make it more difficult?
17:09:29 <Zuu> "import x" needs no change at all
17:09:38 <Alberth> use a REPLACEME prefix in the source, and do a s/REPLACEME/AI/ or s/REPLACEME/GS/
17:10:08 <Zuu> All needed currently is "s/\bAI/GS/" to port from AI to GS with respect of API calls
17:10:26 <Alberth> I can't copy squirrel code anyway, so what's the point to keep import x ?
17:11:02 <Zuu> What do you mean by not being able to copy squirrel code?
17:11:49 <Zuu> So far I have had the idea that manual downloading is too complicated for most users, especially with regard of old libraries. So I have seen that as a lost case and not really even though about that issue with respect of naming.
17:12:02 <Alberth> you consider it import to keep "import x", but I don't see that, I need to change my code everywhere when I move between AI and GS, so what is different about x ???
17:12:42 <Alberth> Zuu: ok, fair enough, just say then that superlib cannot be downloaded manually or from the forum
17:12:43 <Zuu> The only change needed when going from AI to GS is to change "AI" to "GS" at the beginning of all words. That is what s/\bAI/GS/ does.
17:13:10 <Alberth> import AIx will then also change
17:13:51 <Zuu> It would be possible to rename all libraries to include AI or GS at the front and have it easily scriptable.
17:14:07 <frosch123> don't try to fix the wrong problems :)
17:14:07 * Alberth fails to see it completely now
17:14:13 <frosch123> the only issue is the bananas web frontend
17:14:24 <Zuu> But there are multiple library authors, and I can't rename all libraries.
17:14:58 <Zuu> I expect that I additionally need to get the database rights and rename the bananas entries in the dB as IIRC that is not possible in the frontend.
17:15:35 *** chester_ has joined #openttd
17:16:19 <Zuu> Alberth: I see your point, and if we would start from scratch, it could make sense to do like your proposal. But then that would have to be though of when the first AI library was created some 4 years ago or so.
17:16:21 <Alberth> frosch123: the path finder tar is called Pathfinder.Road-4.tar both in AI and in GS. Donwload both and good luck in finding out which one is Pathfinder.Road-4.tar and Pathfinder.Road-4(1).tar
17:17:10 <Alberth> start a new library with a new name????
17:17:20 <frosch123> that won't change if you add a newgrf called "Pathfinder Road" :)
17:17:38 <frosch123> bananas does not include the name into the filename
17:17:56 <Zuu> I guess what we could do is to change bananas to add the content type into the tar name. That will not break any dependencies but help users who download manually.
17:18:07 <Alberth> frosch123: so give them a unique name???
17:18:36 <Zuu> eg <name>-<version>-<type>.tar
17:18:50 <Alberth> Zuu: keep the old one around, nothing breaks. Make a new one, people using it will update the import one time.
17:23:23 <Zuu> So you propose that I should duplicate 14 libraries under a different name?
17:23:46 <frosch123> you cannot have them with duplicate unique ids
17:23:49 <Zuu> Additionally there are 6 libraries which I have no rights to do it.
17:23:54 <frosch123> you would have to hack the database to change all names
17:24:36 *** andythenorth has joined #openttd
17:24:49 <frosch123> i guess the best bet is still to make bananas include the type itself
17:26:16 <Zuu> that would also make the two "Beginner tutorial" possible to distinguish :-)
17:31:20 <Rubidium> adding the type to the file name is not as trivial as it might sound
17:31:59 <Rubidium> the file name is generated during upload and checked for uniqueness. It is then stored in the database and everything uses that name to go back from the numbered files in the disk on the server
17:32:06 <Rubidium> that includes OpenTTD
17:33:21 <Rubidium> since the name may be 31 bytes and the version 15 bytes, you end up with a filename of 47 (name + version + separator). This is exactly the size of the variable containing the file name in OpenTTD
17:33:57 <Rubidium> adding more to the file name means retroactively changing that filename length constant in already released OpenTTDs
17:34:31 <Zuu> That does indeed sound a bit hard to do :-)
17:36:26 <Rubidium> one *might* choose to have a different file name for files served by the web server, but I'm not sure whether the http download uses the filename variable for resolving that location. If it doesn't, then that might be an avenue to take. Otherwise you're probably royally screwed
17:36:46 <Rubidium> (http download from within OpenTTD)
17:45:11 <CIA-2> OpenTTD: translators * r24514 /trunk/src/lang/welsh.txt:
17:45:11 <CIA-2> OpenTTD: -Update from WebTranslator v3.0:
17:45:11 <CIA-2> OpenTTD: welsh - 3 changes by kazzie
18:12:58 *** sla_ro|vista has joined #openttd
18:17:42 <Eddi|zuHause> i have a distinct lack of doctor who...
18:17:56 <Supercheese> I hear he's on first. ;)
18:18:16 <Supercheese> (With the esteemed Dr. What on second)
18:18:52 * Supercheese wonders if anyone will get that joke
18:18:58 *** andythenorth has joined #openttd
18:24:47 <Rubidium> did you miss the last like 6 episodes?
18:25:21 <Eddi|zuHause> i'm missing todays episode, mostly
18:25:56 <Rubidium> but apparantly it starts in 10 minutes
18:29:49 <Rubidium> so... Eddi|zuHause, thanks for reminding me ;)
18:39:01 <Eddi|zuHause> Rubidium: well there was an episode already last week
18:39:39 <andythenorth> Silicon Valley game?
18:46:02 <frosch123> planetmaker: Zuu: Alberth: Terkhen: ?
18:46:14 <frosch123> (leaving out rb watching dw :p )
18:46:37 <Alberth> rb is watching tracks :)
18:50:35 <frosch123> is building materials a good choice?
18:54:24 <Kitty> I am wondering if anyone can provide some guidance. I am struggling to get profitable industries other than oil rig/oil well/oil refinery, or coal mine/powerstation
18:56:23 <Alberth> always have a train loading
18:56:42 <Eddi|zuHause> turn off inflation
18:56:49 <Alberth> try to move cargo rapidly
18:56:50 <Kitty> is it better to have lots of shorter trains visiting frequently, or a big train filling rarely ?
18:56:58 <Alberth> don't build long trains
18:57:00 <Kitty> Eddi|zuHause: allready did that
18:57:18 <Eddi|zuHause> the trains should not be too long, because the waiting time will lower your income
18:57:30 <Kitty> how long is a good length? 4 cars? 6 ?
18:57:40 <Eddi|zuHause> try to have the train ready when the next train comes back
18:58:03 <Eddi|zuHause> "good" length heavily depends on the production rate
18:58:25 <Eddi|zuHause> the longest i once had was 15 tiles
18:58:43 <frosch123> a normal train is 4 to 7 tiles
18:58:57 <Kitty> I normally use 5 tiles for most of my train s
18:59:03 <frosch123> a train should not take longer than a month to load
18:59:29 <Eddi|zuHause> i have 2 tiles for really low production
18:59:50 <Alberth> also depends on the engine, early in the game you want shorter trains due to limited poer of the engines
18:59:57 <andythenorth> frosch123: building materials is quite interesting
19:03:52 <Kitty> It's been ages since I played, and the display chain thing is a really nice new feature
19:04:24 <Alberth> it needs improvement :)
19:04:34 <andythenorth> what's the url to the compile farm?
19:04:38 <andythenorth> I need an older nightly
19:05:03 <Alberth> main page, at the top somewhere
19:06:39 <frosch123> though might not be the best mirror for you
19:08:01 <Kitty> also, is ther an AI that can actually play ?
19:08:33 <Alberth> sure, download one from the ingame content
19:09:12 <Kitty> which ones actually work? I have choochoo trains or sum such, that builds some track, goes bank rupt, and then gets bought out by me...
19:10:26 <Alberth> I don't know, I never bother about having an AI
19:13:57 <Kitty> hmm, that was a waste of 5 million. trying to build an oil rig...
19:15:08 <frosch123> Kitty: usually admiral, simpleai and nocab are good choices
19:17:18 <andythenorth> Zuu: Terkhen planetmaker Silicon Valley?
19:17:23 <andythenorth> we're on the nightly server
19:17:42 <Zuu> Nice, I'll come in a minute.
19:22:18 <Kitty> can anyone suggest good junction designs for 3 way junctions with 4 tracks ?
19:23:43 <Alberth> the wiki has a few, else look at openttdcoop.org
19:23:55 <Alberth> I have never build anything that complicated
19:28:55 *** kais58_ has joined #openttd
19:31:26 <Zuu> Kitty: I tend to build organically, and solve the problems as they come rather than ending up in a symetrical solution.
19:33:29 <Terkhen> andythenorth: I'll be away for the day
19:37:36 *** Alberth is now known as Guest6292
19:43:51 *** valhallasw has joined #openttd
19:46:10 <Rubidium> Eddi|zuHause: but I saw last week's episode and Pond's life
20:06:55 *** Alberth has joined #openttd
20:06:55 *** ChanServ sets mode: +o Alberth
21:20:18 *** Cybertinus has joined #openttd
21:22:49 *** Progman has joined #openttd
21:27:09 *** Cybertinus has joined #openttd
21:41:16 *** kais58_ has joined #openttd
22:00:20 *** Cybertinus has joined #openttd
22:14:35 *** andythenorth has left #openttd
22:21:19 <Kylie> so my road coal trucks - looks like it is time for them to go. they are only making $10,000 a year (24100 - 3630 - 10310 = ~ 10,000) while my trains thaat connect to almost the same coal stations are making $110,000 a year
22:34:53 *** FLHerne has joined #openttd
22:40:44 <Kitty> Did the thing to allow you to disable an airport make it into a mainstream release ?
22:44:09 <Supercheese> "Close airport" is in trunk...
22:44:27 <Kitty> what does that mean to a lay person ?
22:44:49 <Supercheese> As of the most recent nightly, the function "close airport" is available
22:44:58 <Supercheese> Meaning no planes will lane at that airport until reopened
22:46:03 <Supercheese> well, as of a while ago, but available in the most recent nightly
22:46:20 <planetmaker> Supercheese, oh, it is? ok :-)
22:46:30 <planetmaker> I guess I play too little with airports
22:46:43 <Supercheese> I'm running r24514 and it's there :)
22:46:51 <planetmaker> then it was one of the things from the last big assortment of small patches, which is also in 1.2 then
22:50:34 <FLHerne> I never understood why that wasn't added two years ago :P
22:55:09 <Supercheese> Hm, seems the discussion of adding more frames to 4x zoom level is getting a bit out of hand...
22:56:08 <Supercheese> It would be better for animation purposes, but seems untenable from a code perspective
22:56:58 <planetmaker> some people don't take "no" as an answer. And accuse you of being arrogant when giving "no" as an answer. Really sad
22:59:55 <FLHerne> Well, it is a valid problem...
23:09:57 *** Rhamphoryncus has joined #openttd
23:14:41 <planetmaker> FLHerne, yes, possibly it is. But that doesn't change the facts
23:15:21 <planetmaker> and the single sad fact is - again! - that personalities and animosities play the far bigger role than the real subject
23:17:06 <planetmaker> those people should select their own devs to develop the game for them which give them always pampered answers
23:18:00 *** dotwaffle has joined #openttd
23:18:00 *** tparker has joined #openttd
23:18:00 *** Belugas has joined #openttd
23:18:00 *** Rubidium has joined #openttd
23:18:00 *** graviton.oftc.net sets mode: +vov glx Belugas Belugas
23:18:03 <planetmaker> or a "yes, of course we will. And then show a screenshot. Of a patch which might heavily modified make it into OpenTTD in 10 years". Similar to certain NewGRF devs which actually complain about these kind of answers which are just brief
23:19:30 <planetmaker> and I should also just shut up and go to bed. Good night
23:29:06 <Kitty> is there an easy way to replace a group of trains ?
23:35:22 <Kitty> oh wow, that makes things easier
23:35:28 <Kitty> how long has that been in the feature list ?
23:35:49 <Kitty> Wish I had known about it
23:36:09 <Supercheese> It is indeed an immensely helpful feature
23:39:31 *** kais58_ has joined #openttd
continue to next day ⏵