IRC logs for #openttd on OFTC at 2009-03-07
⏴ go to previous day
00:10:23 <Yexo> Belugas: "I do not have neither the power neither the will to punish you." <- that's most likely not what you wanted to say
00:10:49 <Yexo> I think you're double negation it
00:11:01 <SpComb> "I have neither the power nor the will to ..."
00:12:20 *** Yeggstry is now known as Yeggzzz
00:28:19 <Eddi|zuHause> SpComb: anything specific causing this spontaneous outburst?
00:28:35 <SpComb> the double negations :)
00:29:05 <Eddi|zuHause> what do the germans have to do with those?
00:29:28 <NukeBuster> Belugas isn't German...
00:32:25 <SpComb> oh, I should assume less
00:33:16 *** Eddi|zuHause has joined #openttd
00:40:32 *** NukeBuster has left #openttd
00:53:07 *** stillunknown has joined #openttd
01:08:11 *** goodger has joined #openttd
01:35:42 *** goodger has joined #openttd
01:49:18 *** goodger has joined #openttd
02:10:08 <Belugas> i guess i should have better let that one out
02:10:34 <Belugas> or even better: i think i'll simpy leave out the forums's life for a very long time
02:10:44 <Belugas> it's way too depressing, these days
02:14:17 *** goodger has joined #openttd
02:48:44 * Belugas goes plaing some chords
02:52:52 *** kd5pbo1 has joined #openttd
03:12:25 *** TinoDidriksen has joined #openttd
03:42:41 *** TinoDidriksen has joined #openttd
04:03:04 *** KenjiE20|LT has joined #openttd
05:19:08 *** Singaporekid has joined #openttd
05:29:57 *** goodger has joined #openttd
05:45:19 *** lobster has joined #openttd
06:33:19 *** Alberth has joined #openttd
06:50:00 <el_en> Saturday, the seventh of March, two thousand and nine, Anno Domini.
06:51:09 *** maristo has joined #openttd
07:12:31 <Alberth> at least for some of us
08:10:31 *** bubersson has joined #openttd
08:57:24 *** Progman has joined #openttd
09:22:11 *** Cybertinus has joined #openttd
09:26:31 *** energetic has joined #openttd
09:36:01 *** energetic has left #openttd
09:36:24 *** energetic has joined #openttd
09:38:03 *** flowOver has joined #openttd
09:44:14 *** Meincfe has joined #openttd
09:45:18 <Meincfe> can ne 1 tell me how topacth Open TTD
09:52:10 *** Illegal_Alien has joined #openttd
09:58:32 <Eddi|zuHause> i have not understood a single word.
10:19:11 *** PeCeT_full has joined #openttd
10:25:23 *** stillunknown has joined #openttd
10:26:43 <Alberth> He expected people to react very quickly :)
10:27:20 <planetmaker> oh, and a wonderful good morning to all :)
10:33:35 <Splex> any big upgrades in openttd in the last couple months?
10:33:39 <Splex> and good morning to you
10:33:53 <Splex> or good evening...if you are in asia
10:35:23 <|Japa|> I'm in india, and it's 16:05
10:35:33 *** goodger has joined #openttd
10:37:00 <Eddi|zuHause> i guess in japan it's 18:36
10:37:26 <[wito]> more like 19:36, I think
10:39:06 * welshdragon has built a single track line. but isn't sure if pbs will break
10:39:41 <Eddi|zuHause> PBS usually works fine with single track
10:39:53 <Eddi|zuHause> as long as you put signals only on the double track sections
10:41:23 <welshdragon> there are 'dynamic' (passing) loops
10:42:04 <Eddi|zuHause> you can put a signal halfway through the single track section, that increases throughput when two trains want to go into the same direction, but reduces overall capacity, because it can block with 3 trains
10:42:23 <Alberth> Splex: we have 0.7-beta 1, we can now download goodies in-game (newgrfs, scenario's, save-games)
10:42:30 <Eddi|zuHause> blocking risk can be reduced with one way signals, but that again reduces throughput
10:44:30 <welshdragon> thanks Eddi|zuHause
10:48:40 <Splex> alberth: cool, hows the progress on the new graphics?
10:49:04 *** JapaMala has joined #openttd
10:49:58 <bubersson> Splex: OpenGFX is done in moderate climate ;)
10:50:55 <Darkvater> it's a pity devving is halted on that
10:51:59 *** JapaMala is now known as |Japa|
10:54:18 <Splex> opengfx is drop-in replacement?
10:54:24 <Splex> how about the 32bpp gfx?
10:54:48 <Darkvater> mwa I'm not too fond of 32bpp
10:55:09 <Darkvater> there is no coherent style, no coordination, i don't see it becoming anything in the near future
10:55:16 <Darkvater> especially the BIG-tile version
11:05:48 <|Japa|> I want an OTTD version of the cartoon set from simutrans
11:08:33 <Splex> would be nice if there was a 3d viewer... then all the objects/buildings could begin as simple 3d objects, people could create the upgrades over time.
11:08:43 <Splex> i don't see the reason for going to 32bbp... why not just go 3d
11:10:09 <Eddi|zuHause> i'm going away for probably about a month.
11:10:44 <Darkvater> don't bother coming back, ey
11:11:42 <Splex> actually that comic set looks pretty cool
11:12:07 <Splex> does simutrans aim to compete with openttd?
11:13:35 <Ammler> can't, it would have alreadyl lost ;-)
11:13:50 <Darkvater> both games have their respective strengths
11:16:48 <Darkvater> "My bees are in California pollinating almonds," he said. "In the middle of March they are going to be trucked all the way across the United States all the way back to Florida to pollinate oranges then they are trucked another thousand miles north to pollinate apples in Pennsylvania.
11:16:53 <Darkvater> that's fucking sick
11:18:40 *** |Jeroen| has joined #openttd
11:18:48 <[wito]> So who'll be making the newGRF with beehives, apple farms and orange farms? ;D
11:23:35 *** Vikthor has joined #openttd
11:24:08 <Splex> they talk about the bees having various ailments... did they ever consider that they need to STOP SPRAYING?
11:24:16 <Splex> like their deaths are natural... hah
11:24:44 <Darkvater> but that would mean less profit
11:25:15 <[wito]> also, animals DO do weird things. :P
11:28:25 <Splex> surely not as weird as the animals called humans.
11:28:47 <Splex> Darkvater.... yeah... less profit, they would rather make more profit at the expense of everyone else
11:29:11 <Splex> just like the mountaintop removal coal mining
11:30:05 <Splex> ""Robert F. Kennedy, Jr. said this about the Coal River Valley: I flew over these mountains and I saw what [the coal companies] were doing and if the American people could see what I saw there would be a revolution in this country""
11:34:28 <Alberth> Oh, how much we care about the environment.
11:40:24 *** lewymati has joined #openttd
11:41:17 <Splex> lol.... its funny when people say 'SAVE THE EARTH'.... the earth doesn't need saving... we are nothing but fleas on this earth... its up to us if we want to survive on this earth.
12:07:23 *** smeding has joined #openttd
12:08:30 *** Dred_furst has joined #openttd
12:09:20 *** DaleStan has joined #openttd
12:26:18 *** NukeBuster has joined #openttd
12:28:29 *** wollollo has joined #openttd
12:28:49 *** stillunknown has joined #openttd
12:34:08 <[wito]> anyone know of any newGRFs that adds late-1800s trains?
12:35:28 <[wito]> I know the Japanese train set goes back to around 1920ish, no?
12:36:07 <Ammler> the feature is a young OpenTTD only thing, so there aren't many sets, yet.
12:39:44 *** KenjiE20 has joined #openttd
12:39:44 *** Swallow has joined #openttd
12:41:08 *** Singaporekid has joined #openttd
12:44:34 <[wito]> 2cc only goes back to 1913, tho'
12:52:12 <[wito]> except monorail, which goes all the way back to 1880ish. :P
12:56:05 *** Phoenix_the_II has quit IRC
12:56:06 *** Phoenix_the_II has joined #openttd
12:57:29 <Ammler> [wito]: no monorail in 2cc
12:57:49 <[wito]> it's used for metro lines
12:58:13 <Ammler> which means, no monorail :P
12:58:25 <petern> i have a version of that that works properly ;)
13:01:12 <[wito]> so the tracs are supposed to look like metro tracks, I take it? :P
13:01:39 <Alberth> The C++ compilers nowadays give way too many warnings :P , now I have to make my patch a whole lot bigger.
13:01:46 <Ammler> or using monorail and metro the same time, I guess ;-)
13:03:23 <Alberth> I like the idea of a monorail running through the city streets and a tram driving 300+ km/h.
13:04:53 *** Brokkoli has joined #openttd
13:05:11 *** ChanServ sets mode: +v tokai
13:06:22 <[wito]> Alberth: 300km/h+ trams are simply uneconomic
13:06:28 <Alberth> I'd like to extend the WindowDesc data structure with two fields. What would be better, modifying all instances to a constructor call with optional additional fields, or adding "NULL, 0" to each of the instances?
13:06:31 <[wito]> the start-stopping makes it silly. :P
13:07:13 <Alberth> I have seen cities that cover a large part of the map. There it may be useful
13:09:23 *** HerzogDeXtEr has joined #openttd
13:12:01 *** Timitry has joined #openttd
13:17:11 <Alberth> you won't need any form of heating in the houses, the air friction of trams driving through the tunnels would generate enough heat probably :)
13:17:50 <Alberth> Blowing people off their feet at the stations :)
13:46:09 *** smeding has joined #openttd
13:58:37 *** bubersson has joined #openttd
14:08:55 *** pavel1269 has joined #openttd
14:12:47 *** Phoenix_the_II has quit IRC
14:14:37 *** maristo has joined #openttd
14:18:26 *** Zuu is now known as Guest1017
14:19:16 *** Guest1017 is now known as Zuu
14:25:15 *** stuffcorpse has joined #openttd
14:28:39 *** Phoenix_the_II has joined #openttd
14:29:13 <[wito]> are self-plopping industries tied to the date?
14:43:38 *** PhoenixII has joined #openttd
14:43:38 *** Phoenix_the_II has quit IRC
14:46:27 *** SHRIKEE has joined #openttd
14:51:29 *** frosch123 has joined #openttd
14:52:27 *** NightKhaos has joined #openttd
15:01:38 *** Hirundo has joined #openttd
15:03:49 *** Hirundo is now known as Swallow
15:04:41 *** bubersson has left #openttd
15:20:31 <Wolf01> the bananas' signup terms must be updated since now there are scenarios and heightmaps too
15:39:30 *** Ralphis has joined #openttd
15:53:25 *** Belugas has joined #openttd
15:53:25 *** ChanServ sets mode: +o Belugas
15:54:36 *** Phoenix_the_II has joined #openttd
15:55:30 *** DaleStan_ has joined #openttd
15:55:30 *** DaleStan is now known as Guest1034
15:55:30 *** DaleStan_ is now known as DaleStan
16:04:06 *** bubersson has joined #openttd
16:10:12 *** bubersson has left #openttd
16:11:04 *** PeCeT_full has joined #openttd
16:11:51 <[wito]> What is the £300 'Other' expense?
16:13:58 <Alberth> or rent for the HQ, we are not sure.
16:14:12 <[wito]> Well, I don't have a HQ yet, so can't be that. :P
16:14:26 <Rubidium> fees for the chamber of commerce ;)
16:14:30 <goodger> Alberth: you already paid to build the sodding HQ, you can't rent it ¬.¬
16:15:56 <Alberth> Maintenance costs for all the statues then?
16:16:09 <[wito]> Alberth: no statues. :P
16:16:47 <|Japa|> when I start a game and immediateley pay off my loan, there's still a monthly other payment
16:16:48 <goodger> well, the expense is being generated by code
16:17:00 <goodger> it must be possible to find out what it is
16:17:20 <Alberth> goodger: sure, it is just that nobody bothered to find out exactly.
16:17:23 <[wito]> it could be an adjustor to avoid passive companies
16:17:33 <[wito]> designed to do exactly that and nothing else
16:18:15 <Darkvater> |Japa|: don't you know it costs to borrow money? Even if you had it for one day?
16:18:22 <goodger> though it'd take a long time to prevent a passive company with a £300pa cost
16:18:23 <Darkvater> banks are no charity institutions mind you
16:18:41 <goodger> specifically, 334 years before slipping into the read
16:19:17 <|Japa|> well, a long as it stays that, I can still watch an AI game
16:19:47 <petern> isn't it £300 a month?
16:20:00 <Rubidium> banks are thieves, con artists and a legalised form of gambling
16:21:02 <Darkvater> hehe, frustrated? ;)
16:21:25 <[wito]> because in order to get your expenses as low as possible you'd have to repay all loan
16:21:35 <[wito]> so after the first year you'd be at least -300
16:22:07 <Rubidium> Darkvater: no, I just know too much about how banks actually work
16:22:31 <Darkvater> Rubidium: exactly like you described :)
16:23:00 <Darkvater> it's incredible that you can make so much profits, even percentually for doing nothing more than pushing OTHER people's money around
16:23:07 <Darkvater> something's really fucked up in the world
16:23:18 <Rubidium> like did you know that if you transfer money from one savings account to another that even though it seems to be on the other savings account, withdrawing it the same day will cause debit interest?
16:23:51 <Darkvater> yes...I noticed. I was furious by the way (it was within the same bank even!)
16:24:01 <Darkvater> I called them, gave'em hell and got the money back :)
16:24:10 <Rubidium> for the interest money gets basically withdrawn at 0:00 and money gets added 23:59, i.e. you lose a day of interest
16:25:41 <db48x> banks earn money by making loans to businesses
16:25:50 <db48x> you could do that directly, if you wanted
16:26:28 <Darkvater> db48x: hehe, funny. If they really just did that, there would be no worldwide crisis right now
16:27:33 <Rubidium> did you also know that if I open a bank account and put 1 million on it that the bank then may (and will!) lend out several millions? And if they 'rinse' and 'repeat' that they lend out several tens of millions for that single million that is actually in the bank?
16:28:30 <Rubidium> which is why they get into HUGE problems when everybody withdraws a few percent of their savings OR when a few percent can't repay their loans
16:28:34 <db48x> they can only do that by borrowing the extra from somewhere else thoguh
16:28:48 <db48x> they don't magically create the money
16:29:18 <Darkvater> ah, the innocent mind ;), what joyful times those were
16:29:42 <Darkvater> hmm, question: anyone know of an easy, simple and fast rsync program for windows?
16:30:16 <Darkvater> eh, I don't want ubuntu on windows :)
16:30:17 <Alberth> What bothers me big time is that a simple money transfer from Holland to the UK costs 20 euro's just for the transfer.
16:30:37 <Darkvater> the UK should just introduce the EURO
16:31:03 <stillunknown> Alberth: transaction costs?
16:31:10 <Darkvater> hmm, deltacopy looks interesting
16:33:48 <stillunknown> Alberth: i thought money transfers within the EEG were free of transaction costs if done proplerly?
16:33:58 <Darkvater> stillunknown: only euro countries
16:34:46 <frosch123> did you know that lehmann paid boni of average $600 000 per staff in 2007 :)
16:34:48 <goodger> yeah, I was once charged £10 to receive a transfer of £26.99 from Germany
16:35:18 <Darkvater> frosch123: did you know that Morgan Stanley paid a few billion in bonuses with government bailout money?
16:35:52 <Rubidium> db48x: you should watch "Money as debt"
16:36:22 <Alberth> I don't know how much money owen gets from main-land Europe, but it may be useful to setup a bank account at the other side of the channel just to avoid these costs.
16:36:24 <frosch123> Darkvater: yes, but those were for the past.
16:36:29 <stillunknown> Darkvater: well, my online banking system says otherwise, the UK is part of the European Economic Space (or whatever it's called)
16:36:54 <Darkvater> yes but with public money paid to prevent MS from going bankrupt? that's just wrong, really, really wrong
16:37:16 <Alberth> stillunknown: do a test, give owen some money, and see how much the transaction costs
16:37:29 <orudge> I don't know what the transaction costs are on the other side
16:37:37 <Darkvater> I usually pay through paypal though...it never cost me a thing
16:37:39 <orudge> my bank tends to charge me £1 to receive from Europe these day
16:37:43 <orudge> (I'm sure they didn't used to, but oh well)
16:37:56 <orudge> as for PayPal, I get charged fees to receive money from anywhere
16:38:09 <orudge> with the separation of my business from my personal stuff though, I will likely get a personal PayPal account, which will be fee-free
16:38:13 <orudge> which would be better for OpenTTD, etc
16:38:16 <Alberth> orudge: would that mean it is cheaper to have you pay the transaction costs?
16:38:33 <orudge> I get charged the £1 anyway by my bank
16:39:56 <orudge> [16:30:15] <Alberth> What bothers me big time is that a simple money transfer from Holland to the UK costs 20 euro's just for the transfer. <-- €20, really?
16:39:59 <orudge> that is a rather silly amount
16:40:14 <orudge> I think most people who transfer money to me from Europe pay very little fees, if anything
16:40:17 <orudge> as in a couple of euros at most
16:40:39 <Alberth> yeah, I thought so too. 15 euro fixed cost, and some small percentage with a minimum of 5 euro :(
16:40:51 <stillunknown> I think it can be fee-less, if you have an IBAN and BIC number.
16:40:54 <orudge> I think it costs me £20 or something to send money to Australia
16:41:02 <orudge> but not to Europe, heh
16:41:05 <orudge> I do of course have an IBAN and BIC
16:41:08 * goodger pays about fifteen quid a time to pay his european royalty-payees
16:41:12 <Alberth> stillunknown: no, i used those
16:41:33 <goodger> fortunately they are contracted to have said amount deducted from said royalties, but nonetheless
16:42:24 <stillunknown> Alberth: which bank?
16:42:59 <Alberth> stillunknown: Postbank, ING nowadays
16:43:24 <stillunknown> Alberth: check the online system, it clearly says there is no fee if done properly
16:43:37 *** Hirundo has joined #openttd
16:44:02 *** Hirundo is now known as Swallow
17:04:15 <frosch123> yeah, usually I type "=" instead of "==". Now I managed to do it the other way around \o/
17:13:15 *** Mortomes has joined #openttd
17:14:20 *** Timitry has joined #openttd
17:18:29 <Zuu> In delphi you have := instead of = for assignments and = instead of == for comparsion. That I handle quite fine. But VB(A) is different, there you need to use = for both cases, which comfuses me sometimes :)
17:23:01 <goodger> Zuu: that's because it's a bastardised BASIC, and BASIC was designed to be simple like that
17:31:08 <Alberth> (local copy of clean trunk)
17:31:38 *** Dred_furst has joined #openttd
17:35:06 <frosch123> cool attribute, never heard of it :)
17:36:00 <SmatZ> frosch123: Alberth __attribute__((warn_unused_result)) or so
17:36:16 <SmatZ> I was thinking about use it for MallocT-like functions...
17:36:24 <frosch123> SmatZ: that is also written in the paste :)
17:37:15 <frosch123> who would be so stupid to not use the result of malloc?
17:37:42 <SmatZ> frosch123: yeah, that was the reason why I didn't apply that ;)
17:37:49 <SmatZ> though, someone could misuse ReallocT...
17:43:20 *** lewymati has joined #openttd
18:01:42 *** Swallow has joined #openttd
18:02:05 <SmatZ> Alberth: update to most recent revision
18:03:04 <Rubidium> 1) don't treat system responses with asserts because in most cases bad system responses happen when asserts are disabled (as per Murphy's law)
18:03:23 <Alberth> ah, much nicer indeed.
18:04:09 <Rubidium> 2) someone might intentionally create a file where the return value would be 0 thus asserting unneededly for lots of people when it gets distributed via bananas or the forum
18:08:07 <CIA-1> OpenTTD: rubidium * r15634 /trunk/src/fios.cpp: -Fix (r15632): some compilers seem to warn about the return value of fscanf.
18:08:38 <el_en> 2.5) assert is for catching conditions that are supposed to be impossible, things that should never happen. an input file containing too few integers *is not* an impossible situation, it *can* happen easily.
18:09:33 <Alberth> right, less assert(), more 'return failed' :)
18:10:25 *** Dred_furst has joined #openttd
18:12:09 *** welshdragon has joined #openttd
18:12:45 <Darkvater> el_en: well, we've used this idiom for a very long time; basically: "do not trust the user"
18:13:06 <Darkvater> eg, in the beginning all CMD_ commands were checked using asserts :O
18:13:15 <Darkvater> we used to believe in the good of man
18:13:37 <Alberth> that didn't last long, I suppose :)
18:21:43 <|Japa|> #1 in the list of people I don't trust: me
18:22:14 *** goodger has joined #openttd
18:23:00 <CIA-1> OpenTTD: rubidium * r15635 /trunk/src/ (roadveh_cmd.cpp station_base.h): -Fix [FS#2707]: don't allow more than 64 road vehicles to reserve a slot at a single road stop. 255 + 1 gives trouble, but 64 is even more than the roadstop would be able to handle within the slot timeout time.
18:50:42 *** Brianetta has joined #openttd
18:58:49 <CIA-1> OpenTTD: translators * r15636 /trunk/src/lang/ (korean.txt unfinished/greek.txt):
18:58:49 <CIA-1> OpenTTD: -Update: WebTranslator2 update to 2009-03-07 18:58:32
18:58:49 <CIA-1> OpenTTD: greek - 211 fixed, 4 changed by hellis (215)
18:58:49 <CIA-1> OpenTTD: korean - 2 changed by dlunch (2)
19:06:47 <Alberth> (or rather, the let the games begin)
19:12:58 *** smeding has joined #openttd
19:23:23 *** energetic is now known as energetic|afk
19:38:30 *** wollollo has joined #openttd
19:38:58 <frosch123> trying to sync a pre c++ patch?
19:39:21 <Sacro> if you open an old PBS game will it convert it to YAPP?
19:40:36 <frosch123> the readme says, old savegame should be send to sacro, he would convert them by hand
19:44:35 <SmatZ> Sacro: the old-PBS design is very different from YAPP
19:45:44 <Sacro> it should transfer over fine
19:45:59 <Sacro> you can replace YAPP quite easily i'm sure
19:46:22 <SmatZ> how would you like to convert bidir PBS?
19:49:00 <Sacro> i don't recall there ever being bidir PBS
19:52:28 <petern> i am so glad i got it removed ;p
19:53:48 <Sacro> so how did bidir pbs work...
19:54:35 *** energetic|afk is now known as energetic
20:07:20 <[wito]> any way to upgrade town-built road bridges with tramway?
20:14:35 <frosch123> use the usual tram-build tool and click on the bridgehead
20:14:59 <frosch123> same for tunnels and competitor's stuff
20:20:11 <[wito]> dunno why it didn't work before
20:25:40 *** Dred_furst has joined #openttd
20:40:37 *** PeCeT_full has joined #openttd
20:53:46 <CIA-1> OpenTTD: truebrain * r15637 /trunk/src/gfx.cpp:
20:53:46 <CIA-1> OpenTTD: -Fix: when you try to protect something from doing something it shouldn't be
20:53:46 <CIA-1> OpenTTD: doing, make sure that the actions happening after that, doesn't start doing what
20:53:46 <CIA-1> OpenTTD: it shouldn't be doing. This of course would only happen in rare corner cases.
20:55:36 <CIA-1> OpenTTD: truebrain * r15638 /trunk/src/ai/ai_config.cpp: -Fix: there is a time in your life you should stop coding for a project. You know that time is there when you make two huge mistakes in just 10 lines of code. I wish you all the best of luck :) (ps: HUGE mistakes!)
20:58:19 <frosch123> comparing unsigned stuff with zero is also my most favorite mistake :)
21:09:33 *** smeding has joined #openttd
21:18:17 *** Maarten is now known as Guest1085
21:31:57 *** stillunknown has joined #openttd
21:36:25 <el_en> seven of nine acting as the doctor
21:36:26 <CIA-1> OpenTTD: yexo * r15639 /trunk/src/ (lang/english.txt town_cmd.cpp): -Change: When you disable "Towns are allowed to build roads", pressing "Expand" in the town window in the scenario editor will no longer build roads.
21:38:08 *** maristo has joined #openttd
21:41:28 <el_en> succeeding well in doing so.
21:50:38 *** roboboy has joined #openttd
21:58:07 <Rubidium> petern: have you seen FS#2708?
22:03:32 *** Belugas has joined #openttd
22:03:32 *** ChanServ sets mode: +o Belugas
22:07:28 <Aali> any progress on the raging memleak issue?
22:07:49 <frosch123> IIRC noone could reproduce it
22:10:12 <Rubidium> the major problem is that it's hard to measure 'leaking' against normal building up caches, especially in the first few minutes
22:10:41 <Aali> on fastforward I can see the memory usage going up by the second
22:11:03 <Aali> we'll see if it slows down
22:11:43 <Aali> this is a windows machine, maybe that has something to do with it?
22:12:13 <Rubidium> that's something I've been saying for weeks
22:13:18 <Aali> if I keep it minimized, it keeps increasing forever
22:13:53 <Aali> open the window, then minimize it again, and mem usage drops by 1-5MB
22:14:03 <Sacro> that's quite an impossible thing to test
22:14:33 <frosch123> Aali: then try to locate the problem. alternate dedicated server/client, all news/no news, different video/sound drivers (esp. the 'null' ones), ...
22:40:59 <CIA-1> OpenTTD: smatz * r15640 /trunk/src/network/network.cpp: -Fix [FS#2712]: game crashes when network pools are empty, so always allocate at least one pool block
22:41:21 *** PeCeT_full has joined #openttd
22:41:25 <heffer> anyone here for some quick debugging?
22:41:56 <heffer> i'll try to get the compile error in english
22:42:04 <PeCeT_full> hey, who has the newest (I mean 0.6.0-0.7.0 versions) version of OpenTTD for PSP?
22:42:48 <Rubidium> last I know of is 0.5-ish
22:42:57 *** stuffcorpse has joined #openttd
22:42:58 <heffer> fileio.cpp: In function 'void ChangeWorkingDirectory(const char*)':
22:43:06 <heffer> fileio.cpp:810: error: invalid conversion from 'const char*' to 'char*'
22:43:47 <glx> Aali: leaks detection on windows is quite hard
22:44:08 <SmatZ> heffer: update to trunk
22:44:08 <Rubidium> heffer: what're you using? compiler, os, etc?
22:44:33 <heffer> Fedora rawhide (to become Fedora 11) using gcc 4.4 with trunk from a few minutes ago
22:45:02 <heffer> gcc-c++-4.4.0-0.22.x86_64
22:45:51 <SmatZ> heffer: ottd is regularly tested for compilation under variaous versions of gcc ranging from gcc 2.95 to gcc4.4
22:47:58 <heffer> well for me it does not seem to work
22:48:21 <SmatZ> [23:44:17] <SmatZ> heffer: update to trunk
22:48:22 <SmatZ> [23:44:17] <Rubidium> heffer: what're you using? compiler, os, etc?
22:48:24 <SmatZ> [23:44:17] <glx> heffer:revision?
22:48:51 <Rubidium> so what mutilations does fedora do to gccc
22:49:30 <Rubidium> oh... and fedora core 4 sucks
22:49:32 <heffer> well we switched to gcc 4.4 about a week ago and since in fedora it is set to be strict it revealed some bugs
22:49:36 <glx> heffer: "from a few minutes ago" is not a revision ;)
22:50:29 <Rubidium> especially because a few could (with a not-so-liberal view) mean anything between 0 and 5 minutes
22:50:32 <heffer> well if nothing in fileio.cpp changed within the last 20 minutes my revision could be assumed as being 15640
22:50:58 <heffer> but if it helps i could check out the newest trunk right now and try again
22:52:29 <heffer> Rubidium, if you hate FC4 you should try F11 when it's out :P
22:52:49 <heffer> i'm sure it can break your system, too
22:53:06 <Rubidium> that doesn't have the same compiler etc. as FC4
22:53:21 <Rubidium> which is kinda needed to debug FS#2703
22:53:51 *** stuffcor1se has joined #openttd
22:56:17 <Rubidium> Patch0: gcc44-hack.patch
22:56:41 <Rubidium> that sounds very promising when it's in gcc.spec
22:59:31 <heffer> should i file a bug or is this something that can be solved quickly?
23:01:31 <SmatZ> I wonder where is defined strrchr should return "const char *"
23:02:39 <SmatZ> so there are different definitions for C-like strrchr and C++-like strrchr?
23:02:45 <CIA-1> OpenTTD: rubidium * r15641 /trunk/src/fileio.cpp: -Fix: some gcc compilers seem to be chosing the C++ standard for strrchr, others the C standard...
23:03:39 *** Sionide has joined #openttd
23:06:08 <SmatZ> const char * strstr ( const char * s1 , const char * s2 );
23:06:10 <SmatZ> char * strstr ( char * s1 , const char * s2 );
23:06:36 <Rubidium> char *strstr(const char *s1, const char *s2) <- in C
23:07:04 <SmatZ> char *strrchr(const char *s, int c);
23:07:15 <SmatZ> pasted the wrong function, sorry :)
23:07:38 <SmatZ> but it seems if strrchr(char*,int) was used, it would return char* too
23:09:03 <heffer> ok. i'm off then. thank you for your help. it's much appreciated
23:09:06 <SmatZ> hmm, filename is char*...
23:09:13 <SmatZ> I am failing to understand this behaviour then
23:10:23 <SmatZ> indeed it seems to work as expected
23:12:02 * Rubidium is going to break FC4 to the point that it doesn't openttd to crash in it's own libs
23:12:18 <Rubidium> or that the fracking thing does want to find out what packages it can update
23:18:50 <energetic> `how would anyone suggest implementing a goalserver?
23:19:32 <energetic> simply patch ottd? or...?
23:22:17 <Aali> the minimize/maximize trick to kill the memleak works with the console window for a dedicated server
23:24:41 *** stuffcor1se has joined #openttd
23:25:34 <Rubidium> so what does minimize/maximize of the console window change?
23:27:53 <SmatZ> so even minimising+maximising console window in Windows fixes the memory problem?
23:27:53 <Rubidium> or rather, what does it trigger in OpenTTD? glx do you got a clue 'bout that?
23:28:11 <Aali> it doesn't "fix" the memory problem
23:28:17 <SmatZ> dihedral's nighly server has the same problem
23:28:31 <Aali> but memory usage drops to "normal" levels
23:28:37 <Aali> and then starts building up again
23:28:45 <Rubidium> SmatZ: that's his VPS that's acting up!
23:28:47 <glx> SmatZ: hmm he can use valgrind
23:29:04 <Aali> it seems the longer you wait, the more it drops
23:29:30 <Aali> if you let it go up to like 50MB, you can drop it to 15 in one go, but around 20-24 nothing happens
23:30:58 <Aali> actually, this seems really random
23:31:57 <petern> i'm fairly sure minimizing putty has absolutely no affect on openttd within screen...
23:32:44 <Aali> this is obviously some stupid windows thing
23:32:55 <Aali> the question is what openttd is doing "wrong"
23:33:28 <Aali> but the window thing "fixing" it is
23:34:20 <SmatZ> Aali: maximising/minimising helps? and what about just giving focus to that window?
23:35:49 <Aali> hmm, can you fastforward with -v null?
23:36:13 <SmatZ> Aali: it ffs automatically iirc
23:36:42 <Aali> oh, -v null doesn't even work
23:36:47 <SmatZ> and stops after 1000 ticks I think
23:37:07 <Aali> right, thats what I'm seeing
23:37:25 <SmatZ> it's only useful for benchmarking
23:38:00 <SmatZ> but if you mass something like -v null:ticks=1000000
23:38:05 <SmatZ> it will run for 1000000 ticks
23:39:52 <Aali> .. and now I'm getting desyncs
23:40:54 <Aali> dedicated server running the save from FS#2706, minus the AIs, connect with a GUI client
23:41:26 <glx> real desync or lost connection?
23:41:29 <SmatZ> yeah, server runs much faster than client...
23:41:53 <Aali> server is not using -v null
23:42:05 <SmatZ> I think you get "desync" if server sends two sync frames during client's connecting procedure
23:46:44 <Aali> so, anyway, made some progress, the intro screen does not leak as much as that savegame :P
23:47:47 <SmatZ> Aali: does that happen only in multiplayer?
23:48:00 <Aali> leaks in single player all the same
23:48:24 <SmatZ> so SP leaks, but intro doesn't? ...
23:48:31 <SmatZ> and furthermore, those are not leaks
23:48:46 <SmatZ> but something like waiting events or whatever
23:48:50 <Aali> the savegame from 2706 leaks
23:49:04 <SmatZ> because it gets freed when you deminimise the window...
23:49:57 <Aali> is noai/squirrel throwing around a bunch of exceptions or something?
23:52:16 <glx> yes, suspend/resume is done with exceptions
23:52:46 <Aali> well, no AI's (no pun intended) are running, so that shouldn't happen, right?
23:53:25 <Aali> intro does not leak even when run as a standard SP game
23:54:13 <SmatZ> AI events are created even when no AIs are running... but are not dispatched to anyone :)
23:54:33 <SmatZ> can in theory it has something to do with Squirrel leaking?
23:54:49 <SmatZ> is GC called even when no AIs are running?
23:55:14 <Aali> let me start some AI's on the intro game, see what happens
23:55:22 <glx> squirel does nothing if there are no AIs
23:56:22 <Zahl> Aali: just read your flyspray comment, minimizing does the trick here too
23:56:55 <glx> hmm minimizing means no drawing
23:57:10 <Aali> minimizing doesn't stop the leak
23:57:30 <SmatZ> Aali: does minimising start the leak?
23:58:00 <Zahl> isn't it that windows pages out stuff from applications if you minimize them? so it might be the "dead" memory just goes to the pagefile and won't be taken into account by the taskmanager
23:58:14 <SmatZ> so it starts "leaking" while playing, but when you minimise+unminimise it, it releases memory
23:58:16 <Yexo> <@SmatZ> is GC called even when no AIs are running? <- a seperate vm is created per AI that is running, so the GC can't even run without an AI
23:58:53 <Zahl> or is the memory really _freed_ when you minimize
23:59:43 <Zahl> because if it is, it might be a gdi issue
23:59:56 <Aali> it happens to dedicated servers
continue to next day ⏵
