IRC logs for #openttd on OFTC at 2011-10-25
⏴ go to previous day
00:47:58 <Eddi|zuHause> kernel: [3308222.895852] ata1.00: device reported invalid CHS sector 0 <-- err, do i have to worry?
00:53:17 *** michi_cc has joined #openttd
00:53:18 *** ChanServ sets mode: +v michi_cc
01:31:58 <z-MaTRiX> Eddi|zuHause<< well you could do a full test using smartctl
01:32:21 <z-MaTRiX> having some similar things with a 3 year old samsung 400GB PATA
01:32:30 <Eddi|zuHause> DMA: mdma0 mdma1 *mdma2 udma0 udma1 udma2 udma3 udma4 udma5 <-- that doesn't look right either
01:32:39 <z-MaTRiX> drivenotready seekcomplete errors
01:32:59 <z-MaTRiX> do you have ata133 cable?
01:33:13 <z-MaTRiX> old 40wire cables doesnt support UDMA
01:33:41 <Eddi|zuHause> it's an 80 wire cable
01:33:51 <Eddi|zuHause> and it used to do UDMA
01:34:01 <Eddi|zuHause> up to the above error
01:34:06 <Eddi|zuHause> which was apparently 3 days ago
01:34:22 <Eddi|zuHause> no, it's just slower than usual
01:34:43 <z-MaTRiX> smart may tell whats goung on
01:34:45 <Eddi|zuHause> at least from what i can tell
01:35:27 <Eddi|zuHause> smartctl not saying anything interesting
01:35:58 <z-MaTRiX> smartctl -a /dev/sda
01:36:49 <z-MaTRiX> reallocated sector cound?
01:36:52 <z-MaTRiX> Hardware_ECC_Recovered ?
01:37:02 <Eddi|zuHause> 5 Reallocated_Sector_Ct 0x0033 199 199 140 Pre-fail Always - 1
01:37:23 <z-MaTRiX> that means badsector
01:37:30 <Eddi|zuHause> doesn't have Hardware_ECC_Recovered
01:37:48 <Eddi|zuHause> "1" is hardly an alarming figure
01:38:04 <Eddi|zuHause> i mean, the drive is really old
01:38:13 <Eddi|zuHause> something 8 years-ish
01:38:22 <z-MaTRiX> well doesn mean much, my samsung seems to hide every reallocation
01:38:45 <z-MaTRiX> only using for temp storage for videos with md5sum
01:38:55 <Eddi|zuHause> 9 Power_On_Hours 0x0032 036 036 000 Old_age Always - 47436
01:39:11 <z-MaTRiX> Power_On_Hours 0x0032 100 100 000 Old_age Always - 32068
01:39:31 <Eddi|zuHause> @calc 47436/24/365
01:39:31 <DorpsGek> Eddi|zuHause: 5.41506849315
01:39:33 <z-MaTRiX> Hardware_ECC_Recovered 0x001a 100 100 000 Old_age Always - 4373817
01:39:46 <z-MaTRiX> and this means something noncool too
01:39:55 <Eddi|zuHause> 100 means everything is alright
01:40:08 <z-MaTRiX> UDMA_CRC_Error_Count 0x003e 200 200 000 Old_age Always - 2
01:40:21 <z-MaTRiX> well it may be allright for samsung but not me
01:40:58 <z-MaTRiX> has some things like this in report
01:40:59 <z-MaTRiX> 84 51 78 47 2b 40 e7 Error: ICRC, ABRT 120 sectors at LBA = 0x07402b47 = 121645895
01:41:29 <z-MaTRiX> Extended offline Completed: read failure 90% 14303 30159187
01:41:44 <z-MaTRiX> but its magically gone...
01:52:48 <z-MaTRiX> anyway you should get hitachi if you want quality
01:53:09 <z-MaTRiX> they have new superior technology since 2010
01:55:50 *** Biolunar has joined #openttd
01:57:08 <z-MaTRiX> dont know if a power out can cause this during a read
01:57:10 <z-MaTRiX> 40 51 0c 11 60 c4 e9 Error: UNC 12 sectors at LBA = 0x09c46011 = 163864593
02:26:07 *** rhaeder has joined #openttd
02:44:12 *** Eddi|zuHause2 has joined #openttd
04:27:36 <DDR_> Hi. So, I just bought 1100 ships in a multiplayer game, and now my copy of OpenTTD won't run fast enough to keep the network connection up. Anyone got any tips?
04:27:54 <DDR_> That doesn't rhyme with Shell the Clips'?
04:29:54 <DDR_> It's like, the dev version of OpenTTD comes with rivers, and they looked so empty... :P
04:32:59 <Mazur> THey do take up a shitload of processor time.
04:33:17 <Mazur> THew ships, rather than the rivers.
04:56:27 *** Eddi|zuHause2 has joined #openttd
05:02:42 <FFLaguna> Mazur - I'm running the client, there's nothing excessive going on on my end
05:03:05 <FFLaguna> 60 trains, 5 road vehicles, and 1146 ships :P
05:03:56 *** George|2 has joined #openttd
05:03:56 *** George is now known as Guest14659
05:03:57 *** George|2 is now known as George
05:49:56 *** JVassie has joined #openttd
05:53:22 *** JVassie_ has joined #openttd
06:07:28 *** Cybertinus has joined #openttd
06:14:27 *** sla_ro|master has joined #openttd
06:17:07 <TyrHeimdall> DDR_: I read that using bouys helps a lot, it reduces the ammount of possible routes a lot
06:18:23 <TyrHeimdall> I have another problem though; OpenTTD won't run over windows remote desktop :P It just crashes on startup
06:20:01 <planetmaker> DDR_: ship performance has been greatly improved in trunk. Mind that you then change the pathfinder to (the now default) yapf (away from npf) if you just update your openttd.
06:20:09 <planetmaker> still.... 1k ships is A LOT
06:20:41 <TyrHeimdall> would also like to se the possibility to close airports
06:20:53 <TyrHeimdall> its a lot of work to change from one type to the next :/
06:48:11 *** Celestar has joined #openttd
06:48:24 *** TWerkhoven has joined #openttd
06:49:04 * Celestar goes getting openttd again
06:51:33 *** andythenorth has joined #openttd
06:53:26 <Celestar> home partition full O_o
06:59:07 *** Celestar has joined #openttd
07:05:39 *** Brianetta has joined #openttd
07:07:33 *** andythenorth is now known as Guest14672
07:07:33 *** andythenorth_ has joined #openttd
07:07:33 *** andythenorth_ is now known as andythenorth
07:20:53 *** Progman has joined #openttd
07:21:12 *** valhallasw has joined #openttd
07:48:27 <planetmaker> hola andythenorth
07:49:15 <blathijs> Celestar: OpenTTD has gotten awesome nowadays!
07:49:39 <blathijs> (I recently spent a day "testing" my Debian build... ;-)
07:55:00 <planetmaker> Rb yesterday compiled a changelog since 1.1 branch... At least half of the lines had a "[NewGRF]" in it :-)
07:59:04 *** DayDreamer has joined #openttd
08:00:06 <Ammler> is there something else than rivers?
08:00:39 <peter1138> OpenTTD is awesome now? Wasn't it? :S
08:13:35 <Ammler> planetmaker: which say 3 would you mention for users?
08:15:22 *** andythenorth has joined #openttd
08:15:36 <Ammler> IMO, 1.2 will become the most boring upgrade :-)
08:17:22 <Sacro> now we need DarkVator back
08:17:33 * andythenorth votes for DaleStan
08:21:46 <Celestar> what'S the status on cargodist? :P
08:23:51 * Sacro tries to recall what Celestar was last doing
08:24:45 <Ammler> Celestar: you missed yacd, but that seems stalled too
08:25:05 <planetmaker> Ammler, I'd mention each
08:25:12 <TyrHeimdall> oh btw; when trying to share orders between vehicles, why do I some times get an error saying its not possible? (don't remember the exact message now)
08:25:49 <planetmaker> And I'm working on adding "Feature: Transparency setting for grid lines"
08:26:13 <planetmaker> it's 90% finished ;-)
08:26:39 <planetmaker> it needs a bit more quaking :-P
08:40:51 *** LordAro has joined #openttd
08:43:02 <LordAro> phew, that wiki work was fun :)
08:48:57 <Celestar> Ammler: so cargodist died and yacd died?
08:49:21 <Ammler> well, I used "stalled", not died :-P
08:50:07 <Ammler> rather yacd killed cargodist but is stalled now
08:50:48 <LordAro> Celestar: you probably won't know me, but i do a lot of delving into OTTD history :)
08:52:26 <Ammler> or maybe just "hidden" developing like we know from michi
08:54:57 <andythenorth> but stuck on the CPU use issue
08:55:36 <peter1138> but then i didn't play big games. hmm
08:55:58 *** Eddi|zuHause2 is now known as Eddi|zuHause
08:56:20 <DDR_> LordAro: Got any interesting historical tidbits?
08:57:25 <LordAro> well, i found the forum thread about the server 'loss' last night
08:57:36 <LordAro> and the versions went from 0.4.0.1 -> 0.4.5
08:57:44 <LordAro> which is a little odd
08:59:36 <LordAro> can you explain that, o ancient one? ;P
08:59:41 <Celestar> I still cannot believe we lost a whole repo :P
09:00:02 <DDR_> Explanation: Hard drugs.
09:00:12 <LordAro> well, according to the thread, it got hacked, and was sending out spam
09:00:55 <Celestar> I think it was 970ish revisions
09:01:11 <LordAro> first commit say 974/5 methinks
09:01:21 <planetmaker> that's way before 0.4.x
09:01:28 <planetmaker> iirc around 0.2 / 0.1 or so
09:01:40 <Celestar> I have some VERY old code somewhere
09:01:57 <LordAro> oh, and tb 'hid' commit 10k - so no one got the 'honour' :)
09:01:58 <planetmaker> Probably my oldest code is around your newest code :-P
09:02:19 <planetmaker> commit 20k was fun :-)
09:02:39 <planetmaker> I guess half the dev team watched :-)
09:03:24 <planetmaker> the r20k party was quite fun :-)
09:03:53 <DDR_> Not being a dev and all. :P
09:04:17 <planetmaker> everyone was invited :-) you just didn't pay attention
09:04:27 <Celestar> I think 0.4.0.1 went to 0.4.5 because of Great New Feature (TM)
09:04:41 <LordAro> i didn't see anything in the changelog...
09:04:59 <planetmaker> LordAro, it's a release tag of 1.0.x
09:05:14 <planetmaker> i.e. found in the 1.0 branch
09:05:38 <Celestar> LordAro: Vehicle Cloning :P
09:06:16 <DDR_> OH SWEET MERCY that was worth the bump in version number.
09:06:31 <DDR_> Said the guy who just ended a game with 1100 ships.
09:06:33 <Eddi|zuHause> i don't remember the reasoning behind 0.4.5, but certainly "whoa, glad we finally got rid of these buggy PBS" was involved
09:06:38 <Celestar> it took a while to get it fixed, right?
09:06:42 *** Zeknurn has joined #openttd
09:06:56 <Celestar> the CPU manages that? :P
09:07:07 <LordAro> ships use YAPF noaw ;)
09:07:11 <DDR_> No. That's why I ended it. :P
09:07:20 <DDR_> Well, technically, the server's still going.
09:07:31 <DDR_> it's just not *my* cpu which can handle it.
09:08:10 * LordAro should do a wiki page: Notable events in OTTD History
09:08:23 <DDR_> The rivers in the dev version make ships so much more accessible. :)
09:08:47 <planetmaker> LordAro, the road map?
09:08:53 <LordAro> shhh! don't mention rivers to andy!
09:08:57 <DDR_> Though I really could stand for a level dock. Building an inner-city dock takes up a good 9 or 10 tiles. :(
09:09:06 <planetmaker> LordAro, but... he drew them!
09:09:27 <LordAro> he did complain a bit :)
09:09:58 <planetmaker> luckily in an amiable fashion.
09:10:07 <planetmaker> And he works on overcoming his complaints code-wise :-)
09:10:19 <DDR_> He's not happy with rivers?
09:10:37 <planetmaker> DDR_, you need locks for every river slope to have ships pass there
09:10:43 <planetmaker> even at sea level to level one.
09:10:49 <DDR_> They are rather SC2K in their logic... *shrug*
09:10:51 <planetmaker> Which can be considered a bit unfortunate
09:11:08 <LordAro> [10:08:47] <planetmaker> LordAro, the road map? <-- no thiings that _have_ happened, e.g. repo loss, r10k, etc
09:11:29 <planetmaker> I wonder whether there isn't somewhere a history page.
09:11:33 <planetmaker> But sure, go right ahead
09:11:58 <DDR_> I just routed 700 ships through a river and canal system. They're awesome.
09:12:27 <DDR_> If you're fixing them up, I'm looking forward to the improvements. :)
09:12:31 <peter1138> ah "just make it" ...
09:13:50 <LordAro> planetmaker: there is a "release history" page, which i'll probably link to
09:15:01 <LordAro> like i said, "i do a lot of delving into OTTD history :)"
09:15:13 <planetmaker> seems quite like :-)
09:16:24 <LordAro> when i get _really_ bored, i might do a similar "history" page for tt-forums :)
09:16:45 <planetmaker> that's WAY more complex, I guess
09:17:16 <planetmaker> though it makes one wonder: what are the noticable 'events' for a forum?
09:17:58 <planetmaker> flame wars? Addition of <whatever> section? Mod appoint / leave. Ban of <whomever>?
09:18:03 <LordAro> well, the 'events' that mostly get talked about
09:18:26 <planetmaker> bah. off-topic. I don't read it ;-)
09:18:37 <LordAro> r's departure thread, for instance
09:18:56 <Eddi|zuHause> never looked into that forum either
09:19:10 <peter1138> who cares about someone leaving
09:19:45 <planetmaker> it's mostly a "I leave because you all suck" - and in reality just craving for attention
09:20:01 <LordAro> it wasn't quite like that though
09:20:02 <DDR_> Hm, even mod appointments are just the normal noise of a forum ticking over contentedly.
09:20:09 <planetmaker> at least those"departure discussions" I've ever seen
09:21:01 <DDR_> I guess addition of a category could be important. If half the community leaves, that's notable. But really, come to think of it, forums usually aren't that... interesting, as a thing.
09:23:07 <DDR_> LordAro: Yes, but a fairly nice quit message.
09:28:22 <MINM> hmmmh, I wish it was possible to make heightmaps from scenarios
09:29:02 <planetmaker> :-) I love that advice, Terkhen :-)
09:29:12 <Eddi|zuHause> what was that again? "locate the existing implementation: 300$"
09:29:23 <Terkhen> I don't remember where, IIRC it is available from the scenario editor
09:31:31 * MINM finally gets around to getting a nightly ottd
09:31:35 <andythenorth> whenever /me feels like 'quitting' it's just attention whoring
09:31:44 <andythenorth> usually solved by shipping something instead
09:32:39 <planetmaker> yeah, that usually helps me, too ;-)
09:32:45 <Eddi|zuHause> i thought you just say "i should remove supplies" in such cases :p
09:33:14 <Eddi|zuHause> you get at least 15 replies "please don't" each time :p
09:33:18 <andythenorth> that would be equally effective
09:33:26 <andythenorth> although my thinking about that was genuine :P
09:34:58 <planetmaker> I should find motivation to finally finish the new supply mechanism.... but I haven't yet found any
09:35:23 <planetmaker> nor did I find any in fixing parameters of swedish rails...
09:35:31 <MINM> wait, nightly means I need to compile, don't i?
09:35:32 <andythenorth> planetmaker: I have written very little code recently
09:35:44 <andythenorth> maybe 'OTTD is dying'
09:35:46 <planetmaker> andythenorth, newgrf code I haven't written much either
09:36:07 <planetmaker> But then... doesn't matter. Interest in improving / continuing something goes and comes :-)
09:36:19 <Eddi|zuHause> MINM: no, you don't need to compile
09:36:19 <planetmaker> some hiatus on some newgrf projects is not that bad
09:36:35 <planetmaker> MINM, that's why it's called "nightly"
09:36:44 <planetmaker> because they're built every night...
09:36:51 <planetmaker> (or rather evening)
09:37:24 <MINM> then why did I get source when I clicked download?
09:37:47 <planetmaker> because you clicked wrongly?
09:39:14 <planetmaker> download-nightly gives you iirc grfcodec's nightlies
09:40:12 <andythenorth> planetmaker: (1) I should learn nml :P
09:40:20 <andythenorth> (2) we should release 0.7.x or 0.8.x for FIRS
09:40:28 <MINM> well, I got that page, and I have a tarball with 64x ottd in it.
09:41:04 <Terkhen> MINM: there is a dropdown that allows you to select what you want, it should be set automatically to the best download for your OS, but if it's not, choose it manually
09:41:28 <planetmaker> if you, of course, expect a morphos binary or so, you're out of luck ;-)
09:41:31 <andythenorth> planetmaker: releasing on bananas means doing your suggestion of changing grfid
09:41:37 <andythenorth> that is well known now to be a sin
09:41:45 <planetmaker> not every esoteric OS is compiled for for nightlies
09:41:51 <andythenorth> sometimes good people have to do bad things :P
09:42:08 <MINM> does it contain a ready-built running ottd in the tarball?
09:42:16 <planetmaker> andythenorth, currently I'd advocate a bit patience, waiting for tb's re-design of the webpage :-)
09:43:14 <andythenorth> :o something is changing where? Bananas? Or openttd.org?
09:43:18 <Eddi|zuHause> MINM: that depends on what you actually downloaded (FULL name)
09:43:22 <MINM> so, Im in ubuntu, I've got openttd-trunk-r23050-linux-generic-amd64.tar.xz
09:43:27 <andythenorth> incidentally my site is down due to dead box
09:43:38 <Terkhen> MINM: that's not souce, that's a binary
09:43:53 <Eddi|zuHause> yes, that should contain an executable
09:44:01 <MINM> seriously, bear with me, I musn't have had enough tea yet today.
09:44:07 <Terkhen> ubuntu deb packages are only compiled for releases
09:44:14 <LordAro> MINM: i expected better of you ;)
09:44:15 <Terkhen> but the generic linux binary should work fine too
09:46:12 <MINM> ...so, it *is* a running version of OTTD in that folder. why did you say it isn't?
09:46:31 <peter1138> you said you had the source
09:47:49 <MINM> no, I thought that. I saw a tarball containing folders, I think I got a little too used to apt-get and software manager.
09:48:10 <MINM> as I said, not enough tea!
09:48:22 <andythenorth> usual cause of failure
09:48:28 <andythenorth> people who've had enough tea don't screw up
09:48:49 <LordAro> my ottd nightly gets auto-updated, 'cos i'm cool :D
09:49:25 <MINM> when Im home alone I usually make a whole pot of tea and drink it all during the day
09:49:32 <MINM> ...i think I may have a problem.
09:53:28 <Sacro> i hate getting the ttd music stuck inm my head
09:58:01 <MINM> just thought Id throw it in here.
09:58:30 <MINM> it's a map Ive made so Ill have more use for aircraft and ships.
09:59:24 <MINM> should also provide for some interesting railway construction too, as there's some height differences in there
10:02:19 <peter1138> ah, k, thought you had a problem with it :)
10:02:40 <MINM> what do you think about it?
10:10:09 <Eddi|zuHause> i used some snowline multiplicator hack
10:10:32 <Eddi|zuHause> to get the newgrf snowline in some sensible ranges
10:10:43 <planetmaker> MINM, it's relatively smooth. Too smooth for my taste
10:10:45 <Eddi|zuHause> problem is, the snowline "jumps" several heightlevels
10:11:09 <planetmaker> Eddi|zuHause, what for do you need a "snowline multiplicator"?
10:11:11 <Eddi|zuHause> so tiles go from no snow to full snow
10:11:32 *** Strid__ has joined #openttd
10:13:57 <Eddi|zuHause> planetmaker: the hacky part is really simple, in GetSnowLine(), HighestSnowLine() and LowestSnowLine() you add a factor like "return _snow_line->table[ymd.month][ymd.day]*4-6*8;"
10:14:49 <planetmaker> Eddi|zuHause, yes.... but what for? We have snowline level adjustment by the level...?
10:15:01 <Eddi|zuHause> planetmaker: but not for the newgrf snowline
10:15:35 <planetmaker> I can also via newgrf adjust the snowline by the level
10:15:47 <Eddi|zuHause> planetmaker: yes, but not to "more heightlevels"
10:15:50 <planetmaker> or are you talking about some patch? oh
10:15:56 <planetmaker> you should say so :-)
10:16:18 <Eddi|zuHause> planetmaker: i referred to the above picture, which obviously used that
10:16:40 <planetmaker> I don't see anything unusual there
10:16:44 <Eddi|zuHause> it looked more than usual
10:16:50 <Eddi|zuHause> maybe it was just my imagination
10:18:29 <Eddi|zuHause> btw, there's still a bug that the highest possible heightlevel is not actually generated
10:18:34 <Eddi|zuHause> due to some weird rounding issues
10:19:06 <Eddi|zuHause> causing mountain tops to be more flat than they have to
10:21:41 <Eddi|zuHause> some other oddities about snow lines: alpine climate farms, which stop production in winter, will never get generated in places where that matters, because on game start it's always january, so the initial industry generation always gets the lowest possible snow line
10:22:31 <Elukka> andythenorth: ran into some stuff about an industrial tramway and thought you might be interested
10:29:34 *** KenjiE20 has joined #openttd
10:30:33 <Eddi|zuHause> that is totally photoshopped! i can tell by the pixels!
10:41:58 <MINM> it would be nice to have a treeline above which trees won't grow.
10:43:58 <V453000> empty land without trees is ugly :P
10:45:29 <andythenorth> Eddi|zuHause: voxels
10:46:41 <Eddi|zuHause> there was a patch for that once
10:47:32 <Eddi|zuHause> but it would be way more "realistic" if the trees got smaller and smaller, not a "clean" cut
10:48:25 <peter1138> yeah it the wrong way
10:48:40 <peter1138> start of the game shuld be heavily forested
10:49:03 <peter1138> alas, the penalties for felling would be the wrong way around
10:54:15 <Celestar> peter1138: did we ever complete custombridgeheads? :P
10:54:29 <planetmaker> you didn't, Celestar :-(
10:54:47 <Celestar> I know that I hit some blocker at some point.
10:54:55 <peter1138> something to do with reversing iirc
10:55:00 <Eddi|zuHause> afair you started the c++ port because cbh met a dead end
10:55:05 <planetmaker> alegedly it's one of the features which keeps two of the remaining three ttdp players from switching :-P
10:55:06 <Eddi|zuHause> but then you never picked it up again
10:55:22 <peter1138> when i wrote it it was a tiny patch :S
10:55:27 <Celestar> something went really wrong with CBH + crossing bridges.
10:55:34 <Eddi|zuHause> SAC and who else?
10:55:43 <planetmaker> crossing bridges is a totally different topic
10:55:53 <peter1138> maybe i should look at it again
10:55:57 <planetmaker> Eddi|zuHause, I don't recall. But iirc someone else wrote that, too
10:56:14 <blathijs> Celestar: Maybe we should have another try at rewriting the map arrays! ;-)
10:56:21 <blathijs> (not completely kidding)
10:56:31 <planetmaker> look at michi's implementation thereof :-P
10:56:37 <peter1138> there's a 3d map array somewhere
10:56:59 <Celestar> blathijs: ... what did we plan to do there? :D
10:57:01 <blathijs> peter1138: Somewhere meaning the old map-rewrite branch?
10:57:13 *** andythenorth has joined #openttd
10:57:18 <peter1138> something smatz did iirc
10:57:44 <blathijs> Celestar: Separating the ground from the stuff built on top, and allowing multiple stacked tile at one (x,y)
10:57:58 <Celestar> stations on bridges :D
10:58:13 <peter1138> bridges that exist ;)
10:58:28 <Celestar> didn't we remove them? :P
10:58:48 <Celestar> peter1138: where exactly is that from?
10:59:09 <V453000> oh noes SmatZ's tunnel stuff :D
10:59:19 <blathijs> Celestar: Also, generally clean up the map array and add accessors, which turned out to be too much to do at once
10:59:21 <peter1138> it's from that link ;)
10:59:30 <peter1138> we did the accessors
10:59:45 <blathijs> peter1138: Yeah, that's why it might be good to do a second attempt :-)
11:00:24 <Celestar> we could at some point, yeah...
11:00:53 <blathijs> The screenshot peter1138 linked looks nice
11:01:47 <peter1138> it's 3 years old, so it's pretty dead
11:01:49 <LordAro> "some point" < 12 months?
11:01:50 <V453000> I dare to say it is amazing :D
11:02:08 <peter1138> i don't know how well it worked, but it looks cool :)
11:02:10 <blathijs> Though a quick look at the code from the git link suggests that it isn't a very generic solution
11:07:14 <Korenn> peter1138: that's beyond cool!
11:11:38 <Sacro> I remember that screenshot
11:12:00 <LordAro> i have seen it before also
11:16:38 <V453000> there was even a video of it somewhere :)
11:19:11 <Celestar> I might have some time
11:20:39 <peter1138> hmm, galaxy s2 £30 a month
11:24:17 <MINM> nutracks 1.1.2 has some huge issues with the new planning track
11:26:29 <peter1138> good to see 32bpp's going well
11:27:07 <Celestar> LordAro: yeah I'd love that.
11:27:15 <Celestar> but it'll take some time ramping up :P
11:27:35 <planetmaker> yeah... the 32bpp community seems "very active"
11:28:10 <peter1138> can anything be done with those sprites zephyris posted?
11:28:29 <planetmaker> yes. They'll complement ogfx+landscape
11:28:53 <planetmaker> at least partially. It's not sufficient to replace all
11:28:54 *** SirSquidness has joined #openttd
11:29:13 <Celestar> blathijs: so ... what about the map thing needs to be ... refactored?
11:31:02 <LordAro> wait, those are 32bpp? they look... normal
11:31:20 <LordAro> too used to only seeing 32bpp-ez i guess
11:32:26 <Eddi|zuHause> nobody said 32bpp had to be "rendered"
11:32:43 <planetmaker> I expected them to look slightly different, tbh. But Zeph used to draw everything 32bpp and then convert to 8bpp only
11:33:00 <planetmaker> and the 32bpp-ez ground tiles should rather be called "ground" tiles
11:33:11 <planetmaker> as they exceed the allowed dimensions. Which make them behave quite ugly
11:33:46 <planetmaker> they're non-flat as the grass goes beyond the tile border
11:33:55 <planetmaker> which is a big no-no for ground
11:34:38 <planetmaker> I told them that. The reply I got was "looks better. We keep it". Oh well
11:35:15 <planetmaker> results in things like grass on the road between bridge head and road
11:35:27 <LordAro> project for me when readme viewer finished: sort out 32bpp-ez stuff
11:35:46 <planetmaker> LordAro, first go for 32bpp. W/o ez
11:36:02 <Yexo> but if jupix repo is anything to go by: there is not much useful stuff there
11:36:07 <planetmaker> it could be the added sugar. But 32bpp in itself would bring it along a looong way
11:36:13 * Celestar ponders about unittesting openttd
11:36:37 <Yexo> not saying there are no nice graphics, but the vast majority are missing sources, are missing a licence, are incomplete, or are missing the "normal zoom" sprites
11:36:40 <Celestar> using boost::test or google-test
11:36:49 <planetmaker> people usually imply 32bpp = ez. Which is as wrong as it can get :-)
11:37:19 <peter1138> 12:35 < planetmaker> results in things like grass on the road between bridge head and road
11:37:26 <Yexo> one of the problems is that the ez patch has never been complete so it has always forced a 32bpp blitter
11:37:30 <peter1138> ^ which they then considered our bug, iirc
11:37:30 <Yexo> while there is technically no reason for that
11:38:08 <planetmaker> Yexo, yes... and it should (IMHO) give means to be used with 8bpp, too
11:38:15 <planetmaker> there's no reason to require 32bpp
11:38:27 <peter1138> hell yeah, i'd love an extra zoom level with 8bpp
11:38:30 <LordAro> so, a 'proper' ez patch would be made irrespective of 32bpp, yes?
11:40:26 <LordAro> but since now ottd can detect a 32bpp grf/base set, forcing (should) now not be needed
11:40:29 <peter1138> the different recolour stuff is totally irrelevant to zooming
11:40:48 <LordAro> indeed, should be split off, at least
11:40:52 <peter1138> i still don't know what problem that solves
11:41:16 <Yexo> better gradients for company colours
11:41:34 <Yexo> if you use the current system you can only use the 8bpp company colours, which limits you to 8 gradients
11:41:46 <Yexo> the different recolour stuff allows for a lot more gradients
11:41:56 <planetmaker> which in itself would be a nice patch
11:42:04 <planetmaker> and also doesen't require ez in the least
11:42:20 <Yexo> but it's way more visible with ez
11:42:24 <Yexo> the effect without that isn't so big
11:42:36 <LordAro> from memory, yes - although as of latest patch, that still didn't work correctly - white was still grey, etc
11:42:50 <peter1138> yeah, i was thinking it doesn't work properly
11:43:13 <peter1138> does it still use the mask?
11:43:22 <peter1138> i never looked at the algorithm
11:43:32 <Yexo> no clue how it works, I never looked at that code
11:44:03 <Yexo> the zoom algorithm is broken too, it causes glitches when it zooms in 8bpp sprites
11:44:52 <peter1138> and people wonder why it's not added :)
11:46:37 *** andythenorth has left #openttd
11:46:56 <LordAro> oh good, i'm already a 'manager' on the 32bpp-ez-patches project @ #coop
11:48:22 <planetmaker> you wanted to work on it... :-)
11:48:24 <V453000> I thought you meant the irc channel for a second :P
11:49:41 <Celestar> blathijs: do you have any of our old drafts for the map shit?
11:50:52 <peter1138> the ez patch add a palette_modifier_shadow
11:51:14 <peter1138> but... there's already palette_modifier_transparent. hmm
11:54:42 <peter1138> right, i'm not looking any more
11:54:46 <peter1138> i still have biggui to finish off
11:56:24 <LordAro> nope, doesn't do that :)
11:57:51 <DorpsGek> Celestar: help [<plugin>] [<command>]
11:58:02 <DorpsGek> Celestar: Error: There is no command "bug".
11:58:23 <Celestar> it took him 16 seconds for figure that out? :P
12:00:35 <Eddi|zuHause> "Unter den Blinden ist der Einäugige König"
12:00:38 <planetmaker> LordAro, no reason. I guess no-one really looked at it properly
12:02:34 <planetmaker> now: work work ;-)
12:03:36 <LordAro> i can report it still applies without error
12:03:55 <LordAro> but then i don't think find_version.sh has changed since :)
12:08:12 <planetmaker> it actually has ;-)
12:08:33 <planetmaker> the FS entry is from 11 April. The last change on that file from 7 May ;-)
12:08:53 *** welshdragon has joined #openttd
12:10:03 <planetmaker> and one of the "reasons" is that the svn revision detection is not failsafe: one could mimic an svn version by a specifically crafted commit messages to the hg repo.
12:13:19 <Eddi|zuHause> how's that relevant? you can override the version anyway
12:44:50 <CIA-6> OpenTTD: truebrain * r23055 /extra/website/ (10 files in 6 dirs): [Website] -Fix: long pending changes of website; various.
12:47:22 <LordAro> " Do not need donations at the moment!" huh?
12:48:09 <TrueBrain> you sound like that is new to you ;) We haven't been actively looking for donations for the last 2 years :P
12:48:57 <LordAro> how come? is the project being swamped with money?
12:49:18 <TrueBrain> There is enough flow of donations without us actively looking for it
12:49:28 <KenjiE20> didn't you know, everyone here is like "take my money, I don't need it"
12:49:45 <TrueBrain> attitude towards donating has changed a lot over the last few years, that is for sure :)
12:50:12 <TrueBrain> from: "you donated? Why? It is free!" to "What? You play that game for free and never donated anything?", basically :P
12:50:56 <KenjiE20> I wonder if that has anything to do with a certain block game.....
12:51:07 <TrueBrain> not from what I have seen
12:51:22 <TrueBrain> I more think that that certain game got that much money because of the change in attitude
12:51:56 <TrueBrain> then again, 60 euro for a non-indie game is a lot of money .........
12:52:38 <KenjiE20> yeah, they do seem to be getting to the "charging this much, cause we can" area
12:52:51 <TrueBrain> and then they also force you to be online when you play it
12:52:55 <TrueBrain> which makes little sense to me ...
12:53:06 <TrueBrain> and the worst, for me: then you also have to pay 15 euro per 4 maps extra ...
12:53:12 <TrueBrain> I cannot understand people seriously pay for that
12:53:20 <TrueBrain> I remember the days I could download 100 maps FOR FREE
12:53:33 <KenjiE20> AA-title-4: The Rest of the Game DLC Now Avaiable!!11eleven
12:53:36 <TrueBrain> every game Terkhen :P
12:54:11 <TrueBrain> Call of Duty: Black Ops costs in total 120 euro to get the game + 4x 4 maps atm .... counterstrike I downloaded for free (because I had Half-Life) and it came with 8 maps, and I downloaded ...... MANY over the years
12:54:21 <TrueBrain> greed is good, I guess
13:00:06 <Terkhen> with valve games you still get the new content for free
13:00:21 <TrueBrain> it is one of the only that keeps true to his statements, yeah
13:00:41 <KenjiE20> Terkhen: but you game is borked until you finish the 20bilion gb update :P
13:00:47 <TrueBrain> I jsut wonder why game industries want to rob me blind .. I rather spend some money on indie games, and have equal amount of fun with it
13:00:48 <Terkhen> mostly because they want to drag people into their online gaming thing
13:01:36 <TrueBrain> I bought games like Achron. Lovely games. Not as pretty as others, sure, but equal amount of fun tbh
13:01:39 <Arafangion> They're also pretty horrid to work for.
13:02:08 <TrueBrain> I installed Cities XL 2012 a few days ago ... omfg ... what a waste of money is that game. The amount of bugs I found in the first 10 minutes exceeded the amount of memory in my head I wanted to spend on it :P
13:02:25 <TrueBrain> can't understand why anyone would pay such amounts of money on such games ..
13:02:32 <TrueBrain> ugh, enough ranting against the gaming industry :P
13:02:48 <TrueBrain> I am just happy we can give you guys a game like OpenTTD free of charge
13:02:50 <Terkhen> yes... many companies follow the philosophy of "release something broken now, fix in a few months"
13:03:09 <KenjiE20> s/fix/fix and finish/
13:03:15 <TrueBrain> indeed: if at all :D
13:03:37 <TrueBrain> Cities XL 2011 never got fixed of his memory problems. sadly, 2012 version has EXACTLY the same issues
13:03:55 <Arafangion> Computers have 16 GB of RAM, anyway.
13:04:01 <KenjiE20> I miss big transport infrastructure in cities
13:04:03 <Terkhen> if the previous version made money anyways, why bother? :P
13:04:13 <KenjiE20> at least SC4 had highways and stuff
13:04:20 <TrueBrain> which brings me to the rant of for example Roller Coaster Tycoon or Call Of Duty: the identical same game, the same weapons, the same maps, the same everything .... yet it is a new version and they ask again money from you
13:04:44 <TrueBrain> I laughed my ass off that the creator of CoD MW3 said it was not a copy/paste of MW2 ... "it now had better lighting, better textures and better looking maps"
13:04:52 <TrueBrain> so .. how was the gameplay not a copy/paste?
13:05:07 <Terkhen> if you only care about graphics, it is not a copypaste :P
13:05:15 <Arafangion> All modern games are just graphics. :(
13:05:24 <TrueBrain> which moron cares about gfx when playing a game .....
13:05:33 <Arafangion> Most game players..
13:05:34 <TrueBrain> I played BF3 yesterday .. first on the XBox ...
13:05:37 <TrueBrain> the controls made me cry
13:05:49 <TrueBrain> I was so fucking annoyed
13:05:53 <TrueBrain> I can only go from A to B
13:06:02 <TrueBrain> as then I am "leaving the mission area"
13:06:11 <TrueBrain> just another shooter
13:06:13 <TrueBrain> running from A to B
13:06:16 <TrueBrain> shooting what comes at you
13:06:23 <TrueBrain> owh, yeah, "and it looks pretty"
13:06:28 <KenjiE20> no more sneaky gits on karkand?
13:06:30 <Terkhen> All modern games are just graphics <--- that's good for me, if a lot of good games came out I would be broken most months :P
13:06:32 <TrueBrain> worth my 60 euros? A hooker is money better spend
13:07:21 <planetmaker> suddenly it feels dirty here :-P
13:07:29 <Arafangion> TrueBrain: Unlike hookers, though, when you're done with the game, you have no adverse issues other than a slight lack of sleep.
13:08:09 <KenjiE20> same nagging feeling of disappointment later too :p
13:08:50 <TrueBrain> and you still hav ethe puke next time you pick it up
13:08:53 <TrueBrain> so I see no difference :P
13:11:23 * Arafangion is suddenly unable to distinguish.
13:12:03 <TrueBrain> but, to end with some positive: I do like that Steam is allowing indy-developers, and they are doing a good job there
13:12:13 <blathijs> Celestar: There is a branch in the SVN repository, I think
13:12:15 <TrueBrain> I proudly bought a few good indy games, which really were worth their money :)
13:12:22 <KenjiE20> also, bullethell on steam
13:12:24 <blathijs> Celestar: Not sure if we actually wrote a lot of docs up front
13:12:45 <Celestar> blathijs: yeah. I'm trying to find out what we actually tried to do :P
13:12:45 <Celestar> I know there were about 120 opinions of how to do it :P
13:13:12 <TrueBrain> and now, now I need a graphic artist
13:13:15 <TrueBrain> Osai: where are you buddy? :)
13:13:39 <KenjiE20> you need a minion whistle :p
13:14:51 <TrueBrain> Osai did all the imagery on the OpenTTD website for me :D He is that awesome :)
13:15:00 <TrueBrain> just it has been 2 years or something we last worked on it :P
13:15:19 <KenjiE20> it ain't broke yet :)
13:17:12 <Celestar> blathijs: I'm still wondering about those unittests :P
13:18:35 <blathijs> Celestar: unittests?
13:19:26 <Celestar> blathijs: I was wondering whether the basic map/tile functionality should be unittested. That'd make modfying it easier
13:21:08 *** DayDreamer has joined #openttd
13:21:48 <blathijs> Celestar: Probably makes sense, I guess
13:22:06 <Celestar> blathijs: plus maybe then unionizing the current map struct.
13:22:54 <blathijs> I guess the current map struct isn't regular enough to allow unionizing, but I'm not sure
13:23:20 <blathijs> also, if you introduce the concept of layered tiles, your unittests will probably break everywhere :-)
13:23:41 <Celestar> I have that feeling as well
13:23:45 <blathijs> (read: most accessor functions will need to be updated to include a height parameter, or something)
13:31:11 *** Biolunar has joined #openttd
13:31:38 <Jabol> Can someone give me a link to an article about making NewGRF's?
13:33:17 <Yexo> www.tt-wiki.net/wiki/NMLTutorial
13:33:59 <Jabol> Hmm, there's no article in the wiki.openttd.org about it? I thought there was.
13:34:15 <Yexo> not anything comprehensive
13:36:25 <Jabol> Can you give me the link to that article?
13:36:47 <Yexo> I don't know of there is one
13:36:54 <Yexo> if there is one, it's small and unimportant
13:37:58 <Yexo> why do you want an article on wiki.openttd.org so badly?
13:38:13 <Yexo> tt-wiki.net hosts the official nfo and nml specs
13:38:22 <Yexo> and also the best tutorials
13:42:39 *** Starhero has joined #openttd
13:44:19 <Starhero> i am having an issue with some NewGRF. One is throwing and error and I am not sure how to go about debugging (I believe it is incompatiable with something else but the error is not very informitive) Could someone possibly help?
13:45:11 <LordAro> the error report is not very informative if we don;t know the error message
13:45:15 <Jabol> Yexo: It explains the coding, but not the required sprites.
13:45:24 <Jabol> I want to make a new-looking train.
13:45:31 <Jabol> But I don't know how big the sprites are supposed to be.
13:46:05 <Yexo> or take a look at an existing project from dev.openttdcoop.org and copy the sizes from there
13:46:20 <Yexo> Starhero: which grfs / which error message?
13:46:23 <Starhero> This is an SS of the error..let me also SS the packs...I am new so don't eat me even if i am a tasty noob! :P
13:46:53 <Yexo> can you also make a screenshot of your newgrf window?
13:48:22 <Starhero> I turned on debugging GRFs via windows CMD but the text is so jumbled i don't want to even comb through it.
13:48:59 <Yexo> set the right parameter to override that safeguard
13:50:19 <Sacro> Running a grep -r from / is quite slow
13:50:23 <Yexo> blathijs / Celestar: did you look at the link to michi_cc's repository that planetmaker posted earlier?
13:50:37 <Celestar> where is it, I have a bit of a shaky connection here
13:50:38 <Yexo> that already allows layered tiles IIRC
13:50:51 <Starhero> ahh! So how many of these GRF's have wikis? Since i wish i could get some info with the cargos and such...and vehicle intro dates...Also is there ever going to be URL LINKED (like actually clickable) links in the discriptions of the GRFs in the content browers...
13:51:04 <Belugas> mmh.. searching a needle in a <hai??> stack
13:51:09 <Starhero> I found myself blindly d/ling them since they sounded cool...wiht out going to the forums..
13:51:40 <Yexo> Starhero: a few have. This link to this one is visible in the screenshot you posted
13:51:40 <Starhero> because of the fact that i would have to manaully type in the link...and that is annoying when you are on a d/l crazy for new content.
13:51:54 <Starhero> yes but i can't click in IN game...
13:51:59 <Celestar> Yexo: I'll have a look ;)
13:51:59 <Starhero> to open my browser :P
13:52:11 <Yexo> that's a feature request that's been open for quite some time
13:52:27 <Yexo> I do have a half-finished implementation for windows
13:52:49 <Starhero> Ah! That is the OS of choice for me (for gaming..:P)
13:53:22 <Starhero> AH! Thank you again!...wow this is more imformitive then i thought (coming into the IRC that is)
13:54:32 <Starhero> Ok now to set this GRF settings...Tho i ask you this..before i do that..might you review those GRFs in the screen shot and if you know of any problems i might have with those all on? I actually had disabled say FIRS because of alos problems with that one..(it was with ECS base_
13:55:15 <Celestar> Yexo: is that just the framework or are there some testable features?
13:55:41 <Yexo> not too sure about the last state. But I've seen some screenshots with multiple railtypes on one tile (both diagonal and not crossing eachother)
13:55:50 <Yexo> and different snow amounts of tiles with foundations
13:59:59 <Celestar> at least it compiles :D
14:00:36 <Starhero> btw i swear this game looked different at least a few years ago....now it actually LOOKS LIKE TTD! which brought back sooo many hours of wasted life memories :P and now i am wasting more :P
14:00:56 <Starhero> I played this game when I was like 9
14:01:37 * KenjiE20 hasn't really stopped since '94
14:02:23 <Starhero> Also IIRC ttd was the start of the tycoon crazy..and no game has done exactly what ttd has done (tbh all other tycoon games such compaired to this game)
14:02:53 <Starhero> Would i be correct with that statement?
14:03:01 <KenjiE20> TT was renamed TT to fit in with microproses existing tycoon brands, so.....
14:03:56 <Starhero> eh never liked those...i guess cuz i rather have full transport control..not just trains...tho i use trains the most in ttd
14:03:57 <peter1138> Starhero, it always looked like ttd :S
14:04:27 *** Doorslammer has joined #openttd
14:05:16 <Starhero> maybe it was that when i first heard of ottd it wasn't exactly playable? idk...something put me off from it ..and made me get ttd the orgin and played that on windows..or something but got annoyed by somethine about it and stopped playing the game...now ottd seems to just replace ttd for me...rather play this then the orginal...esp with addons.
14:06:54 <blathijs> Yexo: Yeah, I saw it, but it looks like it hacks on layers a bit. Not entirely sure, though.
14:07:49 <Starhero> oh and one other thing i wish i could have a better understanding over.. Horsepower..yes i have seen the wiki page and i have see the forums post on it..but ..is there no way of say..you have 100t to 50hp or 100hp? what is a BASIC rule for making sure my trains run at LEAST on flat land ...and accelerate decently? Trying to figure out Newtons and all that crap hurt my brain :P
14:10:21 <blathijs> Celestar: Don't find what?
14:10:40 <Celestar> blathijs: his changes :P
14:11:35 <Celestar> hm.. there's a submap head
14:14:33 <blathijs> Celestar: newmap.git
14:15:08 *** welshdragon has joined #openttd
14:15:23 <blathijs> Celestar: Then click the ... under the summary of changes to see all of his changes (reading them in chronological order makes sense :-p)
14:20:11 <blathijs> Celestar: Apparently, he changes the map array to be size_y arrays of each size_x elements, and an array of size_x * size_y integers (the offsets)
14:20:45 <blathijs> the offset is the position within the array of size_x size that the "base" tile for a given (x,y) is located
14:21:41 <blathijs> then, he stacks multiple tiles by making the size_x array size_x + 1 long and shifting a bunch of offsets
14:22:09 <blathijs> Not sure if my explaination is helping ;-p
14:25:35 <Celestar> I'm still wondering whether this is the way to go.
14:26:09 <blathijs> Though it seems that these "associated tiles" are not really layered tiles, more like a means to separate the ground from the railway tracks on top
14:26:44 <Celestar> and just have _map[x][y][z] :P
14:27:07 <blathijs> e.g, there is a GetTileType(TileIndex) like before, but also a HasTileByType(TileIndex, TileType)
14:27:25 <LordAro> arrays of arrays of arrays...
14:27:35 <blathijs> but that doesn't seem to handle crossing bridges, for example
14:28:19 <Celestar> why not store a std::vector<Tile> for each X,Y?
14:28:29 <Celestar> with many (most?) of them just having one element.
14:28:31 <blathijs> Celestar: I agree that this way of storing extra tiles is a bit cumbersome, but the exact way of storing these extra tiles is easy to swap out (seems like it's properly abstracted away)
14:28:43 <blathijs> Celestar: Performance! (I guess)
14:29:11 <Celestar> blathijs: well, that could be easily tested I guess. Just make a prototype where each vector has only one element ..
14:29:17 <Celestar> blathijs: and see how it changes ..
14:29:56 <planetmaker> celestar, blathijs, iirc michi made some performance tests on that. you might want to ask him for it
14:30:16 *** Devroush has joined #openttd
14:30:20 <Celestar> planetmaker: thought so
14:30:39 <blathijs> I suspect that a vector per tile is very memory intensive as well
14:31:15 <blathijs> but well, I'm curious about what the plan for these associated tiles is?
14:31:48 <blathijs> It looks a bit like michi_cc is going in the right direction, but is not going far enough ;-)
14:32:18 <blathijs> His commits are nice and contained, though :-D
14:32:24 <Celestar> a vector isn't that big :P
14:32:32 <planetmaker> iirc the idea was to only store additional levels where actually needed
14:32:41 <planetmaker> thus keeping memory consumption in check
14:32:55 <Celestar> blathijs: you could even boil it down a bit further. only use a vector if num_elements > 1
14:33:03 <Celestar> blathijs: otherwise store the data directly.
14:33:34 <Yexo> Celestar: that requires at least one byte extra per tile for num_elements
14:34:13 <Yexo> but there is no more space in the current map array, so that means extending it by a byte
14:34:36 <planetmaker> I guess that needs doing anyway at some stage
14:34:50 <Yexo> that depends on the implementation
14:34:52 <Celestar> not really, because for example the "has bridge" bits are no longer needed ...
14:35:19 <Yexo> currently we use 9 bytes per tile, with michi_cc's implementation you can make it 8 and use stacked tiles when you need more
14:35:58 <planetmaker> hm, what did he remove to reach 8?
14:36:11 <Yexo> not sure he already did, but that was the plan at least at some point
14:36:25 <Yexo> but you don't actually have to remove anything, that's the nice part
14:36:35 <blathijs> Yexo: Also, this one bit "vector or direct" must be stored _outside_ of the Tile struct, of course
14:37:30 <blathijs> Celestar: Though "only use a vector if num_elements > 1" means to store a pointer instead of a Tile struct in the main map array, I guess
14:37:41 <Yexo> still don't really see how that would work
14:37:56 <Yexo> a vector consumes at least 12 bytes of memory, most likely a bit more
14:37:58 <blathijs> Messy, I'm afraid :-)
14:38:04 <Yexo> 8 bytes for a pointer and 4 bytes for the size
14:38:32 <blathijs> Yexo: Can't a vector just have its elements allocated inline?
14:38:53 <Yexo> no, each class must have a known size
14:39:11 <Celestar> blathijs: vectors are inline
14:39:18 <peter1138> it's only 600MB for a 2048x2048x16 map
14:39:20 <blathijs> Ah, right. But you could create vector code that does work like that
14:39:46 <peter1138> minecraft-style chunks? heh
14:39:47 <planetmaker> <3 download gamefile on server connect :-P
14:39:52 <Yexo> yes, but that means a fixed number of tiles per vector, which means there is no reason to use a vector anymore
14:40:00 <blathijs> Yexo: Fair point :-)
14:40:12 <TrueBrain> planetmaker: time remaining: 4 hours and 12 minutes :D
14:40:26 <blathijs> Yexo: Uh, wait. How does the current vector manage adding elements? It's not a linked list, right?
14:40:28 <TrueBrain> would be epic to join coop games ... syncing after the download would be awesome! :D
14:40:35 <Celestar> planetmaker: the vector meta information isn'T saved :P
14:40:45 <Yexo> it allocates memory via alloc, if it runs out of memory it uses realloc to allocate more
14:41:01 <blathijs> Yexo: That applies to "inline" vectors as well
14:41:05 <blathijs> Yexo: Those can also realloca
14:41:06 <TrueBrain> Celestar: how did you manage to get blathijs active? :D
14:41:16 <blathijs> TrueBrain: Talking about ancient plan :-)
14:41:26 <Yexo> what exactly do you mean with an inline vector?
14:41:38 <blathijs> TrueBrain: I won't be doing any coding, don't worry ;-p
14:41:51 <LordAro> Truebrain: his minion call works better than yours ;)
14:42:09 <TrueBrain> LordAro: please do make me a new one :P
14:42:28 <blathijs> Yexo: You described struct Vector {int size; Element* elements}; I'm talking about Vector {int size; Element elements[]; }
14:42:31 <blathijs> or something like that
14:42:52 <LordAro> TrueBrain: no can do, i'm a minion myself, see :)
14:43:02 <blathijs> (I'm not convinced the second example is actual code)
14:43:05 <TrueBrain> LordAro: well, at least you come when I ring :P
14:43:06 <Yexo> ok. Now you want to have an array of those vectors. Say Vector[3]. How are you going to access the 2nd element?
14:43:51 <blathijs> Yexo: Ah, right. I was thinking about having a union {Tile t; Vector* v} map[3];
14:43:56 <Yexo> vector[2] == vector + 2 == ((char*)vector) + 2 * sizeof(struct Vector)
14:44:12 <Celestar> techincally it's (void*) isn't it? :P
14:44:17 <blathijs> Yexo: e.g., only refer to the vector using a pointer, not statically allocate it
14:44:19 <Celestar> vectors are inline :P
14:44:50 *** Adambean has joined #openttd
14:44:58 <Yexo> blathijs: ah, that makes sense
14:46:00 <TrueBrain> can't you guys take a separate channel; #openttd is not to talk about code, ieuw
14:47:10 <Celestar> how's this: new tile type MP_EXTENDED. Then you have a std::map that maps an index into a bunch of stacked tiles?
14:47:44 <blathijs> Celestar: I think that really depends on how many tiles will be stacked
14:48:22 <Celestar> actually std::multimap
14:49:19 <blathijs> and in michi_cc's approach of making the ground and the railway on top of it separate tiles, I think a large percentage of the tiles might have extra tiles
14:49:43 <blathijs> (I actually start to think that the offset approach michi_cc is using isn't that bad, really)
14:51:01 <Celestar> a multimap can store (n) elements with a key (TileIndex)
14:52:39 <blathijs> Celestar: Yeah, but if 50% of the map accesses need to access the multimap, performance will probably suffer
14:53:24 <Celestar> at some point, you have something that will not be O(1) :P
14:53:51 <Yexo> but map access really should be O(1)
14:54:03 <blathijs> Even O(1) can be problematic for sufficiently large values of 1
14:54:24 <blathijs> (That was not a joke ;-p)
14:54:28 <Celestar> Yexo: which means you're stuck with vectors, in one form or another :P
14:54:52 <Celestar> having a vector for each (stacked) tile is O(1).
14:55:05 <Yexo> michi_cc's implementation is also O(1)
14:55:09 <Celestar> just the 1 is a bit larger than the current 1 :P
14:55:10 <blathijs> Celestar: Your map approach is also O(1), right?
14:55:11 <Yexo> but requires a lot less memory
14:55:34 <Celestar> Yexo: yes and no. insert is not O(1) :P
14:55:35 <blathijs> (technically, any operation will be O(n), where n is the number of Tiles per TileIndex, of course)
14:55:46 <Yexo> neither is it with vectors
14:55:51 <Celestar> blathijs: which one? multimap?
14:55:52 <Yexo> since you might have to do a realloc, which is not O(1)
14:56:16 <blathijs> I think that modification at > O(1) is perfectly acceptable
14:56:43 <blathijs> since structural modification will mostly happen after user interaction
14:57:01 <blathijs> (modification of the data inside a Tile struct shouldn't be slow, of course)
14:57:29 <peter1138> what does splitting landscape and railway (and presumably road?) offer?
14:57:37 <Yexo> there are quite a few automatic changes, but those are not that common: industry placement / closure, creating houses
14:59:01 <peter1138> do it the java way. individually allocated objects for each thing, sod the memory ;)
14:59:04 <Yexo> peter1138: multiple railtypes per tile (as long as they don't overlap, which is possible with diagonal rails)
14:59:20 <peter1138> that's possibly anyway
14:59:36 <peter1138> (given a few extra bits, probably)
15:00:28 <blathijs> peter1138: I think splitting them makes stuff like bridges more sensible
15:00:33 <Yexo> I think that was the main point for splitting rail and landscape, but that's just one of the things. It also allows proper snow on both halfs of a tile with foundations
15:00:37 <blathijs> e.g. Ground -> Rail -> Bridge -> Rail
15:00:43 <blathijs> instead of Rail -> RailBridge
15:00:55 <Yexo> and instead of having 2 bits for "has bridge above" you could have a pointer to a bridge struct (or way more information about the bridge)
15:00:55 <peter1138> blathijs, that's a more compelling argument :)
15:01:09 <peter1138> Celestar, we suffer from NIH don't forget
15:01:37 <Celestar> you can implement your own hash map :P
15:01:49 <peter1138> there's probably some already
15:02:03 <blathijs> Celestar: When talking about maps, I usually assume a hash map with O(1) access times :-)
15:02:52 <Celestar> hash maps are USUALLY O(1), but worst case is O(n)
15:03:05 <Celestar> maps are O(log n), but never worse :P
15:03:10 <TrueBrain> Celestar: a full collision hash .. would that be a hash? :D
15:04:15 *** DayDreamer has joined #openttd
15:05:09 <MINM> i'm probably doing something dumb, but can anybody tell me why I never have those dutch town names?
15:05:37 <Yexo> you need to select them in your game options
15:05:48 <Rubidium> oh... cool... Celestar runs when I arrive ;)
15:08:02 <michi_cc> peter1138: Splitting rail/road from the base landscape makes implementing "real" tunnels and bridges a lot more sane.
15:08:58 <michi_cc> And by moving the NewGRF animation part of houses and industry tiles, each tile can probably be reduced to 6 or even 5 bytes.
15:09:09 <MINM> ...ahah, that explains a lot.
15:09:18 <MINM> I wish you could use multiple sets, though.
15:10:53 <MINM> oh well, this is already quite useful :D
15:11:17 <peter1138> michi_cc, yes, it's logical for tunnels & bridges
15:13:19 <MINM> it would be nice to have bridges like in locomotion.
15:13:34 <peter1138> but yeah, "real" bridges that can turn, etc...
15:13:51 <peter1138> that would make placing them harder though :p
15:14:03 <MINM> or, alternatively, a way you could construct a bridge/tunnel independently of rail/road
15:14:26 <MINM> with even double width bridges, so you can have a double line bridge that actually looks a bit sensible.
15:15:08 <MINM> personally Id already be happy if the tunnels get redone like bridges are, with downramps where applicable just like bridges have upramps.
15:15:45 <MINM> that would make tunneled junctions much compacter!
15:16:49 <peter1138> you're volunteering? :)
15:19:04 <MINM> if somebody else would make tunnel sprites and somebody would point me to the bridge and tunnel placement code, I suppose I could try and apply a little harmless cargocult programming
15:19:34 <Eddi|zuHause> whoa... too much activity today... can't follow...
15:20:06 <blathijs> michi_cc: Your current code still assumes that all associated tiles are more or less on the same level, right?
15:20:30 <blathijs> E.g., your HasTileByType accessor doesn't make much sense when you want to arbitrary stick different tiles
15:20:50 <michi_cc> Yeah, those functions would need to be extended with a z parameter.
15:22:07 <blathijs> michi_cc: So you thought about it, but are leaving that for later. Coolness :-)
15:23:06 <blathijs> michi_cc: Btw, a "z" parameter is a bit tricky, given that tiles can be sloped (the exact z of a tile might depend on the direction from which you're coming)
15:23:41 <michi_cc> The idea is to have a stack of e.g. '(base + rail + road) + (base + tree) + (base + rail)'. To optimize that access, there would be another bit next to the current has_next indicator to indicate if another height level follows.
15:24:13 *** Prof_Frink has joined #openttd
15:25:14 <michi_cc> Well, most of the really interesting stuff (i.e. rail, road, stations) are mostly accessed by walking from a neighboring tile, so most of the time you have a reference height and direction.
15:25:33 <blathijs> michi_cc: Ah, a stack of stacks sounds sensible :-)
15:26:25 <blathijs> michi_cc: In the original map array rewrite years back, we split every tile into a base and a construction, but one base plus arbitrary bases makes more sense I guess
15:26:34 <blathijs> uh, arbitrary constructions
15:26:51 <blathijs> and ack on the reference height and direction
15:29:02 <Eddi|zuHause> <Yexo> I think that was the main point for splitting rail and landscape, but that's just one of the things. It also allows proper snow on both halfs of a tile with foundations <-- it may also make arbitrary number of roadtypes (tram, trolley, cobble, asphalt), plus railtype (crossing) on any one tile possible
15:29:46 <michi_cc> Splitting some stuff into more than one tile makes a lot of sense, as for example only industry tiles, houses and stations actually need all map bits, just to support the occasional NewGRF animated tile. So by splitting a house tile into (base + anim) or even (base + house + anim) you can reduce the average tile size probably even below the current 9 bytes.
15:31:15 <michi_cc> Which importantly means that more map can be cached by the CPU which might cover some of the more complex tile access.
15:31:27 <Eddi|zuHause> you'll have an individual tile with <roadtype, roadbits> for each roadtype on one tile. and then some validation routine on construction that certain roadtypes exclude or override each other on a given roadbit
15:31:40 <Eddi|zuHause> drawing may be evil, though
15:32:19 <michi_cc> Not more evil than it would be with the current map design and more road types, you just have to properly sort the road types from bottom to top.
15:32:45 <Eddi|zuHause> yes, but how do you decide the order of cobblestone and asphalt roads?
15:33:11 <michi_cc> By not allowing it? ;)
15:34:12 <Eddi|zuHause> then how do you want to handle drag-building asphalt roads across an existing cobblestone road?
15:34:12 <Starhero> So...umm...I think i have came to the conclusion that FIRS is not compat. with ECS? and Vice versa?
15:34:26 <Eddi|zuHause> Starhero: that is correct
15:34:27 *** Celestar has joined #openttd
15:35:03 <michi_cc> By converting the whole intersection to either one of the road types.
15:35:05 <Eddi|zuHause> michi_cc: with the added complexity that both roadtypes might not come from the same newgrf
15:35:09 <Starhero> Hmm..idk which to use then...I'll have to sets for one or the other..but ..I ask you all this..which do you believe i should play first? FIRS game or ECS game?
15:35:33 <michi_cc> Celestar: read a bit back
15:36:30 <planetmaker> Starhero, both have their own style and challanges. None is "better" than the other
15:36:43 <Celestar> michi_cc: yeah, I was thinking among similar lines.
15:36:55 <planetmaker> I'm prejudiced as I wrote some parts of FIRS, though ;-)
15:36:56 <Starhero> Yes, I figured that much, I should have noted I am asking for opionions.
15:37:07 <Celestar> michi_cc: but I'm still wondering whether it makes sense to store the "additional" tiles in a hashmap.
15:37:18 <Celestar> michi_cc: I'll try to conjure up some prototype and benchmark it :P
15:38:00 <Celestar> but with CPU caching ....
15:38:11 <Celestar> when last I looked we had like 99% cache hits
15:38:21 <michi_cc> Celestar: probably not, especially as a lot of the fuctions can be changed to take a Tile* directly (and some are already in the repo) which makes the actual storage totally unimportant.
15:39:18 <Celestar> michi_cc: true, but what IS the most efficient way of storing shit :)
15:39:18 <peter1138> 16:35 <+michi_cc> By converting the whole intersection to either one of the road types.
15:39:53 <michi_cc> Don't compare my current repo with plain trunk though, as I've merged the TileExtended byte into the main Tile struct for easier coding. That alone is like a 15% performance hit in certain scenarios, but only temporary.
15:40:11 <Celestar> due to aliasing issues?
15:40:29 <Celestar> i.e. more bloat blasted over the memory lines?
15:40:39 <michi_cc> peter1138: As long as there's a NewGRF flag for that it is resolvable.
15:41:02 <michi_cc> Celestar: No, because x86 CPUs need an extra instruction to multiply a pointer by 9 instead of 8.
15:41:17 <Celestar> well I'm heading home. guess I'll be on later, or tomorrow
15:41:39 <peter1138> does it turn out to be 16?
15:41:43 <Celestar> michi_cc: tried bloating the thing to 16?
15:41:44 <planetmaker> Celestar, last time I checked, IRC also works via DSL from home ;-)
15:42:02 <Celestar> planetmaker: but not too while while steering a vehicle over the highway :P
15:42:20 <michi_cc> Doesn't matter, the maximum scale factor is *8 for accesses like [ESI+ECX*8]
15:42:41 <peter1138> and that last byte isn't accessed very often, i assume
15:43:33 <michi_cc> That too. After all current tiles are fully split the next step would be to rearrange the bits to free up now unused space which would allow reducing the size of the Tile struct.
15:43:42 <planetmaker> Celestar, also that could work... but it might be out of the limits of 'save driving' ;-)
15:52:05 *** HerzogDeXtEr has joined #openttd
15:52:53 *** supermop has joined #openttd
15:57:56 <Eddi|zuHause> <peter1138> ^ road / tram <-- what about trolley catenary [must have road, may not have tram], overhead monorail [may not have tram, trolley], subway/lightrail [may not have road], lots of other ideas?
15:59:01 <Eddi|zuHause> (although some of these restrictions may be weakened for gameplay purposes)
16:05:11 <MINM> trolley could have tram as far as I care
16:05:23 <MINM> subway is kindof interesting, though.
16:05:33 <MINM> very invisible, but still
16:06:02 <MINM> would it be possible to hack 'submerged' subways?
16:07:16 <MINM> by having a tramway with the vehicles offset so they appear sunken into the ground?
16:12:19 <Eddi|zuHause> MINM: theoretically yes, but that's not really the point
16:19:51 *** sla_ro|master has joined #openttd
16:21:09 *** LordAro has joined #openttd
16:21:57 *** |Jeroen| has joined #openttd
16:30:18 <Eddi|zuHause> we only have 3 "low IDs" left... and i already cheated a bit with VT08/VT12/ET30/ET56
16:31:39 <Eddi|zuHause> but at least now all current engines have articulated parts
16:37:29 <planetmaker> in any case the amount of warnings is meanwhile drastically reduced, Eddi|zuHause :-)
16:37:49 <Eddi|zuHause> i haven't actually counted them :)
16:38:26 <Eddi|zuHause> > scripts/generate.py 2>&1 | wc -l
16:38:36 <Eddi|zuHause> not exactly pretty :p
16:40:54 <Eddi|zuHause> yes, the rest are those sprite offset issues
16:43:38 *** Brianetta has joined #openttd
16:43:57 <Eddi|zuHause> i could fix those myself, but i'm unsure which of the two options to take
16:44:46 <Eddi|zuHause> (add 1 row of transparent pixels at the botton, or move 1 pixel down and add row of transparent pixels at the top)
16:47:27 *** valhallasw has joined #openttd
16:54:23 *** Kurimus has joined #openttd
17:02:39 * MINM retries with clean source
17:06:26 *** TheMask96 has joined #openttd
17:07:30 *** Celestar has joined #openttd
17:07:31 <CIA-6> OpenTTD: rubidium * r23056 /trunk/src/currency.cpp: -Change: use the currency -> euro conversion rate for currencies that have been replaced with the euro, so when the switch happens the conversion rate at that point is roughly that of the real world conversion rate
17:08:55 <Rubidium> ohaj to the closest thing to an astronaut we've had in the channel (or am I mistaken?)
17:09:25 <MINM> hmmmh, clean build works.
17:09:43 <Celestar> Rubidium: about, yes :P
17:10:03 <Rubidium> quite civilian lately
17:10:20 <planetmaker> hm.... which round of selection did you get to, Celestar?
17:10:43 <planetmaker> so they interviewed you? :-)
17:10:51 <Celestar> no, apparently I was too young :P
17:11:15 <planetmaker> Then I guess we're equal :-P
17:11:46 <planetmaker> anyway bbl. sports
17:11:47 <Prof_Frink> We should get sabdfl in here, just to annoy you.
17:13:17 <Prof_Frink> Mark Shuttleworth.
17:18:58 <Rubidium> it didn't like clean trunk and hitched a ride on the patch you applied
17:20:02 <MINM> I have the clean 1.1.3 in the usual place, but that sound is gone too, and Im certain at least that one had sound
17:20:26 <Rubidium> possibly the volume is set to 0 in the config
17:20:46 <MINM> not in the os, Ive checked that already
17:21:19 <MINM> where do I find that setting in openttd.cfg?
17:21:35 <Rubidium> alternatively the "nosound" sound set is chosen?
17:22:11 <MINM> oh, I found music_vol, it's at 127
17:22:19 <Celestar> hm.. trains that can change the gauge during travel O-o
17:32:24 <MINM> Ill assume 127 is hard enough?
17:32:55 <Rubidium> music_vol isn't the volume of the sound
17:33:52 <MINM> well, of the music, of course
17:34:19 <MINM> but effect_vol is at 127 too
17:34:45 <MINM> however, this worries me:
17:35:10 <Terkhen> MINM: www.pastebin.com
17:35:16 <Rubidium> all are autodetected, except the soundset which is forced to "NoSound"
17:35:29 <Rubidium> which incidentally might be a reason why there is no sound
17:35:42 <Rubidium> but I'm known for giving things misleading names
17:35:50 <MINM> Terkhen: I know, but I didn't feel the need for such a small piece
17:36:14 <Terkhen> I wouldn't call that small :)
17:36:43 <MINM> huh, for some reason I can suddenly select sfx/mfx/gfx now, I couldn't before
17:37:58 <LordAro> i don't really understand why :)
17:38:31 *** supermop_ has joined #openttd
17:38:52 <MINM> I seem to have sound again.
17:40:21 <MINM> why is it not possible to change sound/music volume in the game?
17:40:33 <MINM> and wether or not the music plays
17:40:44 <Rubidium> it is from the main menu
17:40:55 <Rubidium> but it's primarily because of NewGRFs
17:41:15 <MINM> I can only change what music, not sound or playing or not
17:41:39 <Terkhen> switch the volume to zero
17:42:54 <MINM> meh, I disabled it again in the .cfg.
17:43:03 <MINM> I was just commenting on the oddness of the fact
17:43:17 <MINM> ...aaanyway, back to the patching.
17:43:35 <MINM> Im trying to apply the chillcore patch pack to nightly source.
17:44:16 <Terkhen> good luck with that, it is a huge patchpack
17:44:19 <Rubidium> do you have a clue of OpenTTD's source and/or developing?
17:44:42 <Terkhen> I wonder why they don't keep patchpacks as a bunch of diff files, I would have gone mad with a single one
17:44:43 <MINM> well, Ive already managed to compile the clean source.
17:44:46 <Rubidium> if the answer to both subquestions is no, then apply it to the version the patch was made from
17:45:05 <Rubidium> do not try to apply it to any other version, it will more than likely fail in many aspects
17:45:41 <Terkhen> MINM: you will get "rejects" that you will need to solve manually, and that requires knowledge of how the code works, how the different patches in the patchpack works, and what were the changes in trunk since the revision that the patchpack uses
17:46:01 <MINM> I suspected that, but do you happen to have a spare r22273 laying around somewhere?
17:46:27 <Terkhen> just check out that revision from the subversion repository
17:47:11 <MINM> how? Ive seen how to get the current source, but I don't know how to get a specific version
17:47:27 <Terkhen> subversion / mercurial / git allow you to do that
17:48:03 <LordAro> downloading the source from openttd.org is fairly useless
17:48:35 <MINM> Ive got subversion, but a commandline application is of little use when you don't know the correct command
17:50:10 <MINM> I mean, I know "$ svn checkout svn://svn.openttd.org/trunk openttd"
17:50:27 <MINM> but what are the correct magic words to get a specific version of that?
17:50:40 <Terkhen> try svn -h checkout (IIRC) or google, subversion is simple to use
17:51:05 <Terkhen> I don't remember if it used mercurial syntax for revisions or not so that will be faster
17:52:52 <Rubidium> MINM: start with svn help ;)
17:53:57 <Rubidium> MINM: then the names of *all* available subcommands and only then determine which one is best to update your checkout to a particular revision
17:54:11 <Rubidium> after that, use svn help <subcommand> to figure out what parameters it has
17:54:45 <Rubidium> read the parameters carefully and then determine which one is the best for updating to a particular revision
17:56:08 *** welshdragon has joined #openttd
17:57:44 *** valhalla1w has joined #openttd
17:58:44 *** Alberth has joined #openttd
17:58:44 *** ChanServ sets mode: +o Alberth
18:01:21 *** mahmoud has joined #openttd
18:01:27 <MINM> m@APC1:/media/NTFS/OTTD_build/openttd-trunk-r23050$ svn update -r 22273
18:02:05 <Rubidium> so . wasn't a svn checkout
18:02:59 *** valhalla1w has joined #openttd
18:03:55 <Yexo> MINM: how did you create that openttd-trunk-r23050 directory in the first place?
18:04:32 <MINM> extracted /media/NTFS/OTTD_build/openttd-trunk-r23050-source.tar.xz
18:04:39 <Yexo> so it's not an svn checkout
18:05:09 <Celestar> Yexo: well THEORETICALLY it might be .p
18:05:28 <Yexo> sure, but..., given the earlier error message :p
18:06:17 <MINM> ahah. not identical, then?
18:06:34 <Celestar> svn saves repository data in an .svn directory.
18:06:47 <Celestar> if it's not there, svn doesn't know watta do.
18:06:59 <Alberth> MINM: source code is the same
18:07:30 <Celestar> hm. did TrueBrain throw me out? :)
18:07:41 <TrueBrain> throw you out of what?
18:07:49 *** JVassie_ has joined #openttd
18:07:51 <DorpsGek> Celestar: I don't recognize you.
18:08:01 <TrueBrain> you do need to authenticate :)
18:08:08 <DorpsGek> Rubidium: I don't recognize you.
18:08:08 <Eddi|zuHause> MINM: a source tarball misses all svn metadata
18:08:16 <Eddi|zuHause> MINM: so you can't just "update" it
18:08:19 <Rubidium> Celestar: it's not only you ;)
18:08:36 *** ChanServ sets mode: +o Celestar
18:08:40 <DorpsGek> LordAro: I don't recognize you.
18:09:00 <MINM> "svn co svn://svn.openttd.org/trunk 22273" seemed to work.
18:09:15 <Celestar> now you have the stuff in a directory called 22273 :P
18:09:23 <Celestar> but it's not revision 22273 :D
18:09:32 <TrueBrain> Celestar: but I doubt DorpsGek ever knew you :P
18:10:01 <Yexo> svn update -r 22273 <- that will now work if you go into that directory 22273
18:10:13 <Celestar> .. and rename the dir
18:10:26 <LordAro> "svn checkout svn://svn.openttd.org/trunk ottdsrc && cd ottdsrc && svn update -r 22273"
18:10:55 <MINM> done and done, in proper order
18:11:07 <MINM> now we're getting somewhere
18:11:52 <LordAro> tbh, for a 'proper' patchpack with lots of users, it should be set up in an mq at openttdcoop
18:14:18 <Rubidium> the biggest OpenTTD patchpack works pretty well without a mq ;)
18:14:48 <Rubidium> (that ofcourse being OpenTTD itself)
18:15:49 <LordAro> s/'proper'/'proper' 3rd-party/ ;)
18:16:46 <Terkhen> the best is to work in a single patch until it is included, and then move to the next one :P
18:18:00 <Celestar> cya guys. some movie time
18:19:49 *** frosch123 has joined #openttd
18:20:11 <Rubidium> patch packs are very effective in destroying getting them into trunk
18:20:56 *** JVassie has joined #openttd
18:22:03 <Eddi|zuHause> <MINM> "svn co svn://svn.openttd.org/trunk 22273" seemed to work. <-- the right syntax is "svn co svn://blah@22273"
18:22:40 <MINM> would that get me r22273?
18:23:12 <Eddi|zuHause> and the best thing is, this works even if the "blah" branch has been deleted already
18:23:54 <Yexo> that's not very likely for "trunk" :p
18:24:10 <Eddi|zuHause> yeah, but if you want MiniIN@8000 or something :=
18:28:20 <Eddi|zuHause> i still have 2 MiniIN checkouts for some reason
18:30:12 <Eddi|zuHause> and a "dodgyhack.diff"
18:30:52 <MINM> it compiled alright, but the exe fails to execute
18:31:03 <MINM> well, it starts and immediately goes boom again
18:31:09 <Eddi|zuHause> executed the right exe?
18:31:20 <Eddi|zuHause> added the additional .grf files?
18:31:55 <Eddi|zuHause> (why does nobody ever read the installing instructions?=
18:32:18 <Eddi|zuHause> why do i always type = instead of )?
18:32:31 <MINM> why would I need additional .grf-s?
18:33:26 <Eddi|zuHause> because i assume you want to compile chills patchpack?
18:36:15 <MINM> who the hell puts installation instructions and required dependencies in a second post?
18:37:58 <Rubidium> MINM: the same person that doesn't use a mq/repository to have all required files in the same place
18:39:05 <LordAro> woo, redvsblue was good this week
18:39:49 <MINM> got .grfs, still won't start
18:40:18 <LordAro> have you tried starting it from the console?
18:40:28 <LordAro> that may tell you why it isn't starting
18:41:50 <Eddi|zuHause> MINM: the installing instructions include renaming a grf
18:43:53 *** valhallasw has joined #openttd
18:46:24 *** erik1984 has joined #openttd
18:47:50 *** valhallasw has joined #openttd
18:54:49 <LordAro> MINM: now the fun part begins ;)
19:08:08 <Belugas> #I'VE GOT NO MEMORY OF ANYTHING
19:11:38 <Eddi|zuHause> your caps is stuck!
19:21:38 <CIA-6> OpenTTD: frosch * r23057 /trunk/src/ai/api/ai_order.cpp: -Fix: [NoAI] AIOrder::GetOrderCount() did not hide implicit orders.
19:26:23 <Alberth> Eddi|zuHause: if only he remembered where that key is located...
19:27:02 <Eddi|zuHause> damn this altzheimer desease must be getting worse... i'd blame the age :p
19:28:47 *** valhallasw has joined #openttd
19:29:47 <Belugas> I was yelling in the office, sorry it transpired on IRC :)
19:31:05 <Alberth> Belugas: your synaptic interface may need a bit more tuning :)
19:32:32 <Belugas> my synapses are tuned to overload!
19:35:26 *** blotek_ has joined #openttd
19:36:54 <TrueBrain> Eddi|zuHause: I think the stuck caps is spreading
19:37:21 <Eddi|zuHause> well he _is_ a thousand years old...
20:25:10 <DorpsGek> planetmaker: planetmaker
20:27:34 <renske> Can anyone point me in the direction of an explanation of OTTD signals that is aimed at, well, dummies?
20:27:51 <Yexo> have you tried the openttd wiki?
20:28:45 <MINM> with that in another window, experiment!
20:29:17 <MINM> signals are things which take understanding, it comes with application
20:29:40 <renske> yeah, application and lots of crashing.
20:29:44 <renske> or maybe that's just me.
20:31:12 <renske> thanks MINM and Rubidium
20:31:20 <MINM> don't change signals when you've got a train driving around the place
20:39:32 <renske> it's mostly junctions and crossings where I get confused. I don't build super huge ones, but the signal placement is tricky.
20:40:10 <renske> because it annoys me when trains are waiting endlessly when they really don't have to.
20:40:20 <MINM> laying out the tracks in a good way is half the job, really
20:40:44 <MINM> quite a lot of junctions can do with just block signals
20:41:23 <planetmaker> though the easiest and more efficient way is: every signal which leads to a junction a path signal. done
20:42:37 <renske> block signals are the most basic ones, right?
20:43:25 <MINM> this one for example, is done entirely with just block signals. works like a charm, and it's possible to maintain massive speed
20:44:00 <MINM> not to mention massive capacity
20:45:52 <planetmaker> MINM: it is failsafe
20:46:00 <planetmaker> if you _only_ use path signals
20:46:24 <MINM> planetmaker: until a train reverses into an already reserved path
20:46:33 <planetmaker> and the junction you show is considerably more complex than any path signal junction you can build
20:47:04 <planetmaker> MINM: that only happens due to wrong signal placement...
20:47:20 <MINM> nope, Ive seen it happen with perfect placement too.
20:47:28 <planetmaker> or alternatively due to the wrong setting.
20:47:37 <planetmaker> MINM: obviously that placement then is NOT perfect
20:47:48 <MINM> but default setting none the less.
20:48:06 <planetmaker> i.e. make sure the train can't reverse
20:48:11 <Elukka> renske: if you're not sure which signals to use, use path signals everywhere
20:48:15 <planetmaker> into the junction. By proper signal distance
20:48:17 <Elukka> 99% of the time they're the best ones
20:48:55 <planetmaker> the rule of placement with path signals is also only one very simple one: Put one where you want a train allow to stop
20:49:08 <renske> aw, but I want fancy signals! :p
20:49:12 <planetmaker> don't put one where a train shouldn't stop
20:49:20 <planetmaker> i.e. too close after a junction or so
20:49:53 <Rubidium> MINM: the default for reversing at signals is false
20:49:59 <Elukka> i think path signals ought to be the default basic signal with a prominent place in the signal menu
20:50:08 <Elukka> because they're what most players will need most of the time
20:50:09 <renske> ah, that one's aesthetically pleasing :)
20:50:09 <Rubidium> (it might have been true a long time ago though)
20:50:10 <MINM> pretty fast and compact, but capacity is slightly lower than the others
20:50:13 <Korenn> MINM: wth, that flyover page has tracks with signals going both ways.
20:50:41 <MINM> the one before transmogriefed?
20:51:01 <Korenn> the bottom left to top right track has both tracks going south-west
20:52:36 <MINM> this? are you sure, I don't see a double.
20:52:53 <Korenn> try going from bottom left to top right
20:52:57 <Korenn> what track do you take
20:53:27 <Korenn> I think that's the strange signal grf
20:53:31 <Korenn> that i'm not used to :)
20:53:44 <MINM> yeah, the signals aren't very clear, Ill give you that
20:53:50 <Korenn> the front facing signals look like they're backwards
20:53:55 <MINM> but it's very easy to discern
20:54:16 <MINM> by default (is a setting) signals are always are on the drive side!
20:55:04 <Korenn> that's true. But apparently I tend to look for the front / back of the signals when I check them
21:03:09 <renske> I barely just moved on from point-to-point tracks. I think I'll hold off on the flyovers until I have enough trains to justify them. even though they're very cool-looking.
21:04:56 <renske> also, that transmogrified junction seems to have signals just after exits. isn't that a bad idea? if trains would stop there, they'd block the path of other trains.
21:06:26 <LordAro> gah, why is banshee using 600MiB of memeory?!
21:06:43 <LordAro> stupid memory leaks...
21:23:14 <MINM> at least it's not as bad as firefox
21:24:01 *** snorre_ has joined #openttd
21:24:29 <MINM> bah, Ive had enough of patching for today
21:34:51 <planetmaker> LordAro: why do you report insulfrog's posting as spam?
21:35:25 <LordAro> i thought i saw spam in the signature
21:35:43 <LordAro> it looked typical, perhaps i didn't look at it closely :L
21:36:32 <planetmaker> it's a link to his homepage ;-)
21:36:55 <planetmaker> and looks suspicious on first sight
21:37:06 <planetmaker> I just happened to have met the guy a few times before :-P
21:38:11 <LordAro> yeah, now that i think about it, i recognise the name too
21:38:30 <LordAro> in fact, i think i recognised it when i clicked on the topic link
21:38:46 <LordAro> just the standard automatic-absent-minded action :L
21:39:46 <planetmaker> well, just a case of 'shit happens' :-) And just closed it
21:44:24 <planetmaker> I guess I'll close my eyes for a few hours now. Good night
22:35:57 *** HerzogDeXtEr1 has joined #openttd
22:42:36 <Arafangion> That reminds me of a friend I had in uni...
22:42:48 <Arafangion> Every single one of his emails would be automatically classified as spam, somehow.
22:43:06 <Arafangion> He didn't have automatic sigs or anything, just a weird way of writing. :)
continue to next day ⏵
