IRC logs for #openttd on OFTC at 2009-08-29
⏴ go to previous day
00:02:58 <welshdragon> i've found a bug, but i don't know if i t relates to IS 2 Beta 3
00:03:01 <OwenS> Does takeoff/landing use special sprites?
00:03:24 <Eddi|zuHause> welshdragon: cool, we don't know either
00:03:27 <Yexo> OwenS: I think so, but I'm not sure about the default sprites
00:04:32 <Eddi|zuHause> OwenS: i believe newgrfs can use varaction2 for detecting wether an aircraft is in the takeoff phase
00:04:36 <welshdragon> Eddi|zuHause: i change the company colour in the new colour scheme window, but the individual vehicle (rail, road, air, sea) don't change to the global setting
00:04:59 <Eddi|zuHause> welshdragon: that might be a newgrf bug
00:05:35 <welshdragon> Eddi|zuHause: but it's worked fine before...
00:05:39 <Eddi|zuHause> or some kind of cache not being invalidated
00:06:06 <Eddi|zuHause> welshdragon: try to reproduce it in trunk
00:06:35 <Eddi|zuHause> i heard about some newgrfs having problems with incorrect company colours
00:06:57 <Eddi|zuHause> especially in vehicle chains, where the first vehicle changed, but the rest did not
00:07:21 <welshdragon> Eddi|zuHause it's a trunk bug too in 0.7.1 (i don't have 0.7.2
00:08:15 <Eddi|zuHause> 0.7 is not trunk
00:08:43 <welshdragon> trunk as in a nightly?
00:08:59 <Ammler> trunk as not 0.7 branch :-)
00:09:31 <Eddi|zuHause> yes. a nightly that is not older than a few days will probably do
00:10:17 <Ammler> well, but I guess, also bugs for 0.7 can be reported.
00:10:19 <Yexo> Ammler: it looks like those planes already have NW<>SE land/takeoff sprites :)
00:10:51 <Yexo> also bugs for 0.7 can be reported. <- in that case use the latest stable, so 0.7.2
00:11:33 <welshdragon> yeah, it'll have to wait
00:11:40 <welshdragon> im on a 512kbps connection
00:11:58 <welshdragon> (supposedly broadband)
00:12:21 <Eddi|zuHause> thst's still better thsn mine
00:13:04 <OwenS> Thats about what I get during peak hours. When my ISP throttles me for using too much bandwidth ;p
00:13:38 <OwenS> Normally I get about 8mbit/s... which I ought to ring them up about since this is a 16mb/s plan
00:13:43 <OwenS> 20mb/s even... wtf brain?
00:20:18 <MyCatVerbs> OwenS: heehee, NARF.
00:51:02 *** roboboy has joined #openttd
00:58:47 *** KenjiE20|LT has joined #openttd
01:03:23 *** KenjiE20|LT has joined #openttd
01:15:04 *** roboboy has joined #openttd
03:13:35 *** TinoDidriksen has joined #openttd
03:43:56 *** TinoDidriksen has joined #openttd
03:46:11 *** Rexxars has joined #openttd
04:37:54 *** Phoenix_the_II has joined #openttd
05:16:31 *** thingwath has joined #openttd
05:25:41 *** roboboy has joined #openttd
06:15:31 *** Cybertinus has joined #openttd
06:18:23 *** Alberth has joined #openttd
06:52:41 *** Terkhen has joined #openttd
07:28:51 *** Progman has joined #openttd
07:47:30 *** Mesolicious has joined #openttd
07:47:35 *** Mesolicious is now known as Aankhen``
07:52:24 *** roboboy has joined #openttd
07:57:51 *** roboboy has joined #openttd
08:04:08 *** roboboy has joined #openttd
08:21:10 *** thingwath has joined #openttd
09:11:35 *** fonsinchen has joined #openttd
09:31:03 <CIA-1> OpenTTD: rubidium * r17304 /trunk/src/rail_gui.cpp: -Codechange: move (more) drawing code out of the rail depot picker's OnPaint.
09:38:03 *** [com]buster has joined #openttd
09:44:43 <planetmaker> paste.openttd.org: Warning: mysql_pconnect() [function.mysql-pconnect]: Lost connection to MySQL server during query in /www/openttd.org/paste/lib/pastebin/db.mysql.class.php on line 49
09:44:45 <planetmaker> Unable to connect to database
09:47:09 <Rubidium> (and it's the only service I can't boot)
09:47:30 <Rubidium> and TB isn't here (AFAIK)
09:48:11 <Rubidium> though I blame spammers
09:50:59 <Alberth> that is always a safe group to blame :p
09:51:19 <planetmaker> nevertheless a good morning to you all :-)
09:53:04 <Alberth> that seems like a safe bet, not much can happen in those last 10 minutes :p
09:53:52 *** fonsinchen has joined #openttd
09:55:37 *** oskari89 has joined #openttd
09:59:55 <planetmaker> depends on your time reference, though, Alberth ;-)
10:03:36 *** Chris_Booth has joined #openttd
10:12:30 *** DJ_Nekkid has joined #openttd
10:27:00 *** tux_mark_5 has joined #openttd
10:28:24 *** |Jeroen| has joined #openttd
10:39:10 *** Phoenix_the_II has quit IRC
10:39:23 *** worldemar has joined #openttd
10:40:02 *** Phoenix_the_II has joined #openttd
10:44:58 *** Chris_Booth is now known as Guest675
10:45:05 *** Chris_Booth has joined #openttd
10:45:54 *** Dred_furst has joined #openttd
10:46:14 *** oskari89 has joined #openttd
10:46:51 <planetmaker> Can the sprites used by an industry be made time dependent?
10:47:25 <planetmaker> Like, show sprite A for dates < 1920, sprite B for 1920 ... 2000 and sprite B for date > 2000 ?
10:47:36 <planetmaker> possibly via callback?
10:48:44 <Eddi|zuHause> why wouldn't it?
10:51:00 <planetmaker> for one and the same industry?
10:51:05 <planetmaker> not for two different ones
10:52:48 <planetmaker> I've never seen it.
10:52:55 <planetmaker> Doesn't need to mean much, but...
10:55:01 *** KenjiE20 has joined #openttd
11:00:47 *** Terkhen has joined #openttd
11:18:17 <CIA-1> OpenTTD: smatz * r17305 /branches/0.7/src/ai/api/ (6 files):
11:18:17 <CIA-1> OpenTTD: [0.7] -Backport from trunk:
11:18:17 <CIA-1> OpenTTD: -Fix: [NoAI] AITown::GetLastMonthTransported didn't work as documented at all, make it return what AITown::GetLastMonthProduction did
11:18:17 <CIA-1> OpenTTD: -Change: [NoAI] AITown::GetLastMonthProduction now returns the same value as AITown::GetMaxProduction
11:18:17 <CIA-1> OpenTTD: -Add: [NoAI] AITown::GetLastMonthTransportedPercentage and AIIndustry::GetLastMonthTransportedPercentage
11:44:23 <Eddi|zuHause> man, it's totally cold here...
11:47:04 <Alberth> Give your CPUs more work.
11:53:02 *** Azrael- has joined #openttd
11:56:41 *** roboboy has joined #openttd
12:00:42 *** DJNekkid has joined #openttd
12:03:49 *** roboboy has joined #openttd
12:04:43 *** Terkhen has joined #openttd
12:08:46 *** Singaporekid has joined #openttd
12:11:01 *** roboboy has joined #openttd
12:19:43 *** roboboy has joined #openttd
12:47:01 *** fonsinchen has joined #openttd
13:00:02 * Alberth grabs some heat from the room, and sends it to Eddi|zuHause
13:00:16 * SirSquid1ess yoinks some of the heat
13:00:45 <Alberth> No fighting, there is enough for everybody
13:04:05 *** SirSquid1ess is now known as SirSquidness
13:04:09 <SirSquidness> sweet. I could do with a bit.
13:25:22 *** HerzogDeXtEr has joined #openttd
13:37:44 *** KritiK_ has joined #openttd
13:40:36 *** KritiK_ is now known as KritiK
13:44:45 *** HerzogDeXtEr1 has joined #openttd
13:55:13 *** Dred_furst has joined #openttd
13:55:58 *** Grelouk has joined #openttd
14:08:33 *** KritiK_ has joined #openttd
14:14:13 *** KritiK_ is now known as KritiK
14:26:30 *** Chris_Booth has joined #openttd
14:45:51 *** Chris_Booth is now known as Guest702
14:45:57 *** Chris_Booth has joined #openttd
14:46:25 *** ChanServ sets mode: +v glx_
14:46:49 *** glx is now known as Guest703
14:49:57 *** DJ_Nekkid has joined #openttd
14:50:19 *** ChanServ sets mode: +v glx_
14:51:48 *** glx is now known as Guest704
15:06:51 *** Coco-Banana-Man has joined #openttd
15:50:17 *** KritiK_ has joined #openttd
15:53:11 *** KritiK_ is now known as KritiK
16:22:57 <Aankhen``> Man, it would be nice to have a blueprints feature or something like that so that I could discover that my two billion tile station layout is one tile too far to the left before I start building it.
16:23:55 <MyCatVerbs> The whole point is that you do that, then you take a screenshot and post it here or on the wiki or something.
16:24:09 <MyCatVerbs> Then everybody laughs at you and you cry a lot and we all enjoy the schadenfreude.
16:24:27 <MyCatVerbs> But more seriously, just take a savegame before you embark on a big project like that?
16:24:51 <MyCatVerbs> The main issue is the amount of time it takes to build the layout in the editor, not the expenditure in game.
16:25:44 <SpComb> you could write the templates/blueprints as a feature of some AI
16:25:49 <SpComb> and then just let that AI play on your behalf
16:26:04 <Aankhen``> MyCatVerbs: I really meant a general planning feature. So for instance, in this ideal world, I'd select a bunch of tiles and shift them one tile to the right, rather than rebuilding the entire thing.
16:26:51 <MyCatVerbs> Aankhen``: ah, I see what you mean.
16:27:55 <Aankhen``> SpComb: Yeah, great idea— no wait.
16:28:37 <SpComb> need to bring a bit of tension and excitement into the game
16:29:11 <Aankhen``> You get your kicks one way, I get them another…
16:31:00 <SpComb> there's been plenty of controversy re copy-paste features
16:31:31 <SpComb> many people are of the opinion that no such feature should exist
16:31:58 <Aankhen``> Welp, I don't want to add to any controversy. It was just an idle thought.
16:34:25 *** KritiK_ has joined #openttd
16:37:05 *** KritiK_ is now known as KritiK
16:37:19 <SmatZ> + if (OSErr err = Gestalt(gestaltSystemVersion, &systemVersion) != noErr) {
16:37:34 <SmatZ> interesting construct, OSErr is a bool?
16:37:50 <planetmaker> at least I don't think
16:38:07 <planetmaker> IIRC it's an uint or ulongint
16:38:19 <planetmaker> as it reports the type of error
16:38:20 <SmatZ> actually I wonder it's allowed to declare a variable inside if statement :)
16:38:29 <planetmaker> it compiles and works :-P
16:39:01 <planetmaker> I had it first outside, but then moved closest to its use :-)
16:39:35 <SmatZ> OSErr err = (Gestalt(gestaltSystemVersion, &systemVersion) != noErr)
16:39:41 <SmatZ> this is the behaviour it would have in C
16:40:04 <planetmaker> you mean braces like if ((OSErr err = Gestalt(gestaltSystemVersion, &systemVersion)) != noErr) {
16:40:50 <planetmaker> honestly, I didn't look up the type of OSErr. I copied it from the example in the quoted link
16:41:31 <planetmaker> but, of course, you're right... your last version of it looks more sane to me :-)
16:42:42 <SmatZ> I wonder what is the scope of 'err' in this case
16:43:05 <planetmaker> local. Not used elsewhere
16:43:20 <SmatZ> why does it need to be used at all?
16:43:34 <planetmaker> ... it acutally doesn't :-)
16:43:35 <SmatZ> just "if (Gestalt(gestaltSystemVersion, &systemVersion) != noErr)" wouldn't work?
16:44:02 *** Chris_Booth is now known as Guest719
16:44:09 *** Chris_Booth has joined #openttd
16:44:45 <planetmaker> let's see what compile tells me
16:44:58 <SmatZ> + if ((err = Gestalt(gestaltSystemVersionMajor, &versionMajor)) != noErr) versionMajor = 0;
16:45:32 <SmatZ> I suppose err isn't needed here too
16:45:58 <SmatZ> (systemVersion & 0x000F); GB would be nicer :)
16:46:44 <planetmaker> GB instead of the masking?
16:47:26 <SmatZ> (systemVersion & 0xF000) >> 12) ==> GB(systemVersion, 12, 4)
16:47:49 <SmatZ> hmm that file breaks coding style anyway...
16:48:22 <SmatZ> so no comments from me about camelCase or comment style :-x
16:51:29 <planetmaker> my patch actually doesn't change too much there, just the version part - and it moves a few var declarations, mainly as I stumbled upon them...
16:51:57 <planetmaker> I agree that it needs major cleanup.
16:52:15 <planetmaker> also all those magic string lengths. They'd need unification and declaration in one place.
16:52:30 <planetmaker> else, if you change one, you're easily getting seg faults
16:52:31 *** Phoenix_the_II has quit IRC
16:52:39 <planetmaker> been there, seen that :-P
16:52:41 *** Phoenix_the_II has joined #openttd
16:53:32 <planetmaker> I didn't want to do a complete overhaul right now.
16:53:39 <planetmaker> I rather do that piecewise :-)
16:56:28 <SmatZ> planetmaker: what was broken in r8605?
16:56:54 <SmatZ> I think your svn blame doesn't work correctly ;)
16:57:02 <SmatZ> in r8605, the file was just moved
16:57:23 <planetmaker> oh, hm... it was the change which last affected that line
16:57:41 <planetmaker> talks about conversion away from objc to c++
16:58:02 <planetmaker> so... maybe even earlier from a functional POV
16:58:17 *** Aankhen`` has joined #openttd
17:00:45 <CIA-1> OpenTTD: alberth * r17306 /trunk/src/gfx.cpp: -Fix: DrawStringMultiLine() computed available width and height wrongly.
17:01:03 <planetmaker> SmatZ: updated patch uploaded
17:01:52 <CIA-1> OpenTTD: yexo * r17307 /trunk/src/ai/api/ai_rail.cpp: -Fix [NoAI]: AIRail::RemoveRailTrack returned ERR_PRECONDITION_ERROR for road/rail-crossings
17:02:31 <planetmaker> actually, I *think* the version is checked in other places, too.
17:02:52 <planetmaker> So, a proper fix would be to move these version determination to separate functions...
17:03:02 <planetmaker> but that'll need more time...
17:11:46 *** DR_Jekyll has joined #openttd
17:24:04 * AC6000 materializes at 30,000 ft. ...
17:32:21 <planetmaker> within your own interest I hope that you either materialize within the pressurized compartement of a plane or with a propper jumpsuit and a working oxygen supply
17:32:32 *** Doorslammer has joined #openttd
17:37:29 *** JFBelugas has joined #openttd
17:39:16 *** roboboy has joined #openttd
17:45:33 <CIA-1> OpenTTD: translators * r17308 /trunk/src/lang/ (french.txt german.txt indonesian.txt italian.txt russian.txt):
17:45:33 <CIA-1> OpenTTD: -Update from WebTranslator v3.0:
17:45:33 <CIA-1> OpenTTD: french - 22 changes by glx
17:45:33 <CIA-1> OpenTTD: german - 28 changes by Roujin
17:45:33 <CIA-1> OpenTTD: indonesian - 7 changes by fanioz
17:45:34 <CIA-1> OpenTTD: italian - 44 changes by lorenzodv
17:45:34 <CIA-1> OpenTTD: russian - 7 changes by MajestiC
17:46:10 *** roboboy has joined #openttd
18:52:35 *** Chris_Booth has joined #openttd
18:57:06 <Grelouk> any admin of ex's city mania ?
18:58:49 *** Phoenix_the_II has quit IRC
18:59:11 *** Phoenix_the_II has joined #openttd
19:01:53 <Grelouk> yeah, i could use one right now..
19:02:03 <Grelouk> a guy is blocking me everywhere
19:22:56 *** KritiK_ has joined #openttd
19:25:07 *** KritiK_ is now known as KritiK
19:27:57 <asilv> question: if the callback 2E custom cargo generation for houses is used, does the normal cargo generation for passengers/mail happen too or just the callback?
19:49:53 *** Default_ has joined #openttd
20:01:33 <CIA-1> OpenTTD: smatz * r17309 /trunk/src/misc_gui.cpp: -Fix: not all non-ASCII characters were entered with escapes in the About window
20:06:33 <planetmaker> hm... the version detection for macos can indeed be more unified.
20:06:44 <planetmaker> let's see how it works
20:07:18 *** Default_ has joined #openttd
20:08:57 <SmatZ> planetmaker: about FS#3156, Gestalt sounds very germanish :) are you sure it's available in non-German OSX?
20:09:24 <SmatZ> seems so, interesting :)
20:09:27 <planetmaker> SmatZ: I'm pretty sure. It seems to be part of the official API
20:09:48 <planetmaker> :-) But I agree that it sounds very German
20:10:14 <z-matrica> someone know of error correction layer for hdd-s below raid/filesystem level for linux?
20:11:01 *** z-matrica is now known as z-MaTRiX
20:11:26 <DorpsGek> SmatZ: Bjarni was last seen in #openttd 5 weeks, 4 days, 20 hours, 22 minutes, and 15 seconds ago: <Bjarni> as long as it doesn't happen every 5th minute
20:11:46 <SmatZ> z-MaTRiX: every HDD has its own error-recovery system
20:12:03 <SmatZ> data on disk are encoded with some ECC
20:12:19 <z-MaTRiX> sure, but they are likely to go to badsector land
20:12:56 <z-MaTRiX> had some hdds that did it
20:13:01 <SmatZ> bad blocks are listed in the filesystem (afaik)
20:13:13 <SmatZ> eg. you can add list of bad blocks to mkfs
20:13:24 <z-MaTRiX> well, you use a hdd, write data on it, then you cannot read back
20:13:49 <SmatZ> try some utility from HDD manufaturer
20:13:54 <z-MaTRiX> no matter if it relocates it after
20:14:09 <SmatZ> it should be relocated before it becomes unreadable
20:14:19 <SmatZ> 5 Reallocated_Sector_Ct 0x0033 099 099 036 Pre-fail Always - 69
20:14:24 <SmatZ> and no files are broken
20:14:29 <z-MaTRiX> yes, that i think is not really possible
20:14:32 <SmatZ> 195 Hardware_ECC_Recovered 0x001a 024 024 000 Old_age Always - 89151901
20:15:04 <SmatZ> that ECC can have code distance, say, 4
20:15:29 <SmatZ> and if there are two or more fixed bits in given record, the sector may be reallocated
20:15:33 <z-MaTRiX> how about, you have a sector you wrote 2 years ago, and never read out, goes bad, and it is never discovered?
20:15:57 <SmatZ> but HDDs do some "offline" tests too
20:16:18 <z-MaTRiX> sure, that's why was wondering to add 16% of error correction code in a transparent device layer
20:16:29 *** roboboy has joined #openttd
20:16:45 <SmatZ> who is adding "16% of error correction code" where?
20:16:57 <z-MaTRiX> they can do offline tests only offline, not 24/7, and not without being powered
20:17:31 <SmatZ> well, maybe the HDD does tests even when it's being used
20:18:35 <z-MaTRiX> c1&c2 works nice on a DVD disc for example
20:19:35 <SmatZ> sorry, I don't know what c1 and c2 is
20:20:23 <SmatZ> but you get some c1 and c2 rates when reading the disc
20:21:15 <SmatZ> hmm yeah I was taught that...
20:23:20 *** roboboy has joined #openttd
20:23:28 *** ChanServ sets mode: +v tokai
20:23:30 <SmatZ> I guess it has something to do with re-encoding already encoded data
20:23:51 <SmatZ> like, ECC on "byte" levevel and ECC on "sector" level
20:24:03 <SmatZ> so C1 is when it can be corrected on the byte level
20:24:11 <SmatZ> and C2 when it's corrected on the sector level
20:26:21 <z-MaTRiX> Cross Interleave Reed-Solomon Code is a method of error detection and correction employing Reed-Solomon parity bytes together with different interleaving, or delay, patterns that assists in error correction by distributing concentrated read errors over multiple frames that then form the input to a CIRC decoder. After the first deinterleave, CD discs correct small read errors at the C1 level, followed by a second deinterleave and correction of larg
20:26:21 <z-MaTRiX> e read errors at the C2 level.
20:30:16 <SmatZ> I was wondering you are so alive and full of ideas that late in the night :)
20:30:37 <z-MaTRiX> btw time.nist.gov has accurate ntp
20:31:08 <Gold24> so does XX.pool.ntp.org
20:31:15 <SmatZ> 29 Aug 22:31:06 ntpdate[29266]: step time server 192.43.244.18 offset -2.028262 sec
20:31:26 <Xaroth> pool.ntp.org will give a random anyhow :P
20:31:28 <Gold24> where XX is country code
20:32:24 <Xaroth> and country code is irrelevant, if i connect to our datacenter 40km from here the connection will actually cross 7 borders.
20:32:31 <SmatZ> I have local ntpserver running though
20:32:50 <SmatZ> 29 Aug 22:32:20 ntpdate[29283]: adjust time server 192.168.0.5 offset -0.000625 sec
20:44:51 *** |Terkhen| has joined #openttd
20:47:02 <planetmaker> why would I make a function static inline?
20:51:49 <MyCatVerbs> If you make static functions then the compiler will pick whether or not to inline them for you. Modern compilers will usually do a better job of making this decision than you owuld've.
20:52:28 <MyCatVerbs> Yes. I can demonstrate gcc doing it if you like.
20:52:59 <MyCatVerbs> That's an interesting position to take.
20:53:25 * planetmaker wonders whether GetMacOSVersion is time critical, though
20:53:39 <planetmaker> rather GetMacOSVersionAtLeast
20:53:57 <planetmaker> doesn't matter I guess
20:54:11 <SmatZ> gcc doesn't take into account constant parameters
20:54:42 <SmatZ> it just guesses how many ops given function will generate
20:54:43 <CIA-1> OpenTTD: yexo * r17310 /trunk/src/misc_gui.cpp: -Codechange: about window uses nested widget tree
20:55:02 <SmatZ> in OTTD, there are quite often functions called with constant parameters
20:55:16 <SmatZ> or parameters/variables that would become constant after inlining
20:55:22 <SmatZ> gcc doesn't take that into account
20:55:32 <SmatZ> also, gcc stops inling after some point
20:55:49 <SmatZ> (some --param-*-growth or so)
20:55:50 <MyCatVerbs> Yeah, all compilers have to. Otherwise they go into infinite loops.
20:56:05 <SmatZ> so it doesn't compile even the smallest possible function
20:56:09 <MyCatVerbs> SmatZ: How come the constant-folding doesn't catch that?
20:56:34 <MyCatVerbs> These are functions with no side effects, right?
20:58:14 <MyCatVerbs> Annotate all the side-effect-less ones with __attribute__((pure)) or (where the stronger condition holds) __attribute__((const)) and it should in theory get better
20:59:23 <SmatZ> I think this isn't the most common issue
21:00:51 <MyCatVerbs> SmatZ: hrmn. Oh well. I've never had to force gcc to inline short functions, but then I've not written anything nearly as large as OTTD in C.
21:01:54 <SmatZ> MyCatVerbs: one type of problems: there is a switch in "inline" functions that depends on value in map array
21:02:10 <SmatZ> when those functions are inlined, only one access to map array is needed
21:02:26 <SmatZ> when they are not, more accesses and more conditional jumps are needed
21:02:37 <SmatZ> this happens quite often
21:03:15 <SmatZ> other common problem is the above mentioned, gcc stops inlining even very small fuctnions after some point
21:03:32 <SmatZ> causing code grows and slwer execution
21:04:57 <SmatZ> other problem is non-inlining functions like
21:05:40 <SmatZ> which has many parameters
21:05:49 <SmatZ> that have to be passed via stack
21:07:36 <MyCatVerbs> Wait, how'd TTD end up with functions of more than six-ish parameters?
21:08:29 <SmatZ> x86 linux ABI specifies 6 registers for passing parameters?
21:10:43 <SmatZ> well, if you are talking about x86_64...
21:11:51 *** Vikthor has joined #openttd
21:13:04 <MyCatVerbs> SmatZ: IIRC fastcall gets you at least four in registers?
21:13:34 <SmatZ> maybe... sounds windows-ish though :)
21:13:56 <SmatZ> gcc supports __attribute__((regparam=n)) or so too
21:14:12 <SmatZ> adding attributes isn't the way to go (code niceness and such)
21:14:55 <MyCatVerbs> SmatZ: dunno, I'm pretty fond of __attribute__((const)) and unused.
21:15:26 <Yexo> MyCatVerbs: the compiler should be able to figure that out on it's own
21:15:26 <Vikthor> SmatZ: Regarding the r17309, I think paskys second name is Baudiš
21:16:01 <SmatZ> Vikthor: :) I noticed that too late :(
21:16:17 <MyCatVerbs> Yexo: depends how involved the function is. The halting problem is still a real cow. :)
21:17:00 <SmatZ> MyCatVerbs: OTTD uses just __attribute__((noreturn)) and __attribute__((format())) for functions...
21:17:14 <SmatZ> I prefer too add "inline" instead ;)
21:17:32 <SmatZ> FORCEINLINE should be transformed to "inline __attribute__((always_inline))" in future
21:18:03 <SmatZ> someone (tm) has to check its usage before that though
21:19:27 <SmatZ> thing I am thinking about is un-virtualizing functions
21:19:36 <SmatZ> I wonder why C++ doesn't specify that :-x
21:20:48 * MyCatVerbs hands SmatZ a ™ symbol.
21:21:11 <SmatZ> way to un-virtualize functions
21:21:20 <MyCatVerbs> You're going to make a whole pile of currently polymorphically overloaded functions statically resolved instead?
21:21:42 <MyCatVerbs> Because Bjarne was smoking something heavy when he designed the language.
21:22:02 <SmatZ> doing t->Train::IsPrimaryVehicle() is ugly
21:22:04 <MyCatVerbs> Fifteen grams of crack cocaine, wrapped around a lead brick.
21:22:20 <SmatZ> when un-virtualizing IsPrimaryVehicle() in Train would do the same job
21:23:18 <MyCatVerbs> Er, why's it virtual in the first place?
21:23:47 <SmatZ> so Vehicle *v; v->IsPrimaryVehicle() can be used?
21:23:54 <MyCatVerbs> You get the right behavoir either way, right? I presume that Train subclasses must never override IsPrimaryVehicle()?
21:24:08 <Yexo> MyCatVerbs: there are no train subclasses
21:24:24 <MyCatVerbs> You're leaving it virtual in the root but not in the leaves.
21:24:42 <SmatZ> I really do hope the LTO branch will solve all these problems
21:24:57 <planetmaker> new version of the version detection out
21:25:00 <SmatZ> yeah, but its virtuality is inherited
21:25:23 <Rubidium> ghehe... let them first compile OpenTTD without crashing and then wonder about the result of the compilation :)
21:26:07 <SmatZ> planetmaker: gcc-lto :)
21:26:19 * planetmaker has no knowledge of it.
21:26:44 <MyCatVerbs> Link Time Optimization.
21:27:04 <planetmaker> sounds... neat and dodgy at the same time
21:27:09 <MyCatVerbs> Have you ever tried concatenating OpenTTD's entire sources? :D
21:27:22 <planetmaker> no need, is there?
21:27:34 <SmatZ> and failed, of course :)
21:27:39 <planetmaker> why do you do that, SmatZ?
21:27:40 <MyCatVerbs> Then compile with -fwhole-program. :)
21:28:02 <MyCatVerbs> planetmaker: Because otherwise you only get optimization up to the boundaries of .o files, so to speak.
21:28:14 <SmatZ> MyCatVerbs: somehow it failed
21:28:19 <MyCatVerbs> Any optimization that would require analysis of more than one module simultaneously can't be done.
21:28:49 <CIA-1> OpenTTD: yexo * r17311 /trunk/src/ (55 files in 3 dirs): -Change: rename STR_WHITE_STRINGN to STR_WHITE_STRING
21:29:02 <MyCatVerbs> Like inlining functions defined in module A which are used in module B. AFAIK that's the most common issue?
21:29:26 <planetmaker> writing this version-fix patch, btw, was good :-)
21:29:30 <SmatZ> I don't remember why, I tried it quite long time ago
21:29:46 <planetmaker> I learnt something about passing parameters by reference vs. by value ;-)
21:30:02 <MyCatVerbs> SmatZ: Pity, but not surprising. Most projects with interesting build systems can't easily be just catted together. :/
21:32:06 <glx> MyCatVerbs: just use MSVC, it does LTO :)
21:32:29 <MyCatVerbs> glx: but that's good to know. ^^
21:33:02 *** Brianetta has joined #openttd
21:33:28 <planetmaker> that's interesting to know :-)
21:34:01 <CIA-1> OpenTTD: rubidium * r17312 /trunk/src/road_gui.cpp: -Codechange: make the road toolbars and depot picker nested
21:46:40 *** Nite_Owl has joined #openttd
21:47:23 *** [com]buster is now known as Combuster
22:18:05 <Eddi|zuHause> # da simmer dabei
22:18:14 <Eddi|zuHause> # das ist priiiima
22:22:56 <Eddi|zuHause> why must it be that whenever you come from a place where people are, you smell like you smoked?
22:23:27 <Aankhen``> Erm, because you did?
22:24:24 <Eddi|zuHause> i would be a militant non-smoker, if i wasn't so lazy
22:24:38 *** Progman has joined #openttd
22:24:44 <CIA-1> OpenTTD: rubidium * r17313 /trunk/src/road_gui.cpp: -Codechange: make the road station picker nested
22:25:05 <Aankhen``> Eddi|zuHause: So you just laze around on a hammock and smoke all day instead? :-P
22:25:31 <SmatZ> I used to be tolerant to smell of cigarets in pubs... but recently, I am having feelings of vomit when I smell that :-x
22:25:33 <Eddi|zuHause> someone should forbid smoking in all public places
22:25:44 <SmatZ> still I am tolerant, I just don't go to pubs...
22:26:16 <Aankhen``> It's illegal where I live, but laws here are more like guidelines.
22:26:29 <Eddi|zuHause> someone showed me a photo of a "smoking cabin"
22:26:30 <SmatZ> I am against laws of this kind
22:26:49 <SmatZ> the owner of pub should decide if, or when, people can smoke
22:27:19 <Aankhen``> Why? They don't exist in a vacuum.
22:27:35 <Aankhen``> That would make sense if they had their own private atmosphere.
22:27:49 <Rubidium> SmatZ: so the owner of a pub should decide if, or when, people can spit eachother in the face?
22:28:17 * Rubidium is happy smoking is generally banned now
22:28:41 <SmatZ> Rubidium: spitting in face degrades people... when you go in box ring, you can expect being hit
22:28:53 <SmatZ> that's what you expect when you go to pub where smoking is allowed
22:28:54 <Eddi|zuHause> it is not banned in "open" places here, which sadly includes beer tents...
22:29:03 <Aankhen``> Uh, smoking *kills* people. I think that's a little more serious, don't you?
22:29:18 <SmatZ> when you don't want to be hit/smell smoke, you got o pub/restaurant where smoking is forbidden
22:29:28 <SmatZ> Aankhen``: we are born to die
22:29:41 <Aankhen``> Wonder why everyone makes such a fuss about murder then.
22:29:44 <SmatZ> life without any dangers and risks would be boring
22:30:06 <Rubidium> SmatZ: how often are you to actually choose which restaurant to go to? Most of the time I get invited, as such I have no way to choose restaurants
22:30:10 <Eddi|zuHause> SmatZ: we had this siutation for decades... if there is no law against smoking, there are no pubs that forbid smoking
22:30:33 <Eddi|zuHause> no sane pub owner would do that, and push out a large part of his customer base
22:30:57 <SmatZ> many pubs have non-smoking and smoking areas
22:31:03 <Rubidium> on the other hand, the lots of smoking stuff made me not go to pubs/restaurants/whatever
22:31:10 <SmatZ> also, you can sit outside
22:31:12 <Eddi|zuHause> and if you know smokers, you are forced to go to a smoking area then
22:31:21 <Rubidium> SmatZ: sit outside where people are smoking :(
22:31:41 <Eddi|zuHause> instead if smoking is generally forbidden, you can sit with the smokers, and when they "feel the urge", they go outside
22:31:47 <SmatZ> Rubidium: you have to hope in good wind direction then :(
22:31:57 <Rubidium> it's annoying enough that the person 3/4 houses down the road occasionally smokes outside
22:32:25 <Eddi|zuHause> what's a 3/4 house? :p
22:32:41 <Eddi|zuHause> you have quadruple-houses? :p
22:32:55 <Aankhen``> It's where they forgot to put in one wall.
22:33:37 <SmatZ> well... I am occasional-smoker... I haven't smoked for almost a year though
22:33:38 <Eddi|zuHause> i'm going to shower...
22:33:41 <Rubidium> it only results in me coughing like hell, which in effect means that I better not open my windows in summer to cool the house
22:33:43 <SmatZ> I dislike the smell now :-p
22:34:33 <SmatZ> though I sometimes smoke a cigar
22:34:51 <glx> not good for the others :)
22:35:11 <SmatZ> we go outside because of that though
22:36:02 <glx> here we have to go outside too
22:36:23 <SmatZ> I can't imagine such laws in France
22:36:44 <glx> it's to protect employees
22:48:26 <Eddi|zuHause> the "protection" is of not much use, when you have to walk through a (now much higher concentrated) cloud of smoke when you walk through the door
22:57:13 <Yexo> Rubidium / DaleStan: I'm compiling nforenum and it looks like Makefile overwrites the CFLAGS from Makefile.local
22:57:36 <Yexo> shouldn't the "-include ${MAKEFILELOCAL}" be moved a few lines down?
23:02:31 *** Phoenix_the_II has quit IRC
23:02:39 *** Phoenix_the_II has joined #openttd
23:04:56 <Nite_Owl> is it possible in terms of NFO coding for a tram to have one speed when its tracks are off road and another speed when its tracks are on road ?
23:05:20 <Yexo> Nite_Owl: I don't think that's currently possible
23:05:53 <Eddi|zuHause> that should rather be a feature of the road types
23:05:59 <Nite_Owl> I did not think so but it cannot hurt to ask
23:07:09 <Eddi|zuHause> like a town would generally build roads that have a speed limit of 50km/h
23:07:24 <Eddi|zuHause> and faster roads cannot be used to grow houses next to them
23:08:03 <Eddi|zuHause> naybe the town could "downgrade" existing roads
23:09:04 <Nite_Owl> need to feed - later all
23:10:01 <MyCatVerbs> Eddi|zuHause: that would make me want to bulldoze every town off the map.
23:10:23 <MyCatVerbs> I mean sure you could implement that, but not unless you're going to implement anonymous tactical nukes too.
23:11:55 <Eddi|zuHause> i'm not sure what your problem is. that is exactly how every town in the world works
23:13:13 <Eddi|zuHause> if a town grows along an existing road, it limits that road to town speed limit
23:13:43 <MyCatVerbs> My problem is that my road networks would spontaneously become mind-bendingly slow and useless - at exactly the same time as the towns get large enough for it to be profitable to make use of them.
23:14:31 <MyCatVerbs> Fun over realism and all that jazz.
23:15:19 <MyCatVerbs> Oh and also it'd muck the game balance up quite violently - making road vehicles really embarrassingly unprofitable. It'd have to be balanced out some other way.
23:16:38 <Eddi|zuHause> the balance would be "you can't bulldoze half the town to build a train station"
23:17:26 <Eddi|zuHause> and there could be a way to have expressways through a town
23:17:39 <Eddi|zuHause> but the town won't grow along these
23:18:59 <MyCatVerbs> Meh, town councils are annoying.
23:19:21 <MyCatVerbs> I'm telling you, tactical nuclear warhead left anonymously in a suitcase somewhere.
23:22:07 <Eddi|zuHause> i really don't understand you... you say "X must be balanced" and then you say "but that balance is annoying"
23:22:20 <Yexo> it'd muck the game balance up <- what balance?
23:22:45 <Eddi|zuHause> if you want a nuclear warhead, go play command and conquer...
23:23:08 <MyCatVerbs> Eddi|zuHause: I should be clearer. I mean town councils are annoying. Making them more annoying would probably warrant adding a tac-nuke option.
23:23:52 <Eddi|zuHause> and there exists such an option, it's called "magic bulldozer"
23:25:11 <MyCatVerbs> You're right, I should play Command & Conquer. :/
23:33:18 *** Eddi|zuHause has joined #openttd
23:34:25 *** Nicholas has joined #openttd
23:35:19 <Nicholas> It seems that OpenTTD is broken in fullscreen only on Mac OSX Snow Leopard
23:36:48 <Nicholas> Works fine in windowed though
23:37:31 <Eddi|zuHause> because macos is shit
23:38:28 <Eddi|zuHause> i read that as a "yes, i agree."
23:38:49 <MyCatVerbs> That could be construed as a little bit hostile.
23:39:47 <Eddi|zuHause> i'm the least hostile person in this entire channel
23:40:15 <MyCatVerbs> ...I could start hugging everybody who comse in, and then that wouldn't be true any more.
23:40:20 *** Coco-Banana-Man has quit IRC
23:41:39 <Eddi|zuHause> if an unprovoked hug can not be considered an act of hostility by a person unfamiliar with the culture
23:42:11 <MyCatVerbs> You raise a good point. Hrmn.
23:42:31 <MyCatVerbs> How about we dose them with MDMA first so they'll get all cuddly and appreciate it?
23:43:18 <Eddi|zuHause> i have no idea what that is, but it sounds not very comforting
23:46:28 <Eddi|zuHause> is that how you meet your "girlfriends"?
23:49:41 <MyCatVerbs> Eddi|zuHause: you seem to assume some really alarming things about me. oO
23:51:24 <glx> we didn't had enough time to suggest the usual solutions
23:53:14 <glx> [01:35:30] <Nicholas> It seems that OpenTTD is broken in fullscreen only on Mac OSX Snow Leopard <-- though it's a new version so it's not surprising they break our stuff again ;)
23:53:34 <Eddi|zuHause> MyCatVerbs: you mean that from your presence in this channel i derive that you are probably male, are into women, and have trouble approaching them?
23:54:43 <MyCatVerbs> Eddi|zuHause: that much was obvious. The alarming bit was where you semi-accused me of date-raping strangers. oO
23:55:21 <Eddi|zuHause> well _i_ was not the person coming up with the idea of party drugs...
23:55:21 <glx> I understood the same thing
23:55:47 *** TinoDidriksen has joined #openttd
23:56:34 <MyCatVerbs> Eddi|zuHause: what's wrong with giving out free drugs on IRC?
23:57:22 <Eddi|zuHause> probably as much as free viruses or free spam
23:58:12 *** Dred_furst has joined #openttd
continue to next day ⏵
