IRC logs for #openttd.dev on OFTC at 2013-02-24
⏴ go to previous day
01:13:59 *** Supercheese has joined #openttd.dev
09:02:31 *** Alberth has joined #openttd.dev
09:02:31 *** ChanServ sets mode: +v Alberth
10:57:40 *** frosch123 has joined #openttd.dev
10:57:40 *** ChanServ sets mode: +v frosch123
11:31:33 *** fonsinchen has joined #openttd.dev
11:31:33 *** ChanServ sets mode: +v fonsinchen
12:57:53 *** frosch123 has joined #openttd.dev
12:57:53 *** ChanServ sets mode: +v frosch123
13:50:47 *** ntoskrnl has joined #openttd.dev
13:58:13 <frosch123> but it should work find for now
13:59:10 <frosch123> there might be obscure newgrfs out there which use subtypes in a weird way, so i guess this is only for 1.4
13:59:24 <frosch123> so we have one year time to get those weird grfs reported :p
14:01:18 *** LordAro has joined #openttd.dev
14:01:18 *** ChanServ sets mode: +v LordAro
14:06:31 <Alberth> 20, commit message: -Remove: Ordered refit with subtypes, since the cases where it worked were corner cases rather than the main case. (added , moved 'rather') you may want to consider s/main/general/
14:19:01 <Alberth> 30 inline bool operator != ( <-- s/ (/(/ (ie remove the space in front of the parenthesis ? (also with ==)
14:19:50 <Alberth> I somewhat wonder whether it would be better to express one of them in the negation of the other one
14:23:15 <frosch123> the operator space coding style is quite inconsistent (as expected), but the space on both sides seems to be the majority
14:25:10 <Alberth> 40 last + line needed?
14:28:53 <Alberth> 50 change of 'sel' is not really reflected in the doxygen text, eg "\c -1 if nothing is selected" is sufficient? I'd expect an index, or does it hold for each index separately?
14:36:54 <Alberth> 50 some tests do sel[0] < 0 others do sel[0] == -1
14:37:59 <Alberth> looks good, although I don't understand this subtypes stuff :)
15:10:29 <frosch123> over-optimisation :)
15:19:59 <michi_cc> Not totally, in fact. There is a difference if the SmallVector stores something with a custom assignment operator that does surprising things, but then std::list/vector would probably be the better choice anyway.
15:22:25 <frosch123> haha, i meant it the other way around :p
15:22:34 <frosch123> afaik memmove always allocated memory
15:22:47 <frosch123> but since i know in which way the memory overlaps, i can avoid that
15:22:50 <frosch123> and move without conflictr
15:25:29 <frosch123> anyway, updated the diffs
15:27:34 <Alberth> memmove does that? pretty stupid then, imho
15:27:45 <michi_cc> Huh, no, memmove doesn't allocate anything. The one and only difference between memcpy and memmove is that memcpy always copies from low address to high while memmove has a simple of to decide up or down.
15:28:18 <Alberth> that's what I'd expect indeed
15:28:57 <michi_cc> AFAIR in newer glibcs memcpy is simply aliased to memmove anyway.
15:28:59 <frosch123> oh, hmm, the documentation only makes the analogy
15:29:11 <frosch123> it behaves as if it was copied
15:29:37 <frosch123> i guess the conjunctive somehow got lost in my memory :p
15:29:52 <Alberth> the interface of memcpy is probably designed so it can take advantage of hardware features if they exist
15:37:25 <michi_cc> If you'd really want to optimize stuff you could replace ErasePreservingOrder by my Extract and http://paste.openttdcoop.org/show/2134/ (it could potentially save a few array lookups, but other than that it is pretty pointless) :)
15:38:42 <frosch123> ah, thanks for pointing out the wrong & coding style :p
15:39:02 *** fonsinchen has joined #openttd.dev
15:39:02 *** ChanServ sets mode: +v fonsinchen
18:11:50 <frosch123> tamil is quite glitchy :s
18:12:04 <frosch123> the characters are always drawn in slightly different positions
18:12:41 <frosch123> i would think we always draw complete liens
18:25:02 <michi_cc> Languages with strange characters are always a problem. Try testing Japanese, last time I looked it had a ten mile long string right in the game options window, which means switching back to English is a PITA.
18:25:28 <frosch123> that's not what i mean
18:25:43 <frosch123> when i switch to tamil some characters are not drawn every time
18:25:49 <frosch123> they disappear and appear again
20:25:11 *** Supercheese has joined #openttd.dev
20:54:00 <frosch123> hmm, there are no ratings for transfer cargos anymore
20:54:10 <frosch123> is that intentional?
20:54:24 <frosch123> i thought they would vanish if the rating is crap just as well
21:40:44 <frosch123> Alberth: true, transfer credits are no longer displayed
21:41:04 <frosch123> just logging all bugs in this channel :p
21:41:32 <Alberth> k, I wondered why you discuss it here :)
21:42:49 <Alberth> wouldn't a 0 leg percentage do that too ?
21:44:12 <frosch123> no idea, but it is set to 75%
22:13:24 *** Alberth has left #openttd.dev
22:28:59 *** Webster has joined #openttd.dev
22:28:59 *** ChanServ sets mode: +v Webster
continue to next day ⏵