Config
Log for #openttd.dev on 23rd April 2014:
Times are UTC Toggle Colours
01:34:37  *** Supercheese has joined #openttd.dev
01:34:37  *** ChanServ sets mode: +v Supercheese
01:48:59  *** Supercheese has quit IRC
01:49:31  *** Supercheese has joined #openttd.dev
01:49:31  *** ChanServ sets mode: +v Supercheese
02:08:14  *** Supercheese has quit IRC
02:34:04  *** Supercheese has joined #openttd.dev
02:34:04  *** ChanServ sets mode: +v Supercheese
07:39:57  *** Supercheese has quit IRC
15:17:40  *** Alberth has joined #openttd.dev
15:17:40  *** ChanServ sets mode: +v Alberth
16:40:04  *** frosch123 has joined #openttd.dev
16:40:04  *** ChanServ sets mode: +v frosch123
17:30:51  <fonsinchen> planetmaker: no, I want to prefer the cargo with more AvailableCount there. The problem was that it would always load coal, even if only 10t were available, because that has the lowest ID.
19:34:14  <fonsinchen> https://github.com/ulfhermann/openttd/commit/02ec5c9467218b54874c59a44cd6cdc1a987896d
19:34:26  <fonsinchen> No more tile demands.
19:34:34  <fonsinchen> andythenorth convinced me.
19:39:10  <frosch123> looks fancy
19:46:41  *** frosch123 has quit IRC
19:47:37  *** frosch123 has joined #openttd.dev
19:47:37  *** ChanServ sets mode: +v frosch123
19:47:50  <frosch123> what has the stuff in strings.cpp to do with a string class?
19:48:31  <fonsinchen> It does similar things to what other string classes do.
19:48:52  <frosch123> i would compare it to gettext, but not to a string class :)
19:49:08  <frosch123> it does not do string manipulation
19:49:51  <frosch123> string.cpp is the string class stuff
19:49:53  <fonsinchen> It does some sort of printf lookalike. The internationalization is arguably the only application of that, but still.
19:50:07  <frosch123> btw. we should get rid of ttd_strlcpy and ttd_strlcat
19:52:09  <frosch123> i'll take a look at ttd_strl*
19:54:47  <frosch123> hmm, is "<" a meaningful operation for pointers?
19:55:12  <fonsinchen> it's desync-prone
19:55:20  <fonsinchen> where did you find it?
19:55:36  <frosch123> no, i mean to check string/array bound
19:55:47  <frosch123> p < lastof(buffer)
19:56:07  <frosch123> i guess only signedness could mess it up
19:56:46  <fonsinchen> signedness of what?
19:56:54  <fonsinchen> Pointers are unsigned
19:57:26  <fonsinchen> You mean it gets casted to int? That may be ...
19:57:27  <frosch123> oh, is that defined?
19:57:50  <frosch123> so NULL <= any pointer?
19:57:50  <fonsinchen> operator< should be defined on pointers
19:57:55  * fonsinchen looks it up
19:59:20  <frosch123> ah, found the first buffer overrun :)
19:59:40  <frosch123> ttd_strlcpy(r, ".lng", (size_t)(r - pathbuf));
19:59:51  <frosch123> so, strecpy is already better :)
20:04:53  <fonsinchen> nice
20:06:40  <Alberth> afaik you may only compare pointers that point into the same array
20:07:08  <fonsinchen> http://stackoverflow.com/questions/4607418/how-do-the-operators-and-work-with-pointers has a quote from the standard
20:07:18  <frosch123> Alberth: that definition is not particulary useful for a range check :)
20:07:26  <frosch123> it's like the gcc 4.5 enum optimisation
20:07:42  <frosch123> which optimised away the range check for enum values, since it can never be true
20:07:46  <frosch123> resp. false
20:08:55  <fonsinchen> The quote says you can also compare the one-past-the-end pointer of an array to members of the array
20:09:17  <Alberth> fonsinchen: indeed, one further than the end is explicitly supported
20:09:56  <Alberth> frosch123: it has to do with  a - b   which does an implicit  "/ sizeof(*a)"
20:11:43  *** Alberth has left #openttd.dev
20:13:35  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26482 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
20:21:40  <frosch123> where is stredup?
20:22:18  <Rubidium> it's not there yet ;)
20:22:28  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26483 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
20:22:43  <frosch123> ok, i'll continue to use strndup then :p
20:23:23  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26484 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
20:41:45  <frosch123> http://devs.openttd.org/~frosch/diffs/ttd_strl.diff <- removing ttd_strlcpy and ttd_strlcat
20:41:54  <frosch123> ttd_strnlen remains, since there is no replacement
20:43:25  <Rubidium> looks okay to me
20:43:35  <Rubidium> I'm working on s(n)printf now
20:44:43  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26485 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
20:44:54  <frosch123> i replaced one snprintf
20:45:24  <Rubidium> that's fine ;)
20:45:35  <frosch123> just beware of conflicts :)
20:46:06  <frosch123> most of the Fio functions use buflen instead of last
20:46:20  <frosch123> but most of the underlying os functions do as well
20:48:16  <Rubidium> yeah... that bit is a PITA
20:53:13  <frosch123> hmm, some places look like we may need a strdupcat
20:53:30  <frosch123> to allocate a new copy of a concatenation
20:54:02  <frosch123> hmm, actually, no
20:54:22  <frosch123> those were alloca, not allocm
20:55:52  <frosch123> night
20:55:57  *** frosch123 has quit IRC
21:12:10  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26486 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:16:59  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26487 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:19:55  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26488 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:23:22  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26489 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:30:32  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26490 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:32:41  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26491 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:33:29  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r26492 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"

Powered by YARRSTE version: svn-trunk