[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: %datadir or %datadir/games for games?

On Sat, 2006-03-04 at 09:45 +0100, Hans de Goede wrote:
> Ralf Corsepius wrote:
> > On Sat, 2006-03-04 at 09:27 +0100, Hans de Goede wrote:
> >> Hi all,
> >>
> >> I just read on the Games SIG page that it is advised to put data in 
> >> %datadir/<gamename> but shouldn't this be %datadir/games/<gamename> ?
> > 
> > Cf. http://www.pathname.com/fhs/pub/fhs-2.3.html#AEN1550
> > 
> Agreed, this only is for static / data files, is that not what you meant?
I am referring to the general specification/definition
of /usr/share, /usr/lib, /usr/share/<package> rsp. /usr/lib/<package>
etc. Here, in particular to the /usr/share/<package> and static data.

> >> See:
> >> http://www.pathname.com/fhs/pub/fhs-2.3.html#SPECIFICOPTIONS15
> > Note:
> > /usr/share/games .. Static data files for /usr/games (optional)
> > 
> > => I read this as /usr/share/games corresponds to packages having been
> > installed to /usr/games
> > 
> Ern, /usr/games isn't mentioned elsewhere in the doc and is very 
> deprecated. I believe they forgot to update this part of the doc when 
> /usr/games got removed.
Well, I think /usr/games and /usr/share/games essentially are historic
artifacts, and an LSB typical compromise to cater those systems who have
a tradition in using them.

IIRC, there had been 2 motivations for /usr/games and /usr/share/games:
1. Keeping games out of /usr/bin to keep $PATH clean and lean.

2. Cater games which want to play uid/gid tricks on data files.
IMO, this argument is void on shared/static data and if at all, only is
applicable to score-files and the like (which nowadays are supposed to
live somewhere below /var).

> I vote for /use/share/games/xxx because /usr/share is already rather 
> crowded.
> What do you / others vote?
I vote for "treat games as ordinary applications"
i.e. /usr/bin, /usr/share/<package>, /usr/lib/<package>, /var/lib, etc.

Actually I feel any attempt to install files based on any
"classification" to be questionable in general and likely to fail in
longer terms (Due to what classification theory/AI calls "Symbol
grounding problem")


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]