Re: filename translation implementation - how to decide

[Jamie McCracken <jamiemcc blueyonder co uk>]

On Wed, 2004-12-15 at 23:07 +0100, Alexander Larsson wrote:
> On Wed, 2004-12-15 at 21:58 +0000, Jamie McCracken wrote:
> > On Wed, 2004-12-15 at 22:25 +0100, Alexander Larsson wrote:
> > 
> > > 
> > > On the contrary, the API for a library is highly dependant on how the
> > > standard works. It should be obvious to anyone that actually read my
> > > initial mail that the two solutions have a very different API.
> > > 
> > 
> > Yes thats true but it could still be defined to cover both. 
> > 
> > EG you could have:
> > 
> > fdo_get_desktop_dir - returns ~/Desktop with option1 else custom path 
> > fdo_get_desktop_displayname - returns translated name if option 1 etc
> 
> So, for every filename, you'd call fdo_get_desktop_dir() and compare
> with that, and if its true you'd call fdo_get_desktop_displayname(),
> repeat for all special folders. Doesn't sound very efficient.

No. 

If you want to write to desktop you always use  fdo_get_desktop_dir (as
you would with g_get_home_dir)

If you want to display to the user in Nautilus the desktop name you use
fdo_get_desktop_displayname (instead of g_filename_display_basename).
There should be no need to combine or compare the two.


> It could be done better, but basically you'd end up with doing both
> systems, and then only supporting one anyway. Increasing complexity,
> lowering performance, and generally confusing things.

No cause both would be always supported - its just an alternative to
hardcoding the desktop path.

jamie.

> 
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>  Alexander Larsson                                            Red Hat, Inc 
>                    alexl redhat com    alla lysator liu se 
> He's a witless soccer-playing cowboy from a doomed world. She's a foxy 
> renegade wrestler living on borrowed time. They fight crime! 
> 
>