[lsb-discuss] tux assistant idea

rahul at reno.cis.upenn.edu rahul at reno.cis.upenn.edu
Fri Jan 31 17:31:02 PST 2003


To play the devil, and go completely on the other side, the argument could
be made that for a desktop, the LSB, more precisely the filesystem spec 
might even be the wrong solution.

Take the Mac for example. Its filesystem structure derives from the old NeXT,
and for users, is way more intuitive. The appliocation structure has more
in common with old unix (/pkg/appname/[bin/lib, etc]) than with present day
linux systems and their preponderant usage of /usr, which is a consequence of rpm and deb, which in turn is a consequence of the mid 90's where disks were not
too large, and repeated libraries were a no-no.

Both the Mac and .Net have moved ahead to including libraries in app bundles
but defaulting to system areas if they exist. Now the LSB is perfectly compatible with this from the filesystem perspective, but the practice ( which is formalized in LSB paths) isnt really conducive to this.

Anyway, my more general point is, standards are useful if they help you move forward. In standardizing core structures, the LSB helped us move forward. Its better to apply the standards to core parts of the linux system, ratrher than trying
to bring more under the standard umbrella. Ie define what core linux means, so
that everybody can use it..further than that, i would say that even something like package management ought to be outside the scope, since it stifles package innovation.

Or at the very least layer standards so that compliance can be at a certyain level. But lets not prematurely standardize ever...
Rahul

On Fri, Jan 31, 2003 at 03:40:24PM -0600, George Kraft wrote:
> Adrian,
> 
> I don't want to appear to be ignoring you, because we are interested in
> what you are saying; however, we are working at a much lower level than
> that.  In fact we have not yet defined Gtk2 or Qt in the LSB.  We need
> specifications and testing for them, so we are not ready to pursue
> usability.
> 
> I did a quick search and found a GNOME Usability Project (GUP).  For
> now, I think you might have better success with them because we are
> distracted with the API and ABI details.
> 
> http://developer.gnome.org/projects/gup/
> 
> Best regards,
> 
> George (gk4)
> 
> 
> _______________________________________________
> lsb-discuss mailing list
> lsb-discuss at freestandards.org
> http://freestandards.org/mailman/listinfo/lsb-discuss




More information about the lsb-discuss mailing list