On Wed, Aug 06, 2003 at 05:42:56PM +0200, Francois Gouget wrote: > On Mon, 4 Aug 2003, Debian Bug Tracking System wrote: > > On Mon, Aug 04, 2003 at 09:49:25AM +1000, Daniel Stone wrote: > > > No, this is incorrect. $REALSTARTUP, which may be ~/.[Xx]session, if it > > > exists, is only executed at the *end* of the Xsession.d chain, after > > > all the system Xsession scripts have been run. > > I apologize for the delay in my reply, I don't like restarting my X > server<g>. > > ~/.xsession is started at the end of the Xsession.d, that's correct. > However, in that case it's up to the ~/.xsession script to select a > window manager to run, i.e. the user has to duplicate the functionality > of 50xfree86-common_determine-startup. Note that: > * ~/.xsession does not inherit variables such as REALSTARTUP > * ~/.xsession cannot just call 50xfree86-common_determine-startup > because all it would get is REALSTARTUP="... ~/.xsession"... again. I don't see why this is such a problem - why not just put in "exec gnome-session" or whatever? You're only going to have to write that line *once* ... > > I am closing this bug due to the above analysis. > > So I don't think the bug should be closed: > * forcing each user to duplicate 50xfree86-common_determine-startup > functionality just so he can customize his mouse acceleration or have an > xterm on startup does not seem like a good solution. If you know what you want, it's only one line (e.g, "exec startkde"), which is IMO a lot better than the resulting confusion. > * there's an interoperability issue with RedHat systems (e.g. HOME > shared via NFS). There are a lot of things we (shock horror!) do differently. This is one, and I'm standing by it. -- Daniel Stone <daniel@fooishbar.org> http://www.kde.org - http://www.debian.org - http://www.xwin.org "Configurability is always the best choice when it's pretty simple to implement" -- Havoc Pennington, gnome-list
Attachment:
pgpPd2V89AWgK.pgp
Description: PGP signature