On Fri, May 10, 2013 at 11:48:37AM +0200, Marco d'Itri wrote:
[ reiterating the same arguments seen numerous times before ]
I suggest that we leave practical implications of the /usr merge aside
for a moment. The pros and cons have been discussed at lengths. If there
is value in further arguments, please explain that to me.
What issues would Debian face if it were to support both models (for a
time, e.g. jessie)? During that time frame we can gather actual data on
whether this feature is actually used by users and what kinds of
workflows it breaks. Is it really useful to continue discussing whether
it is desirable instead of just doing it? I can try to phrase it as a
release goal to see, whether we can get consensus on it.
Make sure that Debian can be installed with merged /usr and with
separated /usr.
As far as I can see it meets the requirement to be "smart" and affects a
significant number of packages (due to file conflicts with /bin/$foo
/usr/bin/$foo). The first two (independent) steps to getting there is
updating the initramfs to mount /usr (in the works) and to drop
compatibility symlinks /usr/bin/$foo -> /bin/$foo.
This target may be more work that just switching to merged /usr, but
there is no (perceived) consensus on the latter. If there is no
consensus on this (or a similar) goal either, we can stop the discussion
immediately as no change is going to happen. Note that I am not an
advocate for this goal.
Helmut
--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20130510104454.ga28...@alf.mars
