Re: removal of packages from testing.

2010-07-22 Thread Neil Williams 
On Thu, 22 Jul 2010 12:13:41 -0700
Mike Bird mgb-deb...@yosemite.net wrote:

 On Thu July 22 2010 11:28:49 Neil Williams wrote:
  Removing packages from testing does not remove them from any
  existing installation, so it's hard to see how the removal of
  packages which are plainly not suitable for release in stable
  supports an assertion that testing is somehow not intended for real
  users.
 
 Real users have to be able to recreate a dead system or install
 new systems at any time.

Really?? That seems far beyond the requirements of most ordinary
users. Exceptional users could always arrange a local mirror, which
doesn't have to remove packages.

 We actually have a few Testing packages (e.g. WordPress) in our
 mostly-Stable servers and we backup copies of those Testing packages
 both on-site and off-site against the vagaries of the Testing masters.

These are not vagaries - packages are only removed from testing for
essential reasons.
 
 We used to use Testing for desktops and laptops, but the hassle
 of disappearing packages was not worth it. 

Any stats on that? Just how many packages were affected? Before or
after a release freeze was announced?

-- 


Neil Williams
=
http://www.data-freedom.org/
http://www.linux.codehelp.co.uk/
http://e-mail.is-not-s.ms/



pgpFxk2gfpoup.pgp
Description: PGP signature

Re: removal of packages from testing.

2010-07-22 Thread Mike Bird 
On Thu July 22 2010 12:44:39 Neil Williams wrote:
  We used to use Testing for desktops and laptops, but the hassle
  of disappearing packages was not worth it.

 Any stats on that? Just how many packages were affected? Before or
 after a release freeze was announced?

Real users have to be able to clone or recreate systems
at any time.  Not merely at the most auspicious moments
of the release cycle.

Check our Joey's post for why Continuously Usable Testing
would be a good idea:
http://kitenet.net/~joey/code/debian/cut/

It does seem that Debian is wasting an awful amount of
effort by not making Testing as useful as it could be.

--Mike Bird


-- 
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/201007221308.51340.mgb-deb...@yosemite.net