| 1 |
Hi all |
| 2 |
no flame bait or anything but IMHO some packages are way too much in the |
| 3 |
masked state. I don't know the exact QA policy but it's probably do not |
| 4 |
annoy users by installing bleeding-edge software let the brave ones |
| 5 |
unmask explicitely. This is fine but I think it's applied to generically |
| 6 |
While I understand that few people would want the latest glibc, XFree |
| 7 |
snapshot or KDE form CVS, masking also applies to standalone programs |
| 8 |
which don't really affect general system stability. Many of these are |
| 9 |
programs in active development with pretty frequent releases where the |
| 10 |
developers are looking for feedback and where they generally release |
| 11 |
when they consider they improved.There are probably counter-examples |
| 12 |
too though... |
| 13 |
Ex: ebuilds for subversion , distcc, valgrind, scons are either |
| 14 |
entirely masked or generally lagging behind a couple of releases wrt the |
| 15 |
'unstable' ebuild. |
| 16 |
The policy that 'if for an amount of time there are no bugs reported |
| 17 |
against' they are made stable is again two-edged: there's less testing |
| 18 |
of latest releases so only a smaller procent of the gentoo crowd |
| 19 |
actually provides feedback to their development. More like those RPM |
| 20 |
based distros :) |
| 21 |
|
| 22 |
Especially because many tools I find are undeservingly masked are |
| 23 |
developer oriented so there's a greater chance feedback will be sane and |
| 24 |
prompt. |
| 25 |
|
| 26 |
I'd propose to loosen this policy a bit, but then again who am I? |
| 27 |
|
| 28 |
Jani |
| 29 |
|
| 30 |
cc: please |
| 31 |
|
| 32 |
|
| 33 |
-- |
| 34 |
gentoo-dev@g.o mailing list |
Archives