| 1 |
On Tue, Jul 23, 2002 at 11:31:36PM +0200, Christian Bartl wrote: |
| 2 |
> I don't think that it would be such a good idea. |
| 3 |
|
| 4 |
I think it's an excellent idea. :) |
| 5 |
|
| 6 |
> I can only speak for myself: As I love Gentoo, but don't have the |
| 7 |
> time to test known unstable software I'm sure I would be the first |
| 8 |
> user who never touches the unstable branch. I think Gentoo would get |
| 9 |
> the same problems like GNU/Linux or Debian: Everyone waits for it |
| 10 |
> getting officially stable and testing is has to be done by a few |
| 11 |
> freaks. |
| 12 |
|
| 13 |
I decided to check out Gentoo 3 days ago; and still haven't got my |
| 14 |
system up running as I want it. This is not a surprise to me, I know |
| 15 |
gentoo is bleeding edge and frankly I'm delightfully surprised how |
| 16 |
well it all went. The compilation failures I got is already in |
| 17 |
bugs.gentoo.org, with fixes or acknowledges from the developers that |
| 18 |
they're looking into the problem. I found the people on #gentoo |
| 19 |
extremly helpful and pleasant to deal with, so there really isn't |
| 20 |
anything to complain about regarding the experience I have with gentoo |
| 21 |
and its community. |
| 22 |
|
| 23 |
Does this mean that I am wiping the desktop I use for workrelated |
| 24 |
issues (now running Debian/testing) or my servers (Debian/stable) in |
| 25 |
favour of gentoo? No, and I will not until there are a way for me to |
| 26 |
be relative sure that after an 'emerge rsync', I still can use the |
| 27 |
machines. I run Debian/testing on my workstation because I know that |
| 28 |
the most obvious bugs have been caught in testing, and Debian/stable |
| 29 |
on my servers since I can't under any circumstances afford any |
| 30 |
downtime on them. |
| 31 |
|
| 32 |
Of course, my laptop used to run Debian/unstable. This was so I could |
| 33 |
use the latest packages and know what my Debian/testing systems would |
| 34 |
get in a short amount of time. It is now converted to gentoo for |
| 35 |
evaluation purposes. |
| 36 |
|
| 37 |
> Whoever needs a stable Gentoo now only has to read the mailing-lists |
| 38 |
> an knows about common problems. |
| 39 |
|
| 40 |
No, this is not under any circumstances enough. You couldn't get a |
| 41 |
stable Debian from only reading the mailing-lists and using the |
| 42 |
unstable branch. It is impossible to follow a branch that changes |
| 43 |
every few hours and not run into dependency problems. Especially not |
| 44 |
when you are compiling the sources for yourself. You are bound to run |
| 45 |
into a problem noone else has encountered before, because the release |
| 46 |
hasn't been out long enough for anyone to test it with various |
| 47 |
configurations. |
| 48 |
|
| 49 |
As I see it, the package-masking system is already used today for |
| 50 |
hiding the most broken packages from the user. I think it would melt |
| 51 |
in well if that masking system was used for the seperation of |
| 52 |
unstable/stable branches. That way, a user could decide to have KDE as |
| 53 |
bleeding edge, but the base system as 'stable'. It would need a lot of |
| 54 |
thinking of course, but the basis is there. :) |
| 55 |
|
| 56 |
Sorry for ranting, |
| 57 |
//Humming |
Archives