[nSLUG] Finding linux products and software
jdkenney at gmail.com
Thu Apr 12 17:25:30 ADT 2007
On 4/12/07, Ian Campbell <ian at slu.ms> wrote:
> On Thu, Apr 12, 2007 at 04:33:49PM -0300, D G Teed wrote:
> > On 4/12/07, Jason Kenney <jdkenney at gmail.com> wrote:
> > In concept, it is a potentially solid way to get things done. BSD has
> > been doing this for years with ports. The difference is, BSD allows
> > for more QA and more careful release paths. There was a time that
> > Gentoo talked about having a more conservative enterprise repository,
> > but this has died as their isn't enough people to make it happen.
> > So Gentoo has remained a ricer/hobby thing.
Your language is very vague. Many things are "potentially" something.
I mean, you might say I could be wrong or correct. That would be quite
hard to argue with!
> I'm not sure that's fair, Gentoo's reputation isn't entirely deserved.
> Port quality and overall system stability have improved greatly over
> the last couple years, although there's still a ways to go. A lot of
> the stigma of Gentoo comes from people using unsafe compiler flags,
> which a lot of packages (binutils, glibc) don't like. If you pick
> something sane, it's no more or less stable than anything else.
My view of Gentoo being more of a toy isn't that it is necessarily
less solid than other distributions; that is probably dependant mostly
on the mainters of the distribution, and as they change, it changes
over time. It's more that except in extremely specific hardware cases,
I don't see any advantage to wasting time (and energy...) having every
user recompile things themselves. Even for those specific cases
though, I am only open to the possibility of them in theory; computing
is so ubiquitous now that I can't think of an architecture where there
is a *real*, *measurable* difference (for the better of course) from
using pre-compiled packages. Does such a thing exist?
More information about the nSLUG