Re: New APT version
email@example.com (Thomas Lakofski) wrote on 11.05.98 in <Pine.LNX.3.96.980511092813.12507Bfirstname.lastname@example.org>:
> On 10 May 1998, Gregory S. Stark wrote:
> > > i agree. if apt can do it and is ready in time for testing then we
> > > should certainly drop autoup and use apt instead.
> > I disagree strongly. We know what autoup does, it's a simple shell script
> > with a fixed sequence of commands that have been carefully thought out and
> > tested widely. APT is still in beta, and while it's good that it's fairly
> > predictable it cannot be never be quite as reliable as autoup.
> I concur. As a user I followed the discussion of using apt instead of
> autoup for the bo-hamm upgrade, and tried it out on a non-production bo
> machine. libreadlineg2 didn't upgrade successfully, and so lots of useful
> stuff like bash segfaults. I think apt needs a little more testing for
> these purposes (although it's an *excellent* tool for maintenance of a
> current system).
I (strongly) disagree. apt seems to _already_ do a better job than autoup,
and in contrast to autoup, apt doesn't need to be changed when we decide
to change dependencies.
I recently tried a "install bo until it starts dselect, then install apt
and upgrade to hamm", and it was nearly perfect, the only problems being
package bugs outside of apt. (Just like that getline problem you mention
was a well-known problem in another package.)
And in sharp contrast to dselect, *no* "try until it works"!
To UNSUBSCRIBE, email to email@example.com
with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org