Compatability with FreeBSD Ports

Andreas Hauser andy at
Mon Aug 15 14:33:48 PDT 2005

joerg wrote @ Mon, 15 Aug 2005 22:31:02 +0200:
> On Mon, Aug 15, 2005 at 04:51:02PM -0000, Andreas Hauser wrote:
> > joerg wrote @ Mon, 15 Aug 2005 21:10:19 +0200:
> > > On Mon, Aug 15, 2005 at 11:39:59AM -0000, Andreas Hauser wrote:
> > > > - no portupgrade
> > > 
> > > I stated before that I don't agree with this FOR TECHNICAL REASON.
> > > portupgrade is a hack, it sometimes works and sometimes doesn't. See
> > > other comments for examples.
> > 
> > But you have no solution. portupgrade works fine here.
> > It is a mature tool and is up to the task.
> LOL. It's mature because it works for some cases but randomly fails?
> That's why it does quality as hack and nothing more. But the complete
> ignorance of problems is introduced by "minor" updates is too common.

You can laugh a lot, using your imaginary, non-existing better tool,
that you fail to mention.

> > > > - deinstalls before compiling
> > > 
> > > That's the update target, which makes perfect sense to allow a *clean*,
> > > *reproducable* build. Use pkg_comp if you don't like it.
> > 
> > I don't care. It's not what i want.
> Well, don't complain about it then.

If it is so very mature why can i not influence that ?
At least it is not well documented how to turn it off.

> Use ports.

I will.

> It seems we have different requirements for what is considered mature vs.
> working.

I have points and you have polemics.
If it so good, why is noone coming up with a representation of pkgsrc
that is so compelling that it makes me shut up ?

Let me guess this is like with citrus ?
You say you will analyze both and provide benchmarks.
But what you really do is choose without presenting technical reasons
and no benchmarks ever. Then about a year and an ABI breakage later
it's still not fully working. Some where along you must have dropped
your high requirments, that you so readily apply to others.

When will pkgsrc be ready ?

Let me guess, the answer is either "LOL" or silence.


More information about the Users mailing list