The Hermit Hacker wrote:
>
> On Thu, 6 Apr 2000, Doug Barton wrote:
>
> > Christopher Masto wrote:
> > >
> > > On Mon, Apr 03, 2000 at 10:52:13AM +0100, Nick Hibma wrote:
> > > > Are there actually any good reasons why we _should_ upgrade in the first
> > > > place?
> > >
> > > Of course. We now have an obsolete version of Perl. That should be
> > > reason enough to upgrade.
> >
> > You haven't given a sufficiently compelling definition of "obsolete"
> > yet. I think that what people are really asking is, "What does this new
> > perl get us that we don't already have?" Once we've answered that, then
> > we can balance the benefits against the costs (which are pretty high,
> > considering the complexity of integrating perl into the berkeley make
> > environment) and then we can try and apply those arguments in the search
> > for someone who is willing and able to do the work.
>
> My experiences with perl tend to be that as soon as a new release comes
> out, all the module maintainers tend to adopt it as standard and start to
> deprecate the older versions. I'm not saying that this is an overnight
> sort of thing, but it does pose a problem ...
>
> My stupid question, though, is why is this such a big issue? Would it be
> too hard to extend our /usr/src build process so that it is smart enough
> to do an install out of ports, and just build the ports version of 5.6.0,
> vs trying to integrate it into our build tree? Create a symlink to
> /usr/ports/devel/perl560 so that when it cd's to the perl directory and
> does a 'make', it builds that?
Whoah! That would mean that I'd have to install the ports tree on every
machine I intend to build world on.. That's just not acceptable to me.
The source tree should remain independant to external sources.
Regards,
Edwin Mons
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
