[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: [PATCH] svn-config installedness test fails when symlinked

From: Marcus Comstedt <marcus_at_mc.pp.se>
Date: 2002-09-11 15:31:40 CEST

Justin Erenkrantz <jerenkrantz@apache.org> writes:

> On Wed, Sep 11, 2002 at 12:19:00PM +0200, Marcus Comstedt wrote:
> > Yeah, the problem is that it doesn't "just work". And to me it seems
> > _much_ more robust to just toggle a flag when installing, than trying
> > to do magic on $0.
>
> No, I'm not sure that's any better. To me, resolving the symlink
> seems a nicer way than tweaking the file as we install it.

But flipping a flag when installing can be done easily and portably.
Resolving symlinks cannot. And there may be more problems with the
current code than symlinks. What about if someone makes a hardlink
for example? It just seems less robust.

> > You mean simply remove support for "location=source" and
> > "location=build" altogether? Well, I'm fine with that. Maybe
> > somebody else want to use them though.
>
> No, I think Greg means having an explicit 'location=install' option
> as well. And, we need to keep svn-config working when it isn't yet
> installed. -- justin

I'm not sure I understand either what you mean or how that relates to
what Greg said. With "explicit 'location=install' option", do you
mean explicitly telling the installed script that it is installed when
you invoke it? That sounds awfully awkward.

  // Marcus

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Sep 11 15:33:19 2002

This is an archived mail posted to the Subversion Dev mailing list.