Re: [ANNOUNCEMENT]: Important change to symbolic link functionali ty

[Christopher Faylor <cgf at redhat dot com>]

On Thu, Feb 22, 2001 at 01:24:18PM -0500, Charles S. Wilson wrote:
>Corinna Vinschen wrote:
>
>> The POSIX path in the shortcut which is used by Cygwin is saved EXACTLY
>> as it's given. So there is no change in behaviour as far as Cygwin is
>> concerned!!! OTOH the symlinks now have a meaning for DOS tools as well.
>> 
>> I can't see a disadvantage here.
>
>Sounds like a good system to me.  The only drawback I can see -- and I
>don't mean this as a criticism, it's simply unavoidable -- is the
>following:
>
>If you create a symlink that points to an object and spans a mount
>entry, both the "cygwin" path and the "dos" path will 'do the right
>thing' and point where you think they should.
>
>Then, change the underlying mount entry.
>
>The "cygwin" path will point to the new location (since it is
>interpreted using the new mount entry) but the "dos" path will still
>point to the old (possibly non-existant, now) location because it *was*
>interpreted at symlink-creation-time using the old mount table.
>
>IMO, this slight and rare inconvenience is worth the price -- at least
>now, symlinks will mostly work from Explorer.  Let's be honest, now: how
>often do you really rearrange your mount table?

I don't change my mount table very often but this is still potentially
very surprising behavior for a user.  I don't think we should dismiss
it lightly, as much as I like the idea of using .lnk files.

cgf

--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple