On 7/25/2011 9:24 AM, Mark Phippard wrote:
> Von: Les Mikesell [mailto:lesmikesell_at_gmail.com
> On 7/25/11 3:22 AM, Stefan Sperling wrote:
> I am not aware of any existing developer who expressed
> interest in
> working on 'svn detach'.
> Just curious: what drove the original WC design where each
> was inherently independent? And is this the same set of
> My guess is that this Design was "inherited" from CVS.
> And AFAICS, the initial set of developers had some common
> members with
> the CVS developers, and it changed gradually, there was no
> replacing all existing developers with new ones.
> The tradeoffs are sort of obvious, but it seems like a very drastic
> change to make with no option to maintain the old behavior.
> In SVN 1.6 and earlier it is not like someone wrote code specifically to
> allow people to copy/move folders out of their WC. It was just
> something that fell out of the design.
Agreed, but it is a logical design that falls out of the way
subdirectories work and the way people use them.
> There are plenty of instances
> where people did not want this behavior, and for them it is now fixed.
Agreed again. Although the change I would have preferred would have
been an option to not keep a pristine copy at all for the circumstances
where it doesn't work out well.
> For example, in Eclipse this has always been a huge problem. Users
> would copy and paste a folder and this would silently bring the .svn
> folder with it which was not what was desired. I mention Eclipse simply
> because this is one of the only events where it does not provide a hook
> for plugins to get involved so the SVN plugins for Eclipse are not able
> to detect and fix this situation and it is perceived as a bug.
> The opinion was that if users really want to be able to easily detach
> folders from their working copy someone will step up with scripts and/or
> patches to add a feature to SVN to do it.
I'm not arguing that the change is bad or shouldn't have been done, just
that it is a very surprising change in design philosophy, and projects
that make surprising design changes without concern for existing use
patterns make me nervous about what other surprises may be lurking in them.
Received on 2011-07-25 16:50:01 CEST