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

Re: Working copy corruptions due to dual use of SVN on Linux and Windows against same working copy

From: Ben Fritz <fritzophrenic_at_gmail.com>
Date: Tue, 20 Nov 2012 12:31:16 -0600

On Tue, Nov 20, 2012 at 2:02 AM, Oto BREZINA <otik_at_printflow.eu> wrote:
> 1.6 was file based WC making it easy to copy and share - not really safe
> through.
> 1.7 is DB based. Do you expect you DB. (e.g. MySQL) databases data can
> be safely read from multiple platform, or even same platform in multiple
> instances?
> It may but it is not recommended.
> When switching OS/or user many states may be unhandled, rights may get
> wrong ...
> DB versions may not match, have different int size, or have whats ever
> difference.
> You are free to try but was informed that it is not supported - tested,
> was not intend to make it work - so don't be surprised.

This may be true, there may be implementation decisions made that
prevent using a working copy from one system on another. But if this
is the case, I think this is a bad design decision; or at the very
least, bad documentation. Consider a person who keeps a working copy
on a thumb drive to cart around with them to whatever computer they
happen to be working on. Shouldn't they be able to commit, update,
whatever, on any of these systems, as long as a compatible version of
SVN is installed? At the very least, the fact that SVN does not
operate well across OSes should be documented. Using git or Hg, for
example, I would certainly expect this to Just Work, I see no reason
why SVN should be different in this regard. A program should be able
to read and write its own data regardless of where the program was
running at the time the data was created.

I have noticed some quirks when working between OSes on the same WC,
but they are little things like the Windows SVN not being able to set
the execute bit when the SVN properties say it should be set. I expect
things like that. WC corruption would be totally unexpected to
basically anybody I think. I don't think you can fairly call it user
error when the user was never warned not to do that.


To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_tortoisesvn.tigris.org].
Received on 2012-11-20 19:31:43 CET

This is an archived mail posted to the TortoiseSVN Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.