W3C home > Mailing lists > Public > ietf-dav-versioning@w3.org > October to December 2000

Re: DAV:checked-out

From: Greg Stein <gstein@lyra.org>
Date: Fri, 6 Oct 2000 12:16:50 -0700
To: ietf-dav-versioning@w3.org
Message-ID: <20001006121650.P9396@lyra.org>
Woah. Where did that come from? Multiple checkouts of a version selector
were something that I had specifically requested a couple months ago during
a review. The answer at that time, IIRC, was "sure, it is done like <so>"

The scenario is:

  - Joe creates and activity, checkouts out /foo into that activity, and
    begins updating it.
  - Jane does likewise.
  - One or both merge the activity

Obviously, they can both have it checked out at the same time. Why prevent
Jane from checking the thing out, when Joe may simply end up cancelling his
checkout. That is just wrong.

Let's take a look at two of the version control systems that I'm familiar
with:
  - SourceSafe: allows multiple checkouts (config option)
  - CVS: default/typical operation is multiple checkouts (with some work, I
    think you can prevent them)

Subversion (a CVS replacement, which I'm adding DAV support to) will operate
similarly: multiple checkouts.

DeltaV *must* allow this. It is a very standard mechanism, and I can't
imagine being able to tell people, "well DeltaV is supposed to handle
versioning, but it can't even reach the level of CVS"

Cheers,
-g

On Fri, Oct 06, 2000 at 03:54:14PM +0100, Tim_Ellison@uk.ibm.com wrote:
> 
> You can only checkout a version selector once (the precondition failure, in
> 10.2 CHECKOUT, would be <DAV:must-not-be-checked-out/>)
> 
> Tim
> p.s. comments based on versioning-8.2
> 
> 
> Greg Stein <gstein@lyra.org> on 2000-10-06 03:37:52 PM
> 
> Please respond to Greg Stein <gstein@lyra.org>
> 
> To:   ietf-dav-versioning@w3.org
> cc:    (bcc: Tim Ellison/UK/IBM)
> Subject:  DAV:checked-out
> 
> 
> 
> 
> I was just perusing v10 of the draft... The Version Selector property named
> DAV:checked-out does not account for the case where multiple checkouts of a
> resource may occur.
> 
> Seems it would be possible to do the following:
> 
> Joe:  CHECKOUT /foo
> Jane: CHECKOUT /foo
> Jane: PUT /foo
> Jane: CHECKIN /foo
> Jane: CHECKOUT /foo
> 
> In Jane's second checkout, the /foo Version Selector's target points to
> something different from Joe's (active) CHECKOUT.
> 
> What is the purpose of this property? Thinking on it... I'm not sure
> whether
> its value is even "interesting".
> 
> Cheers,
> -g
> 
> --
> Greg Stein, http://www.lyra.org/
> 
> 
> 

-- 
Greg Stein, http://www.lyra.org/
Received on Friday, 6 October 2000 15:16:55 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 13:57:39 GMT