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

Re: delta in SVN is reversible or not

From: Branko Čibej <brane_at_wandisco.com>
Date: Thu, 14 Mar 2013 01:20:41 +0100

On 14.03.2013 01:03, Daniel Shahaf wrote:
> Branko Čibej wrote on Thu, Mar 14, 2013 at 00:43:24 +0100:
>> On 14.03.2013 00:07, Bo Chen wrote:
>>> Can anyone help me make clear this question? In SVN, if I have two
>>> versions A and B. I compute a delta of B against A, i.e., I can
>>> recover B by A and this delta. My question is, can I recover A by
>>> using B and the same delta?
>> No. It's theoretically possible to reverse a delta, but to do that, you
>> must have both A and B available.
> If you have A and B, you can just do (A-B) to get a reversed delta.
>
> Or maybe you mean that (A-B) can be computed more efficiently given
> (A,B,B-A) than given (A,B)?

I believe it can be more efficient in terms of the cost for producing
the inverse delta in some cases, but I never researched that in detail
because its clear that the /size/ of such a delta would not be optimal.

-- Brane

-- 
Branko Čibej
Director of Subversion | WANdisco | www.wandisco.com
Received on 2013-03-14 01:21:18 CET

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

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