[cvsnt] Re: Does atomic commit work in 2.0 stable?

Tony Hoyle tmh at nodomain.org
Tue Apr 15 09:16:53 BST 2003


Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.


Ori Berger wrote:

 Tony, Can you elaborate a little on how this works? Searching
> through the Wiki I found two short mentions, one in the CvsClient
> page, and one in the history page, neither which is really informative.

See the thread at:

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=cf333042.0301231424.53e67cbd%40posting.google.com&rnum=1&prev=/groups%3Fq%3Dcvs%2Batomic%2Bcommits%2Bhard%2Blinks%26hl%3Den%26lr%3D%26ie%3DUTF-8%26selm%3Dcf333042.0301231424.53e67cbd%2540posting.google.com%26rnum%3D1

Derek Price originally brought it to my attention, and I stuck an
implementation in, which seems to work.

> And is there any reason why it shouldn't work on a Unix/Linux
> system? (hard links have recently been added to it <wink> )

Yes, it works fine under Unix.

TBH Atomic Commits are solving a problem that really isn't one.  CVS is
already atomic at the file level, so there's no risk of corruption from
power failure (the worst you'll probably get is a few lock files hanging
around).  Commit-level atomicity sounds cool but really doesn't buy you
anything.  I implemented it because some people seem to want it, but have
never actually used it myself beyond the testing stage.

Tony



More information about the cvsnt mailing list
Download the latest CVSNT, TortosieCVS, WinCVS etc. for Windows 8 etc.
@CVSNT on Twitter   CVSNT on Facebook