This mail missed dev@ list. Forwarding it.
attached mail follows:
On Tuesday 28 June 2011 03:01 PM, Noorul Islam K M wrote:
> arwin_at_apache.org writes:
>
>> Author: arwin
>> Date: Tue Jun 28 08:43:40 2011
>> New Revision: 1140482
>>
>> URL: http://svn.apache.org/viewvc?rev=1140482&view=rev
>> Log:
>> Added a branch readme.
>> * BRANCH-README : New file.
>>
>> Added:
>> subversion/branches/svn-bisect/BRANCH-README
>>
>> Added: subversion/branches/svn-bisect/BRANCH-README
>> URL: http://svn.apache.org/viewvc/subversion/branches/svn-bisect/BRANCH-README?rev=1140482&view=auto
>> ==============================================================================
>> --- subversion/branches/svn-bisect/BRANCH-README (added)
>> +++ subversion/branches/svn-bisect/BRANCH-README Tue Jun 28 08:43:40 2011
>> @@ -0,0 +1,70 @@
>> +This branch exists for the implementation of the svn bisect command.
>> +It is maintained as a reintegrate-able branch with regular catch-up
>> +merges from trunk.
>> +
>> +PURPOSE
>> +=======
>> +
>> +For years, subversion users have used third-party scripts to do what
>> +should rightfully be a part of subversion itself. This branch aims
>> +to implement bisect functionality and make it available via the
>> +public API.
>> +
>> +
>> +SPEC
>> +====
>> +
>> +svn bisect start [-rN[:M]]
>> +
> When we discussed you had a concern that above syntax is different from
> the normal svn sub command syntax. Is this finalized?
>
I wouldn't say it's finalized.. I simply wrote down a spec as a rough draft.
I'm sure the community will have some ideas about this. (Like implementing
a sub-subcommand interface of some sort.)
>> +Initialize the bisection state.
>> +If a state already exists, clean it up and create a fresh one.
>> +If a revision range is specified, limit bisection to this range.
>> +If a single revision is specified, bisect from that revision to HEAD.
>> +If no revision is specified, bisect from revision 1 to HEAD.
> Does this mean that N will be set a bad and M will be good revision
> initially?
>
That's correct.. except, N will be set as good and M will be set as bad..
i.e the lower revision is always the good revision. A later (higher)
revision
introduces the bad behaviour and we use bisect to pick this out.
> Thanks and Regards
> Noorul
Received on 2011-06-28 11:50:32 CEST