Perl's not the only TAP source

[Andy Armstrong]

On 16 Jun 2009, at 15:14, Andy Armstrong wrote:
> On 16 Jun 2009, at 11:07, Andy Armstrong wrote:
>> I'm going to do another run through of the git conversion and use
>> whether I can successfully merge your branch as some kind initiation
>> test for the git conversion. If that works out I'll move the repo to
>> github.
>
> Here's the repo:
>
>  http://github.com/AndyA/Test-Harness/tree/master
>
> I haven't done the merge of sourcedetector yet.


Hmm. Well, the merge is there now. Anyone who's already cloned it  
should probably clone it again because I had to rewrite master's  
history to drop the commit where I deleted Steve's work from trunk.

Because Steve started out in trunk and I deleted his work after  
copying it to a branch there's a delete in the history which gets  
reflected in the subsequent merge. So I cherry picked everything from  
master into a new branch based on sourcedetector and then renamed that  
new branch as trunk. Clearly that's not ideal. Can any git experts  
suggest a different approach?

------[delete]-[A]-[B]-[C]-[D]  <-- trunk
   \
    \--[E]-[F]-[G]-[H] <-- source detector

I pretended history looked like:

------[A]-[B]-[C]-[D] <-- trunk
   \
    \--[E]-[F]-[G]-[H] <-- source detector

-- 
Andy Armstrong, Hexten