[Bioperl-l] Bio-FeatureIO

[Chris Fields]

On May 16, 2010, at 12:32 PM, Jay Hannah wrote:

> On May 16, 2010, at 11:16 AM, Chris Fields wrote:
>> Just a heads-up.  I recently (Jan 2010) moved Bio::FeatureIO to it's own repository for refactoring.  However, the original Bio::FeatureIO code is still within bioperl-live and the 1.6 release branch, and the branch where these were removed was no longer clean, so I removed it.  I'm in the process of syncing the 1.6 branch with master soon (where it will remain unmodified), and then will remove Bio-FeatureIO code from master and pull it into the master branch of the separate Bio-FeatureIO repo, as the current (significantly refactored) code is only partly refactored and needs more work and integration.
> 
> I'm curious about how this works in terms of git storage.
> 
> Does this mean that the separate Bio-FeatureIO repo will have the entire history of BioPerl inside it? (Making git clones of Bio-FeatureIO 189MB?)
> 
> In the recent past I have attempted pulling certain files across git repos before, and ended up with the full history of repo1 inside repo2. I'm unclear if this is just how life is, or if I did it wrong.
> 
> You could, of course, always just cp text files in, but then you lose the history of those files. 
> 
> Is there some way to get all the history of a handful of files from massive repo1 into tiny repo2 without making repo1 massive?
> 
> I don't know if any of these considerations are important for the eventual de-monolithification of BioPerl, I was just generally curious. git does that to me.  :)
> 
> Thanks,
> 
> Jay Hannah
> http://biodoc.ist.unomaha.edu/wiki/User:Jhannah

I'm just planning on having something to the effect of 'Bio-FeatureIO is a set of modules developed by author X that once was part of bioperl-live, but was removed at point XYZ to significantly refactor the code,' then point back to bioperl-live if anyone is interested in software archaeology.  Not sure we would need to go beyond that.

chris