[development] Git best practices for client codebases

Fahri Reza dozymoe at gmail.com
Tue Mar 1 21:03:47 UTC 2011


> On 3/1/11 7:53 AM, Fahri Reza wrote:
>> how about doing shallow clone with
>> $git-clone --depth=1
>> then run recursive-diff on the sub-directories (n=1 is just an example)
>> 
>> can someone gives comment whether this approach is recommended?
>> 
> Clone with depth parameters will help reduce size, yep. Not sure what

I just thought that if I use this, it's pretty much like fetching drupal 
without the tar.gz attribute (compression), and the benefits from incremental-
diffs from git is not well implemented.

> you're going for with a diff on subdirs - if you mean diffing modules,
> then diffing subdirs really depends on how you've chosen to compose your
> tree - submodules, subtree merges, or plain nested repos.

he he, my kmail doesn't support quoting for now, it's related to the OP's 
question:
  .. our usual practice with clients was to checkout core and contrib using
  CVS, so that we can easily monitor any patches that have been applied,  ..


-- fireh --


More information about the development mailing list