On Sun, Apr 26, 2015 at 10:26:22PM +0200, Michael Haggerty wrote:
> After that, there is one patch for each callsite, rewriting it to use
> for_each_ref natively (which usually entails modifying my_function to
> take an object_id parameter then undoing the wrapper). These patches
> involve a little bit of thought, but not too much. And the results are
> very bisectable because each patch makes a single small change. I also
> suspect it might be easier to rebase and/or merge my patch series, for
> the same reason.
> 
> The end result was very similar to yours, so I am confident that the net
> result of your patch series is correct. But the remaining differences in
> the end results are also interesting. I made a few more changes in the
> neighborhood of the patches, not to mention a few formatting
> improvements in code that I touched. If you compare the tip of my
> branch, above, to the tip of yours (I uploaded that to my repo too, as
> branch "bc-oid-refs"), it may give you some ideas for other code that
> can be changed to object_id.

This is a very interesting approach.  I've only just had time to look at
it, but I like it.

I agree that it's much more bisectable, although your series has a much
larger quantity of patches.  I feel like sending an 83-patch series to
the list may frighten reviewers away.  That's really the only thing
preventing me from replacing my series with yours.

Junio, what would you think about such a series?  Would you prefer
slightly larger patches, one per file, if it meant that we had many
fewer patches, or would you prefer a large number of patches touching a
single function each?
-- 
brian m. carlson / brian with sandals: Houston, Texas, US
+1 832 623 2791 | http://www.crustytoothpaste.net/~bmc | My opinion only
OpenPGP: RSA v4 4096b: 88AC E9B2 9196 305B A994 7552 F1BA 225C 0223 B187