Jeff Hostetler [off-list ref] writes:
On 3/22/2017 4:54 PM, Junio C Hamano wrote:
quoted
quoted
Thank you for an update.
One notable difference I noticed since the previous round is that
this no longer adds precomputed hash to "struct cache_entry". As
you are aiming to manage an index with a large number of entries,
this is a welcome change that makes sense.
Yes, this completely isolates the changes inside the name-hash.c
code. And it eliminates the need to update/invalidate the precomputed
hash values as entries are changed.
Thanks. Having a summary like that for differences since the last
round in the cover letter would help during the review, but this is
an example that lack of such a summary would be one way to make sure
that reviewer(s) are paying attention ;-)