Re: What's cooking in git.git (Nov 2009, #03; Sun, 15)
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Nov 16, 2009, 00:26 UTC
- Message-ID
- <7vtywvb9hs.fsf@alter.siamese.dyndns.org>
- In-Reply-To
- <loom.20091116T003914-635@post.gmane.org>
Peter Weseloh <Peter.Weseloh@gmail.com> writes:
> Does that mean this feature will not make it into 'master' any time soon?
It does not mean anything about this _feature_, but it does mean this particular _implementation_ is not likely to.
> Having the ability to do sparse checkouts only of the repo sounds like a > way out.
There is a big difference between (a) a feature is desirable and (b) a particular implementation of that feature does things right.
The change in the topic is very intrusive and I am hesitant to merge it into even 'next'. Before considering inclusion to 'next', it does not matter that much if a topic realizes the feature as it advertises, but it does matter a lot if it breaks things for people who do not need the feature. While 'pu' has carried this topic for quite a long time, we haven't heard much success report from folks like you with the need for the feature if this implementation worked well for them without breaking things.
I am not convinced that it won't regress fundamental things for the common codepaths when "sparse" is not used, and the latest update patch posted to the list (I do not think I picked it up) seemed very likely to regress things for the normal codepath.
> Another question: What's the timeline for 1.7.0? I couldn't find anything > about it neither here nor in the wiki.
When it is ready. But it is likely that there won't be 1.6.7.