Re: What's cooking in git.git (May 2017, #03; Wed, 10)

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: May 11, 2017, 13:08 UTC
Message-ID: <alpine.DEB.2.21.1.1705111506030.146734@virtualbox>
In-Reply-To: <xmqqinl9xpb8.fsf@gitster.mtv.corp.google.com>

Hi Junio,

On Wed, 10 May 2017, Junio C Hamano wrote:

Show 19 quoted lines

> * jc/bundle (2016-03-03) 6 commits
>  - index-pack: --clone-bundle option
>  - Merge branch 'jc/index-pack' into jc/bundle
>  - bundle v3: the beginning
>  - bundle: keep a copy of bundle file name in the in-core bundle header
>  - bundle: plug resource leak
>  - bundle doc: 'verify' is not about verifying the bundle
> 
>  The beginning of "split bundle", which could be one of the
>  ingredients to allow "git clone" traffic off of the core server
>  network to CDN.
> 
>  This was surrected from a "to be discarded" pile, as from time to
>  time people wonder about resumable clone that can be primed without
>  bothering Git servers with dynamic packfile creation, and some
>  people seem to think that the topic could serve as a useful
>  building block for that goal.  But nothing seem to have happend.
>  Unless people really want it, I am inclined to discard this topic.
>  Opinions?

The primary concern that wants to be solved by these patches is the resumable clone, right?

If so, I think that we may want to rethink that approach. If your bandwidth is flakey and your repository is large, the upcoming work to support fetching objects incrementally (there are three competing proposals about this IIUC, hopefully they will settle into a single approach soon) may actually be the better way forward.

Ciao, Dscho