threads / discuss / 32120

use cases for git namespaces

Subject: use cases for git namespaces

## tl;dr

2 messages between Nov 15, 2012 and Nov 15, 2012.

replies: 1people: 1as markdown or json

Sitaram Chamarty· Nov 15, 2012, 08:33 UTC · lore
Hi,

It seems to me that whatever namespaces can do, can functionally be done using just a subdirectory of branches. The only real differences I can see are (a) a client sees less branch clutter, and (b) a fetch/clone pulls down less if the big stuff is in another namespace.

I would like to understand what other uses/reasons were thought of.

I looked for discussion on the ml archives. I found the patch series but could not easily find much *discussion* of the feature and its design. I found one post [1] that indicated that "part of the rationale..." (being what I described above), but I would like to understand the *rest* of the rationale.

Pointers to gmane are also fine, or brief descriptions of uses [being] made of this.

[1]: http://article.gmane.org/gmane.comp.version-control.git/175832/match=namespace
Thanks
-- 
Sitaram
Sitaram Chamarty· Nov 15, 2012, 12:39 UTC · re: Sitaram Chamarty · lore

Re: use cases for git namespaces

On Thu, Nov 15, 2012 at 2:03 PM, Sitaram Chamarty <sitaramc@gmail.com> wrote:
Show 9 quoted lines
> Hi,
>
> It seems to me that whatever namespaces can do, can functionally be
> done using just a subdirectory of branches.   The only real
> differences I can see are (a) a client sees less branch clutter, and
> (b) a fetch/clone pulls down less if the big stuff is in another
> namespace.
>
> I would like to understand what other uses/reasons were thought of.

(I should mention that I am asking from the client perspective. I know that on the server this has potential to save a lot of disk space).

Show 15 quoted lines
> I looked for discussion on the ml archives.  I found the patch series
> but could not easily find much *discussion* of the feature and its
> design.  I found one post [1] that indicated that "part of the
> rationale..." (being what I described above), but I would like to
> understand the *rest* of the rationale.
>
> Pointers to gmane are also fine, or brief descriptions of uses [being]
> made of this.
>
> [1]: http://article.gmane.org/gmane.comp.version-control.git/175832/match=namespace
>
> Thanks
>
> --
> Sitaram
-- 
Sitaram

← back to recent threads