Web lists-archives.com

Re: [PATCH v1 0/8] Introducing odb remote

Christian Couder <christian.couder@xxxxxxxxx> writes:

> This is a follow up from the patch series called "Promisor remotes and
> external ODB support" that I sent earlier this year.
> Following discussions of these patch series, where Junio said "a
> minimum s/ext/remote/ would clarify what it is", I decided to rename
> "external odb" to "odb remote".

In "external odb", "odb" is a noun, decorated by adjective
"external".  I would expect the same would apply to "remote" and

> I am still open to another name, but I
> think that "odb remote" works well with "odb helper" that was already
> used in the series and is as good or perhaps better than "remote odb",
> as a "remote odb" I think would be easier to confuse with a regular
> "remote".

As long as our use of "remote" as a noun is understood as omitting
(an obvious) "repository" from a noun phrase "remote repository", I
do not think there is any confusion.

"odb remote" sounds like the phrase is about a noun "remote" (we do
use "remote" as a noun when talking about fetching and pushing, as a
short hand for "remote repository") that uses "odb" as an adjective
somehow (e.g. "this is one of those remotes defined but not for the
usual fetching and pushing, but for accessing its odb", or something
ilke that), which somehow does not sit well on my tongue nor in ears.
Perhaps that is because I am not well versed in languages whose natural
word order in a noun phrase is noun followed by adjective, but I dunno.

> Another obvious difference with the previous series is that this
> series is only about integrating with the promisor/narrow clone work
> and showing that it makes it possible to use more than one promisor
> remote.

Sounds like a sensible "incremental" approach.

Will read the patches and commet on them in another sitting.