[PATCH 8 of 9] util: support multi-threaded zstandard compression for bundles

[Gregory Szorc]

OK then. Thanks, Yann.

Does anyone see any value in adding it as an experimental config option in
Mercurial? Or should we drop the feature for now? I'd kinda like it there
to make it easier to experiment with. But if others aren't comfortable with
it, I totally understand.

On Thu, Apr 6, 2017 at 10:26 AM, Yann Collet <cyan at fb.com> wrote:

> Yes, the multi-threading API is still considered experimental, hence not
> yet recommended for production use.
>
> We are actually working on that currently, and expect to change this
> status in a future release.
>
>
> On 06/04/2017, 10:21, "Gregory Szorc" <gregory.szorc at gmail.com> wrote:
>
>
>
>     > On Apr 6, 2017, at 06:10, Yuya Nishihara <yuya at tcha.org> wrote:
>     >
>     >> On Sat, 01 Apr 2017 15:31:59 -0700, Gregory Szorc wrote:
>     >> # HG changeset patch
>     >> # User Gregory Szorc <gregory.szorc at gmail.com>
>     >> # Date 1491085720 25200
>     >> #      Sat Apr 01 15:28:40 2017 -0700
>     >> # Node ID c171ff452d36d4caeab1887df4da43e45a61022a
>     >> # Parent  7f469db35d4a303d53b4c7dc141a8378fa67d9e1
>     >> util: support multi-threaded zstandard compression for bundles
>     >
>     > Is multi-threading support stable enough? I heard someone saying the
> code
>     > quality wasn't great.
>     >
>     > https://github.com/facebook/zstd/issues/491#issuecomment-277780684
>
>     "experimental" in zstd land typically means "unstable API." This isn't
> an issue for python-zstandard since it bundles it's on zstd copy and
> statically links.
>
>     But just in case the code may not be suitable to run in production
> just yet, I've CCd Yann who give an opinion.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mercurial-scm.org/pipermail/mercurial-devel/attachments/20170406/c7610e9a/attachment.html>