New candidate JEP: 357: Migrate from Mercurial to Git

David Lloyd david.lloyd at redhat.com
Tue Jul 16 15:41:38 UTC 2019


FWIW you can "seed" your clone if you have another one handy by
cloning the local directory e.g. "git clone -o local ../other-checkout
new-checkout".  Then fetching the upstream will only fetch objects
that weren't in "other-checkout".  I do this frequently for large
repositories, which is especially useful in laptop-in-coffee-shop
kinds of situations.

On Tue, Jul 16, 2019 at 10:37 AM Aleksey Shipilev <shade at redhat.com> wrote:
>
> On 7/16/19 4:23 PM, Aleksey Shipilev wrote:
> > On 7/16/19 2:19 PM, Severin Gehwolf wrote:
> >> Fresh clone times: Git has fewer metadata to transfer (above). JDK-
> >> 8211383 could help for Mercurial how does it compare?
> >
> > Last I checked, jdk-jdk zstd bundle is about 260 MB. This is roughly how much data transfer would
> > happen with clonebundles.
>
> For fun, redid the cloning experiment with clonebundles vs current Git mirror here:
>   https://bugs.openjdk.java.net/browse/JDK-8211383?focusedCommentId=14278459#comment-14278459
>
> Amusingly, Mercurial with clonebundles gets me less total bandwidth than Git checkout, even after
> Mercurial client did two transfers: the bundle itself, and several months of changesets on top. So
> if we do care about transfer overhead, then migrating from (improved) Mercurial to Git is the
> pessimisation, especially for slow links.
>
> See how "Alternatives" section gets curiouser and curiouser.
>
> --
> Thanks,
> -Aleksey
>


-- 
- DML


More information about the discuss mailing list