On Thu, Jun 11, 2015 at 10:32 AM, Jeff King <peff@xxxxxxxx> wrote:
> On Thu, Jun 11, 2015 at 08:02:33PM +0700, Duy Nguyen wrote:
>
>> > I see that do_fetch_pack checks server_supports("shallow"). Is that
>> > enough to cover all fetch cases? And if it is, why does it not cover the
>> > matching clone cases?
>>
>> I think this replacement check would do
>>
>> if ((args->depth > 0 || is_repository_shallow()) && !server_supports("shallow"))
>> die("Server does not support shallow clients");
>
> Oh, indeed, there is the depth flag I was looking for. :)
>
> And from some rudimentary testing, I believe that:
>
> git init
> git fetch --depth=1 ...
>
> is currently broken in the same way as clone (we are not shallow yet, so
> it does not complain when the server does not support it). I think the
> patch above fixes both that and the clone case.
Great point! Duy's proposed fix worked for me testing against the Git
server that
led me to the original bug (code.google.com), for both clones and init/fetch.
Shall I send that out as a revised patch for review? (This is my first
stab at a Git
patch)
> Of course it's hard to add to the test suite, since we do not have a way
> of hitting a server that does not understand shallow (I simply fudged
> server_supports() to return false on the client).
I noticed this, and figured it would be more dangerous for users to include a
testing backdoor (eg a flag overriding the advertised capabilities). Is that an
accurate assessment, or is there a safer way to test these sorts of interop
cases?
Cheers!
--
Michael Edgar | Software Engineer | adgar@xxxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html