From: Junio C Hamano <gitster@pobox.com>
Date: Fri, 31 Jul 2015 00:58:04 GMT
Subject: Re: [PATCH v2 6/6] clone: add tests for cloning with empty path
Message-ID: <xmqqzj2d5ejn.fsf@gitster.dls.corp.google.com>
In-Reply-To: <CAPig+cQ5wR8xp8erAHBzEMEs62ddULrrML53GNVuuie3kd9EvA@mail.gmail.com>

Eric Sunshine <sunshine@sunshineco.com> writes:

> On Wed, Jul 29, 2015 at 11:51 AM, Patrick Steinhardt <ps@pks.im> wrote:
>> Test behavior of `git clone` when working with an empty path
>> component. This may be the case when cloning a file system's root
>> directory or from a remote server's root.
>>
>> Signed-off-by: Patrick Steinhardt <ps@pks.im>
>> ---
>> diff --git a/t/t1509-root-worktree.sh b/t/t1509-root-worktree.sh
>> index 553a3f6..acfa133 100755
>> --- a/t/t1509-root-worktree.sh
>> +++ b/t/t1509-root-worktree.sh
>> @@ -237,6 +237,45 @@ test_foobar_foobar

All true, but a more interesting question is why add more to this
test, which is known to be skipped by everybody?  The issue being
corrected is that any "<scheme>://<user>@<pass>:<site>/" that says
"the whole site serves a single repository" is problematic.

Surely, file:// and ssh:// may be examples of schemes that require
the filesystem root to be usable as the trash directory to test,
requiring a dedicated VM (causing most people to skip t1509), but
wouldn't "http://<user>@<pass>:<site>/" be easier to arrange to make
the whole site serve a single repository?