Mailing-List: contact dev-help@subversion.apache.org; run by ezmlm
Precedence: bulk
MIME-Version: 1.0
In-Reply-To: <20161012142805.njen7ikgmc67ykyw@pks-xps>
References: <20161012142805.njen7ikgmc67ykyw@pks-xps>
From: Ivan Zhakov <ivan@visualsvn.com>
Date: Fri, 14 Oct 2016 14:40:25 +0200
Message-ID: <CABw-3YeeMWADG5y3Y17TPYMTZ4ikn4Zzug7grX6mF5Ve0OXDcg@mail.gmail.com>
Subject: Re: [PATCH] Reject checkouts to existing directory
To: Patrick Steinhardt <patrick.steinhardt@elegosoft.com>
Cc: Subversion <dev@subversion.apache.org>
Content-Type: text/plain; charset=UTF-8
archived-at: Fri, 14 Oct 2016 12:40:55 -0000

On 12 October 2016 at 16:28, Patrick Steinhardt
<patrick.steinhardt@elegosoft.com> wrote:
> Hi,
>
> attached is a patch to reject checkouts to already existing
> directories when `--force` is not given. This is according to
> `svn co --help`.
>
> [[
> Reject checkout to existing paths without force
>
> * subversion/svn/checkout-cmd.c:
>   - (svn_cl__checkout): Reject checkout to existing directory
>     without --force
> ]]
>
As far I see we have four different cases of checking out to existing directory:
a. Checkout of non-empty directory to non-empty local directory
b. Checkout of non-empty directory to empty local directory
c. Checkout of empty directory to non-empty local directory
d. Checkout of empty directory to empty local directory

I think that cases (b),(c),(d) should not require '--force'. It seems
to be pretty safe operations. Also (c) is common way to 'takeover'
(version) existing repository. I meant:
1. Create empty directory in repository
2. Check it to existing directory with source code
3. Add all requested files/directories
4. Review all changes and commit.

But it makes sense for to require '--force' flag for case (a): it's
very easy mess up local changes in this situation.

-- 
Ivan Zhakov