>> 6) delete dir1 + add dir1/bla.txt
>> Like in every shell, I would think that this would delete all dir1/*
>> files first, and then adds a new file bla.txt in dir1.
>
> Tthe question is whether it is missing an
>
> add dir1/
>
> Do we create directories implicit?
In my opinion, yes. If you look
> We currently have no mv in the ChangeSet and I would propose that we
> start with delete and add operations (testcases below) for version
> 0.1.
Well, that's where it gets interesting ;)
> "Also how do we deal with directories?
> delete dir1
> add dir1/test.txt
>
> mv dir1/test.text dir2/test.t
Well. Looks like archivers can handle empty files. But this just
modifies slightly the testcases below, just wanted to tell
Christian
On Mon, Jan 12, 2009 at 7:32 AM, Christian Grobmeier
wrote:
> On Sun, Jan 11, 2009 at 6:36 PM, Torsten Curdt wrote:
>>> OK, please feel free to give me your idea
On Sun, Jan 11, 2009 at 6:36 PM, Torsten Curdt wrote:
>> OK, please feel free to give me your ideas here. I gladly refactor the
>> code now, before it becomes to big.
>
> I've already removed the worker :)
Yes, just checked it. I'll sound like the guys at "who want to be a
millionaire", but this
> OK, please feel free to give me your ideas here. I gladly refactor the
> code now, before it becomes to big.
I've already removed the worker :)
>> The implementation does cover the most simple changes but I guess it
>> can become more complicated than that. Again see the comments on the
>> iss
> Applied a slightly changed version. See my comments on the issue.
> (Essentially I wasn't a big fan of the worker class in this case.)
OK, please feel free to give me your ideas here. I gladly refactor the
code now, before it becomes to big.
> The implementation does cover the most simple chang
Applied a slightly changed version. See my comments on the issue.
(Essentially I wasn't a big fan of the worker class in this case.)
The implementation does cover the most simple changes but I guess it
can become more complicated than that. Again see the comments on the
issue.
I think we should s
Hi,
i made a patch of a first idea of the ChangeSet design which has been
proposed by Torsten.
https://issues.apache.org/jira/browse/SANDBOX-183
The testcase fails with the ArOutputStream. Problem here is located in
the Ar-Implementation. Torsten,
would be quite cool if you can check this out, I
