Nullbackend definition

[Vlad Skvortsov]

Stefan Reichör wrote:
> Hi!
>
> In the next weeks I would like to invest some time for ditrack development.
> And I want to see some progress on the Nullbackend.
>
> However, I need some help from the core developers to go in the correct direction.
>
>
> I would like to make dt-createdb fully working with the Nullbackend
>
>
> The next step would be to support dt new.
>   

We are currently in the middle of the release cycle for version 0.6. So 
it makes sense to hold on until the 0.6 branch is cut. The next thing 
would be to bring null-backend branch in sync with the trunk.

>
>
> So I did the following experiment (using the Nullbackend branch):
>
>
> # dt-createdb t1
>
> # ls -la t1
> total 4
> drwxr-xr-x 5 stefan stefan 152 2007-08-28 21:38 ./
> drwxr-xr-x 9 stefan stefan 440 2007-08-28 21:38 ../
> drwxr-xr-x 2 stefan stefan  48 2007-08-28 21:38 data/
> drwxr-xr-x 2 stefan stefan 152 2007-08-28 21:38 etc/
> drwxr-xr-x 2 stefan stefan  72 2007-08-28 21:38 meta/
> -rw-r--r-- 1 stefan stefan 105 2007-08-28 21:38 README.txt
>
> # dt new
> svn: '.' is not a working copy
> '.' (default path) is not an issue database root
>
>
> So my question is now, how should dt new work when using the Nullbackend?
> Should it be called from the t1 directory?
>   

That's one option. You can also use '-d' switch or DITRACK_ROOT 
environment variable.

> Which files will be generated?
>   

'dt new' should create files/directories under t1/data, if that's what 
you are asking.

> Any other things I need to consider?
>   


-- 
Vlad Skvortsov, vss at 73rus.com, http://vss.73rus.com