[Insight-developers] New workflow to add test data

[Cory Quammen]

Hi Brad,

The new workflow for adding test data is very nice. You are too modest
about the quality of your instructions in the email (they were quite
detailed).

I have two comments about the workflow.

One thing that caused me some headaches was the hiding of my original
test data after running CMake. When I made a mistake while going
through this process (which happened a number of times), I deleted the
MD5 files and started the process again from a clean slate. Each time,
I had to copy the test data back into the repository from some
external place on my machine. It would be nice if the test data could
be left in place during development. Perhaps a pre-commit check could
be added to prevent this original test data from being committed.

Another point of concern: It seems possible that the gerrit builds
could begin before the data objects are transferred to the web server
if the processes are not synchronized. Is it set up so that the
transfer of the objects always happens prior to running the gerrit
builds?

Thanks,
Cory

On Mon, Jun 20, 2011 at 10:20 AM, Brad King <brad.king at kitware.com> wrote:
> Hi Folks,
>
> I've documented the new workflow for adding test data here:
>
>  http://www.itk.org/Wiki/ITK/Git/Develop/Data
>
> and linked to it from the appropriate points in the main development
> instructions here:
>
>  http://www.itk.org/Wiki/ITK/Git/Develop
>
> The instructions are a refined version of what I outlined here:
>
>  http://www.itk.org/mailman/private/insight-developers/2011-June/019048.html
>
> One no longer needs to add or remove any options by hand to use it.
>
> For the moment we will use this approach for new data and keep the
> Testing/Data submodule (as read-only) for existing data.  After we
> "kick the tires" on the new approach we can look at converting all
> the existing data to work the new way and remove the submodule.
>
> -Brad
>
> P.S.  Thanks to Cory and Hans for stumbling through this without any
> real instructions to help me work out some kinks.
>



-- 
Cory Quammen
Research Associate
Department of Computer Science
University of North Carolina at Chapel Hill
http://www.cs.unc.edu/~cquammen