Here is what I see how we can place in SVN what we have or potentially can
have some time in the future:
SVN:
 /testing
   /vm
       /<some_test_suite>    //dedicated vm tests, if any
           /build                     //test suite build
           /tests                     //test suite tests
   /api
       /autogen
           /rmi                      //ITC test generator for unit rmi
tests
           /math                    //ITC test generator for unit math
tests
       /api_sigtest                //scripts to run JAPI vs Harmony API
       /<another test suite>   //another API suite – stress, for example
   /tools
       /junit_ext                    //here we store JUnit extensions if
and when we have them
       /coverage                   //scripts for getting coverage
information for API, VM
             /api
             /vm
       /<another tool>
   /misc
       /scenarios                   //automated scenarios
             /eclipse3.1            //Eclipse 3.1 automated scenarios
             /eclipse3.2
             /jedit

Thanks, Vladimir

On 6/8/06, Geir Magnusson Jr <[EMAIL PROTECTED]> wrote:

As we start to flesh out the project more and more, one of the things
we've talked about is testing (and I don't mean unit tests..).  I think
that our testing infrastructure and results tracking is going to be a
very significant undertaking in the project, and the time feels right to
seriously make it a part of the project.

I've thought about this a bit, and while I'm sure we'll probably get
something setup here at the ASF,  I honestly don't think a single
resource like the ASF (or anyone) will be able to match what we can do
together as a distributed community, in terms of numbers or variety.

Therefore, given that have been significant achievements in this already
(like what IBM has running...), and I'm sure that there are many others
interested with contributions of code as well as time,  I'd like to
propose...

* Add /testing (or some other appropriate name) off of root, as a peer
to /classlib, /jchevm, /drlvm etc

* Bring together what various people are using into a unified and
easy-to-use setup in SVN that anyone can just download, configure and
run  (Hey, Mark!)

* Setup a place on the website to track who is doing what platform.

This way, no one entity has any "responsibility" for doing this for the
community - we all do it - and we all share in contributing to the
infrastructure.

I'd certainly setup a machine at home to do this, and I'd work to get
something setup inside my employer, Intel.  I bet a lot of other people
would as well.

Thoughts?

geir




---------------------------------------------------------------------
Terms of use : http://incubator.apache.org/harmony/mailing.html
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


Reply via email to