Did just Cedric lost interest and left it as an incomplete experiment,
or there were implementation hurdles? My understanding (without
looking at the code yet) is that just a little effort is needed to
complete it, so I'm wondering why (since the initial experiments in
2006) nobody came forward and completed it. I would like to try to
complete this implementation myself, but before I waste my time in
something that has been already tried (and maybe can't be done quickly/
easily enough), I'd like to get feedback from others that did it in
the past (if any).
My interest in this feature is more for a regression kind of testing
on different environments than workload distribution to accelerate
runtime. In fact, another fine way to achieve what I need, is simply
to re-run the tests on the various platform. And that is working fine,
except that I end up with quite a lot of reports, which then became a
pain to read.
So my next question is: is there any way to consolidate reports from
separate runs? I know I could write my own Reporter and use it to save
(maybe serialize) the data instead of creating the HTML. And then I
may write something that reads all that data from different runs (on
different platforms), consolidates it and uses the original testNG
Reporter to create the HTML in that format I love. But again, this
sounds like a need others had (since the original
http://beust.com/weblog2/archives/000362.html post, people were
salivating for runs on different platform) so I'm wondering if there
is a way to do it without changing/adding code to testNG or if there
is much much more work to do than it seems.
Thanks and have a nice weekend,
You received this message because you are subscribed to the Google Groups "testng-dev" group.
To post to this group, send email to testn...@googlegroups.com.
To unsubscribe from this group, send email to testng-dev+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/testng-dev?hl=en.
2011/11/19 Cédric Beust ♔ <ced...@beust.com>: