|Re: [p2-dev] end 2 endl test release|
This is great. An important thing to test. Would it make sense to use
a smaller test data set that we can run as part of the p2 tests? This
makes it easier (and thus more likely) that people will run it. The
full SDK test might be good as part of the SDK packaging effort but it
seems we should have a reduced set for p2 itself.|
Related, is the test data for this fixed? We have had trouble in the past where the test data changes as the code evolves. This has led to both false positives and negatives. This should evolve as one of the best practices for Equinox testing in general.
Pascal Rapicault wrote: