>Yes, but AFAIK the Solaris test suites and the ksh93/AST test suites are
>designed to do different things - the ksh93 tests were created to find
>bugs which were hit once (to prevent that they ever surface again) while
>the Solaris test suites are more designed to find violations of
>standards... or am I wrong here ?
You are wrong.
The Solaris test suites serve multiple purposes; there is a big
patch of standards tests *and* there's a big batch of regression
testing:
- making sure stuff works as it should
- test cases for particular nasty regressions/security bugs
which were hit.
Not all bugs get a regression test but some do.
>But would these test suites be part of which single build - and break
>the build if a test failure occurs ?
Not sure I understand that question. The test suites report all failures
which occur and don't terminate on the first failure.
Casper
