abdera-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "joe kim" <joe....@gmail.com>
Subject Re: Testing Requirements
Date Wed, 05 Jul 2006 21:50:44 GMT

It sounds like alternatives to JUnit is not a priority to you.  I am
not convinced I should care either.  But, I think your insistence
that, "The existing stuff works just fine", is not a valid argument
for not doing an evaluation.

ant and maven are de facto standards because enough people started
using them that they became the way to build java projects.  Before
the advent of ant and maven, a person had to actually go out and
"evaluate" that ant was better than make.  Of course, someone could
make the claim that "Hey, the existing stuff works just fine.  Let's
not do an eval."  That would be a valid argument if they 1) knew
enough about alternatives to know that they  didn't need any of their
features 2) had more important things to worry about.

> Well, there will obviously be new test cases, but I don't see how that
> leads to "so we should dive into an evaluation of competing testing
> frameworks".  The existing stuff works just fine.

If there is a compelling reason to change frameworks, then the sooner
it happens the sooner the test cases can take advantage of them.  It
doesn't make sense to change frameworks after the immeninent test
cases are written.  Whether or not to change frameworks and whether or
not to even consider different frameworks are two separate arguements.

> Ant and Maven provide abilities /that we need/ that you cannot easily
> do in make.  They're also defacto standards in the Java world.
> Neither of those are true for JUnit 4 or TestNG at this time, AFAIK.

Can you clarify what "AFAIK" entails?  Did you already evaluate
whether or not features from JUnit 4 and TestNG are not useful for our
purposes?  If so, you should have just said so, no need to argue about
something that's already done.  If not, how can you make such blanket


View raw message