|
|
> Ok, thanks. What I don't understand is what is this Resources.java file in
> test06?
It's not test06, it's testIncludedResources :-)
Felipe, any idea? It's not commented and its use is not obvious to
> me. The fact that there is no sourceDirectory defined and that the test
> still passes seem to indicate it's not required and it should be deleted. Is
> this true? ;-)
Yes, I guess it's not necessary.
> WAR are there!). Would it be possible to comment all files that are
> committed so that we have at least the same level of comment quality than
> what exists in the rest of Cactus? We don't need to overdo it, but anywhere
> where it's not obvious would be much welcome... :-)
I do commit every file, but sometimes I use the same comment for a recursive
commit (that's what happened in this case). So, I think we should either
document the class itself or, better yet, document what the test case does in
the <description> (at the POM level).
> This seems completely useless to me and we've paid great attention in all
> the rest of the Cactus code not to have this (because it's not useful IMO).
> Let me know if there's a reason I don't see! :-)
Ok.
> Last, the rule we have followed so far in Cactus was to name JUnit tests
> with the following pattern: TestXXXX.java. I know lots of people prefer the
> pattern XXXTest.java (I also do prefer it now - I didn't some time back).
Ok, I'm fine with that.
> The problem is that we have to be consistent and follow the existing
> pattern. If we want to change it, fine, I'm all for it, but we would need to
> change it for all files (which can be done in a scattered way).
Agreed - convention is important. BTW, what about defining a default package
for testcase classes (for instance, org.apache.cactus.testcases.maven)?
-- Felipe
---------------------------------------------------------------------
To unsubscribe, e-mail: cactus-dev-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: cactus-dev-help@xxxxxxxxxxxxxxxxxx
|
|