On Apr 29, 2006, at 1:03 AM, ant elder wrote:
On 4/26/06, Jim Marino <[EMAIL PROTECTED]> wrote:
<snip/>
Ant, you -1 this stating that you weren't sure what this entails and
that it was buried in another email. Now that Dan pulled it out and
included a description of what checkstyle is, do you have specific
concerns about the proposal?
For the record yes I'm no longer -1. I do agree with Dan that having a
consistent style does make code look better and makes it easier to
understand. It does seem a bit odd that we looking at having bad
style fail
a build and checking this with every build but at the same time
still don't
test things that really matter like the samples.
I believe there is violent agreement that our test coverage is
lacking (sucks). It was decided a while back that this should be
fixed by writing test cases which exercise our APIs and integration
with deployment platforms. I don't think we need to continue to
rehash this - let's just go about doing it.
Another concern i have is I
expect this will cause more build breaks - I'm fine with that,
builds break
- but right now people seem to be taking build breaks a bit more
seriously
than I'd like.
I'm not sure how it will cause more build breaks except on
developers' machines when they perform a build prior to checkin. I
think this is a good thing since it will stop code from being checked
in that shouldn't be.
Could you be more specific about your concern over the seriousness of
build breaks? Our policy has been that when they happen, the person
responsible needs to fix it ASAP since it holds everyone else up.
There was an occasion when we purposely decided to break the build
for two days, but that was a deliberate exception and discussed on
the list prior to it being done. Once we have an integration server
set up, this means that developers will need to wait around five
minutes or so after a checkin to receive an all-clear email. I'm not
concerned about this as it is standard practice, are you?
Since there are no longer -1s on the checkstyle issue, I would like
to go ahead with implementing it after JavaOne. Jeremy, I assume you
are o.k. with the proposal now on the table?
Jim
...ant
