Findbug count is starting to rise again.  Us fellas doing the committing
have let a few slip through.  We need to be more vigilant.

What is the best way evaluating a patch post commit when findbugs warnings
is not zero?  As soon as one findbugs item comes in, because findbugs still
reports many issues masked by the tolerated count over in
dev-tools/test.patch#OK_FINDBUGS_WARNINGS, it is hard to figure the culprit
given findbugs:findbugs dumps out xml with OK_FINDBUGS_WARNINGS issues.
 Any pointers?

Meantime let me set the warnings back to zero.

St.Ack

Sorry, this is new.

On Wed, Jan 16, 2013 at 10:38 AM, Stack <[EMAIL PROTECTED]> wrote:

> Findbug count is starting to rise again.  Us fellas doing the committing
> have let a few slip through.
>
--
Best regards,

   - Andy

Problems worthy of attack prove their worth by hitting back. - Piet Hein
(via Tom White)

I think the only simple solution is to put it to zero, then it's easy to
see what's new. But it requires some work. Not much. But some :-(
On Wed, Jan 16, 2013 at 7:53 PM, Andrew Purtell <[EMAIL PROTECTED]> wrote:

> Sorry, this is new.
>
> On Wed, Jan 16, 2013 at 10:38 AM, Stack <[EMAIL PROTECTED]> wrote:
>
> > Findbug count is starting to rise again.  Us fellas doing the committing
> > have let a few slip through.
> >
>
>
>
>
> --
> Best regards,
>
>    - Andy
>
> Problems worthy of attack prove their worth by hitting back. - Piet Hein
> (via Tom White)
>

On Wed, Jan 16, 2013 at 10:57 AM, Nicolas Liochon <[EMAIL PROTECTED]> wrote:

> I think the only simple solution is to put it to zero, then it's easy to
> see what's new. But it requires some work. Not much. But some :-(
>
>
Figured.  You are right N.  Thanks.  Lets squash this last set...
St.Ack

One issue with the current situation is that it's pretty hard to
understand what warnings are new, and in some JIRAs the report seems
to be wrong (e.g. if you download the reports from old and new build
you see the difference is only in dates and timing).
I filed a JIRA to make it better (so you could rely on it and see what
to fix) but I won't be able to get around to it at least until next
week :(

On Wed, Jan 16, 2013 at 12:16 PM, Stack <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 16, 2013 at 10:57 AM, Nicolas Liochon <[EMAIL PROTECTED]> wrote:
>
>> I think the only simple solution is to put it to zero, then it's easy to
>> see what's new. But it requires some work. Not much. But some :-(
>>
>>
> Figured.  You are right N.  Thanks.  Lets squash this last set...
> St.Ack

Yes, I was also surprised some time ago, that our findbugs reports does not
show the diff, but the whole, and you need dark skills to find those the
patch has caused.

The only reasonable way to proceed is to get them to zero.
On Thu, Jan 17, 2013 at 10:41 AM, Sergey Shelukhin
<[EMAIL PROTECTED]>wrote:

> One issue with the current situation is that it's pretty hard to
> understand what warnings are new, and in some JIRAs the report seems
> to be wrong (e.g. if you download the reports from old and new build
> you see the difference is only in dates and timing).
> I filed a JIRA to make it better (so you could rely on it and see what
> to fix) but I won't be able to get around to it at least until next
> week :(
>
> On Wed, Jan 16, 2013 at 12:16 PM, Stack <[EMAIL PROTECTED]> wrote:
> > On Wed, Jan 16, 2013 at 10:57 AM, Nicolas Liochon <[EMAIL PROTECTED]>
> wrote:
> >
> >> I think the only simple solution is to put it to zero, then it's easy to
> >> see what's new. But it requires some work. Not much. But some :-(
> >>
> >>
> > Figured.  You are right N.  Thanks.  Lets squash this last set...
> > St.Ack
>

I have started to look at the warnings one by one. Most of them are
easy fixes. Submitted one patch, more to come.

JM

2013/1/17, Enis Söztutar <[EMAIL PROTECTED]>:
> Yes, I was also surprised some time ago, that our findbugs reports does not
> show the diff, but the whole, and you need dark skills to find those the
> patch has caused.
>
> The only reasonable way to proceed is to get them to zero.
>
>
> On Thu, Jan 17, 2013 at 10:41 AM, Sergey Shelukhin
> <[EMAIL PROTECTED]>wrote:
>
>> One issue with the current situation is that it's pretty hard to
>> understand what warnings are new, and in some JIRAs the report seems
>> to be wrong (e.g. if you download the reports from old and new build
>> you see the difference is only in dates and timing).
>> I filed a JIRA to make it better (so you could rely on it and see what
>> to fix) but I won't be able to get around to it at least until next
>> week :(
>>
>> On Wed, Jan 16, 2013 at 12:16 PM, Stack <[EMAIL PROTECTED]> wrote:
>> > On Wed, Jan 16, 2013 at 10:57 AM, Nicolas Liochon <[EMAIL PROTECTED]>
>> wrote:
>> >
>> >> I think the only simple solution is to put it to zero, then it's easy
>> >> to
>> >> see what's new. But it requires some work. Not much. But some :-(
>> >>
>> >>
>> > Figured.  You are right N.  Thanks.  Lets squash this last set...
>> > St.Ack
>>
>

I agree we should get findbugs down to zero.

In the interim, what I did for HBASE-7329 was to diff a known good
newPatchFindbugsWarningshbase-server.xml with the one generated by Hadoop
QA and look for warnings on the classes touched by the patch.

Cheers

On Thu, Jan 17, 2013 at 5:16 PM, Enis Söztutar <[EMAIL PROTECTED]> wrote:

> Yes, I was also surprised some time ago, that our findbugs reports does not
> show the diff, but the whole, and you need dark skills to find those the
> patch has caused.
>
> The only reasonable way to proceed is to get them to zero.
>
>
> On Thu, Jan 17, 2013 at 10:41 AM, Sergey Shelukhin
> <[EMAIL PROTECTED]>wrote:
>
> > One issue with the current situation is that it's pretty hard to
> > understand what warnings are new, and in some JIRAs the report seems
> > to be wrong (e.g. if you download the reports from old and new build
> > you see the difference is only in dates and timing).
> > I filed a JIRA to make it better (so you could rely on it and see what
> > to fix) but I won't be able to get around to it at least until next
> > week :(
> >
> > On Wed, Jan 16, 2013 at 12:16 PM, Stack <[EMAIL PROTECTED]> wrote:
> > > On Wed, Jan 16, 2013 at 10:57 AM, Nicolas Liochon <[EMAIL PROTECTED]>
> > wrote:
> > >
> > >> I think the only simple solution is to put it to zero, then it's easy
> to
> > >> see what's new. But it requires some work. Not much. But some :-(
> > >>
> > >>
> > > Figured.  You are right N.  Thanks.  Lets squash this last set...
> > > St.Ack
> >
>

On Thu, Jan 17, 2013 at 5:19 PM, Jean-Marc Spaggiari <
[EMAIL PROTECTED]> wrote:

> I have started to look at the warnings one by one. Most of them are
> easy fixes. Submitted one patch, more to come.
>
Thanks JM.  I fixed two a few days ago (Hurray for me!)

St.Ack

Hi,

I wuld like to configure my local findbugs the same way the build
machine is. Is there a place I can copy the configuration from and
apply it to the eclipse plugin?

Thanks,

JM

Are you using http://findbugs.sourceforge.net/manual/eclipse.html ?

We have dev-support/findbugs-exclude.xml in trunk.

Cheers

On Fri, Jan 18, 2013 at 12:40 PM, Jean-Marc Spaggiari <
[EMAIL PROTECTED]> wrote:

> Hi,
>
> I wuld like to configure my local findbugs the same way the build
> machine is. Is there a place I can copy the configuration from and
> apply it to the eclipse plugin?
>
> Thanks,
>
> JM
>

Hi Ted,

Yes, I installed the plugin from FindBugs -
http://findbugs.cs.umd.edu/eclipse. I got the URL on the link you
sent.

I will take a look at the dev-support/findbugs-exclude.xml and see
what I can get from it.

Thanks,

JM

2013/1/18, Ted Yu <[EMAIL PROTECTED]>:
> Are you using http://findbugs.sourceforge.net/manual/eclipse.html ?
>
> We have dev-support/findbugs-exclude.xml in trunk.
>
> Cheers
>
> On Fri, Jan 18, 2013 at 12:40 PM, Jean-Marc Spaggiari <
> [EMAIL PROTECTED]> wrote:
>
>> Hi,
>>
>> I wuld like to configure my local findbugs the same way the build
>> machine is. Is there a place I can copy the configuration from and
>> apply it to the eclipse plugin?
>>
>> Thanks,
>>
>> JM
>>
>