May I suggest finishing 0.7 and then doing a 0.7.1 that is only style/pmd/findbug/etc. changes?
On Mon, Jun 11, 2012 at 6:06 PM, Drew Farris <[email protected]> wrote: > FWIW, It's my understanding that there is a fair amount of > configuration that can be done with both pmd and findbugs that will > let us tweak it to suit our needs. > > Of course someone has to >do< that configuration, but nevertheless it > can be done. > > I think it would be a great first step to get pmd/findbugs/checkstyle > integrated with our maven builds. This would at least give us some > level enforcement pre-checkin and via jenkins. > > From my point of view, Eclipse integration is a nice to have, with the > eventual goal of having the command-line tools, eclipse and intellij > agreeing on what's broken and the right way to fix it. I'm strongly > against >only< enforcing these things via an IDE, but I don't think > anyone's suggesting this at this point. > > Full disclosure: as an Eclipse user, I've lived for years with > half-baked maven eclipse integration, indigo being a worse offender > than other. I sort of expect to bounce between the terminal and my ide > for both builds and checkins to make sure everything is truly as it is > when presented to me via the IDE. > > Drew > > On Mon, Jun 11, 2012 at 7:47 PM, Dmitriy Lyubimov <[email protected]> wrote: >> "In practice, the rate of false warnings reported by FindBugs is >> generally less than 50%" >> >> which probably disqualifies it as a hard build stopper if 30-ish% is noise.. >> >> On Mon, Jun 11, 2012 at 4:40 PM, Robin Anil <[email protected]> wrote: >>> pmd does have static analysis >>> http://maven.apache.org/plugins/maven-pmd-plugin/ >>> use findbugs as well http://mojo.codehaus.org/findbugs-maven-plugin/ >>> ------ >>> Robin Anil >>> >>> >>> On Mon, Jun 11, 2012 at 6:36 PM, Dmitriy Lyubimov <[email protected]> wrote: >>> >>>> Robin, >>>> what maven plugin does that? PMD? >>>> >>>> On Mon, Jun 11, 2012 at 3:52 PM, Robin Anil <[email protected]> wrote: >>>> > Can we just focus on the bugs first. Its more useful to get static >>>> analysis >>>> > in and make that a build blocker. Checkstyle is secondary and is a >>>> > sensitive topic. >>>> > >>>> > Robin >>>> > >>>> > On Mon, Jun 11, 2012 at 5:49 PM, Jeff Eastman < >>>> [email protected]>wrote: >>>> > >>>> >> +1 I feel the same way as Dmitriy >>>> >> >>>> >> >>>> >> >>>> >> On 6/11/12 6:39 PM, Dmitriy Lyubimov wrote: >>>> >> >>>> >>> which is why i am saying it would be easier to decouple code tools >>>> >>> from environment. That is, put them into the build process . >>>> >>> >>>> >>> if we could relax some of the checks on top of it to degree they >>>> >>> actually agree with sun style autoformatter in eclipse, that would be >>>> >>> super-great (those are mostly things like spaces in comments and some >>>> >>> wrapping choices which are for some reason construed differently by >>>> >>> checkstyle and eclipse styles). I don't think there are many, if, >>>> >>> perhaps, any at any given checkstyle settings. >>>> >>> >>>> >>> On Mon, Jun 11, 2012 at 3:21 PM, Ted Dunning<[email protected]> >>>> >>> wrote: >>>> >>> >>>> >>>> I don't want to ignite a flame war since (a) and (b) are sufficient >>>> >>>> grounds >>>> >>>> to stay with whatever you like to use, but >>>> >>>> >>>> >>>> c) I am not convinced of this >>>> >>>> >>>> >>>> d) IntelliJ has a free edition that works great. You can also get the >>>> >>>> full >>>> >>>> version for free for use on open source projects. >>>> >>>> >>>> >>>> Like I said, this doesn't change your deep motor training. I would >>>> point >>>> >>>> out that I have been through more generations of editing environments >>>> >>>> than >>>> >>>> I can count over the last 35 years and while it has always been >>>> >>>> unpleasant >>>> >>>> to change, changing to the next big thing has generally been >>>> worthwhile. >>>> >>>> >>>> >>>> On Mon, Jun 11, 2012 at 2:14 PM, Dmitriy Lyubimov<[email protected]> >>>> >>>> wrote: >>>> >>>> >>>> >>>> And no, i am not ready to part with eclipse yet. a) years of reflexes >>>> >>>>> hard to beat. b) it is the only environment that has a lot of other >>>> >>>>> integrations that I use such as StatET (integrated R step-by-step >>>> >>>>> debugger& help). c) faster native visuals. d) free. >>>> >>>>> >>>> >>>>> >>>> >>>>> On Mon, Jun 11, 2012 at 2:09 PM, Dmitriy Lyubimov<[email protected]> >>>> >>>>> wrote: >>>> >>>>> >>>> >>>>>> Yeah eclipse is a problem. Or checkstyle configuration. or both. >>>> >>>>>> >>>> >>>>>> I used to set up the checkstyle plugin that highlights the problems >>>> >>>>>> but i never managed to get it to 0 since it contradicts sun >>>> >>>>>> recommended formatting as set in eclipse styles (or even with >>>> >>>>>> adjustments). So both tools aimed at style actually contradict each >>>> >>>>>> other on what the right style is. Which tells me both are probably >>>> >>>>>> overzealous in their assertion since apparently those assertions >>>> are a >>>> >>>>>> subject of contention among authors of those tools. >>>> >>>>>> >>>> >>>>>> Eclipse concerns can be partially set aside if the maven build fails >>>> >>>>>> and reports why it fails so I don't have to check in and wait for >>>> >>>>>> Jenkins to shove it into my face. >>>> >>>>>> >>>> >>>>>> On Mon, Jun 11, 2012 at 7:31 AM, Jeff Eastman >>>> >>>>>> <[email protected]> wrote: >>>> >>>>>> >>>> >>>>>>> On 6/11/12 7:39 AM, Benson Margulies wrote: >>>> >>>>>>> >>>> >>>>>>>> It may take me a few days. >>>> >>>>>>>> >>>> >>>>>>>> How much do people care about Eclipse integration? Can I ignore >>>> that >>>> >>>>>>>> for the moment? >>>> >>>>>>>> >>>> >>>>>>>> >>>> >>>>>>>> -1 No, I think Eclipse integration is important but I will devote >>>> >>>>>>> some >>>> >>>>>>> energy to helping here if I can >>>> >>>>>>> >>>> >>>>>> >>>> >>> >>>> >> >>>> -- Lance Norskog [email protected]
