Re: [Codenarc-user] new GMetrics rule about software stability?

[Hamlet D. <ham...@ca...>]

> Are you thinking that a heuristic would be based on relative
> comparison? Perhaps # commits versus average or beyond one standard
> deviation -- 

Does Sonar do anything like this? Perhaps we are reinventing the wheel. 
To start, it would be good to list the absolute numbers sorted highest to lowest. 
Then you might want to list the # of commits per day per file sorted highest to lowest. 
After that maybe start looking at standard deviation stuff. 

> BTW, My intent was to switch back to working on GMetrics for a while
> and give that some attention. 

It's your decision. I am excited about the momentum we have with CodeNarc, especially to see that Cedric created an IntelliJ IDEA plugin. In my opinion, adoption will come once we have a ton of rules. There is plenty of low hanging fruit left to pick (ie. easy rules to implement). I'm going to stick with more CodeNarc rules for now. 

> Sigh. Never enough hours in the day...

There are plenty of hours, we just have day jobs :( 

--
Hamlet D'Arcy
ham...@ca...

----- Original Message -----
> That is quite interesting. I agree that could be a valuable metric.
> And GMetrics does seem like a reasonable place to put it.
> 
> Are you thinking that a heuristic would be based on relative
> comparison? Perhaps # commits versus average or beyond one standard
> deviation -- figure out a good way to determine "outliers" on the high
> side.
> 
> BTW, My intent was to switch back to working on GMetrics for a while
> and give that some attention. There is a long to-do list for that
> project as well. Sigh. Never enough hours in the day...
> 
> Chris
> 
> Hamlet DArcy <ham...@ca...> wrote on 01/25/2011 03:20:02 AM:
> 
> > Hi all,
> >
> > I ran across an interesting blog post this morning from Michael
> > Feathers:
> > http://michaelfeathers.typepad.
> > com/michael_feathers_blog/2011/01/measuring-the-closure-of-code.html
> >
> > As a measure of risk, he graphed the number of version control
> > commits against each file in the codebase. A file at revision 3 is
> > less risky than a file at revision 300. The file at revision 300 is
> > a code smell because all changes filter through that file.
> >
> > I have long wanted to see metrics on this version control revision
> > number. A high number shows unstable code that, at a minimum, needs
> > extensive test coverage, and possibly refactoring.
> >
> > Do you think GMetrics is the place for this code? Could we produce
> > some sort of report showing the risk areas of a project based on
> > revision number? And could you define heuristics for a CodeNarc rule
> > based on this info?
> >
> > Or perhaps a different open source project is a better fit for this
> > functionality.
> >
> > --
> > Hamlet D'Arcy
> > ham...@ca...
> >
> >
> > ------------------------------------------------------------------------------
> > Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
> > Finally, a world-class log management solution at an even better
> > price-free!
> > Download using promo code Free_Logger_4_Dev2Dev. Offer expires
> > February 28th, so secure your free ArcSight Logger TODAY!
> > http://p.sf.net/sfu/arcsight-sfd2d
> > _______________________________________________
> > Codenarc-user mailing list
> > Cod...@li...
> > https://lists.sourceforge.net/lists/listinfo/codenarc-user
> 
> > ForwardSourceID:NT000F9C52