harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Richard Liang" <richard.lian...@gmail.com>
Subject Re: [classlib][test]Test Coverage
Date Mon, 05 Feb 2007 03:04:35 GMT
Hello Paulex,

On 2/5/07, Yang Paulex <paulex.yang@gmail.com> wrote:
> Seems this mail didn't reach the mailing list, resend it...please ignore if
> you have got it.
>
> And if no other volunteers, I'd like to update the detail information on my
> apache home.

I'd like to take a try. ;-)  But could you please say more about the
"cannot be instrumented classes"? And I have no idea about the legal
issues.  (Should be OK to run GNU Classpath's tests.)

Best regards,
Richard


>
> 2007/1/30, Yang Paulex <paulex.yang@gmail.com>:
> >
> > I noticed that the test coverage data wiki[1] hasn't been updated since
> > Oct. 2, and the detailed information page[2] has disappeared for some time,
> > any volunteer to recover and maintain this information?
> >
> > Further, I'm thinking about the classes cannot be instrumented by
> > Emma(say, j.u.Properties, j.l.Class, etc), I'd like try to find some way
> > to generate coverage data for them.
> >
> > I think it is time for classlib guys to pay more attention on the code
> > quality/performance, etc, improve the test coverage rate may be a good
> > start. Currently, seems the average method coverage is less than 60%, i.e.,
> > almost 40% methods are not tested at all, I cannot sleep everytime I think
> > about this;-). And as a rough comparison, GNU Classpath has 50K test
> > cases[3], while Harmony only has 20K. (But, possible for us to run GNU
> > Classpath's Mauve test cases on Harmony? any legal and technical issues?)
> >
> > [1] http://wiki.apache.org/harmony/Coverage_information
> > [2] http://viv.byethost15.com/
> > [3] http://www.object-refinery.com/classpath/mauve/report/
> >
> > --
> > Paulex Yang
> > China Software Development Labotary
> > IBM
>
>
>
>
> --
> Paulex Yang
> China Software Development Labotary
> IBM
>
>


-- 
Richard Liang
China Development Lab, IBM

Mime
View raw message