commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From robert burrell donkin <>
Subject Re: [logging] demonstration code
Date Mon, 21 Mar 2005 21:11:21 GMT
On Sun, 2005-03-20 at 22:22 -0800, Brian Stansberry wrote:
> --- robert burrell donkin
> <> wrote:
> > inspired by ceki's examples, i've been working on an
> > extended systematic
> > analysis of the various classloader cases. there's
> > demonstration code
> > for them as well as documentation. it's not complete
> > but most of the
> > substantial cases address by ceki's document are
> > covered. (the
> > unconventional context classloader cases highlighted
> > by simon are
> > missing but they need a little more consideration.)
> > i think that
> > developers should find it interesting and maybe a
> > little educational. it
> > certainly help me think a lot more clearly about
> > these issues. 
> Look forward to seeing this.  I was thinking of
> getting going on some unit tests based on the earlier
> analyses, but think I'll wait a bit to see what you've
> done.

i would recommend waiting.

it should be quite easy to add extra analysis cases in the same style to
the code base when it's done. i haven't covered much of the ground you
had started to explore (nor the ground suggested by simon concerning
unconventional classloaders). it would great if these could be added.

it may be a day or three before it's ready. i don't particularly want to
rush. it's easy to make mistakes when analysing these issues and it's
easy to then build misleading arguments upon shaky foundations. i hope
that people will bring any mistakes i make to our attention as quickly
as possible. 

i haven't heard any feedback about: i'll be relying on
the concepts and terminology outlined there in the analysis and
demonstration code. again, i'd like to encourage anyone who finds any
inaccuracies to speak up as soon as possible.

- robert

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message