Re: Working on...

["Eric B. Ridge" <ebr@tcdi.com>]

we use CVS also.  we have CVS configured to autocompile our .java files when
we commit them.  works perfectly for us.  you might want to look into that
yourself.  :)

eric

----- Original Message -----
From: Hans Gerwitz <java@phobia.com>
To: <gnujsp@gjt.org>
Sent: Monday, September 27, 1999 4:30 PM
Subject: Re: Working on...


> As a user, I just have to say that this would make us VERY happy. :-)
>
> Off topic, does anyone have any great development setups they'd like to
> share?  We are managing our JSPs through a CVS repository that
automatically
> exports to the web server's directories, managing the Java source of our
> beans through CVS, and manually copying the compiled .class files to the
> server.  Needless to say, that final manual copy annoys me... but no
better
> approach has panned out.
>
> Automatic bean source compilation would enable me to work around this...
> local compilation would be for "validation", while GNUJSP kept the object
> code in sync with the source tree.
> --
> hans@gerwitz.com
>
> ----------
> >From: "Noel J. Bergman" <noel@ibm.net>
> >To: "Carsten Heyl" <Heyl@nads.de>
> >Subject: RE: Working on...
> >Date: Mon, Sep 27, 1999, 12:18 PM
> >
>
> > Carsten,
> >
> > I recommend against introducing JSP extensions as a normative issue, but
if
> > there is a compelling reason, we could consider it and suggest it to
Sun.
> >
> > In this case, I actually think that it is might be a bad idea to add new
> > interfaces to a JSP page.  JSP pages are supposed to have a specific,
> > generic, interface.  The place to have additional interfaces would be in
> > beans used BY the pages, not the pages, themselves.  Beans can span
pages as
> > well as requests.
> >
> > As for the other, I was actually thinking of a source issue, so that if
> > someone has a helper bean for a page, that if they change the SOURCE for
the
> > bean, the bean and the page are recompiled.  And if we do that, why
can't we
> > act as the classloader for the bean?  I don't believe that there is
anything
> > in the specification which precludes this semantic, even if it isn't
> > implemented by other JSP engines.
> >
> >  --- Noel
> >
> >