About Java applets sizes

[Koos Vriezen]

On Sat, 22 Feb 2003, George Staikos wrote:

> On Saturday 22 February 2003 07:25, Koos Vriezen wrote:
> > On Sat, 22 Feb 2003, George Staikos wrote:
> > > On Friday 21 February 2003 18:30, Koos Vriezen wrote:
> > > > most of it :) In my tree this makes kjavaappletviewer work more nicely
> > > > with kjas.
> > >
> > >   Do you have an example site where I can reproduce this behaviour?  I
> > > haven't been able to reproduce these problems recently.
> >
> > Currently the kjavaappletvieuwer is only used for OBJECT based applets. In
> > my tree I use it for APPLETS too and there all sites have sometimes
> > (mostly first time loading) this problem :-).
> >
> > But the point of the patch is that it's wrong to resize applets manually
> > this way.
>
>   Ok well I think you understand this better than most of us.  If you think
> it's correct, then apply it to HEAD and we'll work with it for a little
> while.  If it works well, then we can backport (unless you are absolutely
> confident with it, then you might as well backport it right now too).

Ok thanks. Btw. ackporting is only needed if kjavaappletviewer is
backported too.
I've a local kde-3.1 branch where it's backported and it works fine. I can
backport it as a quick fix for OBJECT based applets. However, in the
unlikely case, pages having both OBJECT as APPLET based applets, two jvm
are started. For that reason I'm trying to get rid of the old
widget based mechanism in favor of the kpart one. This needs changes in
KHTMLPart for the onload event (already in HEAD), in ecma/kjs_html for
applet.toString (already done in my tree), somehow share
HTMLAppletElementImpl::attach which HTMLObjectElementImpl (have something
working in my tree) and don't link libkjava.a and render_applet.o with
khtml (work in my tree also, saves about 150 kb for libkhtml.so).

Anyway, this might be too much to backport :-)

Koos

 >
> --
>
> George Staikos
>
>