Re: widget size independant to screen resolution

Well, the reason I brought up scaling windows in the first place was
when I mentioned that all the widgets in Y could be rendered using
vector graphics.
Looking at this browser window here, the whole thing could be rendered
with vector graphics, including most of the webpage (except for the
images, but these would only need to be scaled once per page update).
So, for an application that is designed for Y, which leaves ALL
rendering to the server (this is how it should be, right?), then
scaling would be nearly no extra cost at all, for any properly
designed Y program.
Even programs that use bitmaps should fit into this category, since
all you'd have to do is put it into a bitmap widget, which should be
arbitrarily scalable by either the client or the server (I dunno if
"should" is the right term here, but that's the basic idea).

If this was done correctly (doesn't seem like it would be too hard to
do, really), it would be possible to scale any Y window arbitrarily,
and it would be just as fast to render as any regular window (unless a
window had a constantly updating bitmap widget, I suppose), and I mean
without any sort of hardware acceleration, too.

On Wed, 20 Oct 2004 11:28:19 +0100, Phil <[email protected]> wrote:
> Paul Bonser wrote:
> > Am I wrong in these assumptions? Scaling a bitmap would be slow, yes,
> > but scaling fonts shouldn't be since they only need be scaled once, or
> > will be rendered with less detail, which would be faster...
> 
> If you only want it for terminals, that's fine.  However you asked for
> it to be possible to shrink windows, and the majority of windows have
> more than just text in them.
> 
> If you want to make a shrinking terminal, go for it; however, shrinking
> of arbitrary windows is not going to happen.
> 
> Phil
> 


-- 
"C++ also supports the notion of *friends*: cooperative classes that
are permitted to see each other's private parts." - Grady Booch