nanogui: GTK+ making progress towards 2.0
Subject:
Re: GTK+ making progress towards 2.0
From:
Jordan Crouse ####@####.####
Date:
5 Apr 2001 16:56:27 -0000
Message-Id: <3ACCA3B9.FA9CDEA9@censoft.com>
Porting GTK+ to Nano-X has few advantages. The size and scope of the
GTK+ project immediately discounts it from consideration in many of the
embedded projects I work on every day. Whats more, I think that it is
just too powerful - Its like putting a V6 into a lawn mower. Embedded
applications need to be simple, quick and effective. They often run in
4 and 8 bit spaces on relatively slow processors. The very nature of
touchscreens makes them more adaptable to 2D widgets than 3D widgets.
GTK+ is written to take full advantage of the desktop space with 16
million colors and a 800 MHZ processor. Shrinking a desktop toolkit is
the same folly that Microsoft has encountered with WinCE. You just
can't take your desktop and shrink it to 200x320 and call it good. The
embedded space requires its own specific design in order to be
successful.
Now, having said that, we have a whole slew of clients who have GTK+
applications that they want to embed for whatever reason. But in this
case, it is the API that they want, not the particular engine that GTK
provides. So, if we take a very good widget set that was developed for
the embedded space (TinyWidget) and combine it with the GTK+ API, then
we can get the engine and performance that we want with the API that the
the customers crave.
Whats more, its a consideration of effort. We have 30, maybe 40 people
who can jump right in and help with such a porting project. Of those,
only a few have enough free cycles to dedicate toward porting. That's
alot of work for a few people to handle, especially if everyone is going
to turn it down because of size.
Like I said, I would rather that if people have some free time, that
they dedicate it toward the two existing native toolkits. FLNX is good
in its current state, but a port to the latest version is badly overdue,
and TinyWidgets is in the starting stages, and I am sure that Amit would
love the extra help (configure script is on its way, Amit... I
promise!!!). First we need a good engine that we can depend on. Then,
we can make API conversion layers at the drop of a hat for whatever
toolkit we want.
As far as the Microwindows community is concerned, our biggest obstacle
is the development environment. When we have two very good toolkits
with good GUI development environments, then people really won't care
about the API, as long as they can easily make a good looking app
without a PhD.
Jordan
Alex Holder wrote:
>
> I suspect that would probably be a much bigger job than simply updating
> the GTK+ port to Nano-X actually. I see the two toolkits as serving
> different purposes; the GTK+ port would cater for people who have existing
> GTK+ applications that they want to bring to Nano-X and don't mind about
> the large size of GTK+, whilst TinyWidgets will provide the basis for the
> native "PDA/embedded system/IA environment" that I'm planning to develop.
>
> Ideally I think we should be aiming to eventually have Nano-X ports of
> all the most popular popular X toolkits (GTK+, QT, TK, the Java AWT,
> perhaps Wine, maybe even Motif) so as to provide people with an easy
> migration path from X11. At the same time we'll be developing our own
> integrated environment and application suite designed specifically for
> small systems (PDAs, internet appliances, embedded systems, etc.) which
> people can use when writing natively for the system.
>
> --
> ------- Alex Holden -------
> http://www.linuxhacker.org/
> http://www.robogeeks.org/
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ####@####.####
> For additional commands, e-mail: ####@####.####
