🚀 go-pugleaf

In article <7331@boring.mcvax.cwi.nl> guido@cwi.nl (Guido van Rossum) writes:
>Long, long ago (a year?) somebody started a net-wide project whose
>purpose was to design and implement an interface to window managers
>which would allow applications to be written that could easily be ported
>to other machines such as the Atari or the Amiga.  I believe I've even
>seen a status report on the project, which mentioned some delay but
>promised results in the end.
>
>[...]
>
>Guido van Rossum, Centre for Mathematics and Computer Science (CWI), Amsterdam
>guido@cwi.nl or mcvax!guido or guido%cwi.nl@seismo.css.gov (from ARPAnet)

Since others may be interested in the current state of this, I'm posting
rather than mailing out an individual reply.

The project, which is/was not net-wide but was announced and discussed
on the net, and whose working name is PL (portability library), is
still around.  The goal of the project is to allow people (like me) to
write programs once and compile them on bunches of different machines.
The idea was to have something that was roughly conceptually similar
to Curses, but for window systems instead of CRTs. 

It's been educational, to say the least.

As you may guess from the fact that it hasn't been released yet, it's not
complete.  There are several reasons for this, to wit:

1.  It's about halfway down on my priority list.  Important, but not as
    important as supporting existing software (which can take a lot of
    time and energy) and doing development on systems that will be used
    directly for instructional purposes (this is a university).  It gets
    worked on when I am inspired or get some free time (free time? what's
    free time?).  Since it's not a product (it will be placed into the
    public domain), it has no funding behind it.  That's how life goes
    sometimes.

2.  Doing it right (as opposed to a set of quick hacks) is one <expletive
    deleted> of a lot harder that it seems at first glance.  This is where
    the "educational" part mentioned above comes in.  The basic problem
    is that each window system presents its own model of display
    manipulation and user interaction.  These models tend not to map well
    to each other in fairly significant ways.  This means that
    factoring out a common model requires a lot of thought.

3.  C may not be the best language to implement this in (I admit it;
    I'm a LISP bigot from way back).  Unfortunately, not everyone has C++,
    so the library ends up doing a lot of low-level shuffling.  X is
    proving instructive, however.

In summary, it's on its way, but it may still be a while yet.  I hope
it will be worth the wait.


-=-


"The greatest warriors are				Clayton Elwell
the ones who fight for peace."			Elwell@Ohio-State.ARPA
		--Holly Near		   ...!cbosgd!osu-eddie!elwell