lynx-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: LYNX-DEV extension languages for Lynx


From: Andrew Kuchling
Subject: Re: LYNX-DEV extension languages for Lynx
Date: Mon, 5 Jan 1998 11:27:41 -0500 (EST)

address@hidden (Larry W. Virden, x2487) wrote:
>There have been requests over the past years for a scripting language.
>The primary one, as folk can imagine, is JavaScript.  Tcl has also been
>requested several times, and I can imagine the Perl and Python camps
>also having some interest.

        In practice, JavaScript is the only one that matters; no one
is going to write Web pages using Tcl or Perl or Python just for Lynx.
That capability would be an amusing curiosity, nothing more.  Now, one
could implement JavaScript on top of those other languages; I know
someone who has a pet project, sadly now in limbo, to convert
JavaScript to Python bytecodes.

        Therefore, Tim Pierce's Guile work is interesting, since a
JavaScript parser & execution engine could be implemented in Guile.
For those unfamiliar with it, Guile is GNU's Ubiquitous Intelligent
Language for Extension, an implementation of the Scheme language.  See
http://swissnet.ai.mit.edu/~jaffer/GUILE.html .
        
>Is there any way that the varying interests could all be served by first
>a redesign of the innards of lynx, and THEN folk splintering off to implement
>their favorite feature?

        Yes.  Lynx was originally written long before there were such
things, and there comes a point where adding them to the code simply
becomes too ugly.  People also won't work on freeware projects that
aren't fun; if making code changes is extremely painful, there's a
strong urge to go and work on something more interesting.

        However, there are still some hard questions to answer.  

        * If there aren't enough people to implement simpler things,
who's going to implement this massive restructuring?  For example,
who's going to implement Javascript in Guile?  (Or are there Guile
developers already working on it?)

        * The constraint of DOS compatibility is a problem; as Lynx
gets bigger and bigger, it gets harder to make things work on DOS.
Should DOS users be abandoned, and told "Well, you've got the DOS port
of Lynx 2.7, and that's it for you."?  Then development could focus on
32-bit platforms: Win95, MacOS, Unix, and VMS if anyone has access to
it.

        For example, have Guile or Tcl even been ported to DOS?  I
know Perl and Python have, but those ports aren't considered very
important by the user communities.


        Andrew Kuchling
        address@hidden
        http://starship.skyport.net/crew/amk/

reply via email to

[Prev in Thread] Current Thread [Next in Thread]