I'm not too clear on how it works, but there is a lesser-known
server-side JavaScript language [or dialect, if you want to be
picky about it]. Since I have to assume the engine is built for
UNIX-like HTTP servers, why not install this on the system and
write a small, separate application which implements it. The
program could in turn act as a filter [downloader, HTTP proxy
or something like that], parsing scripts in web pages and passing
a kind of "pre-chewed" HTML to Lynx?
This could be a little like the system-wide gettext issue earlier:
instead of each application wasting space using a separate and
disparate version of the gettext package internally, you could
have them all running off of one installed in the system. ie:
DON'T redesign Lynx from the ground up, but instead create an
assistant for it.
Maybe that's the wrong comparison. More like the HTML-izing script
which creates a document based on lynx.cfg?
The only serious risk I can think of is that CLIENT-side JavaScript
is intentionally crippled, preventing rude web authors from doing
nasty things to your browser, files, or system merely by letting
you go to their web pages. Server-side JavaScript could have
a bit more leeway, so I leave the hazards of using it with a client
to your imagination.
Actually, I'm leaving a lot to your imagination here: sorry, this
idea is only the vaguest kind of guess, but it seemed worth mentioning.
I'll go looking for more info, posting URLs here if I find anything
that seems at all useful. And to whoever's reading this; I invite
you to do the same.
Patrick
<mailto:[EMAIL PROTECTED]>
; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to [EMAIL PROTECTED]
