Re: lynx-dev Javascript

[David Woolley]

> 
> <#include "not-subscribed-to-this-list-please-CC-replies.h">

This is a hassle the way the list is set up, and your article is
likely to generate a long thread; almost certainly someone will 
forget to copy your address back.

> GPL. It is already available, but NPL-only; as soon as the GPL
> is effective, it becomes an useful addition to lynx.

Having the Javascript interpreter code is only half the story.  First it
has to be matched to way that Lynx calls extensions.  Then the link
object model has to be modified to match a superset of the IE4 and
NS object models.   Note that IE and NS implementations of the object
model in version 4 browsers are frustratingly inconsistent and require
extensive testing or user agent checks; the latter would probably just
reject Lynx outright.  My guess is that you are talking about upwards
of a 20% re-write of Lynx.

From what I gather, the person who has been doing most of the Lynx
coding recently has other commitments at the moment, so you are
going to have find volunteers to do the required major overhaul.

> 
> Is anyone already thinking of this?
> 
> The sources can be found at your favourite mozilla mirror at
> .../pub/js/ (try ftp://ftp.mozilla.org/pub/js/ for example). It
> has the library, a "javascript shell" (???) and what seems to be
> a perl module.
> 
> Some guys with enought guts to call themselves webmasters are
> acquiring the horrendous habit of writing forms without submit
> buttons - they add some other button (or worse, an anchor) that
> triggers a script that does something (usually verifying the
> data in the form, something that everyone knows should be done
> by the server) before submitting. Among a lot of

Actually, I think this is the most legitimate use of Javascript, although
the way I've done it is to intercept a normal submit with an onClick
event handler, so that the unvalidated data goes through in default.
Most of the other programmers where I work, currently writing, intranet
applications for sale, would probably have done it the way that causes
you problems, because the marketing brief is:  work well with IE4 and 5,
work with NS4, in both cases assume that Javascript has not been blocked
(anyone using JS on IE4 on the internet had better be subscribing to all
the security bug mailing lists!) and spend no company time on supporting
anything else.

Validation is legitimate, because you can validate field by field, easily
highlight the bad field, avoid expensive multiple transmissions of 
uncacheable material, etc.  Client side validation is a very old concept
in IT and is generally in all client server applications.

In the case of an internet application, the server needs to revalidate,
but needn't give very useful error messages.  Intranet applications
may well assume that people will not try and force unvalidated material
through.

> javascript-related usability issues, this last one is the most
> annoying, and unfortunately seems to be a growing trend. If
> Mozilla.org's GPLed libjavascript is sane enought, I might even
> put aside some time to help lynx use it.

Doing it right will be a major job and a continuing one, as the sort
of person who discounts useability when writing web pages, tends to be
the sort of person striving to always use the latest gimmick to keep
ahead of the competition.

If you don't need Lynx because of a disability, my advice would be 
to use IE4 (it is at least a year ahead of NS4) when accessing commercial
sites; most of them are designed to be almost unuseable without graphics.
If you do need it for a disability, then you either have to rely on their
being teeth in the relevant legislation (this sort of legislation tends
to be window dressing in the UK with no budget for the required
inspectors, etc.), or you have to apply market forces in a way which comes
to the attention of the senior management.  Lynx users are not a very
good market for this as they tend to be people not prepared to pay for,
at least software, or people who do not have a lot of disposable income.
Generally, any site with a form on it is there to make its owner money.

Most "web" applications are not really HTML, they are really IE4 or NS4
applications.

Note there are a lot of other ways that JS is used to make pages
inaccessible, and these are often combined with graphics.  I would
estimate that these are much more common than validation, although it
is possible that what you are seeing is now is a change in the way that
validation is done, so that you become aware of it.