Re: lynx-dev Java script support.

[brian j pardy]

On Tue, Jun 29, 1999, Heather Stern wrote:
> brian j pardy wrote:
> > On Mon, Jun 28, 1999, Heather Stern wrote:
> >> [...]
> >> I think these features are -
> >>    * mouseovers 
> >>      - some sites don't bother with ALT, but they do have
> >>        a good mouseover status line
> > 
> > How about, in lieu of Javascript (not saying never Javascript, but in
> > the indefinite time until we have it), something to just fake an ALT
> > tag for a mouseover, if there is no ALT tag?
> > 
> > (If this is what you were saying, then I agree it'd be a good idea.)
> > 
> > Shouldn't be too difficult to do, and just as easy to rip out upon
> > getting real javascript.
> 
> Yes, this would be a good thing.
> 
> I was sort of vaguely thinking, if we had the "lynx can do a limited
> number or type of pseudo-graphical things" stub, and something to spot 
> that the HTML/js author is trying to do such a thing, we could fake it.  

I don't know anything about js -- is the syntax for a mouseover pretty
much just something in the <IMG> tag somewhere?  If it's something
like that, can try to find the 'mouseover=' text and splice out what's
in quotes after it (dealing with backslashed quotes, etc).

> To be "real" mouseover (?!) for the scenario I described, lemme see, I'd
> imagine:  The image doens't display (of course) and there's no ALT so 
> assuming you have the feature on, we show the graphic's filename.  When
> it become hot, we'd honor the mouseover and redraw the page?  I don't
> think so.  More likely, "fake javascript" would steal yet another screen
> line for its status,

As an easier (and less demanding on screen realestate, and in keeping
with how annoying mouseover truly IS (heh)), how about if a mouseover
just did a temporary message over the existing status line, using the
facilities we currently have?  Could do something like have it write
with the same precedence as informational messages (JSSECS=0 ? Ouch.)
or something. 

Didn't someone recently add a history mechanism for statusline
messages?  If the fake-js is displayed on screen for a configurable
amount of time, and recallable, I think that'd be a nice way to do it
without eating up another line of screen.

> and might be initially stubbed to show what it is
> guessing rather than try to actually do anything new with js code.  That
> way at least we avoid what really annoys many Big Two users, that it's
> great having the status bar say something, but the actual status, which
> you might have liked to know, was overwritten in the process.

I know it annoys me whenever I'm unlucky enough to use one of those
Other Browsers.  

> Hope that helps - I'm not much of a C hack these days...

Ideas always help.

-- 
"Taxes should hurt.  I just mailed my own tax return last night and I
am prepared to say `ouch!' as loud as anyone."
                -- Ronald Reagan