lynx-dev
[Top][All Lists]
Advanced

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

Re: lynx-dev javascript? was Re: [...] Lynx will be soon an old ...


From: David Woolley
Subject: Re: lynx-dev javascript? was Re: [...] Lynx will be soon an old ...
Date: Mon, 29 Apr 2002 09:24:25 +0100 (BST)

>       When I go to the Domino-produced site, the first link is
> absolute and lynx generates a GET request.  When I click on any
> one of the next layer of links that appears, no traffic is
> generated.

The first four links on the Projects by Project Number page have 
A HREF="", which means they are links to the same page.  Browsers 
should do nothing on following such links, in particular they shouldn't
reload the page.  They only make sense in a dynamic HTML context, although
there is absolutely no reason why the Javascript processing could not
have been done server side.

Those links with non-trivial HREFs appear to be interpreted correctly.
They are relative links, but the mildest form of relative link (site
relative, rather than directory relative).  Being relative is completely
irrelevant as far as I can see.

The HREF's for the tree expansion (ALT text = "- ") look fairly severely
broken.  & is not allowed in HTML documents in this context (must be &)
and I'd be surprised if & and possibly even ! were permitted in URLs in this
context.   I am pretty sure that fragments can't start with a digit.
Nonetheless, it looks to me like Lynx may have interpreted this as intended.

Back to the first four non-links.  This seems a fairly unusual use of
Javascript.  At first it looked like a simple popup, but it's not.
I'd wonder whether they are generally pushing the bounds of what you
can assume about even fully JS implemneting browsers.

What they appear to do is to look and see if the element that invoked 
the function already contains an HREF attribute.  If it does, they compose
a URL from a base URL in the funcition and a parameter, which appears to
form a parameter value in the URL (certainly not a relative component).
They then update the URL in the element and let the browser follow their
now updated link normally.

If the element doesn't contain a URL, they use window.open to create a new
window, with its frame target name taken from the third parameter and
defaulting to the current one.  They still try to take any normal action
associated with the element, as, for both paths, they let the return status
from the last operation fall through, rather than setting it explicitly.


; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden

reply via email to

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