lynx-dev
[Top][All Lists]
Advanced

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

Re: lynx-dev Re: Another TagSoup problem


From: Henry Nelson
Subject: Re: lynx-dev Re: Another TagSoup problem
Date: Wed, 4 Feb 2004 10:12:13 +0900
User-agent: Mutt/1.5.4i-ja.1

On Tue, Feb 03, 2004 at 11:35:48PM +0100, Arclele wrote:
> Il 03/02/04 14.44.35, Henry Nelson  ha scritto:

[If possible please set your mailer so that it will not insert the
 quoted author's email address.  Particularly, the lynxdev list is
 being hit hard by address harvesters.]

> > run the page in question through a validator.
> >
> I suggest to add a new hot-key (or what you prefer) to 
> w3c-validator site (like Opera browser) to test the code, 

The problem with "hot-key"s, as convenient as they are, is that there
are a limited number of them.  Just press "K" to see.  Since Tom (at least
I think he did it) added ":" for entering a command sequence, conceivably
it could be done, but _someone_ has to write the code.

Experienced users of Lynx know that a choice of validator link_s_ is
available from Lynx's default H)elp page.  I have a customized 21-line
version of that page which fits on one screen with numbers available
for navigation.  Thus for me 4 validators are only a couple of key
strokes away, namely "H" (or "?") => "25" (or "26" "27" "28") => [Enter].
Anyone can do this quite easily.  Up until about 3 years ago before public
validators could handle all languages (charsets), I even maintained a
lynxcgi link to a local validator that could do checks on Japanese pages.
That method would still be useful to people with very slow or costly
Internet connections.

There is almost nothing you cannot do with Lynx if you are running it on
a Unix system (maybe Windows or OS/2 for all I know).  It only takes
imagination and a little study.  If you are interested, look into the
lynxcgi and lynxexec functions.  The pseudoproxy, cern rules, and external
mechanisms are all very powerful, too.

I recommend anyone who uses Lynx on a daily basis that they take the
time to read the documentation (including the huge distribution lynx.cfg).
Lynx has so many features that it's almost impossible for any one person
to know and use them all, I know, but if you don't at least make an effort
to use them, there isn't much incentive or satisfaction for developers to
debug old features or add new useful features.  If all a person wants from
the Internet is "pretty", then quite frankly I cannot see how in the world
that person can justify using Lynx other than as a novelty.  Lynx flatly
cannot do javascript at all, and still cannot render some very simple
tables intelligibly (snippet example of inability to line-up cells appended).

BTW, as I discussed recently, I plan on submitting, for the next development
round, an I)ndex page so that the search engine links now on the H)elp page
would be available with one hot-key, "I" (which in my opinion has been
wasted for many years).  Perhaps it would be useful to put the validator
links on there, too.  In other words, the "I" keystroke would present you
with a menu having two sections: "Search Engines" and "Validators".  These
sections would then be removed from the H)elp page.  Please express your
opinion one way or the other.

__Henry

Rather common example of a type of table that Lynx has serious problems
in rendering intelligibly.  Particularly the last two fields where there
is no hint from the data (Yes/No or O/X) what the field is, it is imperative
that the browser line the cells up vertically.

<html><body>
<table>
<tr>
<td>Model Number</td>
<td>Number of user accessible cylinders</td>
<td>Number of data heads</td>
<td>Has threading capability</td>
<td>Firmware update available</td>
</tr>
<tr>
<td>MK216</td>
<td>2135</td>
<td>6</td>
<td>O</td>
<td>X</td>
</tr>
</table>
</body></html>


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

reply via email to

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