HTML stuff: validators, metadata and Lynx

[spober at manvax.cc.manhattan.edu]

>From:	MX%"tdowling at ohiolink.edu" 11-JUN-1997 08:00:07.54
>I'm guessing you have an outdated
>version; with Lynx 2.7.1 I was able to search FirstSearch with no problem.

Yes, good suggestion.  It is a version problem, then.  We are running
an older version than 2.7.1.

>>I'm running into problems with longer pages, however.  The
>>longer the text, the more likely I'm going to get them citing me for
>>nonexistant coding errors.  Lines seem to get mysteriously dropped
>>when I look at the output, and I don't know what is causing it.
>>(I can see the text and it looks fine after I've pasted it into
>>my pages.)

>If you're copying and pasting into the WebTechs form, there may be
>copy/paste limitations with your browser.  If you're giving WebTechs a URL,
>however, there should be no problems with them getting the complete source
>of your page.

Yes, but I prefer to validate them before I mount them.  I can't 
mount pages directly onto the server (the computer center doesn't 
allow this) so to mount and then revise, and then ask that the 
revisions be mounted is a bit of a pain.

>Since WebTechs is comparing your page against a given DTD, the formal rules
>of HTML, there's no such thing as a nonexistant coding error.  

Let me see if I can explain this.  

I send up a page using the cut-and-paste submission form.

Let's say line 50 of the page I'm validating is:

<h2 align=center>Important News!</h2>

When I get the results back, the validator will give me an 
error because it "saw" something that was not IN line 50, for
example, a "%" sign.  And in the results from the validator, where
I get the source code back with line numbers, they may print:

50 <h2 align=center>Important News!<%g9>

for line 50, meaning that the validator "saw" code other than what
was submitted.  I have noticed this most often in longer pages.

Another error that is occurring is that sometimes (again mostly in
longer pages) a few lines or portions of one line will be missing
when they give me what the validator saw as the source code, creating
errors because some end tags were the lines that mysteriously 
disappeared.  

Though, yes, I do agree that the validator's response is sometimes
rather cryptic, I'm talking about the validator processing code that 
is slightly different than what is seen on the page.  And I don't know
at what point in the transaction the inserted characters and dropped 
lines occur.  Is it when the pasted text is sent from our server or 
is there actually a bug in the validator that is causing it to do this?

>BTW, is Lynx access really the way you want to go for FirstSearch?  If
>people are already in a terminal environment, why not just use the telnet
>version of FirstSearch?

I haven't used the telnet version in quite a while, but as I recall
it was a little short on instruction.  Okay for librarians and folks
who use it often, but not so friendly for the rest of our users.

Thanks for your response.

//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\
 Stacy Pober			   Internet: spober at manvax.cc.manhattan.edu
 Information Alchemist		   http://www.manhattan.edu/library/mclmenu.html
 Manhattan College Libraries	   Phone: 718-862-7980                    
 Riverdale, NY 10471	     	   Fax: 718-862-7995                    
\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//\\//