LYNX-DEV Re: HREF pedantry

[Klaus Weide]

On Sun, 14 Sep 1997, Al Gilman wrote:

> Klaus>
> There really is no excuse why *an automatic tool* should *generate*
> IDs with embedded spaces.
> 
> Al> 
> That's going too far.  There are good reasons why automatic tools
> should not be _generating_ the NAMEs and IDs used in HTML.

The tool are "generating" the strings in the sense that they are
generating an output file, with (supposedly) HTML, and the NAME
attribute values as part of it.  If they are just taking any string of
bytes the user may have typed in and put it in the attribute and HREF,
without making sure it is valid, then they are not doing their job.

> Consider the fact that the Lynx message screens do not contain
> hotlinks into the User's Guide simply because the User's Guide
> made the unfortunate choice of using serial integers as opposed
> to mnemonic strings for anchor names.

Is that so?  Would Lynx message screens contain hotlinks int the
User's Guide if the User's Guide was using different anchor NAMEs?
Send the diffs, Al.

> Allowing whitespace in names is part and parcel of a _positive_
> trend.  On the other hand, it is appropriate for URLs to be a 
> bit lagging in this regard.
> 
> Just like our beloved Lynx.

I cannot characterize anything that breaks (deliberately or out of
ignorance, but _without need_) older applications which follow the specs
as "a positive trend".  It's the trend of "We think it's a good idea, and
it works with [some version of] Netscape and Microsoft, to hell with the
rest".

Using "What_is_new" instead of "What is new" wouldn't have broken 
anything, AFAIK.

Anyway, Lynx tries to cope; if there are specific problems with the
current code (recent devel or fotemods, should be the same) then let's
talk about those.

  Klaus

;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;