Re: URI not accepted

On Wed, Sep 01, 1999 at 08:13:48AM +0200, Terje Bless wrote:
> On 31.08.99 at 19:00, Gerald Oskoboiny <gerald@w3.org> wrote:
> >On Wed, Aug 25, 1999 at 09:12:39AM +0000, Terje Bless wrote:
> >> The error handling of the validator isnīt really very impressive.
> >I fixed that just now (now it displays the error code and text),
> That's much better.
> >I think this functionality must have been lost with the upgrade
> >to LWP from the old perl4 libwww-perl library.
> 
> Yeah, along with the redirect reporting which currently is a NOOP (in the
> local copy of 1.27 I have have).

Yeah, I noticed that and removed it in 1.28; see

    http://dev.w3.org/cgi-bin/cvsweb/validator/httpd/cgi-bin/check.diff?r1=1.27&r2=1.28

for the diffs.

> That's what you get for writing code that's not -w and "use
> strict" clean! :-)

Ah, the joys of exposing one's source to the world... :)

Anyway, I welcome any clues/patches to make the service better.

> BTW, why are you jumping through such hoops to proxy the
> authentication request?

Because it's so cool!

> Why not just offer a username and password field in the main
> HTML form and authenticate using that? Much easier, and much
> cleaner; and it doesn't require recompiling Apache and opening
> up security holes.

Seriously, doing it this way allows the browser to cache the auth
info so once I've provided it once I can validate other pages
within the same realm without having to type the auth info again.
(I tend to have my browser running for weeks or months at a time.)

> Oh, yeah, and $nice_html40_doctype and $html40_doctype are
> missing a double quote right after the FPI.
>     i.e. ''HTML 4.0//EN\n '' should be ''HTML 4.0//EN"\n ''

Actually, I just realized those variables aren't even used any
more, so I'll just remove them.

-- 
Gerald Oskoboiny       <gerald@w3.org>  +1 617 253 2920
System Administrator   http://www.w3.org/People/Gerald/
World Wide Web Consortium (W3C)      http://www.w3.org/

Received on Sunday, 5 September 1999 23:00:58 UTC