W3C home > Mailing lists > Public > www-validator@w3.org > November 2000

Re: Validator option suggestion

From: Terje Bless <link@tss.no>
Date: Thu, 2 Nov 2000 22:14:52 +0100
To: W3C Validator <www-validator@w3.org>
Message-ID: <20001102233228-r01010600-21890810@10.0.0.2>
On 02.11.00 at 13:48, Gerald Oskoboiny <gerald@w3.org> wrote:

>One problem with this though is that there may be some 401s that are a
>result of the draft being (temporarily) in member space, and others that
>are actually links to other member-internal resources that need to be
>reported.

Y'all seem awfully hung up on the fact that you use the link checker to
check links in W3C material... Try focusing on making it work right in a
generic (non-W3C specific) fashion and __fix_your_working_methods__! :-)


>How about an option that says "ignore 401s for resources that have the
>same URI prefix as that being checked" or something?

And this would help, how? Are we talking about links from a TR to an
included file like, say, a DTD here? This is impossible to distinguish from
a bad link to a protected area because the link checker's ESPs[0] aren't
yet fully enough advanced to know which files you'll be moving to a Public
area and which will get left behind[1].

A 401 is a 401, try checking your documents in a Public "staging" area
before publication instead of trying to hodge-podge around fundamental
aspects of HTTP. Simply make it a part of the routine to check out a copy
(y'all _do_ have these things under VC, right?) in "Staging" and run the
checker on that. This could probably even be automated...


I think the main problem was false positives though. Patrick said he was
getting a gazillion 401 warnings in Member space. This sounds like internal
links to Appendices, DTDs, and the like. Having to pick up a few leftover
401s o Member areas afterwards has got to be better then a few hundred
false 401s to components of the same comound document.



[0] - "Extra Sensory Powers" for the non-BeeGees fans. :-)

[1] - I'll give you "relative links", but not "same URI prefix".
      Ok, I'm a pedant, so sue me. :-)


-- 
"I don't want to learn to manage my anger; I want to FRANCHISE it!"

                          -- Kevin Martin <brasscannon@bigfoot.com>
Received on Thursday, 2 November 2000 17:32:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:13:54 GMT