W3C home > Mailing lists > Public > public-qa-dev@w3.org > February 2005

Re: Checking Recursion (was: validator/httpd/cgi-bin check,1.391,1.392)

From: olivier Thereaux <ot@w3.org>
Date: Mon, 7 Feb 2005 12:36:40 +0900
Message-Id: <1c52f99cb36ea9007b1698003caefd2f@w3.org>
Cc: QA Dev <public-qa-dev@w3.org>
To: Terje Bless <link@pobox.com>

On Feb 6, 2005, at 6:41, Terje Bless wrote:
> Bjoern Hoehrmann <derhoermi@gmx.net> wrote:
> _Private_ HTTP header fields; they're not supposed to be meaningful 
> outside of
> the instance where /check is talking to itself. I dunno. The way we're 
> using
> them I have few qualms about it.

Would they allow instance B to check instance A, and recursively?
Do we care about such a situation, one way or another?

>> I further think that refusal to check the validator output is a bad
>> idea (we, after all, link to that in the results page), so the default
>> should be to limit two or at least one recursion.
> Well, there's a distinction between what we ship as a default and what 
> we
> configure our installation on v.w3.org to do. The default as shipped 
> should be
> zero because we want to err on the side of caution, but v.w3.org 
> should be
> able to check itself and qa-dev should probably allow infinite 
> recursion.

To the points already stated in this discussion, I would add that we 
often recommend checking a "local document" to test any local 
installation. one recursion thus seems to be a good minimum for every 
installation, and, perhaps also a good maximum. I would lean toward 
making this a default, and at least hardcoded minimum.

-- 
olivier

Received on Monday, 7 February 2005 03:36:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 19 August 2010 18:12:45 GMT