RE: Version 1.11

["Pamster" <pam@pamster.com>]

In a ring of 225 sites, I think I got about 10-15 of those...  While that's
a small percentage, in a really large ring it could be a significant number.
Pam
> -----Original Message-----
> From: owner-ringlinklist@1-host.com
> [mailto:owner-ringlinklist@1-host.com]On Behalf Of Gunnar Hjalmarsson
> Sent: Monday, November 20, 2000 8:07 AM
> To: ringlinklist@gunnar.cc
> Subject: RE: [ringlinklist] Version 1.11
>
>
>
> Pamster wrote:
> >
> > For the problem with saying that links are missing when
> > they're not - I reran the site in the previous example - and
> > it returned: HTTP/1.1 200 OK - is that what you mean?
>
> Well, yes. Actually I realized when I had posted yesterday that
> I shouldn't have needed to ask. The checker doesn't look for any
> links if the return code is something but 200.
>
> Btw, how common is it that the checker reports failing links
> when the fragment seems to be ok?
>
> As I mentioned in the last posting, the checker tolerates
> certain variations. But this means that the regular expression
> which compares the strings with the pages is rather complicated,
> and I suspect that this may be a reason for these incorrect
> results. I.e. an execution error rather than a program logic
> error.
>
> On the other hand, if I simplify the regex it might be possible
> to reduce the number of execution errors, but instead the
> checker would report some incorrect failures due to program
> logic errors...
>
> In other words: Do you think it's worthwhile to play with the
> code?
>
> / Gunnar