Re: Problem with XHTML validator and link checker?

On Thursday 26 July 2007, David Arnold wrote:
> Now when we try to run the link check, it returns a page of HTML code
> prefixed with:
>   fetched in 1.6s
> Content-Type: text/html; charset=utf-8
> Content-Language: en
> If I copy/paste the HTML code into a file and view it, it looks ok, but
> at the end after the 'Results' section it coughs up an "Internal Server
> Error" error.  This is coded into the HTML returned, not from our web
> server.
> Oddly enough, sometimes it returns the link check page properly formatted,
> with no server error.  It seems to be random/sporadic.  Could this have
> some relation to the perlmod bug discussed on the list yesterday and
> reported resolved by olivier Thereaux?

Yep, that's likely.  I suspect a corresponding change to the 
server configs as the latter hunk of this change is the culprit:

That one (probably inadvertently) makes the link checker to run with mod_perl 
too, and apparently it isn't quite ready for that after all.

In addition to the things you describe above, I'm seeing various other 
oddities such as the link checker's front page displaying at the bottom stuff 
like "Checked 6 documents in 2686.8 seconds." just by loading it, ie. without 
actually checking anything yet.

I think this should work around it until the link checker is fixed to work 
properly with mod_perl, Olivier, could you have a look?

Received on Friday, 27 July 2007 21:49:47 UTC