- From: Terje Bless <link@tss.no>
- Date: Tue, 19 Oct 1999 08:03:08 +0200
- To: W3C Validator <www-validator@w3.org>
On 18.10.99 at 13:28, Ulrich Wisser <u.wisser@luna-park.de> wrote: >where can I download a functional copy of validator. I'm afraid that's a bit of a tall order at the moment. The validator has a few prerequisites that you need to install first. As it's written in Perl, you'll need a recent perl installed <URL:http://www.perl.com/>. Perl 5.004 works, but Perl 5.005 is better. Also, the validator uses James Clark's <URL:http://www.jclark.com/> validating SGML parser nsgmls to do it's dirty work, so you need that installed too. It's written in C++ so you'll need a C++ compiler (like gcc 2.7+). If you use the weblint option you'll need weblint installed <URL:http://www.weblint.org/> and if you need charset conversions you'll need "nkf" installed (I haven't gotten around to that bit yet so I dunno where you'd get that). Once that is done, you'll need to get the source for the validator from <URL:http://validator.w3.org/source/>. Either using CVS or by downloading the tared and gzipped CVS tree linked to from that page. You'll need to modify your web server's configuration and change some file paths and URLs in the "check" CGI program. I'm afraid I haven't gotten around to writing a HOWTO for those things yet, but if you have problems you can always send mail to the list or to me off-list. The main issues with the web server config are to make it recognize "check" as a CGI program and to enable content negotiation so it'll return either GIF or PNG versions of the images. In "check" you need to modify file paths for log files and binaries; some string comparisons so you won't get redirected to validator.w3.org; and some HTML output that tries to get images from w3.org.
Received on Tuesday, 19 October 1999 02:04:55 UTC