- From: <Jonsm@aol.com>
- Date: Mon, 12 Feb 1996 17:44:27 -0500
- To: www-html@www10.w3.org
- Cc: eric@gauthier.centre.edu (ericgauthier)
In a message dated 96-02-12 16:25:30 EST, eric@gauthier.centre.edu (Eric Gauthier) writes: >Anyway, it sounds like a great idea and I'd love to get in on it, >but I think the bandwidth and server load that it would take up >at lycos and yahoo would be to large for them to want to do this. >It would be interesting. If anyone out there is interested, I'm >game... The overhead would be minor compared to the computation required to do the full-text indexing. Bandwidth is not an issue since the page had to be fetched to index it. I would suggest these categories: 1) Errors - pages containing severe errors like overlapping tags, quote/comment problems. Any page that can't be parsed by a SGML system gets this rating. This doesn't mean all of the tags/attributes will be understood, it just means that the page is not lexically correct. 2) HTML1 - Is this needed? 3) HTML2 4) Extended HTML2 - tables and extensions common to several browsers. A general rule could be that the extension must exist in at least four browsers. 5) Vendor specific - pages that will only work on specific browsers (frames, Java, VBScript, etc..) Vendors will need to provide a DTD to validate against. A page that is lexicalyl correct yet fails 2-5 would be vendor specific where the vendor is unknown. Sites that generate multiple pages based on the User-Agent field could get more than one rating. Jon Smir, jonsm@aol.com
Received on Monday, 12 February 1996 17:45:21 UTC