Great to see the progress on this paper. Comments follow 1. "To make RDF access easy on a platform which is open to everyone with a website (or, hopefully, a computer) program in. (@@CLARIFY)" Something like maybe: "The scripting environment within a typical web browser provides a programming language [javasrcipt 9], network accesses and a powerful user interface [HTML DOM]. Developers pushing the bounds of web technology tend to be conversant with this platform".? 2. " The documents published by W3C and the organizational structure" s/the/its/ 3. s/sees it was a Saturday, but still don’t /sees it was a Saturday, but still doesn’t/ 4. sect 3.1 s/A se- mantic web browser must have an awareness/However, a se- mantic web browser must still have an awareness/ 5. Space mising before paren in /progress(yellow)/ 6. Missing comma s/sort process/sort, process/ ______ From: timbl@w3.org Subject: Comments 7-23 Re: [Tabulator] SWUI06 Tabulator paper Date: August 10, 2006 13:59:37 EDT To: lkagal@csail.mit.edu Cc: tabulator@csail.mit.edu 7. Section 3.2 s/ In the analysis mode, the user can select certain fields to define a pattern,/ To move to analysis mode, the user can select certain fields (arcs) to define a pattern,/ 8. s/From any table cell/To switch back, from any table cell/ SECTION 3.3 9. In / circle-and-arrow diagrams[6], RDFAuthor, etc./ missing ref for RDFAuthor. The [6] is OK for Isaviz. 10. S/This gives one feel/This gives one a feel/ 11. Link for the Foafnaut or how/why diagrams? Maybe I just don't see the link. Foafnaut could do with a ref anyway. 12. s/The outliner displays links in both directions /In order to keep this independence on a specific direction, the browser must treat forward and backward links wit equal stature. The outliner therefore displays links in both directions/ SECTION 3.4 13. s/Within the query results it looks for the RDF predicates from iCal [], which specifies vocabulary/ontology for working with other calendars /The calendar view is available when any the query results involve certain RDF predicates from the RDF iCalendar-equivalent ontology [@@@ref to IETF iCalendar spec and namespace URI for the ontology]/ 14. Under SPARQL, s/loops/and cycles in the query pattern/ 15. s/ highlighting the mailbox field in a tabulator developer node/ Browsing from the Tabulator project, opening a developer, and highlighting the mailbox field/ 15. s/This does not of course, define for a given x what graph should be true. / This does not of course, help a server decide, for a given x what graph G would be useful to export. / 16. /resolved and x the sub ject/ the x is not italic where before it was 17. similarly iitalic x in / when x is dereferenced./ 18. SWITCH 4.2 and 4.3 to get HTTP dealt with before we move up to RDF. 19. s/ it its document element id / it its document element is / 20. In old 4.2, the @@CLARIFY is application/xhtml+xml s"representation has content-type (@@CLARIFY) or is application/xml "representation has content-type application/xhtml+xml, or is application/xml" 21, At end of 41, now In /Suc- cessful dereferencing of an HTTP URI gives either a redirection or a represen- tation consisting of met (@@CLARIFY) / /Successful dereferencing of an HTTP URI gives a status code and either a redirection (status 300-303) or (status 200) a represen- tation consisting of content bytes and metadata./ 22. Need a reference for GRDDL on first occurrence. Ask Danc SECTION 4.4 23. s/The protocol, then, is t (@@CLARIFY)/ The protocol, then, is then to load the resource linked by rdfs:seeAlso, and then to merge ('smush') nodes with the same mailbox or mailbox hash. As these are inverse-functional properties, they are dealt with by the inference layer, which we now describe./ To be continued