Improving TUS usability

L2/09-xxxx

I used to be able to look up the algorithm for NFC by going to http://www.unicode.org/reports/tr15/tr15-29.html#Specification and reading it.

I now:

This is part of a very general (and rather extensive) issue of providing links directly to sections from throughout the UAXes. Note that there are dozens of specific section references in UAX #44 that would benefit by being direct links to the text, rather than indirections through [Unicode] references in UAX #41. [This paragraph is extracted from an email discussion with Ken.]

Moreover, because we don't number subsections in the standard, we have to use goofy locutions like: See the subsection "Gorp" in Section 3.11, "Normalization Forms" of [Unicode]."

To get the discussion started, I'll suggest a concrete proposal.

    1. We number subsections as well as chapters and sections, eg "Basic Text Processes" in "2.1 Architectural Context" becomes "2.1.1 Basic Text Processes"

    2. We generate PDF links for all of the chapters, sections, subsections, tables, and figures (I don't know if the last 2 are feasible, but if possible...)

    3. If the PDF links can be reasonable and predictable, like http://www.unicode.org/versions/Unicode5.0.0/ch01.pdf#S2-1-1, then we are set with that.

      1. If not, then we provide a set of reasonable redirects, like ...ch01.pdf#S2-1-1 => ...ch01.pdf#G6600

    4. Because these links are then predictable, we can use them from within UAXes, even if they are on a different production schedule.

    5. We can then provide direct links from UAXes to relevant portions of the book, putting a direct link on the text, and still maintain the [Unicode] link. Eg

As an alternative to #3, we could have a separate page of links, where each is linked to by a URL of the form http://www.unicode.org/versions/Unicode5.0.0/links.html#S2-1-1, and the line it links to has the gorpy PDF link ... ch01.pdf#G6600. That way it would take at most two clicks to get to the right place.