[tldoc] html + \Thanh updates

[Karl Berry]

I agree with Reinhard that it would be desirable to have UTF-8 output in
all the HTML.  I tried to do that for the English version.  It did not
work out -- I could never get the encoding declarations to match the
actual output, and I did not have the time and inclination to create
minimal test files for Eitan.  So while I think it is a goal, it is
something that is going to need time and effort put into it.  Are there
any volunteers?

I also agree with Reinhard that the accumulation of "old stuff" in the
TL doc is a pain.  The Czech and Polish (maybe others) .sty files that
repeat (almost) everything from the English make additions unnecessarily
painful too.

As I wrote before, I would like to factor things out to make it easy for
the translations to change what needs to be changed, but I don't know
what, specifically, to do.

And I agree with Reinhard once more that the time to work on these
"infrastructure" things is (essentially) now, and not at the end when we
are updating the content and the doc is the critical factor keeping us
from making the release.

So ... any volunteers willing to try to rationalize the setup,
delve into encodings and tex4ht, work with Eitan, etc.?

karl