I've started my project to replace htmltext with a stub using qpy.  I
need this to avoid exceptions with Cheetah, both because htmltext is
not recognized as a string  and because it might contain non-ASCII
characters that will cause UnicodeDecodeError when the final page is
output.  Yet Quixote uses htmltext internally all over the place so
it's impossible to avoid it.  Right now I'm using a patched version of
Cheetah with a default filter that recognizes htmltext and carefully
converts it to unicode.

I'm planning on a standalone module that will replace htmltext in
memory, using whichever 'c8' qpy chooses -- C or Python.

I did a diff of Quixote 2.4's html package vs 2.5a1, and it looks like
there are no changes affecting me, just some redefines of "int" to
"Py_ssize_t" in  _c_htmltext.c .

In qpy, the only difference between 1.4 and 1.5a1 is the test suite,
so I don't think that will affect me either.

Are there any other changes planned for Quixote or qpy that would
affect this?  How close are we to final releases?

--
Mike Orr