[plt-scheme] One more SRFI 19 issue

By the way, you can implement an SRFI-19 "date" objects as subtypes of
the PLT ones, with only a couple gotchas.  This was done in
"http://www.neilvandyke.org/srfi19-plt/" (which I no longer plan to
merge).

I agree that adding another field to PLT's "date" structure, rather than
using structure subtyping, is seems the best way to support SRFI-19, if
SRFI-19 is considered desirable.  However, I wouldn't add a field to PLT
"date" *only* to support SRFI-19 -- my first priority would be to
improve PLT "date" functionality independent of any particular SRFI.

For people who enjoy the finer points of ISO8601, I have an over-
engineered R5RS date&time library (akin to Scheme's numeric tower),
called "bigtime.scm", that I'll try to find time to finish and release
by the end of July.

Also, people who just need to do RFC3339 parsing and formatting in R5RS
code have the option of using "http://www.neilvandyke.org/rfc3339-scm/".
(Dorai recently enhanced Pregexp, so "rfc3339.scm" can now be portable
to Schemes other than PLT.  I'll update the code and documentation
shortly.)

-- 
                                             http://www.neilvandyke.org/