Timezones
Brent Pinkney
pinkney_b at aircom.co.za
Wed May 7 05:37:37 UTC 2003
> [Whole timezone thing]
> Has me thoroughly confused.
Me too - that is why the NAMED timezone stuff is the bit I have yet to complete.
>Yes, this is what happens when important commonly required facility is
>ignored for too long - now we have too many incompatible solutions.
agreed - my stuff was published years ago.
Although I admit some unit tests would have gone a long way :)
>I think so - but now I'm confused by Brent's message as I'm not clear
>what relation (if any) his package (which I just learned of) has to the
>ANSICompatibility package put up by Alain Fischer (which I have loaded
>now).
No relation to this package, expcet that my package also implements the ANSI DateAndTime and Duration classes.
I also implement a superset for timespans (Weeks, months, etc) as motivated on the Swiki. This I belieive is a nice illusytration of the power of Smalltalks block semantics bug-less iteration.
>To remain focused on the basic goal - timezone aware squeak - Brent's
>package is more focused (it only deals with dates and times where the
>ANSI package has a whole grab bag of stuff from the ANSI spec) and thus
>is "the simplest thing that could possibly work".
Agreed - the current ANSI package is to monolithic for my liking - perhaps separate SM packages for each ANSI protocol instead ?
>Anybody have an opinion?
This is mine.
Brent
More information about the Squeak-dev
mailing list
|