mxTools (was Re: why no "do : until"?)

Kevin Russell krussll at cc.umanitoba.ca
Wed Jan 10 17:07:14 EST 2001


Jason Cunliffe wrote:

>
> hmmm... let's consider for a moment neither you nor I..
> What about all the newcomers to Python, as beginining programmers, kids,
> students, newbies,
> immigrants from other languages..? What sanity it would be to arrive in
> Pythonia, where DateTime() is actually corresponds to the scope of
> historical time, instead of some *nixed no-life before 1970!
>
> Any kind of datetime needs [beyond system timestamping of files or message
> packets]
> needs mxDateTime(). Any use concerning people and not machine prblems,
> including even basics like date of birth. Anyone over the age of 30+ is out
> luck. Let's see that is maybe 2-3 billion people?
>
> mxDateTime() gets it right the first time. It allows for sane administering
> of all kinds of historical, scientific, statistical, medical, geographic,
> demographic and business data......Or how about wonderful extended features
> of mxDateTime for easier design of more intuitive and useful calendar
> applications.

I have to add my agreement that the standard library needs a module
that actually does dates and times correctly -- which for now means
mxDateTime.

I find the not-enough-people-use-it-so-it's-just-bloat argument utterly
unconvincing in this case.  Are there honestly fewer Python users who
need to deal with dates than there are users who do other things that
the standard library provides for?  (Like interfacing to the Berkeley DB
engine, using the gopher protocol, dealing with image files under
SGI IRIX, interfacing to Sun's NIS?)

If it honestly turns out that there *are* fewer, then by the same logic
the standard library shouldn't include the current broken DateTime.
If there aren't enough Python users who need dates done right, how
many Python users could there possibly be who need dates done
wrong?  If the standard library can't handle dates in the transparent
and correct way that a user would expect (given the rest of the
standard library), then it shouldn't be pretending to handle dates at
all.

-- Kevin





More information about the Python-list mailing list