lost interest?

Chris Barker chrishbarker at attbi.com
Wed Dec 12 13:09:20 EST 2001


Alex Martelli wrote:
> To me, it suggests something different: that authors are duplicating
> functionality or doing without it rather than reusing, say, mx.DateTime or
> other packages offering low-level, reusable functionality -- because nobody
> wants their distributed packages to be harder to install and use, and
> therefore less frequently used, by requiring dependencies on others without
> a cpan-ish infrastructure in place.

I absolutely agree. I make a point of following discussions related to
Numeric on this list, and
in the last few months there have been MANY discussions that went
something like this:

OP: What is the best way to multiply all the elements of a long list of
numbers by all the elements of another long list of numbers?

Me (or someone else) Use Numeric.

OP: I don't want my users to have to install a separate package

ME: It's really not that big a deal, and you can distribute it with your
app.

OP: I really only want to use the standard lib. Besides, Numeric seems
overkill for what I want to do. So is map or list comprehensions a
better way to do this? By the way, I also need to take the sine of all
those numbers, and I may have to work with 2-d arrays in the future.

ME: Then you really, really, want Numeric

OP: I need this code to run on various machines that are centrally
administered, I can't install Numeric on all of them and the sysadmins
are not willing to do it for me...

etc, etc, etc.

Fill in similar conversations for PIL, mxDatetime, wxPython, etc. This
really is an issue! Another example is tkPlotCanvas. It was originally
written using Numeric (which makes lots of sense) and I have seem
various versions that have been adapted to "not require Numeric". This
is sad for me.

Note: There are movements affoot to get both a decent DateTime package
into the lib, and Numeric2, but even if this happens, there will be many
other very useful packages that will never make it inot the standard
lib.


> I'm ready to bet a good lunch at Bologna's best restaurant,
> wine included, that dependencies will become much more common, i.e., that
> reuse will indeed have been encouraged.

There's no way I'm taking that bet! Unless maybe you'll cover airfare to
Italy as well :-)


-Chris


-- 
Christopher Barker,
Ph.D.                                                           
ChrisHBarker at attbi.net                ---           ---           ---
                                     ---@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Oil Spill Modeling                ------   @    ------   @   ------   @
Water Resources Engineering       -------      ---------     --------    
Coastal and Fluvial Hydrodynamics --------------------------------------
------------------------------------------------------------------------



More information about the Python-list mailing list