Why do you use python?

Dan Sommers dan at tombstonezero.net
Mon Mar 21 00:49:58 EDT 2016


On Mon, 21 Mar 2016 15:13:22 +1100, Chris Angelico wrote:

> On Mon, Mar 21, 2016 at 2:59 PM,  <rharding64 at gmail.com> wrote:

>> instead, to be efficient, it is best to combine tools to solve
>> problems that contain complexities where there is nothing available
>> off the shelve that does the job. c# is free, free VS studio, i can
>> run ironpython there, i can do python there, and talk to linux boxes
>> with python, i can run c# on linux boxes using mono(did that back in
>> 2004 and thereafter for a while).  i can run python on my beaglebone
>> black inside of snappy ubuntu, ect.

>> so i ask those employers why not use what is available to solve
>> problems instead of limiting yourself to just one???

> Because you won't be there forever, and they'll have to find someone
> else to maintain your hellspawn hodge-podge of languages, tools, and
> libraries. (And yes, it will be described that way by the next person,
> no matter how careful you are.) It's in their interests to restrict
> its complexity at least a bit. I'm not sure what advantage you gain by
> incorporating C# into the mix, but the *dis*advantage is that, forever
> afterward, Visual Studio and Mono will be necessary to use and develop
> this project. Every new thing needed is another thing that can go
> wrong, another thing people need to learn, etc, etc.

> So instead of treating programming like a plumber at a hardware store,
> treat it like an artist with a canvas. You wouldn't normally see a
> portrait done partly in watercolor and partly in oils - or if it is,
> it's for a VERY deliberate effect. You'd more often see one style used
> for one project, and maybe another one used for another.

Both viewpoints must be tempered in order to create successful systems.

I've worked on jobs where the tool or target operating sytem (singular)
was chosen first, or specified as part of the "system requirements," and
it can work out just as badly as a hellspawn hodge-podge of a solution.
We've all heard the one about all problems looking like nails.

It should, by now, go without saying, but choose the right tool (or
tools) for the job, where "right" includes any number of things *not*
related to its immediate implementation, and often includes some number
of things *counter* to an obviously superior, in one way or another,
implementation.



More information about the Python-list mailing list