I like that you've been thinking about thinking while writing your
framework, Isaac.  I dig that. :-)


The way I'm currently tackling the tests, is to try to work them into
my routine.

That hard part is, as been mentioned, not having to go back a lot to
keep it all in sync.

I wonder if this ties into keeping branches synced up with trunk and
whatnot with SVN... heh.  Somehow similar...

Anyways, where I'd, before, just keep doing stuff in a scribble file,
sorta testing, building, testing (but not saving any tests), I now try
to save the tests, sorta.

It is hard to write tests that are easy to refactor, but it's getting
easier as time goes by.  There are sweeping changes sometimes, and I'm
a sweeping changer, for all my not much changing... but it really does
save time having those old tests, of the various bits.  If just in
having that history/memory refresher...

I'm still working on the testing of the behaviors, I guess, because
I'm quite often including stuff in later tests that are tested
themselves, but should be stubs (instead of the real things), just to
verify the behavior, but that's still sorta beyond me.  Basically I
can see cascading fails, where I don't think you'd have that with
stubs and whatnot.  It still works, but I could see how it would work
better to keep it strictly to interactions, vs. creating and whatnot.
If that makes sense.  Probably didn't explain it too clear since I'm
still twirling the ideas around in my head, sorta.

I get the failure aspect of overthinking, or overengineering, and I
battle those twin demons (or maybe it's just one demon with two
heads), and other, more Rube Goldberg-ish impulses, but I have
actually seen some pretty good return on tests, even with poor style,
as it were.

What sells me, is projects like the javascript toolkit "dojo".  That
thing is verbose, but put together really well.  It's a snap to read
the sources, and there are some good standards, and *various* types of
tests, from your more literal "it passed!" to your, "here a few random
examples".
I've had really good luck refactoring and keeping up-to date with it,
and I attribute the majority there to well written code.  It has been
really interesting to watch it evolve.  Same with projects like
Eclipse and whatnot.  Apache, hell, it's been ages.

I think that part of the reason open source is kicking ass, is because
it's really tackled the problems of the future, you know?  There's a
lot of history, too, that we can learn from,  all the various flavors
of OSes.. Mozilla... from sources open to closed to a mix of the
two... but, heh, the battle of big balls of mud...

Maybe it's fundamental.  Part of nature.  Something about entropy or
whatnot, ya know?  Chaos and order, a spinning yin-yang of seasons.
:]

Balance.  I'm trying to hold it, but it still sorta feels like jumping
from one side to the other, versus just sorta maintaining.  Not
exactly smooth... but not too rough.

Hmmm... maybe it will be easier if I go faster!  HaHA!

-- 
There is no such thing as a 'self-made' man. We are made up of
thousands of others.
George Matthew Adams

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to 
date
Get the Free Trial
http://ad.doubleclick.net/clk;203748912;27390454;j

Archive: 
http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311897
Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4

Reply via email to