goran.hultgren at bluefish.se wrote:
	Wow, I am impressed Richard - this is really interesting reading. May I
	come with a suggestion to all who are following this discussion?
	
	Why don't we start a "rework the whole darn Collection mess"
	project and let Richard take a first hit at it.

Because I really need to spend a bit more time on my paying job.
In the time that's left from that, I need to work on some of the
paying consulting I've been offered (inflation: about 3%. employer's
pay offer: 1.5% + suspension of any union members who strike).
After that, I should spend some more time with my family.

And I haven't even _started_ to read the StrongTalk stuff yet.

I mentioned this whole topic to our sysadmin.  (Yet another person
who expected the operation I'm not going to name to just plain work.)
She asked why the ANSI test suite didn't catch it.

What we really need are good test suites for the "core classes",
whichever they are construed to be.  YAY! for SUnit!  Once again,
a great big Thank You to Richard Harmon for his ANSI test cases.

Hang on, what we _really_ need is better documentation.
Maybe the harvesters could give priority to documentation fixes for a while?