Re: [Phys-L] numbering and versioning -- FCI and otherwise

[John Denker <jsd@av8n.com>]

On 06/24/2013 04:56 AM, Philip Keller wrote:

> ..... should address 4, 15, 16, 25 and 28 on the FCI.  I don't think 17 is a 3rd law question.

I have caused some confusion by referring to FCI questions by number.

This is a problem, because there are different versions of the FCI 
running around, with the questions re-worded and re-ordered, in 
such a way as to change the numbering.

The question I referred to as 17 is 18 in some versions.  It involves
an elevator.  I don't want to be too much more specific, for obvious
reasons.  We don't want details of the test to be googleable.

Sorry for the confusion.

======================

It would be nice if these things had version numbers on them, so that
we could more easily detect and discuss version-related compatibility
issues.

More generally:  It is a good practice to mark version numbers and
format numbers on everything.  If you're writing a program that
logs data in such-and-such format, put a format indicator at the
top of the file, even if you can't (yet) see why it is necessary.
I've seen this done right -- and done wrong -- enough times to know
rather precisely how important it is.

If you can't think of anything better, write "#! foo" or perhaps
"#! rumplestiltskin" at the top of the file ... something you can
recognize, something you can change if/when the format changes.