Ake Hedman wrote:
> Yes, Unicode  makes a lot of things harder and for an OS it must be 
> used. Not as sure for a C/C++ and other program languages though. To 
> restrict variable names etc to be US-ASCII does not feel like big deal 
> to me even if I come from another languages zone. But any deterministic 
> translation would work well also I think.

The key there is *feel* like a big deal to *you*.  Consider a programmer 
  in a society that uses a non-Western language.  Yes, he has to accept 
that the (relatively few) language keywords are going to be written in a 
foreign language using a foreign writing system, but why should he have 
to do that with his variable and function names?  After all, choosing 
good names is a big part of making code readable and maintainable, and 
having to translate as you read and write code really detracts from 
that.  A likely result is going to be a bias toward using arbitrary names.

[_The Psychology of Computer Programming_ by Gerald Weinberg]

> This sounds like a must have book. Will buy it have never heard of it 
> before. Thanks for the tip!

It was written 34 years ago and a new edition was recently released. 
Some of it seems a little "dated"; it was written before the advent of 
personal computers and a lot of the examples deal with things like 
arguments over whether timesharing systems were more productive than 
batch systems, but the basic principles are still valid.  One of 
Weinberg's major contentions is that programming on non-trivial projects 
has to be considered a *social* activity; that's likely to ruffle some 
feathers, but I think he supports it well.

Some highlights:

A consideration of all the factors that determine whether or not a 
program is a "good" program.

The hilarious description of the progress-reporting system where 
progress reports for a given month had to be entered during the 
preceding month.

The observation that "debugging" actually consists of at least three 
distinct activities which use different cognitive abilities and 
personality traits.

All the stuff on psychological set.

The classic observation that "an 'art colony' is a place where everyone 
knows how to look like an artist but few, if any, know how to paint like 
one." (it appeared in the context of a discussion of programming 
practices (and programming management practices) that encourage 
improvement versus practices that merely encourage wannabeism).

The observation that "if builders built buildings the way programmers 
write programs, the first woodpecker to come along would have destroyed 
civilization."

Emphasis on the distinction between how easy a tool is to *learn* and 
how easy it is to *use*.

Lots of illustrations of the psychological phenomena that make extremely 
smart people behave in extremely dumb ways, such as the team that nearly 
signed off on the completion of a project that was running six weeks 
behind, the team that decided that a certain piece of bad output was the 
result of simultaneous hardware glitches affecting two completely 
independent systems, several cases of managers rewarding programmers for 
making mountains out of molehills, the programmer who had to patiently 
explain to group of other programmers that the fact this his solution to 
a problem produced correct results whereas theirs didn't was more 
important than the fact that their solution ran faster than his, and 
many more.
_______________________________________________
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist