Re: Check for redundancy

[Emanuel Berg]

Yuri Khan <yuri.v.khan@gmail.com> writes:

> There are still reasons to use Hungarian notation,
> because even the most strongly typed languages in
> common use are not strongly typed enough. (Haskell
> is or can be persuaded to be, but my employer is not
> ready to embrace Haskell.)

Good call :)

> 1: Every time I have to work with two or more
> coordinate systems that use the same axis names
> (typically X and Y). E.g. a mathematical plane and
> pixel coordinates relative to a window and/or
> a monitor and/or a workspace. Eventually I start
> naming variables starting with “wx, wy” for window,
> “sx, sy” for screen, etc.

Yes, however that falls outside of my definition of
Hungarian notation - I would do the same, only I would
probably spell out "window", "screen", and so on.
So in Lisp it would be window-x, window-y; in C,
win_x, win_y; in C++ perhaps Window would be
a class; etc.

> Without carefully labeling each affected variable,
> after a month of not touching the codebase, it
> becomes impossible to make modifications without
> breaking anything.

Indeed, I do long and descriptive names to the best of
my ability. I consider it a virtue. In C, when you
don't do that typically, I still do them longer and
more descriptive than most C programmers.

> Note that this isn’t the Systems Hungarian many
> people are used to, which encodes the actual type.
> It’s Applications Hungarian, where variables that
> have the same physical type get different prefixes
> after their conceptual type.

In so many words, I agree. Tho to my mind this is
common sense rather than any particular notational
system being adhered to.

-- 
underground experts united
http://user.it.uu.se/~embe8573