Hi there,

I wonder how many people unknowingly create errors by their choice
of a name for something? As well, I quote an example of _good_
error reporting. Very simple, copy/paste an example into post.

"asbesto" wrote to lkml:
>using the character > ' < in the
>
>() Local version - append to kernel release
>
>will cause "make" to fail. Here's an example:
>
>Using:
>(Gesu'Cristo) Local version - append to kernel release
>
>will cause:
>gemini linux # make
>/bin/sh: -c: line 0: unexpected EOF while looking for matching"'
>/bin/sh: -c: line 1: syntax error: unexpected end of file
>make: *** [include/linux/version.h] Error 2


I look at the quoted error message above, see a shell error, and
am reminded of:

"Doctor, it hurts when I do that?" ... "Don't do that!"


So what is safe for a name? At a minimum one could apply programming
variable name rules: start with a letter, follow with alpha + number +
underscore characters ([a-zA-Z][a-zA-Z0-9_]*). Embedded dashes are
safe for file names. Other characters may be safe, but then one must
track exceptions for various contexts. Stuff may happen

Information overload precludes thinking about naming of things
too much, no?

Grant.