Re: Letters with **three** cases?

In article <87k6k8aszh.fsf@xxxxxxxxxxxxxxxxxxxxxxxxxx>,
Lee Sau Dan <danlee@xxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
>>>>>> "Colin" =3D=3D Colin Fine <news@xxxxxxxxxx> writes:

> >> I don't think so. "a" is different from "A". Why should the
> >> computer pretend that they're the same?

> Colin> Because people think they're the same (in some useful sense
> Colin> - in particular, people often have difficulty remembering
> Colin> which is used).

>At least mathematicians don't mix them up. They often write "Let X be
>a set and x be an element of X..."

Not only that, but mathematicians will use characters from
different fonts or styles as different variables. A boldface
"a" or a Fraktur "a" are different variables than a lower
case Roman (or whatever is used for math mode) a. In general,
a variable can be any recognizable symbol or symbolic expression.
This is what makes them so powerful.

Neither do chemists mix them up.
>They write "NaCl", not "NACl". Everyone can make the distinction.

As an example where it is needed, NO is nitric oxide,
while No is nobelium.

> Colin> Indeed. There are contexts where people distinguish them -
> Colin> while knowing all the time that 'a' and 'A' are the same
> Colin> letter.

>Don't think in terms of "letters". Think in terms of "characters".



> >> Remember, computers are stupid and they can't read your mind.
> >> Having the computer guess (or determine, if you like this wrong
> >> word) when you want to distinguish "a" from "A" and when you
> >> don't would be a stupid design decision. If you fix the rule
> >> (e.g. "a"=3D=3D"A" in file names; "a"!=3D"A" in word processor
> >> document contents; "a"=3D=3D"A" in circumstances 1, 3, 7, 18;
> >> "A"!=3D"A" in circumstances 2, 5, 9, 11; ...), then the human
> >> user must learn such rules. This would make it even more
> >> complicated to operate a computer.
> >>=20
> Colin> The human doesn't need to learn those rules because readers
> Colin> of a language with case differences know them, more or
> Colin> less.

But it is necessary to know when the letters are characters,
or letters from an alphabet. BTW, I believe that email
addresses are case insensitive, by convention.

>But that doesn't mean there is an *obvious* rule for whether file
>names should be case sensitive.

However, filenames are strings of characters, with few limits.

> Colin> As I said above 'A' and 'a' are two forms of the same
> Colin> letter.

>Two different "characters".

Most definitely.

> Colin> If I fail to capitalise the first letter of 'Andrew', or
> Colin> capitalise the middle letter of 'Dan', it's an error in the
> Colin> sense that it's departing from standard practice and might
> Colin> raise an eyebrow.

>Computer filenames were something new. There were no conventions
>established for them before computers were invented.

> Colin> But it doesn't change the meaning (words like 'reading' and
> Colin> 'natal' apart, but then there are plenty of homographs in
> Colin> English).

>Yeah. Should chemists then follow suit, and start not to distinguish
>"Cu" from "CU"?

I do not know if CU is a possible compound, but it has a
tentative meaning.

> Colin> It's interesting how often now you hear people telling
> Colin> somebody their email address and making a point of
> Colin> specifying the case, when it's not actually significant in
> Colin> that context.

>That's because of the inconsistency I mentioned. In some contexts,
>case is significant and in others, it isn't. It'd be better to be
>consistent: Either everything in a computer is case-sensitive, or
>everything in a computer is case-insensitive. If find the latter make
>more sense.

Part of the reason for this is that some computers were, and
may still be, caps only. AFAIK, some computer keyboards have
caps locks, not shift locks.

> Colin> People have got used to the fact that they have to make
> Colin> this awkward distinction when they are mucking about with
> Colin> computers (and of course they do have to in passwords).

>So, having that distinction in passwords but not filenames creates
>more confusions than it helps.


> >> Unix goes for simplicity, which indeed results in high
> >> flexibility. Less is more.
> >>=20
> Colin> 'Simplicity' is like 'fairness' - it generally doesn't work
> Colin> in argument because it is often not a single-valued
> Colin> function. To me, brought up in English, case sensitivity is
> Colin> a complexity, not a simplicity.

>To me, case is a complexity in itself in the first place. Having to
>write programs to do case-insensitive matching is bringing this kind
>of nonsense into programs, making programs more error-prone.

Variables were the first "alphabetic" expressions used in
computer languages. Variables are character sensitive; no
mathematician would think of using "cases" in variable names.
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Department of Statistics, Purdue University
hrubin@xxxxxxxxxxxxxxx Phone: (765)494-6054 FAX: (765)494-0558
.

Relevant Pages

  • Re: How many years dedicated to characters learning in China ?
    ... Colin> LEE Sau Dan wrote: ... Peter> radical + phonetic that you didn't learn in school in the ... Colin> school in the first few years". ... No -- that suggests that the non-composite characters are rare in text ...
    (sci.lang)
  • Re: Problem with the strip string method
    ... Colin J. Williams wrote: ... The chars argument is a string specifying the set of characters to be removed. ...
    (comp.lang.python)
  • Re: Hello
    ... Colin E. wrote: ... you shell hackers and your bloatware. ... A *real* coder ... would have done that in under 80 characters. ...
    (rec.juggling)
  • RE: Password complexity
    ... Colin. ... > requires users' passwords to be at least seven characters ... > Tom G. ...
    (microsoft.public.windowsxp.general)
  • Re: Letters with **three** cases?
    ... > Colin> Actually, in my opinion case sensitivity was the worst ... As I said, when computers were the domain of geeks, it was fine for them ... English, case sensitivity is a complexity, not a simplicity. ... But if you're equipped with a good editor, ...
    (sci.lang)