[boost] Re: result of compose discussion

From: Reid Sweatman (reids_at_[hidden])
Date: 1999-06-21 14:08:40

I had no intention of shutting out international users when I suggested this
scheme, and really don't have any objection to using digits instead of
spelled-out English numbers.  However, it seems a silly bit of
hair-splitting when the rest of the words in the names are English, as are
most of the keywords in the C++ language.  For foreign speakers who already
program in C or C++ I can't see that it makes much difference to throw a
couple of more meaningless (for them) English symbols in the mix, especially
since numbers are usually about the first thing anyone learns in a foreign
language.  I can still count in four languages, although I only speak
English with any fluency, and can sort of read German.  If C++ had been
developed in, say, France, (an example I pick because I have nullary command
of French <g>) I'd be programming in it fluently, accepting the funny
keywords as memorized symbols whose meaning I've acquired, and picking my
own variable and function names in my own language.  No problemo.  After
all, all the operator notations are either Greek- or Latin-derived, and
those are generally foreign languages to everyone.  I'm only familiar with
them because I spent years studying math (and I slept through two years of
Latin in High School <g>).  I mean, I have to stop and think to puzzle out
what a "cohomology" is from the word; but I know what the concept behind it
is, so I don't worry about the etymology of the word; it's merely a symbol
attached to a particular referent (or, in computer terms, a key hashed to an
address to a value <g>).

Anyway, where this ramble started was as an attempt to debunk the notion
that it matters whether it's in English or not; I suspect that hard-core
foreign users will just translate the code into their own languages as much
as possible anyway.  After all, templates are merely *patterns*; it doesn't
matter what the symbols are.  It's all machine op codes at the other end of
the mill.

> -----Original Message-----
> From: Andy Glew [mailto:glew_at_[hidden]]
> Sent: Monday, June 21, 1999 10:42 AM
> To: boost_at_[hidden]
> Subject: [boost] Re: result of compose discussion
> 
> 
> >These names are more "fun"<g>, but I'm unconvinced that 
> they're better. The
> >other ones parse like real english.
> 
> 
> When I have defined similar libraries in other programming languages,
> I have been able to say
> 
>     '0ary function'
>     '1ary function'
>     '2ary function'
> 
> etc. - but this was in languages that allow arbitrary 
> characters to be in
> names, and are not limited to alphanumeric strings like Algol 
> derived languages
> such as C++. (The above are actual names.)
> 
> I would be tempted to say
>     
>     _0ary_function
>     _1ary_function
> 
> but I think that leading underscores are discouraged by the 
> C++ standard.
> 
> I kept trying miscellaneous alphanumeric prefixes, such as
> 
>     n_1_ary_function
>     n_2_ary_function
> 
> but this seems clumsy.
> 
> 
> However, as for the clumsiness of the postfix form not being 
> "English-like"
> 
>     function_1_arg
>     function_2_arg
> 
> etc., may I remind you that, if this library is successfull, 
> many non-English speakers
> will use it? Spelling numbers is silly when digits are 
> universally understood.
> 
> 
> --------------------------------------------------------------
> ----------
> FREE email Newsletters delivered right to your in-box.
> CNET, USAToday, RollingStone, and more...
> Click Here Now!  http://clickhere.egroups.com/click/314
> 
> 
> eGroups.com home: http://www.egroups.com/group/boost
> http://www.egroups.com - Simplifying group communications
> 
> 
> 
> 

------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/boost
http://www.egroups.com - Simplifying group communications