simple, yet powerful (was: Re: C++/C is a mess, but there is nothing better around)

Georg Fuellen fuellen at athena.mit.edu
Wed Jan 12 22:06:00 EST 1994


In article <2gs9fe$d2d at Mercury.mcs.com>, tcubed at MCS.COM (James Hanlon) writes:
|> [...] More than quality/productivity as
|> such, managers desire _control_. And latest_exotic_language_X, even if
|> it performs as advertised (not always the case), will certainly leave a manager
|> picking people from a smaller potential pool when it comes time to replace
|> the original developers. Or, the ultimate managerial nightmare: they are
|> held hostage by beards-and-sandals intellectuals, who force them to fork
|> over large sums of money, and embarrassing social concessions like 
|> dispensations from the corporate dress code. Better to do it in C, and
|> hire C programmers (or nominal C++ programmers, who use a little C++
|> syntax in their C-type programs)--you can always get a C programmer. Better,
|> if the C programmer in-house gets a little too self-important, you can
|> always get _another_ C programmer. Control, you see.
[...]
|> meuer (meuer at geom.umn.edu) :
|> : The marketplace is filled with superior products that don't get wide
|> : acceptance for non-technical reasons.  NeXTStep is a prime example.
|> : The sad truth is that often the solutions that are technically the
|> : best are not the best financially.  As a programmer, I get really
|> : frustrated by this.  But it is something I understand even when I wish
|> : the situation was otherwise.
|> 
|> Most frustrating is when the quality/productivity factors are 5X-10X (not
   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|> per cent, real multiples) in favor of the unfamiliar, and still the status quo
   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|> guys torpedo the change.
   ^^^^^^^^^^^^^^^^^^^^^^^^
The success of Mosaic/html in networking [see comp.infosystems.www]
may give some hope that it is at least possible for a superior product
to be successful and unify lots of inferior approaches. Indeed, doing
an ftp session via Mosaic is 10X as fast. Whether "simple, yet powerful"
is possible or desirable for programming languages is another question.

    (( A personal note:
    Well, at least I hate it to spend 50% of my time with 
    unnecessarily complicated gadgets, whether these are programming 
    languages, US bank accounts, health plans, or university 
    administrators. I wanna do research... But dont send me flame mail 
    (aka "Get a life, you moron") because it would just cost me more 
    time to sort 'em out ;-))

Sather [see comp.lang.sather] may be the language closest to the idea of 
"simple, yet powerful",
and the idealist hopes that the more ready-to-use classes are developed, the 
more people will use it. (Personally, I wonder whether another new trend, 
_e.g._  visual (graphical) languages like Prograph or Labview, [see the 
comp.lang.visual FAQ] will have found even more disciples by then... )

Anyway, this is the result regarding the email/follow-ups I got up to now
(note that the numbers are too small to be statistically significant; I'm 
happy they arent :-)

Ada        is favored by 5 people,
Eiffel      "            5   "
Sather      "            5   "
Objective-C "            4   "
C++         "            3   ".

Regards,
  georg
fuellen at mit.edu
The convex hull of all disclaimers made on usenet last year applies to this mess








More information about the Bio-soft mailing list