[comp.org.acm] Programming languages/different machines in contests

dwallach@soda.berkeley.edu (Dan Wallach) (05/13/91)

I competed quite a bit in high school, and I Turbo was the only way to go.
Since contests are always timed, the fast turnaround of Turbo is a godsend.

Most recently, I went to a contest where we were given a choice: Turbo C
or Pascal.  My group chose C and everybody was happy.  It seems to me that
if people really wanted Standard Pascal, that could be offered, too.  I
see no reason to have a contest with ONE language, ESPECIALLY something
icky like BASIC.

Especially since style doesn't really matter in these contests, there's no
reason not to offer both C and Pascal.  If judges can read one, they can
read the other (sans obfuscated C, of course...).  Turbo Pascal's string
libraries are similar to ANSI C's standard library functions, in terms of
utility, so there's not much of a conflict of "but the Pascal guy already
had the whole program written for him!"  I don't know how true this is
for Object Oriented Turbo Pascal.

I'd like to bring up a different topic with regard to contests: should
they always be on PC's?  The last contest I went to had a lab of PS/2
286 boxes using some horrendous networked disk thing which made saving
a modest executable a 2 minute adventure, especially in the final half
hour of the contest...

For H.S., most people have only been exposed to PC's, and that's about
all you can offer them.  For college contests, however, workstations
would be lots nicer, i.e., having GNU programming tools.  How could
you prevent security problems from arising, like 'talk'ing to a friend
at home and asking questions, or even ftp-ing solutions from elsewhere?

Thought for food,

Dan Wallach
dwallach@soda.berkeley.edu

rang@cs.wisc.edu (Anton Rang) (05/15/91)

In article <1991May12.213538.1544@agate.berkeley.edu> dwallach@soda.berkeley.edu (Dan Wallach) writes:
>I'd like to bring up a different topic with regard to contests: should
>they always be on PC's?

  I don't see any real reason why they should or shouldn't be, except
that a bunch of PCs is probably easiest to set up.  The first regional
contest I went to was on a VAX/VMS system, with a little menu-driven
system so if you didn't know VMS commands you weren't at a particular
disadvantage (though knowing the editor helped), and so that you could
only do editing, compiles, prints, and (electronic) submissions.  It
seemed to work better than the PC-based contests I've been at, simply
because of the electronic submission capabilities, and because
compiles and prints both ran at very reasonable speeds.  Also, it made
restricting the contest to the standard languages (Standard Pascal and
FORTRAN 77) trivial.

>For college contests, however, workstations would be lots nicer [ ... ]

  Yes, I think it would be possible to run a contest with some sort of
workstation (say, a net of diskless Sun 3/50s).  It'd be more work
than using PCs, though, and I suspect one can make an argument that
using PCs is more "fair" because more colleges have PCs than have any
particular kind of workstation....

	Anton
   
+---------------------------+------------------+-------------+----------------+
| Anton Rang (grad student) | rang@cs.wisc.edu | UW--Madison | "VMS Forever!" |
+---------------------------+------------------+-------------+----------------+

raymond@cs.vu.nl (Raymond Michiels) (05/15/91)

rang@cs.wisc.edu (Anton Rang) writes:

>In article <1991May12.213538.1544@agate.berkeley.edu> dwallach@soda.berkeley.edu (Dan Wallach) writes:
>>I'd like to bring up a different topic with regard to contests: should
>>they always be on PC's?

>  I don't see any real reason why they should or shouldn't be, except
>that a bunch of PCs is probably easiest to set up.

This year's finals were on AT&T WGS-6386 (386 based) workstations which were
connected to the jury's computer by a network. All submissions were done using
the network. This really works much better than running around with floppies.

>It seemed to work better than the PC-based contests I've been at, simply
>because of the electronic submission capabilities, and because
>compiles and prints both ran at very reasonable speeds.

Even though a 386 is often used as a PC, we too had the e-mail
capabilities, we each had our own printer and the compilation speed of a
25 MHz 386 is reasonable. (VERY reasonable, in my opinion.)
The biggest advantage, however, is that we didn't have to use MS-DOS.

>>For college contests, however, workstations would be lots nicer [ ... ]

>  Yes, I think it would be possible to run a contest with some sort of
>workstation (say, a net of diskless Sun 3/50s).

Just like the finals!

	-Raymond.


PS: Why my enthusiasm about the finals? We came in second!

objtch@extro.ucc.su.OZ.AU (Peter Goodall) (05/16/91)

Why not give the competitors one hour to set up the environment of their choice.I wouldn't touch pascal or c with the proverbial barge pole. So why should that
be a handicap in a competition, if it isn't in producing applications.

-- 
Peter Goodall - Smalltalk Systems Consultant - objtch@extro.ucc.su.oz.au
      ObjecTech Pty. Ltd. - Software Tools, Training, and Advice
162 Burns Bay Rd, LANE COVE, NSW, AUSTRALIA. - Phone/Fax: +61 2 418-7433