Starting Computer Languages...
Eddie Maldonado
am interested in learning a programming language but I'm not sure where to
start. I thought perhaps C++? I don't know, any insight or advice is
greatly appreciated. I plan to major in Computer Science @ the University
of Central Florida, in case you were wondering.
Eddie
manac...@mpinet.net
Greg Comeau
You can succeed with C++ as your first language. Check out some of
the book suggestions at the Comeau web site. C++ has a vast capability
in terms of a high level language. All in all, learn as many languages
as you can though. And don't forget to pace yourself.
Good luck.
- Greg
--
Comeau Computing, 91-34 120th Street, Richmond Hill, NY, 11418-3214
Producers of Comeau C/C++ 4.2.38 -- New Release! We now do Windows too.
Email: com...@comeaucomputing.com / Voice:718-945-0009 / Fax:718-441-2310
*** WEB: http://www.comeaucomputing.com ***
GREG TAYLOR
required to take as part of your major, get all the math you can under your
belt, including algebra, calculus, trig, etc., etc.,...
Eddie Maldonado wrote in message <74v80d$6tq$1...@remarQ.com>...
>I'm a senior in H.S. and I'm getting ready to go to college in the Fall. I
>am interested in learning a programming language but I'm not sure where to
>start. I thought perhaps C++? I don't know, any insight or advice is
>greatly appreciated. I plan to major in Computer Science @ the University
>of Central Florida, in case you were wondering.
>
>Eddie
>manac...@mpinet.net
>
>
Ben Pfaff
Personal opinion only here, but in addition to the languages you will be
required to take as part of your major, get all the math you can under your
belt, including algebra, calculus, trig, etc., etc.,...
I recommend studying linear (matrix) algebra and differential
equations, at the least. The first chapter (all 225 pages worth) of
Knuth's _Art of Computer Programming, Vol. 1_ is a good pointer to the
sort of math useful in computer science, as well.
--
(supporter of the campaign for grumpiness where grumpiness is due in c.l.c)
Please: do not email me copies of your posts to comp.lang.c
do not ask me C questions via email; post them instead
Will Rose
: "GREG TAYLOR" <greg...@akos.net> writes:
: Personal opinion only here, but in addition to the languages you will be
: required to take as part of your major, get all the math you can under your
: belt, including algebra, calculus, trig, etc., etc.,...
: I recommend studying linear (matrix) algebra and differential
: equations, at the least. The first chapter (all 225 pages worth) of
: Knuth's _Art of Computer Programming, Vol. 1_ is a good pointer to the
: sort of math useful in computer science, as well.
The use of computational geometry, such as matrices, seems to be on the
rise with 2D and 3D graphics. I've been scrabbling through a book on
the topic, and not getting very far; modern graphical toolkits are pretty
powerful, and will take you a long way, but when you want to mess with the
underlying primitives it helps to have a grip on what the math is doing.
Sunil Rao
>
> "GREG TAYLOR" <greg...@akos.net> writes:
>
> Personal opinion only here, but in addition to the languages you will be
> required to take as part of your major, get all the math you can under your
> belt, including algebra, calculus, trig, etc., etc.,...
>
> I recommend studying linear (matrix) algebra and differential
> equations, at the least. The first chapter (all 225 pages worth) of
> Knuth's _Art of Computer Programming, Vol. 1_ is a good pointer to the
> sort of math useful in computer science, as well.
A more suitable text for this purpose might be "Concrete Mathematics" by
Ronald Graham, Donald Knuth and Oren Patashnik. It's published by
Addison-wesly and is ISBN0201558025.
--
{ Sunil Rao }
"There is no scorn more profound, or on the whole more justifiable,
than that of the men who make for the men who explain."
-- HARDY, Godfrey Harold.
Ing. Franz Glaser
>
> I'm a senior in H.S. and I'm getting ready to go to college in the Fall. I
> am interested in learning a programming language but I'm not sure where to
> start. I thought perhaps C++? I don't know, any insight or advice is
> greatly appreciated. I plan to major in Computer Science @ the University
> of Central Florida, in case you were wondering.
An analogy:
If you first learn Latin, it is very easy to learn Italian, Spanish,
French, Roman and even English later - you have the basics.
If you learn Nashville-Tennessee - English first, what should be the
followup?
And Pascal is much more alive than Latin!
Regards, Franz Glaser
John E. Carty
Ing. Franz Glaser wrote in message <3673A1BB...@eunet.at>...
Ing. Franz Glaser
>
> Pascal? The guy wants a job afterwards, not just a title. :-0
>
> Ing. Franz Glaser wrote in message <3673A1BB...@eunet.at>...
> >An analogy:
> >If you first learn Latin, it is very easy to learn Italian, Spanish,
> >French, Roman and even English later - you have the basics.
> >
> >If you learn Nashville-Tennessee - English first, what should be the
> >followup?
> >
> >And Pascal is much more alive than Latin!
Correct. When I am looking for an employee, I am not interested in
the title. I test his knowledge, especially the basics.
I am very modest when looking for special knowledges with a new
employee. He/she can learn the specialties in my company, but I do
not want to educate the basics first. And the worst situation is,
when I find out later, that all his/her brilliance was not backed
by in depth knowledge.
But this might come from the fact that I am not selling "windy"
programs, where the ultimate of performance is a pretty and quickly
painted impressive picture on the screen. ;-)
Regards, Franz Glaser
Ben Pfaff
A more suitable text for this purpose might be "Concrete Mathematics" by
Ronald Graham, Donald Knuth and Oren Patashnik. It's published by
Addison-wesly and is ISBN0201558025.
Is that book worth buying when I already have TAoCP Vols. 1-3? I like
Knuth, but his books are expensive... Do you recommend it in general?
i.e., is it on the same level as TAoCP, or is it more of a `beginner's
book'?
Thanks in advance for any comments.
--
"You call this a *C* question? What the hell are you smoking?" --Kaz
Judson E. Knott
>
> Ing. Franz Glaser wrote in message <3673A1BB...@eunet.at>...
> >>
> >> I'm a senior in H.S. and I'm getting ready to go to college in the Fall.
> I
> >> am interested in learning a programming language but I'm not sure where
> to
> >> start. I thought perhaps C++? I don't know, any insight or advice is
> >> greatly appreciated. I plan to major in Computer Science @ the
> University
> >> of Central Florida, in case you were wondering.
> >
> >If you first learn Latin, it is very easy to learn Italian, Spanish,
> >French, Roman and even English later - you have the basics.
> >
> >If you learn Nashville-Tennessee - English first, what should be the
> >followup?
> >
> >And Pascal is much more alive than Latin!
> >
Well, having traveled this route a little time ago, I will tell you how
I got here (being at C++). I first traveled down the road of Pascal, yes,
its not very useful in the real world, but it is a good teaching language.
The I hoped to the C train on my own time, learning much of that, then
came college and I learned C++, which after this little journey wasn't at
all hard.
+--------------------+---------------------------------------+
| Judson E. Knott | http://www4.ncsu.edu/~jeknott/ |
| 222A Bragaw Hall | or http://knott.rh.ncsu.edu/~jeknott/ |
| NCSU Box 15298 | ICQ #: 17605659 |
| Raleigh, NC. 27607 | PHONE: (919) 512-9176 |
+--------------------+---------------------------------------+
Alex Yacoub
<tuOc2.1882$w91.9...@newsread1-mx.centuryinter.net>...
> Pascal? The guy wants a job afterwards, not just a title. :-0
Pascal is probably the best beginning language to learn, IMHO, because of
the fact that Pascal won't let you deviate from a strict structured
methodology. Unlike C/C++, it is almost impossible to use gotos, breaks,
returns, and other keywords to change the direction of your code. Pascal
requires you to use functions, loops, and if-thens to make decisions. (I'm
not saing that break and return are bad, but they must be used only in
moderation. Besides, a couple more if statements won't slow down your code
too much, unless you're writing a game or something.)
Case in point-- I'm in my second year of high school computer science.
Last year, our teacher taught us Pascal, and halfway through the year, I
broke off and started doing the projects in C++. This year, our teacher
has decided to teach the first year students C++ instead of Pascal. The
other day, I went into the lab after school, and it happened to be the day
before that class had a major program due. I looked at their code, and it
was terrible!!! I mean, there was absolutely no proper structure in it
whatsoever. (And this was just a simple program - take grades in as
integers, average them, and output the highest grade, using a menu) If
they had learned Pascal first, their habits would have been developed
correctly and they would have written much better programs (Of course, the
teacher is also slightly at fault, but let's not get into that).
Don't get me wrong, I would never go back to Pascal, but I'm glad I learned
it. Many major corporations also require a rudimentary knowledge of
Pascal. I know that IBM requires knowledge of Pascal before they hire you.
-Alex
mic...@trollope.org
"Eddie Maldonado" <manac...@mpinet.net> wrote:
> I'm a senior in H.S. and I'm getting ready to go to college in the Fall. I
> am interested in learning a programming language but I'm not sure where to
> start. I thought perhaps C++? I don't know, any insight or advice is
> greatly appreciated. I plan to major in Computer Science @ the University
> of Central Florida, in case you were wondering.
In all likelihood, your first language will be chosen for you by the school,
i.e., you'll be required to take some kind of "computer fundamentals" or
"introduction to computing" that will in fact be a programming course --
probably C++, maybe even C, but if you're lucky, you'll get Scheme.
Whatever way the school chooses, your best bet is to learn to <program> and
then the choice of language can be based on your choice of interest.
mp
mic...@trollope.org
http://www.trollope.org
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Sunil Rao
>
> Sunil Rao <suni...@ic.ac.uk> writes:
>
> A more suitable text for this purpose might be "Concrete Mathematics" by
> Ronald Graham, Donald Knuth and Oren Patashnik. It's published by
> Addison-wesly and is ISBN0201558025.
>
> Is that book worth buying when I already have TAoCP Vols. 1-3? I like
> Knuth, but his books are expensive... Do you recommend it in general?
I actually do like the book quite a lot - it certainly goes into more
detail. TAOCP, IMVHO, has less worked *mathematical* examples than it
really should.
> i.e., is it on the same level as TAoCP, or is it more of a `beginner's
> book'?
I'd reckon that it's at about the same level - but it's definitely
pitched more at mathematicians than computer scientists. I quote from
http://www-cs-staff.Stanford.EDU/~knuth/gkp.html - "...the authors
themselves rely heavily upon it". I haven't actually bought myself a
copy yet, but I plan to in the VNF. :)
Michael J. Tobler
says...
> John E. Carty <jec...@NOSPAMcenturyinter.net> wrote in article
> <tuOc2.1882$w91.9...@newsread1-mx.centuryinter.net>...
> > Pascal? The guy wants a job afterwards, not just a title. :-0
>
>
> Pascal is probably the best beginning language to learn, IMHO, because of
> the fact that Pascal won't let you deviate from a strict structured
> methodology. Unlike C/C++, it is almost impossible to use gotos, breaks,
> returns, and other keywords to change the direction of your code.
programming with an OO methodology. So, Pascal is out on this one - many
are using Java for this (and Java has NO goto :) I agree that Pascal is
good for teaching proper structured programming, but in the OO world, you
gotta switch paradigms.
> Pascal
> requires you to use functions, loops, and if-thens to make decisions.
Uhhhh - it doesnt REQUIRE me to, these are constructs that should be used
in any language.
> (I'm not saing that break and return are bad, but they must be used only in
> moderation. Besides, a couple more if statements won't slow down your code
> too much, unless you're writing a game or something.)
>
> Case in point-- I'm in my second year of high school computer science.
> Last year, our teacher taught us Pascal, and halfway through the year, I
> broke off and started doing the projects in C++. This year, our teacher
> has decided to teach the first year students C++ instead of Pascal. The
> other day, I went into the lab after school, and it happened to be the day
> before that class had a major program due. I looked at their code, and it
> was terrible!!! I mean, there was absolutely no proper structure in it
> whatsoever.
I have seen a LOT of Pascal code that looked like $]it, so dont blame it
on the language. It could be the teacher too, or many other things.
Granted, C/C++ are terse languages, but they also allow you to be very
expressive. Heck, I've seen people with 3-4 years of C++ experience write
crap code. It's all context and you and I are not sure where the problem
is with this "class" you're mentioning.
Not trying to be insultive, but a second year high school comp-sci studend
STILL has much to learn, so "grading" other work should be taken with a
grain of salt. I'll bet YOUR code might possibly NOT pass a code review
from me :)
> (And this was just a simple program - take grades in as
> integers, average them, and output the highest grade, using a menu) If
> they had learned Pascal first, their habits would have been developed
> correctly and they would have written much better programs (Of course, the
> teacher is also slightly at fault, but let's not get into that).
Ahhh HA ! Just as I suspected - the teacher!
> Don't get me wrong, I would never go back to Pascal, but I'm glad I learned
> it. Many major corporations also require a rudimentary knowledge of
> Pascal. I know that IBM requires knowledge of Pascal before they hire you.
Really, they never asked me that.
--
<<<<<<<<<< Blue Skies >>>>>>>>>>>
< Michael J. Tobler >
< mto...@no-spam-ibm.net >
< remove "no-spam-" when replying >
<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>
Michael J. Tobler
> Michael J. Tobler schrieb:
> >
> > Depends on what you're teaching. Some schools dive right into teaching
> > programming with an OO methodology. So, Pascal is out on this one - many
> > are using Java for this (and Java has NO goto :)
>
> of it in DOS Pascal and I never felt a lack. And the Windows
> version of Pascal, named Delphi, has migrated to the classes as
> defined with C++ in addition to the objects approach introduced
> by Turbo Pascal, and - how I feel - much more precise and logical.
Pascal is Pascal - a structure language; just as C is. Languages that
extend some parent language dont make the parent language OO. Just because
Delphi is OO (or is it object-based?) doesnt go backward to Pascal.
The language must support the OO features, or it is OO.
Jeffrey C. Dege
>In article <36744B69...@eunet.at>, meg-g...@eunet.at says...
>> Michael J. Tobler schrieb:
>> >
>> > Depends on what you're teaching. Some schools dive right into teaching
>> > programming with an OO methodology. So, Pascal is out on this one - many
>> > are using Java for this (and Java has NO goto :)
>>
>> Are you sure that Pascal lacks OOP methodology? I am doing a lot
>> of it in DOS Pascal and I never felt a lack. And the Windows
>> version of Pascal, named Delphi, has migrated to the classes as
>> defined with C++ in addition to the objects approach introduced
>> by Turbo Pascal, and - how I feel - much more precise and logical.
>
>Pascal is Pascal - a structure language; just as C is. Languages that
>extend some parent language dont make the parent language OO. Just because
>Delphi is OO (or is it object-based?) doesnt go backward to Pascal.
>
>The language must support the OO features, or it is OO.
Turbo Pascal is, in its later versions, is a fully OO language, but
then, Turbo Pascal is a proprietary extension to Pascal, just as
Delphi is, it doesn't come close to complying with the ISO standard.
--
What? Me .sig?
Michael J. Tobler
says...
[snip]
> >The language must support the OO features, or it is OO.
>
> Turbo Pascal is, in its later versions, is a fully OO language, but
> then, Turbo Pascal is a proprietary extension to Pascal, just as
> Delphi is, it doesn't come close to complying with the ISO standard.
I've fallen way away from Pascal and its derivatives. I thought the OO
Pascal is called Object-Pascal?
Eddie Maldonado
in the south, if not the best in the south. Like you said, they do start
with a certain language....and its Pascal. Suprise suprise. Any opinions
on that?
Eddie
manac...@mpinet.net
John E. Carty
basics at college and then leaving without ever getting a degree. In the
Computer Science fields too many employers feel that what's being taught is
out dated and doesn't apply to the 'real' world. Any opinions on that?
Eddie Maldonado wrote in message <751jph$9p3$1...@remarQ.com>...
John E. Carty
Martin Ambuhl
>
> I've heard from alot of people that UCF is one of the top comp sci schools
> in the south, if not the best in the south. Like you said, they do start
> with a certain language....and its Pascal. Suprise suprise. Any opinions
> on that?
The question pretty refutes its own premise.
Here is a nice list I stole from Ashraf El-Hamalawi. See if UCF is on
it:
E-mail Computer Science Schools' List
If you find any errors in this list, or additions or modifications,
please send me a message at one of my above addresses ( ie
aae...@cus.cam.ac.uk or a...@eng.cam.ac.uk) :
Computer Science E-mail List
Last Update 28 October 1994
USA
Alabama: Grad Info, CSD...@UA1VM.UA.EDU
Alabama at Birmingham: Grad Info, grad...@cis.uab.edu
American: Larry Medsker, med...@auvm.bitnet
Arizona: Wendy Swartz, w...@cs.arizona.edu
Arizona State: David Pheanis, phe...@asuvax.eas.asu.edu
Auburn: Kai H. Chang, kch...@eng.auburn.edu
Ball State: Clinton Fuelling, fuel...@bsu-cs.bsu.edu
Boston University: Grad Info, csd...@cs.bu.edu
Bowling Green: Grad Info, computer...@andy.bgsu.edu
Bradley: Kathie Strum, kat...@cs1.bradley.edu
Brandeis University: Myrna Fox, m...@chaos.cs.brandeis.edu
Brown: Grad Info, grad...@cs.brown.edu
Buffalo: Ellie Benzel, ben...@cs.buffalo.edu
California-Chico: Orlando Madrigal, O...@csuchico.EDU
California-Berkeley: Grad Info, gra...@hera.berkeley.edu
California-Davis: Grad Info, grad...@cs.ucdavis.edu
California-Irvine: Nancy Leveson, gcou...@ics.uci.edu
California-Los Angeles: Verra Morgan, ve...@cs.ucla.edu
California-Sacramento: Grad Info, grad...@csus.edu
California-San Diego: Grad Info, grad...@cs.ucsd.edu
California-Santa Barbara: Grad Info, grad-a...@cs.ucsb.edu
Carnegie Mellon: Martha Clarke, Martha...@cs.cmu.edu
Case Western Reserve: Lee White, le...@alpha.ces.cwru.edu
Chicago: Janos Simon, si...@gargoyle.uchcago.edu
Cincinnati: Grad Info, compsci....@uc.edu
Clemson: Grad Info, c...@cs.clemson.edu
Colorado State: Grad Info, csgra...@cs.colostate.EDU
Colorado, Boulder: Grad Info, csgra...@cs.colorado.edu
Columbia: Grad Info, grad...@cs.columbia.edu
Cornell: Grad Info, p...@cs.cornell.edu
Dartmouth: Grad Info, cs...@dartmouth.edu
Delaware: Kathleen McCoy, mc...@cis.udel.edu
Duke Sue Jarrell: grad-admissions@acpub,.duke.edu
Eastern Washington: Steve Simmons, rha...@ewuvms.bitnet
Emory: Grad Info, d...@mathcs.emory.edu
Florida: Ms. Marlene Hughes, mar...@cis.ufl.edu
Florida Atlantic: Grad Info, off...@CS.FAU.EDU
Florida State: Grad Info, in...@cs.fsu.edu
George Mason: Grad Info, gm...@gmuvax.bitnet
George Washington: Grad Info, EE...@seas.gwu.edu
Georgia: Thiab Taha, th...@csunl.cs.uga.edu
Georgia Tech: Grad Info, infore...@cc.gatech.edu
Harvard: Harry Lewis, le...@das.harvard.edu.
Houston: Grad Info, grad...@cs.uh.edu
Illinois-Chicago: Christopher Schejbal, ch...@bert.eecs.uic.edu
Illinois-Urbana: Grad Info, acad...@cs.uiuc.edu
Indiana: Grad Info, admis...@cs.indiana.edu
N. Iowa: Kevin O'Kane, ok...@cs.uni.edu--w
Iowa State: Grad Info, grad...@cs.iastate.edu
James Madison: Diane Spresser, FAC_SP...@JMUVAX1.bitnet
Johns Hopkins: Grad Info, admis...@cs.jhu.edu
Kentucky: Grad Info, gra...@ms.uky.edu
Kansas State: Grad Info, grad...@cis.ksu.edu
Kent State: Grad Info, grad...@mcs.kent.edu
Louisiana Tech: Grad Info, csi...@engr.latech.edu
MIT: N. Lyall, ly...@lcs.mit.edu
Maine: Grad Info, grad...@gandalf.umcs.maine.edu
Maryland - College Park: Grad Info, csgr...@cs.umd.edu
Maryland - Baltimore: Prof. C. Nicholas, grad...@cs.umbc.edu or :
Maryland - Baltimore: Prof. A. Norcio, ifsm-g...@umbc.edu
Massachusetts, Lowell: Charlie Steele, cha...@cs.ulowell.edu
Massachussets: Sharon Mallory, mal...@cs.umass.edu
Michigan: Dawn Freysinger, da...@eecs.umich.edu.
Michigan State: Grad Info, grad...@cps.msu.edu
Michigan Tech: Grad Info, csd...@cs.mtu.edu
Minnesota-Minneapolis: Grad Info, d...@cs.umn.edu
Minnesota-Duluth: Grad Info, c...@ub.d.umn.edu
Mississippi-starkville: Grad Info, off...@cs.msstate.edu
Missouri-Columbia: Gordon Springer, spri...@umvma.bitnet
Missouri-Rolla: Grad Info, csd...@cs.umr.edu
Montana State: Denbigh Starkey, grad...@cs.montana.edu
Nebraska: Joseph Leung, j...@cse.unl.edu
Nevada-Las Vegas: Roy Ogawa, r...@jimi.cs.unlv.edu
New Hampshire: Daniel Bergeron, r...@unh.edu
New Mexico: Bernard Moret, mo...@cmell.cs.unm.edu
New Mexico State: Grad Info, csof...@nmsu.edu
New Mexico Tech: Hamdy Soliman, h...@nmt.edu
New York: Cynthia Mazzant, maz...@cims.nyu.edu
N. Carolina State: Grad Info, te...@adm.csc.ncsu.edu
N. Carolina-Chapel Hill: Grad Info, ad...@cs.unc.edu
N. Carolina-Charlotte: Ken Chen, ch...@unccvax.uncc.edu
N. Dakota: Thomas O'Neil, UD12...@ndsuvm1.bitnet
Northeastern: Grad Info, grads...@corwin.ccs.northeastern
.edu
Northwestern: Grad Info, gr...@eecs.nwu.edu
Notre Dame: Grad Info, c...@cse.nd.edu
Ohio: J. Farrar, far...@ace.cs.ohiou.edu
Oregon: Betty Lockwood, be...@cs.uoregon.edu
Oregon Grad Institute: Grad Info, cse...@cse.ogi.edu
Oregon State: Bernadette Feyerherm, ber...@cs.orst.edu
Pacific Lutheran: Grad Info, cs...@plu.bitnet
Pennsylvania: Grad Info, cis-gra...@cis.upenn.edu
Pennsylvania State: Vicki Keller, vi...@cs.psu.edu
Pittsburgh: Loretta Shabatura, lor...@cs.pitt.edu
Portland State: Leonard Shapiro, l...@cs.pdx.edu
Princeton: Melissa Lawson, m...@cs.princeton.edu
Purdue: Grad Info, grad...@cs.purdue.edu
Rensselaer Poly. Inst:. Sandy Charette, char...@cs.rpi.edu
Rice: Grad Info, com...@owlnet.rice.edu
Rochester: Grad Info, admis...@cs.rochester.edu
Rutgers: Valentine Rolfe, ro...@cs.rutgers.edu
SUNY-Stony Brook: Grad Info, grad...@sbcs.sunysb.edu
South Alabama: David Feinstein, FC...@USOUTHAL.BITNET
South Carolina: Manton Matthews, matt...@cs.scarolina.edu
South Florida: Kate Johnson, k...@sol.csee.usf.edu
Southern California: Grad Info, csd...@pollux.usc.edu
Southwestern Louisiana: Michael Mulder, mul...@cacs.usl.edu
Stanford: Sara Merryman, merr...@cs.stanford.edu
Stephen F. Austin: Craig Wood, CW...@SFAUSTIN.BITNET
SUNY Binghamton: Sudhir Aggarwal, sud...@bingsuns.cc.binghamton.edu
SUNY Utica/Rome: Grad Info, c...@sct60a.sunyct.edu
Syracuse: Barbara Powers, ba...@top.cis.syr.edu
Temple: Jackie Harriz, har...@cis.temple.edu
Tennessee-Knoxville: David Straight, stra...@utkvx.utk.edu
Texas A&M: Grad Info, csd...@cs.tamu.edu
Texas Tech: James Archer, j...@cs.ttu.edu
Texas-Arlington: Bill Carroll, car...@evax.arl.utexas.edu
Texas-Austin: Grad Info, csa...@cs.utexas.edu
Tufts: Grad Info, mas...@cs.tufts.edu
Tulane: Fredrick Petry, pe...@cs.tulane.edu
Tulsa: R.L. Wainwright, rog...@tusun2.mcs.utulsa.edu
Utah: Grad Info, graduate-c...@cs.utah.edu
Utah State: Greg Jones, gr...@cs.usu.edu
Vanderbilt: Grad Info, cs...@vuse.vanderbilt.edu
Vermont: Grad Info, csgra...@uvm.edu
Villanova: Don Goelman, goe...@villanova.edu
Virginia: Grad Info, csadmi...@uvacs.cs.virginia.edu
Washington at St Louis: Jean Groethe, je...@wucs.wustl.edu
Washington-Seattle: Grad Info, grad-ad...@cs.washington.edu
West Virginia: Donald Butcher, d...@a.cs.wvu.wvnet.edu
Western Kentucky: John Crenshaw, cren...@wku.edu
Western Michigan: Grad Info, csa...@gw.wmich.edu
Western Washington: Gary Eerkes, eer...@wwu.edu
Wichita State: Mary Edgington, edgi...@wsuiar.wsu.ukans.edu
William and Mary: Vanessa Lynch, van...@cs.wm.edu
Wisconsin-Madison: Grad Info, admis...@cs.wisc.edu
Worcester Polytechnic: Grad Info, grad...@cs.wpi.edu
Wright State: Grad Info, csd...@wright.edu
Yale: Grad Info, grad...@cs.yale.edu
Australia
Adelaide: Andrew Wendelborn, and...@cs.adelaide.edu.au
Australian National: Brendan McKay, b...@cs.anu.edu.au
Canberra: Mary O'Kane, m...@ise.canberra.edu.au
Flinders: Chris Marlin, mar...@cs.flinders.edu.au
James Cook: Grad Info, com...@coral.cs.jcu.edu.au
Monash: Lloyd Alison, gra...@bruce.cs.monash.edu.au
Newcastle: Bryan Beresford-Smith, b...@cs.newcastle.edu.au
Queensland Brisbane: Ian Holmes, hol...@cs.uq.oz.au
Southern Queensland: M. McFarlane, mcfa...@zeus.usq.edu.au
Western Australia: J. Rohl, je...@cs.uwa.edu.au
Wollongong: John Fulcher, jo...@cs.uow.edu.au
Canada
Alberta: Grad Info, grad...@cs.ualberta.ca
British Columbia: Grad Info, grad...@cs.ubc.ca
Calgary: L. Storey, cp...@cpsc.ucalgary.ca
Carleton: Rosemary Carter, car...@scs.carleton.ca
Dalhousie: Gretchen Smith, gret...@cs.dal.ca
Manitoba: Lynne Romuld, rom...@cs.umanitoba.ca
New Brunswick: Dana Wasson, W...@UNB.CA
Prince Edward Island: Wayne Cutcliffe, cutc...@upei.ca
Saskatchewan: Mostafa Abd-El-Barr, mos...@cs.usask.ca
Simon Fraser: Grad Info, gra...@cs.sfu.ca
Toronto: Grad Info, gra...@cs.utoronto.ca
Victoria: Grad Info, csg...@csr.uvic.ca
Waterloo: Grad Info, csg...@uwaterloo.ca
Finland
Joensuu: Martti Penttonen, Pent...@cs.joensuu.fi
Jyv{skyl{: Kalle Lyytinen, Ka...@jytko.jyu.fi
Oulu: Samuli Saukkonen, Sam...@tolsun.oulu.fi
#Helsinki: Patrik Floreen, flo...@cs.Helsinki.fi
Vaasa: Mats Jakobsson, m...@uwasa.fi
Turku: Timo J{rvi, ja...@kontu.cc.utu.fi
Tampere: Pertti J{rvinen, p...@cs.uta.fi
Kuopio: Jussi Parkkinen, park...@cs.uku.fi
Abo Akademi: Aimo T|rn, at...@finabo.abo.fi
Helsinki U. Tech.: Martti M{ntyl{, m...@cs.hut.fi
Tampere U. tech: Reijo Kurki-Suonio, r...@tut.fi
Lappeenranta U. tech: Erkki Oja, o...@ltkka.lut.fi
New Zealand
Auckland: Bob Doran, b...@cs.aukuni.ac.nz
Canterbury: Bruce McKenzie, h...@cocs.canterbury.ac.nz
Lincoln: A. Mckinnon, a.mck...@lincoln.ac.nz
Massey: John Hudson, J.Hu...@massey.ac.nz
Otago: Anthony Robins, cos...@otago.ac.nz
Waikato: Ian Witten, i...@waikato.ac.nz
Victoria: Grad Info, Grad.En...@comp.vuw.ac.nz
Ireland
Dublin City: Alan Smeaton, SME...@dcu.ie
Trinity: Secretary, ion...@cs.tcd.ie
Hong Kong
Hong Kong: Grad Info, enq...@csd.hku.hk
Chinese U. of Hong Kong: Grad Info, grad...@cs.cuhk.hk
Hong Kong (HKUST:) Grad Info, csd...@uxmail.ust.hk
Norway
Oslo: Rolf Bjerknes, ro...@ifi.uio.no
For any changes or updates to the above list, please contact me at
aae...@cus.cam.ac.uk
Ashraf El-Hamalawi
Engineering Department,
University of Cambridge,
Trumpington Street,
Cambridge CB2 1PZ
ENGLAND, UK
--
Martin Ambuhl (mam...@earthlink.net)
Note: mam...@tiac.net will soon be inactive
David Masters
> > Pascal? The guy wants a job afterwards, not just a title. :-0
Here at the American University of Paris we have recently gone through
this discussion about ITL (initial teaching language). All universities go
through it every four or five years.
Pascal is excellent because:
- It has, as a correspondent to this thread has said, strong support
for structured programming. C/C++ doesn't.
- It performs rigorous runtime checking: index errors, pointer errors
and undefined variables are all detected in ISO compliant compilers. C/C++
doesn't.
What we decided to do, after consultation in this group and others, was
use Java as ITL. It is a good, but safe (it does most, not all, of
Pascal's checks), introduction to C++. So far it's going well and is
popular with students. It's harder than I originally expected though.
I received much email on this topic and if anyone wants to see it then
please email me. I'd be delighted to share.
David Masters
Remove the obvious from: dmas...@ac.theobvious.aup.fr
Eddie Maldonado
of graduate schools. I don't exactly need a directory of graduate schools.
Doesn't impress me nor does it prove any vague point you are attempting to
make here. UCF, for one, stands alone in the south. Do the real research
and you'll find not even GT can match up to the research facilities at UCF.
And another point, you have Florida State University on this list. Talk
about a weak program. Next time you decide to show off, think twice.
Oh yeah and one other thing:
>Computer Science E-mail List
>Last Update 28 October 1994
Try finding something more up-to date.
Eddie
manac...@mpinet.net
Greg Comeau
>Eddie Maldonado wrote:
>>
>> I'm a senior in H.S. and I'm getting ready to go to college in the Fall. I
>> am interested in learning a programming language but I'm not sure where to
>> start. I thought perhaps C++? I don't know, any insight or advice is
>> greatly appreciated. I plan to major in Computer Science @ the University
>> of Central Florida, in case you were wondering.
>
>If you first learn Latin, it is very easy to learn Italian, Spanish,
>French, Roman and even English later - you have the basics.
>
>If you learn Nashville-Tennessee - English first, what should be the
>followup?
>
>And Pascal is much more alive than Latin!
This is all one way to look at it. Another is that language is not
just grammar and words but also culture, linguistics, idioms, even
techniques (say public speaking), etc.
I mention this because your implication is that if you learn Latin first,
then you're on easy street. But this imlies Latin in and of itself will
do that. And that is just not the case. Yes, the roots of Latin WILL BE
important. That cannot be discarded. But too, it's just not that in
isolation. For instance, obtaining proper educational sources is
essential. That said and done, one can go pretty far with
Nashville-Tennessee-English.
Greg Comeau
>I got here (being at C++). I first traveled down the road of Pascal, yes,
>its not very useful in the real world, but it is a good teaching language.
>The I hoped to the C train on my own time, learning much of that, then
>came college and I learned C++, which after this little journey wasn't at
>all hard.
Sigh. To say that Pascal is a good teaching language is mythology IMO.
I say this because it just solemely rules out so many things.
Note my saying this is NOT an attack on Pascal. But a statement that
when one is armed with proper educations mediums, the sky is the only limit.
IOWs, you're implying that say C or C++ are not good teaching languages
and that is simply not so.
Michael J. Tobler
> I don't know what you are trying to prove with this list of e-mail addresses
> of graduate schools. I don't exactly need a directory of graduate schools.
> Doesn't impress me nor does it prove any vague point you are attempting to
> make here. UCF, for one, stands alone in the south. Do the real research
What's UCF?
Greg Comeau
>John E. Carty <jec...@NOSPAMcenturyinter.net> wrote in article
><tuOc2.1882$w91.9...@newsread1-mx.centuryinter.net>...
>
>the fact that Pascal won't let you deviate from a strict structured
>methodology. Unlike C/C++, it is almost impossible to use gotos, breaks,
>requires you to use functions, loops, and if-thens to make decisions. (I'm
>not saing that break and return are bad, but they must be used only in
>moderation. Besides, a couple more if statements won't slow down your code
>too much, unless you're writing a game or something.)
If deviating from a strict unstructured methodology is "bad",
then the methodology is as important as the choice of language.
You just don't jump into Pascal either. As with anything, it needs
to be incrementally learned.
>Case in point-- I'm in my second year of high school computer science.
>Last year, our teacher taught us Pascal, and halfway through the year, I
>broke off and started doing the projects in C++. This year, our teacher
>has decided to teach the first year students C++ instead of Pascal. The
>other day, I went into the lab after school, and it happened to be the day
>before that class had a major program due. I looked at their code, and it
>was terrible!!! I mean, there was absolutely no proper structure in it
>whatsoever. (And this was just a simple program - take grades in as
>integers, average them, and output the highest grade, using a menu) If
>they had learned Pascal first, their habits would have been developed
>correctly and they would have written much better programs (Of course, the
>teacher is also slightly at fault, but let's not get into that).
Ah, but this its home on my previous posts on this matter.
You seem to be blaming C++ for the lack of structure, yet don't even
really want to fault the teacher. Indeed, proper education mediums are
of the utmost important. Even in pascal. When all is said and done,
the language does not a program write, people do. Writing a program
in Pascal is not a prescription of perfection.
Furthermore, when the education mechanism invokes real people and not
just books, then the teachers have duties and responsibilities they
need to uphold, and often don't.
>Don't get me wrong, I would never go back to Pascal, but I'm glad I learned
>it. Many major corporations also require a rudimentary knowledge of
>Pascal. I know that IBM requires knowledge of Pascal before they hire you.
That borders on being somewhat silly.
I think good programmers should know as many languages as possible.
If that includes Pascal, good for them.
If it doesn't, then they should look at it.
Greg Comeau
>Eddie Maldonado wrote in message <751jph$9p3$1...@remarQ.com>...
>>in the south, if not the best in the south. Like you said, they do start
>>with a certain language....and its Pascal. Suprise suprise. Any opinions
>>on that?
>>
>basics at college and then leaving without ever getting a degree. In the
>Computer Science fields too many employers feel that what's being taught is
>out dated and doesn't apply to the 'real' world. Any opinions on that?
There are problems. But then again, you are somewhat/completely implying
that this is only a CS problem and not a general academic issue.
Greg Comeau
>
>> > Pascal? The guy wants a job afterwards, not just a title. :-0
>
>this discussion about ITL (initial teaching language). All universities go
>through it every four or five years.
>
>Pascal is excellent because:
> - It has, as a correspondent to this thread has said, strong support
>for structured programming. C/C++ doesn't.
Sure they do. This is simply not based upon fact.
Or is there some hidden meaning in your use of the word "strong"?
> - It performs rigorous runtime checking: index errors, pointer errors
>and undefined variables are all detected in ISO compliant compilers. C/C++
>doesn't.
This is partially fact.
>What we decided to do, after consultation in this group and others, was
>use Java as ITL. It is a good, but safe (it does most, not all, of
>Pascal's checks), introduction to C++. So far it's going well and is
>popular with students. It's harder than I originally expected though.
So, at what point do you unleash the students into the cruel streets of the
programming world? And upon what premise is that made valid and justified
at that point?
Martin Harvey
If I had to list starter languages in order of preference, I'd go as
follows:
Imperative (Best to Worst)
Modula-2
Pascal
Modula-3
Oberon
Java
C
C++
In particular OO languages may be a bit of a handful at first, and C++
still manages to confuse the living daylights out of some very
experienced programmers... Learn it perhaps as a third or fourth
language..
Functional languages.
Most of these are not used so much in the industry.. but learning a
functional language is very useful... they let you learn more about
formal methods for verifying programs, as well as showing you that there
are two ways to skin a cat.
ML (Based on polymorphically based versions of the lambda calculus)
Prolog
Lisp
MH.
Eddie Maldonado wrote:
>
> I'm a senior in H.S. and I'm getting ready to go to college in the Fall. I
> am interested in learning a programming language but I'm not sure where to
> start. I thought perhaps C++? I don't know, any insight or advice is
> greatly appreciated. I plan to major in Computer Science @ the University
> of Central Florida, in case you were wondering.
>
> Eddie
> manac...@mpinet.net
--
Martin Harvey.
Totally rewritten web pages at:
http://www.harvey27.demon.co.uk/mch24/
"ALGOL 60 was a language so far ahead of its time that it
was not only an improvement on its predecessors but also
on nearly all its successors". C.A.R. Hoare
--------------BEGIN GEEK CODE BLOCK--------------
Version: 3.12
GCS/CC d(+) s-:- a-- C+++$ UL@ P L@>++ E- W++
N+++ o-- K++ w+++$ O--- M-- V-- PS@ Y-- PGP-
t--- 5-- X-- R-- !tv b+ DI+ D+ G e++ h- r z++>---
---------------END GEEK CODE BLOCK---------------
Martin Harvey
>
> Personal opinion only here, but in addition to the languages you will be
> required to take as part of your major, get all the math you can under your
> belt, including algebra, calculus, trig, etc., etc.,...
>
Additional opinion... but if you want to get a *real* grounding in
Computing Math, you are very good at Math on the whole, and you've
started to learn a functional language was well as imperative languages,
try learning Lambda Calculus, and the Church-Rosser theorem :-)
Denotational Semantics is also useful, but I seem to remember finding it
a bit difficult!
MH.
Martin Harvey
>
> A more suitable text for this purpose might be "Concrete Mathematics" by
> Ronald Graham, Donald Knuth and Oren Patashnik. It's published by
> Addison-wesly and is ISBN0201558025.
I used that and found it fairly useful. Another good book is
"Introduction to Algorithms" by Cormen, Leiserson & Rivest.
Unfortunately I don't have my copy here, so I can't give you the ISBN.
Martin Harvey
>
> Pascal? The guy wants a job afterwards, not just a title. :-0
>
John,
I first learnt to program in Modula-2 (v-similar to pascal) at the age
of 12. I then learnt Modula-3 (which is OO modula-2 (ish)) at the age of
18, Delphi (an OO pascal) at 19, Various assemblers, Prolog, ML, and
loads of other theory and practical stuff in my university Computer
science course by the age of 21, and by the time I'd gotten round to
being employed, where I had to learn C, I managed to hack my way through
K&R in the space of about 4 days with no difficulty at all. After a
month in the Job, I started writing stuff in MS Visual C++ .. something
that is not trivial at the best of times, even more so when you're
modifying a large complicated app that is part of the company's core
product set, and after doing that for a few months, I was happy with
most aspects of the language (although the more subtle bits of C++ and
MFC still get a bit scary even today!...). Pascal gives you an exceppent
grounding. Anyone who knows pascal should be able to be prefectly
competent at C in under a month.
Martin Harvey
>
> Well, having traveled this route a little time ago, I will tell you how
> its not very useful in the real world, but it is a good teaching language.
> The I hoped to the C train on my own time, learning much of that, then
> came college and I learned C++, which after this little journey wasn't at
> all hard.
>
You seem to have travelled a similar road to me. However, I must
disagree that pascal isn't useful in the real world. I'm currently doing
a lot of Delphi stuff under Win32, and I have no problem doing stuff at
all levels.... admittedly I have to write device drivers in C... but
that's just due to the functionality of current impelementations.
There's nothing to stop one using something like GNU Pascal, and writing
virtual an entire OS from the ground up in pascal (apart from the bits
that always have to written in ASM).
Martin Harvey
>
> Uhhhh - it doesnt REQUIRE me to, these are constructs that should be used
> in any language.
No, but it encourages it.
To be honest, the best advantage of Pascal over C is that it has a
strict typing system and doesn't allow macros. These two things make all
the difference in my eyes.... (not least the fact that the lack of
#include means that things build at least 5 times as fast... a boon in
large projects!).
>
> I have seen a LOT of Pascal code that looked like $]it, so dont blame it
> on the language. It could be the teacher too, or many other things.
True. You can write both good and bad code in almost any language (apart
from some ultra anal functional languages where it's impossible to write
bad code that works!).
> Granted, C/C++ are terse languages, but they also allow you to be very
> expressive.
I find C++ a little *too* expressive... actually, perhaps I'd better
rephrase that... I don't mind expressive languages, provided they are
consistent. Just think about how many dirty ways the compiler allows you
to copy an object in C++, and the different semantics that these copies
have....
> Heck, I've seen people with 3-4 years of C++ experience write
> crap code. It's all context and you and I are not sure where the problem
> is with this "class" you're mentioning.
>
> Not trying to be insultive, but a second year high school comp-sci studend
> STILL has much to learn, so "grading" other work should be taken with a
> grain of salt. I'll bet YOUR code might possibly NOT pass a code review
> from me :)
Maybe not... but I've seen a fair number of experienced programmers
throw good style out of the window... although some might say that if
they know the rules back to front that might mean that they're allowed
to break them!
> > (And this was just a simple program - take grades in as
> > integers, average them, and output the highest grade, using a menu) If
> > they had learned Pascal first, their habits would have been developed
> > correctly and they would have written much better programs (Of course, the
> > teacher is also slightly at fault, but let's not get into that).
> Ahhh HA ! Just as I suspected - the teacher!
>
> > Don't get me wrong, I would never go back to Pascal, but I'm glad I learned
> > it. Many major corporations also require a rudimentary knowledge of
> > Pascal. I know that IBM requires knowledge of Pascal before they hire you.
>
> Really, they never asked me that.
Me neither... In fact they seemed to actively discourage applicants with
a *pure* comp-sci bias... (at least in this part of the world).
Martin Harvey
>
> i.e., you'll be required to take some kind of "computer fundamentals" or
> "introduction to computing" that will in fact be a programming course
I remember mine being ML. It was a great leveller. It's the kind of
language where those who already know C, C++ or Pascal actually have a
*disadvantage*. Learning a pure functional language (not imperative) to
start with is great... it's just a completely different paradigm... no
statements, no assignments, no variables, no flow of execution..
guaranteed to really piss off the students who already know C :-)
> Whatever way the school chooses, your best bet is to learn to <program> and
> then the choice of language can be based on your choice of interest.
Definitely. In addition, I'd mention that having at least 4 or 5
languages under your belt, preferably differing as much as possible, is
no bad thing.
Michael Powe
Hash: SHA1
>>>>> "Eddie" == Eddie Maldonado <manac...@mpinet.net> writes:
Eddie> I've heard from alot of people that UCF is one of the top
Eddie> comp sci schools in the south, if not the best in the
Eddie> south. Like you said, they do start with a certain
Eddie> language....and its Pascal. Suprise suprise. Any opinions
Eddie> on that?
I don't, particularly. A lot is going to depend on the combination of
student motivation and instructor talent. I just finished a term of
C++ with a mediocre instructor -- a nice guy, day job developing
embedded systems at a local electronics mfg firm & teaching at night;
but just basically out of it while in front of the blackboard.
Motivation was all that got me through the course.
I don't buy into all that black magic about how you should learn C++,
the whole C++ and nothing but the C++, so help you God. I'm studying
it because it's required to get into the interesting courses -- C,
design, theory. But while I'm at it, I will <learn> it.
mp
- --
Michael Powe
mic...@trollope.org http://www.trollope.org
Portland, Oregon USA
-----BEGIN PGP SIGNATURE-----
Version: PGP for Personal Privacy 5.0
Charset: noconv
iQA/AwUBNnS3WLajuNi/6Js3EQJsAQCfYxYFtLSpXwO+WV5B+7uItc099PQAnjM8
JzsLbtK4dGYyLKDTgnyj1CNX
=Jfa7
-----END PGP SIGNATURE-----
Ing. Franz Glaser
>
> Depends on what you're teaching. Some schools dive right into teaching
> programming with an OO methodology. So, Pascal is out on this one - many
> are using Java for this (and Java has NO goto :)
Are you sure that Pascal lacks OOP methodology? I am doing a lot
of it in DOS Pascal and I never felt a lack. And the Windows
version of Pascal, named Delphi, has migrated to the classes as
defined with C++ in addition to the objects approach introduced
by Turbo Pascal, and - how I feel - much more precise and logical.
Regards, Franz Glaser
Ing. Franz Glaser
>
> >Pascal is Pascal - a structure language; just as C is. Languages that
> >extend some parent language dont make the parent language OO. Just because
> >Delphi is OO (or is it object-based?) doesnt go backward to Pascal.
> >
>
> Turbo Pascal is, in its later versions, is a fully OO language, but
> then, Turbo Pascal is a proprietary extension to Pascal, just as
> Delphi is, it doesn't come close to complying with the ISO standard.
When ALGOL was improved to Pascal by Prof. Wirth there was no need for
GotoXY or ClrScr, because the Teletype with 110bd did not support it.
It did not support VGA graphics too.
Of course a language/compiler/tool must be extended to the newer
environments, regardless of any academic standard.
Franz Glaser
Judson E. Knott
> >its not very useful in the real world, but it is a good teaching language.
> >The I hoped to the C train on my own time, learning much of that, then
> >came college and I learned C++, which after this little journey wasn't at
> >all hard.
>
> I say this because it just solemely rules out so many things.
> Note my saying this is NOT an attack on Pascal. But a statement that
> when one is armed with proper educations mediums, the sky is the only limit.
> IOWs, you're implying that say C or C++ are not good teaching languages
> and that is simply not so.
Actually, in most places you look it says that Pascal is a good
teaching language, or should we really start with <a
href=http://www.cwi.nl/~steven/abc/">ABC</a>? And, i don't imply that C
and C++ are not good teaching languages either...just that as with all
languages it requires a planning phase on how to teach it
effectively...i.e. lecture for 2 hrs a week then have the students do 3
hrs of labs (yes, this is what my school did), which to me, and many
other students was just a complete waste (me, i got my A), others just
completely did not get it, and failed horibly.
The most effective way, that I have been through, is to let the
students work at their own pace and when they get stuck the teacher
explains the concept to them, and the students keeps going. Now, why is
this better? Well, the student works at their pace, and not at a pace
they cannot keep up with, and can play around with the code to see what
happens when they do what. (Who hasn't had that bastard teacher?)
-- Besides, this whole thread is a matter of personal preference...and
will never fully satisfy everone.
wal...@my-dejanews.com
"Eddie Maldonado" <manac...@mpinet.net> wrote:
> I've heard from alot of people that UCF is one of the top comp sci schools
> in the south, if not the best in the south. Like you said, they do start
> with a certain language....and its Pascal. Suprise suprise. Any opinions
> on that?
>
> Eddie
> manac...@mpinet.net
>
>
But you have nine months until you start there. You can start learning
today. And then, whether you start with Pascal or C or C++ or Java or
whatever, you'll be a leg up on some other students who won't have *any*
programming knowledge.
Just a thought, but I always look for those who've done more than what was
required. Why hire someone that only does what you tell him? (Note: I've
never been in a position to hire programmers, so this may not appy there.)
--
----------¤
WARNING: Anti-spam fanatic. Keep your ads to yourself!!!
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
wal...@my-dejanews.com
"Eddie Maldonado" <manac...@mpinet.net> wrote:
> I don't know what you are trying to prove with this list of e-mail addresses
> of graduate schools. I don't exactly need a directory of graduate schools.
> Doesn't impress me nor does it prove any vague point you are attempting to
> make here. UCF, for one, stands alone in the south. Do the real research
> And another point, you have Florida State University on this list. Talk
> about a weak program. Next time you decide to show off, think twice.
>
> Oh yeah and one other thing:
>
> >Last Update 28 October 1994
Are we rude or what? You asked for some advice. He's giving it to you.
Don't get upset because you don't like his advice. You are free to follow or
ignore it.
You're about to go to college. It is time to grow up.
> Eddie
> manac...@mpinet.net
Ing. Franz Glaser
> >If you first learn Latin, it is very easy to learn Italian, Spanish,
> >French, Roman and even English later - you have the basics.
> >
> This is all one way to look at it. Another is that language is not
> just grammar and words but also culture, linguistics, idioms, even
> techniques (say public speaking), etc.
Yes, correct. What you do not see in the various C and C++
newsgroups is the fact that I am "famous" here in the Pascal
newsgroups to claim the "beauty" and readability of the
source code. And this is why I sometimes catch flaming from
the multilingual posters to prise Pascal too much for its
readable source code and to neglect the efficiency issues.
So I did not feel that it were necessary to emphasize that
explicitly.
Regards, Franz Glaser
Nathaniel Thomas Cowan
all (in Princeton's review, US News and World Report, or Money Mag).
Not saying it doesn't have a good program, but I will have to say Rice
is a great southern school, especially for engineering and comp sci. Of
course, I'm biased in that respect :) Rice was ranked in all of the
above mentioned mentioned reviews (#18 nation wide in Princeton, and #2
as best buy in the US in US News (or money mag, I forget)). Rice has a
great reputation, especially here in Texas and in the south and provides
a great education. They also have the second largest endowment in the
US, so the facilities are outstanding. Also home of last year's nobel
prize winner in Chemistry, Dick Smalley for his Buckey Ball.
But, to get to the point of starting languages, Rice starts off with
Scheme for a semester then jumps into C++ (or JAVA). But I had no
experience in programming before college and I did just fine learning
C++ and OO methodology. If you're interested in a comp sci degree at a
southern school, check out Rice's program.
Nathan
David M. Cook
>You can succeed with C++ as your first language.
A beginner might be able to struggle thru it. But the language is so crufty
that I believe it would obscure many basic principles. I could imagine even
a bright student getting frustrated. I'd hate to have to teach an
introductory programming class with C++.
I've heard, though, that C++ is now used for Advanced Placement classes.
Dave Cook
David M. Cook
>So, at what point do you unleash the students into the cruel streets of the
>programming world? And upon what premise is that made valid and justified
>at that point?
I think the idea is that students shouldn't be dumped in the deep end first.
If you start with C you have to spend a lot of time with side issues. When
you start with C++ the side issues are greatly compounded.
Unfortunately, I can't think of any language that has both pedagogical
strength and the market share of C or C++ (I don't know VB, but would be
wary of teaching any proprietary/non-portable language as a first language).
I think if you're realistic you either spend a lot of time ramping students
up where they can begin to learn fundamental CS concepts with C, or you use a
different language and leave C for a later course.
Dave Cook
David M. Cook
<mar...@aziraphale.demon.co.uk> wrote:
>To be honest, the best advantage of Pascal over C is that it has a
>strict typing system and doesn't allow macros.
Pascal also has true pass by reference.
Dave Cook
Sunil Rao
>
> mic...@trollope.org wrote:
> >
> > i.e., you'll be required to take some kind of "computer fundamentals" or
> > "introduction to computing" that will in fact be a programming course
> > probably C++, maybe even C, but if you're lucky, you'll get Scheme.
>
> I remember mine being ML. It was a great leveller. It's the kind of
> language where those who already know C, C++ or Pascal actually have a
> *disadvantage*. Learning a pure functional language (not imperative) to
> start with is great... it's just a completely different paradigm... no
> statements, no assignments, no variables, no flow of execution..
> guaranteed to really piss off the students who already know C :-)
It does depend a hell of a lot on the background of the students in
general. A lot of students entering CS at a *good* Uni typically have
some prior programming experience and are at least familiar with basic
ideas such as loops (or are extremely competent mathematicians) and they
tend to find the functional model easy to understand, if hard to see the
point of. The procedural model is definitely easier to grasp if you've
never programmed before.
--
{ Sunil Rao }
"There is no scorn more profound, or on the whole more justifiable,
than that of the men who make for the men who explain."
-- HARDY, Godfrey Harold.
nick.k...@gecm.com
Martin Harvey <mar...@aziraphale.demon.co.uk> wrote:
<snip>
>
> To be honest, the best advantage of Pascal over C is that it has a
> strict typing system and doesn't allow macros. These two things make all
> the difference in my eyes.... (not least the fact that the lack of
> #include means that things build at least 5 times as fast... a boon in
> large projects!).
>
I hope you meant to put a smiley there. How big is "large" to you? I don't
think the 100,000 line application I'm working on would have faster build
times if it were all in one file!
The lack of facilities for separate compilation is one of the major things
that makes Pascal a toy language. Without, non-standard, extensions it is
unsuitable for real work.
<snip>
raw...@my-dejanews.com
nick.k...@gecm.com wrote:
> In article <36745183...@aziraphale.demon.co.uk>,
> Martin Harvey <mar...@aziraphale.demon.co.uk> wrote:
>
> <snip>
> >
> > To be honest, the best advantage of Pascal over C is that it has a
> > strict typing system and doesn't allow macros. These two things make all
> > the difference in my eyes.... (not least the fact that the lack of
> > #include means that things build at least 5 times as fast... a boon in
> > large projects!).
> >
>
> I hope you meant to put a smiley there. How big is "large" to you? I don't
> think the 100,000 line application I'm working on would have faster build
> times if it were all in one file!
>
> The lack of facilities for separate compilation is one of the major things
> that makes Pascal a toy language. Without, non-standard, extensions it is
> unsuitable for real work.
I have never used an implementation of Pascal that didn't allow separate
compilation (most commonly, by allowing function or procedure headers
to be marked external in the same way that the standard allows them to
be marked forward). It's not standard, perhaps for the same reasons that
#include is said to give undefined behavior with non-standard headers;
while this is unfortunate, you will find that some sort of non-standard
extension is needed for real work in C, too.
Generally, comparable implementations of separate compilation I have
used have been faster with Pascal than C because the compiler doesn't
have to rescan header files again and again; the worst instances for C
are header files that are included repeatedly and surrounded by the usual
#ifndef THIS_HEADER
#define THIS_HEADER
...
#endif
where the preprocessor, at least, looks at every line of the header
each time.
--
MJSR
Greg Comeau
>Judson E. Knott wrote:
>>
>> Well, having traveled this route a little time ago, I will tell you how
>> its not very useful in the real world, but it is a good teaching language.
>> The I hoped to the C train on my own time, learning much of that, then
>> came college and I learned C++, which after this little journey wasn't at
>> all hard.
>>
>
>disagree that pascal isn't useful in the real world. I'm currently doing
>a lot of Delphi stuff under Win32, and I have no problem doing stuff at
>all levels.... admittedly I have to write device drivers in C... but
>that's just due to the functionality of current impelementations.
>
>There's nothing to stop one using something like GNU Pascal, and writing
>virtual an entire OS from the ground up in pascal (apart from the bits
>that always have to written in ASM).
This is great, and I agree with you in part, however, is simply does not
address the issue that was raised. To wit, for instance, Delphi is
not pascal. It is just not pascal. So, you cannot disagree, at least
with the argument you gave, because you haven't used pascal in the real work,
you've used delphi. And it was Pascal that was the issue being raised
on the table. You may feel that Delphi has some nice extensions to it,
but that's a seperate issue.
Greg Comeau
> Eddie> south. Like you said, they do start with a certain
> Eddie> language....and its Pascal. Suprise suprise. Any opinions
> Eddie> on that?
>
>I don't, particularly. A lot is going to depend on the combination of
>student motivation and instructor talent.
Indeed, like I've mentioned in other posts, the educational medium used
is essential.
> I just finished a term of
>C++ with a mediocre instructor -- a nice guy, day job developing
>embedded systems at a local electronics mfg firm & teaching at night;
>but just basically out of it while in front of the blackboard.
>Motivation was all that got me through the course.
Then it sounds like that education medium was poor then. If it wasn't the
teacher's fault, then it was the curriculum's fault.
>I don't buy into all that black magic about how you should learn C++,
>the whole C++ and nothing but the C++, so help you God.
I'm a C++ enthusiast, but anybody telling you that is a fraud.
They should be told so and they should be challenged.
>I'm studying
>it because it's required to get into the interesting courses -- C,
>design, theory. But while I'm at it, I will <learn> it.
You have a good attitude. BTW, I have gone into many things that I did not
care for, only to get some AHAs! when other pieces to it became clear. Anyway,
your goal is to learn programming, techniques, and all the interrelated
skills implied in that. The language used does "interfere" with that,
but too, lots or parts can remain neutral as you hone yourself.
Greg Comeau
>In article <751jph$9p3$1...@remarQ.com>,
> "Eddie Maldonado" <manac...@mpinet.net> wrote:
>> I've heard from alot of people that UCF is one of the top comp sci schools
>> in the south, if not the best in the south. Like you said, they do start
>> with a certain language....and its Pascal. Suprise suprise. Any opinions
>> on that?
>>
>> Eddie
>> manac...@mpinet.net
>>
>>
>
>But you have nine months until you start there. You can start learning
>today. And then, whether you start with Pascal or C or C++ or Java or
>whatever, you'll be a leg up on some other students who won't have *any*
>programming knowledge.
Something like that. Just make sure you get good books and such.
>Just a thought, but I always look for those who've done more than what was
>required. Why hire someone that only does what you tell him? (Note: I've
>never been in a position to hire programmers, so this may not appy there.)
Sure, it's good to have people who don't need hand-holding on everything,
and to even take the initiative when necessary. But the other extreme is
sometimes folks getting out of hand. As with many things, this is a compromise.
Greg Comeau
>Greg Comeau schrieb:
>>
>> In article <3673A1BB...@eunet.at> "Ing. Franz Glaser" <meg-g...@eunet.at> writes:
>
>> >If you first learn Latin, it is very easy to learn Italian, Spanish,
>> >French, Roman and even English later - you have the basics.
>> >
>> This is all one way to look at it. Another is that language is not
>> just grammar and words but also culture, linguistics, idioms, even
>> techniques (say public speaking), etc.
>
>Yes, correct. What you do not see in the various C and C++
>newsgroups is the fact that I am "famous" here in the Pascal
>newsgroups to claim the "beauty" and readability of the
>source code. And this is why I sometimes catch flaming from
>the multilingual posters to prise Pascal too much for its
>readable source code and to neglect the efficiency issues.
You're preaching to the preacher. I too take a romantic view
on source code. Have had too many students who simply hated me
for it at first.
>So I did not feel that it were necessary to emphasize that
>explicitly.
My point was that basics, fundamentals, approach, techniques,
even things like hard work are also involved in the success of things.
Your original point implied otherwise, even if it was not your intent,
so I just wanted to bend things this way a tad.
zaphod
Sunil Rao wrote in message <3674351D...@ic.ac.uk>...
>Ben Pfaff wrote:
>>
>> Sunil Rao <suni...@ic.ac.uk> writes:
>>
>> A more suitable text for this purpose might be "Concrete Mathematics"
by
>> Ronald Graham, Donald Knuth and Oren Patashnik. It's published by
>> Addison-wesly and is ISBN0201558025.
>
>I actually do like the book quite a lot - it certainly goes into more
>detail. TAOCP, IMVHO, has less worked *mathematical* examples than it
>really should.
>
>> i.e., is it on the same level as TAoCP, or is it more of a `beginner's
>> book'?
>
>I'd reckon that it's at about the same level - but it's definitely
>pitched more at mathematicians than computer scientists. I quote from
>http://www-cs-staff.Stanford.EDU/~knuth/gkp.html - "...the authors
>themselves rely heavily upon it". I haven't actually bought myself a
>copy yet, but I plan to in the VNF. :)
Would it be a good refresher for someone who has the math but needs his
memory jogged from time to time?
Sunil Rao
This is getting to be a bit off-topic for comp.lang.c - not to mention
the other groups it's being crossposted to - but the answer is, IMHO,
yes, it would most certainly be.
Ing. Franz Glaser
>
> The lack of facilities for separate compilation is one of the major things
> that makes Pascal a toy language. Without, non-standard, extensions it is
> unsuitable for real work.
>
language" and "the other compiler" is on both sides the same. Well,
there
are many Pascal freaks who blame C about points where it is easy to find
out that he/she had absolutely no glue about C programming. This mail
shows that the same applies to the C programmers too. ;-)
What, excuse me, is more standard in C than in Pascal? In C you link a
lot of .h and .c to your project to adapt your application to a
particular
environment or OS. You need already the STDIO.H for the very simple
screen I/O. The same applies to Pascal, eg. Turbo Pascal with its couple
of units. Does that make Pascal more non-standard in another manner than
C?
If you treat the TeleTYpe 110bd as the proper console, you can claim a
standard.
Franz Glaser
Sunil Rao
>
> nick.k...@gecm.com wrote:
> >
> > The lack of facilities for separate compilation is one of the major things
> > that makes Pascal a toy language. Without, non-standard, extensions it is
> > unsuitable for real work.
> >
> It is really exciting to see that the missing knowledge about "the other
> language" and "the other compiler" is on both sides the same. Well,
> there
> are many Pascal freaks who blame C about points where it is easy to find
> out that he/she had absolutely no glue about C programming. This mail
> shows that the same applies to the C programmers too. ;-)
Oh - really?
> What, excuse me, is more standard in C than in Pascal?
The fact that C has an international standard that is in fact widely
adopted and used.
> In C you link a
> lot of .h and .c to your project to adapt your application to a
> particular
> environment or OS. You need already the STDIO.H for the very simple
> screen I/O. The same applies to Pascal, eg. Turbo Pascal with its couple
> of units. Does that make Pascal more non-standard in another manner than
> C?
stdio.h (note the lowercase) is a standard header - the C standard
defines its behaviour quite precisely. And stdio.h is not exactly meant
for screen I/O, it's rather for generic stream-based FILE I/O.
> If you treat the TeleTYpe 110bd as the proper console, you can claim a
> standard.
--
Lawrence Kirby
meg-g...@eunet.at "Ing. Franz Glaser" writes:
>Jeffrey C. Dege schrieb:
>>
>> >Pascal is Pascal - a structure language; just as C is. Languages that
>> >extend some parent language dont make the parent language OO. Just because
>> >Delphi is OO (or is it object-based?) doesnt go backward to Pascal.
>> >
>> >The language must support the OO features, or it is OO.
>>
>> Turbo Pascal is, in its later versions, is a fully OO language, but
>> then, Turbo Pascal is a proprietary extension to Pascal, just as
>> Delphi is, it doesn't come close to complying with the ISO standard.
>
>When ALGOL was improved to Pascal by Prof. Wirth there was no need for
>GotoXY or ClrScr, because the Teletype with 110bd did not support it.
>It did not support VGA graphics too.
It would be absurd for any portable language to support VGA graphics
since these are so heavily tied to one polatform.
>Of course a language/compiler/tool must be extended to the newer
>environments, regardless of any academic standard.
As an extension of the above many, probably most standard languages even
modern ones don't have built-in graphics facilities because however
you define them there will be many platforms where they can't be implemented.
It is generally languages where particular implementations have proved
more significant than standards where support for things like graphics
(and various other types of hardware support) have been included (by
particular implementations). BASIC is probably the classic case but
Pascal is a significant example of this too.
--
-----------------------------------------
Lawrence Kirby | fr...@genesis.demon.co.uk
Wilts, England | 7073...@compuserve.com
-----------------------------------------
Lawrence Kirby
dave...@home.com "David M. Cook" writes:
>On Sun, 13 Dec 1998 23:45:07 +0000, Martin Harvey
><mar...@aziraphale.demon.co.uk> wrote:
>
>>To be honest, the best advantage of Pascal over C is that it has a
>>strict typing system and doesn't allow macros.
>
>Pascal also has true pass by reference.
That is if anything a design flaw in Pascal. Certainly some mechanism is
required to affect more than one value in the caller but pass by reference
has some serious dawbacks. The problem is that you can't tell by looking
at the code in the caller which arguments can be modified by the called
function. C is much better in this respect (although not pefect). You
know in C that no argument can be modified by the caller which leads to
looser coupling and clearer code. C still has some problems in this
respect e.g. array arguments get converted to a pointer to the first
element so a pointer is passed by value and the array can be modified by the
called function. C++ has references and these are important for its OO
constructs. However even in C++ when you are defining a plain function
it is generally better to use pointer rather than reference arguments
when you want to modify something in the caller.
Lawrence Kirby
...
>I have never used an implementation of Pascal that didn't allow separate
>compilation (most commonly, by allowing function or procedure headers
>to be marked external in the same way that the standard allows them to
>be marked forward). It's not standard,
So you can't depend on it for standard code.
>perhaps for the same reasons that
>#include is said to give undefined behavior with non-standard headers;
That is only true of the contents of the headers isn't explicitly part of
the program. The effect of including a header is well-defined if either
the header is a standard one or its text is specified as part of the program.
>while this is unfortunate, you will find that some sort of non-standard
>extension is needed for real work in C, too.
You can do a lot of real work in C without resorting to non-standard
extensions.
>Generally, comparable implementations of separate compilation I have
>used have been faster with Pascal than C because the compiler doesn't
>have to rescan header files again and again; the worst instances for C
>are header files that are included repeatedly and surrounded by the usual
>
>#ifndef THIS_HEADER
>#define THIS_HEADER
>...
>#endif
>
>where the preprocessor, at least, looks at every line of the header
>each time.
This isn't generally a big problem because lexing code is generally a
pretty fast operation. A number of compilers also recognide the #ifndef
contruct above and know to ignore the header if any further attempt is
made to include it in the translation unit. The ability to split
large programs into smaller parts with selective recompilation, edit code
while compilation takes place and the general speed of todays processors
has in most cases made this a non-issue. I've used a lot of different C
compilers with widely varying compilation speeds (also with different
optimisation options which can have a very big effect) and generally
I don't find any real advantage is choosing the faster compilers since the
slower ones are still fast enough not to impact development time noticably.
Of couse there will be some counter-examples, like maybe proving
full source builds for different environments but these aren't that
common.
The main thing to watch out for is of you change a header which many source
files depend on. Possibly a bigger problem is very large system headers such
as <windows.h>.
Bub
> Unfortunately, I can't think of any language that has both pedagogical
> strength and the market share of C or C++ (I don't know VB, but would be
> wary of teaching any proprietary/non-portable language as a first language).
VB would probably be a bad idea as a "first" language.
Although around here this seems to have become rather trendy. IMHO for a
really
basic introductory course, they should use plain old BASIC (though Rexx
also comes
to mind). It introduces you to simple concepts like loops and variable
assignments,
difference between strings and numbers etc...
With a couple of weeks of basic you can tackle something more serious.
> I think if you're realistic you either spend a lot of time ramping students
> up where they can begin to learn fundamental CS concepts with C, or you use a
> different language and leave C for a later course.
>
> Dave Cook
--
Bub
bub at videotron dot ca
http://alcor.concordia.ca/~sd_fort
" Hope is the beginning of unhappiness "
Blake McCurdy
especially in this ng, usually in the context of recommending some newbie
learn VBasic or Modula or C, or Pascal, or *insert language here*. I
started C++ w/ a night school class about a year ago, and my instructor
said I picked it up phenomenally. ( I got a 99% on the course, lost one
point by adding [in my head] 1+4+9+16+25=45 on the exam). I asked a friend
of mine, a nice guy w/ 15+ yrs of experience what he thought of some of my
more recent code. He said it was good, but needed
a) more planning, a little too fat and inefficient
b) better naming conventions and comments; a bit unclear at times
c) A significantly better understanding of OOP.
It would seem he thought it was all bad, except for the structure, flow,
and use of goto; and break;, etc.
But I thought I wouldn't know that stuff, cause I'll I evr learned was
C++. Well, I guess I can't do anything, if I believe you to. Strange that
Every Newspaper in the Halifax Area uses a program that someone with no
talent wrote. Very interesting.
My Question: What would Murphy say?
Micah {-;
Martin Harvey
>
> Could be why statistics report that a lot of students are learning the
> basics at college and then leaving without ever getting a degree. In the
> Computer Science fields too many employers feel that what's being taught is
> out dated and doesn't apply to the 'real' world. Any opinions on that?
>
Too many *managers* feel that what's being taught is out dated <grin>
I learnt my CompSci at Cambridge Uni UK, and a lot of the stuff I did
there was (in various fields) closer to cutting edge than I have ever
found in employment. In fact most business programming is simply drudge
work... You have to be in a research department (whether industry
funded/based or academic) in order to do real cutting edge stuff...
Hey... C is *ancient* <g>.
MH.
--
Martin Harvey.
Totally rewritten web pages at:
http://www.harvey27.demon.co.uk/mch24/
"ALGOL 60 was a language so far ahead of its time that it
was not only an improvement on its predecessors but also
on nearly all its successors". C.A.R. Hoare
--------------BEGIN GEEK CODE BLOCK--------------
Version: 3.12
GCS/CC d(+) s-:- a-- C+++$ UL@ P L@>++ E- W++
N+++ o-- K++ w+++$ O--- M-- V-- PS@ Y-- PGP-
t--- 5-- X-- R-- !tv b+ DI+ D+ G e++ h- r z++>---
---------------END GEEK CODE BLOCK---------------
Eddie Maldonado
out that what Mr. Ambuhl posted was a list of many computer science grad
schools. What was his intention? Beats me. If you look at the so-called
"rankings" I don't believe you would see half of those schools to begin
with. Given MIT, Harvard, and Carnegie Mellon are among the best institutes
in the world to begin with in any intended major but try finding the other
90% of the schools listed. But who cares about rankings? They are mearly
people's biased opinions of a school. (Example, some employer will give
someone a job without thinking if they have a major in Philosophy or
Anthropology IF they are coming out of MIT.)
But now, let's not get off the subject here. I started this thread here to
talk about C++, not the politics of universities. Let's end this please.
Now....back to C++ and the other programming languages.
Sincerely,
Eddie
manac...@mpinet.net
Michael J. Tobler
mar...@aziraphale.demon.co.uk says...
> Michael J. Tobler wrote:
[snip]
> > on the language. It could be the teacher too, or many other things.
>
> True. You can write both good and bad code in almost any language (apart
> from some ultra anal functional languages where it's impossible to write
> bad code that works!).
Cool, what language would THAT be ?
> > Granted, C/C++ are terse languages, but they also allow you to be very
> > expressive.
>
> I find C++ a little *too* expressive... actually, perhaps I'd better
> rephrase that... I don't mind expressive languages, provided they are
> consistent. Just think about how many dirty ways the compiler allows you
> to copy an object in C++, and the different semantics that these copies
> have....
Well, that's the rub. If you want to excel at something, then you have to
learn the inards. If you want to be an excellent Porsche mechanic, you
better know the inside/out of the engine, drive-train, and all the little
nuances of its behaviour - AND, you better know about the 914, 911, 356,
956, etc etc. Same thing with programming - if you want to be proficient,
you better know the going-ons....
> > Heck, I've seen people with 3-4 years of C++ experience write
> > crap code. It's all context and you and I are not sure where the problem
> > is with this "class" you're mentioning.
[snip]
> Maybe not... but I've seen a fair number of experienced programmers
> throw good style out of the window... although some might say that if
> they know the rules back to front that might mean that they're allowed
> to break them!
That's what I'm saying - I see this a lot - being a consultant with a
well-regarded consulting company, I see a lot of incompetent developers.
Excuse me, I shouldnt say "developers" - that implies that they DEVELOP
something. I see a lot of programmers that have, say, 3-5 years under
their belt and they dont know any more than someone straight out of
school. I think it has a lot to do with pride of ownership and a
willingness to excel. No drive.
[snip]
> > > Don't get me wrong, I would never go back to Pascal, but I'm glad I learned
> > > it. Many major corporations also require a rudimentary knowledge of
> > > Pascal. I know that IBM requires knowledge of Pascal before they hire you.
> >
> > Really, they never asked me that.
>
> Me neither... In fact they seemed to actively discourage applicants with
> a *pure* comp-sci bias... (at least in this part of the world).
>
> MH.
>
> --
> Martin Harvey.
> Totally rewritten web pages at:
> http://www.harvey27.demon.co.uk/mch24/
--
<<<<<<<<<< Blue Skies >>>>>>>>>>>
< Michael J. Tobler >
< mto...@no-spam-ibm.net >
< remove "no-spam-" when replying >
<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>
Michael J. Tobler
mar...@aziraphale.demon.co.uk says...
> John E. Carty wrote:
> > Pascal? The guy wants a job afterwards, not just a title. :-0
> John,
>
> I first learnt to program in Modula-2 (v-similar to pascal) at the age
> of 12. I then learnt Modula-3 (which is OO modula-2 (ish)) at the age of
Ooooh, that sparks a memory. I really enjoyed Modula-2/3. Too bad it
really never got any marketing press - I guess since M$ wasnt involved, it
got stomped on.
[snip]
Michael J. Tobler
[snip]
> Yes, correct. What you do not see in the various C and C++
> newsgroups is the fact that I am "famous" here in the Pascal
> newsgroups to claim the "beauty" and readability of the
> source code. And this is why I sometimes catch flaming from
> the multilingual posters to prise Pascal too much for its
> readable source code and to neglect the efficiency issues.
If you're saying that Pascal is much more "readable" and has more "form
and function" to its appearance, I'd have to disagree. It's the programmer
that makes the code that way.
I find that writing software is like writing a book. A program is a story
that the computer reads. So, programs should be written with that in mind.
Michael J. Tobler
mar...@aziraphale.demon.co.uk says...
[snip]
> Imperative (Best to Worst)
> Modula-2
> Pascal
> Modula-3
> Oberon
> Java
> C
> C++
Gee, no Smalltalk or Eiffel - how sad.
[snip]
Michael J. Tobler
mar...@aziraphale.demon.co.uk says...
[snip]
> I learnt my CompSci at Cambridge Uni UK, and a lot of the stuff I did
> there was (in various fields) closer to cutting edge than I have ever
> found in employment. In fact most business programming is simply drudge
> work... You have to be in a research department (whether industry
> funded/based or academic) in order to do real cutting edge stuff...
IT employment? Yuk. Try consulting. You can be doing some cutting edge
stuff there.
Greg Comeau
Murphy doesn't say anything, he waits and hide for opportunities! ;-)
My own perspective is that programming takes a lifetime to learn.
That, with considering that last year you (anyone, not you) were an idiot,
simply means that there is always a lot of work to get done.
When you think you know everything, that is the time to get scared.
greg
>
> I've heard from alot of people that UCF is one of the top comp sci schools
> in the south, if not the best in the south.
I'm sure it has a fine department, but "best" or "top"? Certainly not
by reputation...
Like you said, they do start
> with a certain language....and its Pascal. Suprise suprise. Any opinions
> on that?
Yes. They're probably arguing with each other right now about whether to
teach C++ or Java to the students in the introductory courses, and,
because they cannot decide, they stick with what they've been doing for
the past 15 years...
IMHO, a (respectable, non-proprietary) language with object-oriented
features should be at least one of the languages taught to beginning CS
students... Teaching Pascal is now about as justifiable as teaching
Fortran (which is to say, somewhat justifiable, but not very
justifiable). I do feel that any educated computer scientist should be
familiar with Pascal.
g
>
> Eddie
> manac...@mpinet.net
;alksdjf
Pick a place and start. Here at Minnesota State University I started with
C++, and am now learning Java. I does not take a rocket scientist to do it.
In fact I started to teach my friends 14 year old brother C++ and after 2
months he is right up to speed with the college students. Shure he may not
understand the really hard stuff, but any moron can make a program to sort
an array or output a fibinachi number sequence. That is what most
introductory courses teach you anyway. If you can already do this stuff
when you get to college your life will be much easier. Once there, they
will remove most of your bad habits and you will be on your way. If you
want to know about programming ask students not teachers. Most professors
are teaching cis for a reason. Why else would somone turn down 100,000 for
50,000 to teach unless they are stupid.
Michael Powe
Hash: SHA1
>>>>> "Judson" == Judson E Knott <jek...@unity.ncsu.edu> writes:
>> Sigh. To say that Pascal is a good teaching language is
>> mythology IMO. I say this because it just solemely rules out
>> so many things. Note my saying this is NOT an attack on
>> Pascal. But a statement that when one is armed with proper
>> educations mediums, the sky is the only limit. IOWs, you're
>> implying that say C or C++ are not good teaching languages and
>> that is simply not so.
Judson> Actually, in most places you look it says that Pascal is
Judson> a good teaching language, or should we really start with
Judson> <a href=http://www.cwi.nl/~steven/abc/">ABC</a>? And, i
Judson> don't imply that C and C++ are not good teaching languages
First, I must say that I have strong objections to the notion that the
most important aspect of the introductory course is the language
taught. If anything, a study of a language like Pascal serves a dual
purpose: it broadens the student's understanding of the "programming
universe" and it forcibly prevents the students from focussing on the
language itself, to the exclusion of concepts.
I know it's my prejudice, but I think there's just of much of the "I
want to learn to program in C++" and not enough of "I want to learn to
program." The whole purpose of a "teaching language" ought to be the
the pursuit of the latter at the expense of the former.
mp
- --
Michael Powe
mic...@trollope.org http://www.trollope.org
Portland, Oregon USA
-----BEGIN PGP SIGNATURE-----
Version: PGP for Personal Privacy 5.0
Charset: noconv
iQA/AwUBNnYAurajuNi/6Js3EQLcDQCbB+ifp6gQh9/glBDcVqc/1rZedHcAn3JE
PRueri36QsL2R3bmnS6U8X6g
=e0ex
-----END PGP SIGNATURE-----
Alex Yacoub
>
> > Pascal
> > requires you to use functions, loops, and if-thens to make decisions.
> Uhhhh - it doesnt REQUIRE me to, these are constructs that should be used
> in any language.
Yes, I agree they should be used in all languages, but Pascal, unlike
others doesn't have keywords such as goto, which makes code almost
impossible to read.
>
> I have seen a LOT of Pascal code that looked like $]it, so dont blame it
> on the language. It could be the teacher too, or many other things.
> Granted, C/C++ are terse languages, but they also allow you to be very
> expressive. Heck, I've seen people with 3-4 years of C++ experience
write
> crap code. It's all context and you and I are not sure where the problem
> is with this "class" you're mentioning.
>
Again, I agree, but its a lot harder to make Pascal code look like $]it
then it is in C or C++. With students who are just beginning to learn,
they don't realize why structured programming is so important and they tend
to form bad habits unless they're forced not to. Pascal is the kind of
language that helps you form good habits.
> Not trying to be insultive, but a second year high school comp-sci
studend
> STILL has much to learn, so "grading" other work should be taken with a
> grain of salt. I'll bet YOUR code might possibly NOT pass a code review
> from me :)
>
I know I still have a lot to learn, but like I said earlier, I'm not
following the curriculum and haven't been for a while now. Besides, maybe
I wouldn't pass your standards, but you'd be surprised how high mine are.
(And the teacher specifically asked me to help out students who were stuck,
so maybe I do know what I'm doing. He thinks so).
-Alex
Martin Ambuhl
>. If you
> want to know about programming ask students not teachers. Most professors
> are teaching cis for a reason. Why else would somone turn down 100,000 for
> 50,000 to teach unless they are stupid.
Because someone has to train the next generation out of their idiocy.
The task seems hardly started with you.
--
Martin Ambuhl (mam...@earthlink.net)
Note: mam...@tiac.net will soon be inactive
John E. Carty
English language. Any idiot can learn to spell. ;-)
;alksdjf wrote in message <754rkg$i5$1...@nitrogen.mankato.msus.edu>...
>Any Idiot can learn to program!
>Pick a place and start. Here at Minnesota State University I started with
>C++, and am now learning Java. I does not take a rocket scientist to do
it.
>In fact I started to teach my friends 14 year old brother C++ and after 2
>months he is right up to speed with the college students. Shure he may not
>understand the really hard stuff, but any moron can make a program to sort
>an array or output a fibinachi number sequence. That is what most
>introductory courses teach you anyway. If you can already do this stuff
>when you get to college your life will be much easier. Once there, they
>will remove most of your bad habits and you will be on your way. If you
Martin Harvey
>
>
> This is great, and I agree with you in part, however, is simply does not
> address the issue that was raised. To wit, for instance, Delphi is
> not pascal. It is just not pascal.
No, I have to admit that it's 80% not pascal :-)
> So, you cannot disagree, at least
> with the argument you gave, because you haven't used pascal in the real work,
> you've used delphi.
Yep... I've relied very heavily on OO.... sure is nice tho. Perhaps I
should have said that "Wirth style languages aren't dead!"
> And it was Pascal that was the issue being raised
> on the table. You may feel that Delphi has some nice extensions to it,
> but that's a seperate issue.
Pascal is a bit dysfunctional... However, I suspect if it allowed a bit
of inline asm (perhaps in "unsafe units"), then there's no reason why it
shouldn't be used instead of C.
MH.
--
Martin Harvey.
Totally rewritten web pages at:
http://www.harvey27.demon.co.uk/mch24/
"ALGOL 60 was a language so far ahead of its time that it
Martin Harvey
>
> Well, the student works at their pace, and not at a pace
> they cannot keep up with, and can play around with the code to see what
> happens when they do what. (Who hasn't had that bastard teacher?)
The people that enjoy learning to program most are those that teach
themselves. I spent many a teenage afternoon figuring out Modula-2 (a
pascal derivative). Oh dear what a sad techie I was..... perhaps as a
kid I should have discovered good dating skills instead <g>... :-)
Martin Harvey
> This isn't generally a big problem because lexing code is generally a
> pretty fast operation.
Oh really? I was always told that the two most costly operations in any
compiler were lexing and linking ... all the parsing and optimising in
between is not I/O limited.
> The main thing to watch out for is of you change a header which many source
> files depend on. Possibly a bigger problem is very large system headers such
> as <windows.h>.
Agreed. The great thing about most structured implementations of pascal
is that every header part iof a file is compiled once and only once...
everything else is then done by manipuating the intermediate
representation in the "compiled units" (see my previous post about the
difference between them and objs).
Martin Harvey
>
>
> I hope you meant to put a smiley there. How big is "large" to you? I don't
> think the 100,000 line application I'm working on would have faster build
> times if it were all in one file!
Well... my latest little app is about 17,000 lines. Pray let me copy &
paste the latest compile info for you from my Delphi (OO pascal)
compiler :-)
D:\Delphi Progs\Distributed Raytracer\Code\Server>dcc32 drayserver -B -q
Delphi for Win32 Version 9.0 Copyright (c) 1983,96 Borland
International
17771 lines, 2.91 seconds, 285300 bytes code, 38808 bytes data.
That just about says it all. Just out of interest, the -B switch says
"recompile all units not part of the standard borland library"
My experience with larger projects is that a 100,000 line compile takes
about 15 seconds on my P180 MMX.
The trick here is that unlike some languages, Pascal is one pass, and
the grammar is fairly simple. In addition, most Pascal compilers have a
"compiled unit", which does a lot more than an obj... it not only stores
the code & fixup info, it also stores all the type information which
means that you don't have to recompile much in the way of library files.
Note that in the above compile, I specifically told the compiler *not*
to use the compiled unit info. If I get it to use the compiled units...
I then get (and this is basically just the link time)
D:\Delphi Progs\Distributed Raytracer\Code\Server>dcc32 drayserver -q
Delphi for Win32 Version 9.0 Copyright (c) 1983,96 Borland
International
53 lines, 0.86 seconds, 285300 bytes code, 38808 bytes data.
I stand by the above evidence. Most other pascal users will be able to
post similar compile times for similarly sized projects.
> The lack of facilities for separate compilation is one of the major things
> that makes Pascal a toy language. Without, non-standard, extensions it is
> unsuitable for real work.
The language does not need to specify the exact implementation details.
Most Pascal compilers have mechanisms that handle separate compilation
beautifully.
In addition, because the most common mechanism used is a structured
mechanism (interface imports must follow a directed acyclic graph ...
hidden implementation imports may follow any structure), it's far more
difficult to get one's #includes mixed up or in the wrong order.
David M. Cook
wrote:
>That is if anything a design flaw in Pascal. Certainly some mechanism is
>required to affect more than one value in the caller but pass by reference
>has some serious dawbacks. The problem is that you can't tell by looking
>at the code in the caller which arguments can be modified by the called
>function. C is much better in this respect (although not pefect).
Neither can you tell with C. You know the value of a pointer isn't changed,
but you don't know whether a memory location is being changed.
Something like Ada's IN/OUT keywords would be better. I beleive Eiffel has
something similar, but I haven't gotten to that part is Bertrand Meyer's
massive OOSC2
Dave Cook
Michael Powe
Hash: SHA1
>>>>> "Greg" == Greg Comeau <com...@panix.com> writes:
>> I just finished a term of C++ with a mediocre instructor -- a
>> nice guy, day job developing embedded systems at a local
>> electronics mfg firm & teaching at night; but just basically
>> out of it while in front of the blackboard. Motivation was all
>> that got me through the course.
Greg> Then it sounds like that education medium was poor then. If
Greg> it wasn't the teacher's fault, then it was the curriculum's
Greg> fault.
The texts were sucky, that's a fact. The lab book used the dreadful
`void main()' for all the exercises. Part of the problem (to my mind)
is the schizophrenic nature of the course. It's called "Computer
Fundamentals," but it amounts to "C++ Programming." This seems to be
a fairly normal paradigm, too.
>> I'm studying it because it's required to get into the
>> interesting courses -- C, design, theory. But while I'm at it,
>> I will <learn> it.
Greg> You have a good attitude. BTW, I have gone into many things
Greg> that I did not care for, only to get some AHAs! when other
Greg> pieces to it became clear. Anyway, your goal is to learn
Greg> programming, techniques, and all the interrelated skills
Greg> implied in that. The language used does "interfere" with
Greg> that, but too, lots or parts can remain neutral as you hone
Greg> yourself.
I'm always open-minded about learning experiences. Also, any time I
actually pay for something, I'm motivated to get my money's worth!
;-)
mp
- --
Michael Powe
mic...@trollope.org http://www.trollope.org
Portland, Oregon USA
-----BEGIN PGP SIGNATURE-----
Version: PGP for Personal Privacy 5.0
Charset: noconv
iQA/AwUBNnYjAbajuNi/6Js3EQJCRwCffyd1+6st4d3g4mc5vM7sDOPkg0kAn1nG
wP9ZY4vuyzsfjM8H1YZR6GTu
=XjXM
-----END PGP SIGNATURE-----
paolo...@my-dejanews.com
"Alex Yacoub" <rya...@worldnet.att.net> wrote:
> Michael J. Tobler <mto...@no-spam-ibm.net> wrote
> >
> > > Pascal
> > > requires you to use functions, loops, and if-thens to make decisions.
> > Uhhhh - it doesnt REQUIRE me to, these are constructs that should be used
>
> > in any language.
>
> Yes, I agree they should be used in all languages, but Pascal, unlike
> others doesn't have keywords such as goto, which makes code almost
> impossible to read.
May be I have an old PASCAL manual and things are changed, but GOTO is a
PASCAL istruction. On the other hand, excessive use of GOTO is bad,
occasional use might be useful, otherwise it would have not retained in many
modern languages like C and C++. In practice in C, C++, FORTAN90 goto are
very very rarely used, so that no judgement on the quality of these languages
can be drawn on this basis. Even good programming FORTAN77 makes the use of
GOTO occasional if not in very disciplined places.
> >
> > I have seen a LOT of Pascal code that looked like $]it, so dont blame it
> > on the language. It could be the teacher too, or many other things.
> > Granted, C/C++ are terse languages, but they also allow you to be very
> > expressive. Heck, I've seen people with 3-4 years of C++ experience
> write
> > crap code. It's all context and you and I are not sure where the problem
> > is with this "class" you're mentioning.
> >
>
> Again, I agree, but its a lot harder to make Pascal code look like $]it
> then it is in C or C++. With students who are just beginning to learn,
> they don't realize why structured programming is so important and they tend
> to form bad habits unless they're forced not to. Pascal is the kind of
> language that helps you form good habits.
It is probably even harder in FORTAN77. But that is a weak reason use it in
place of C and C++.
Languages that allow to do many things may become hard to read, more limited
languages are easier to keep them readable.
> -Alex
>
Ciao Paolo
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
greg
>
> Michael J. Tobler <mto...@no-spam-ibm.net> wrote
> >
> > > Pascal
> > > requires you to use functions, loops, and if-thens to make decisions.
> > Uhhhh - it doesnt REQUIRE me to, these are constructs that should be used
>
> > in any language.
>
> Yes, I agree they should be used in all languages, but Pascal, unlike
> others doesn't have keywords such as goto, which makes code almost
> impossible to read.
Which Pascal are you talking about? I haven't used Pascal much, but the
goto keyword is a part of ISO pascal (though it is actively discouraged)
>
> >
> > I have seen a LOT of Pascal code that looked like $]it, so dont blame it
> > on the language. It could be the teacher too, or many other things.
> > Granted, C/C++ are terse languages, but they also allow you to be very
> > expressive. Heck, I've seen people with 3-4 years of C++ experience
> write
> > crap code. It's all context and you and I are not sure where the problem
> > is with this "class" you're mentioning.
> >
>
> Again, I agree, but its a lot harder to make Pascal code look like $]it
> then it is in C or C++. With students who are just beginning to learn,
> they don't realize why structured programming is so important and they tend
> to form bad habits unless they're forced not to. Pascal is the kind of
> language that helps you form good habits.
>
nick.k...@gecm.com
"Alex Yacoub" <rya...@worldnet.att.net> wrote:
> Michael J. Tobler <mto...@no-spam-ibm.net> wrote
> >
> > > Pascal
> > > requires you to use functions, loops, and if-thens to make decisions.
> > Uhhhh - it doesnt REQUIRE me to, these are constructs that should be used
> > in any language.
>
> Yes, I agree they should be used in all languages, but Pascal, unlike
> others doesn't have keywords such as goto, which makes code almost
> impossible to read.
actually Pascal *does* have a goto.
<snip>
> With students who are just beginning to learn,
> they don't realize why structured programming is so important and they tend
> to form bad habits unless they're forced not to. Pascal is the kind of
> language that helps you form good habits.
I agree. Pascal is a good choice as a teaching language.
<snip>
raw...@my-dejanews.com
fr...@genesis.demon.co.uk wrote:
> In article <7535pr$tp1$1...@nnrp1.dejanews.com> raw...@my-dejanews.com writes:
>
> >I have never used an implementation of Pascal that didn't allow separate
> >compilation (most commonly, by allowing function or procedure headers
> >to be marked external in the same way that the standard allows them to
> >be marked forward). It's not standard,
>
> So you can't depend on it for standard code.
If you're also using graphics extensions, or whatever, then you're already
out of luck.
> >perhaps for the same reasons that
> >#include is said to give undefined behavior with non-standard headers;
>
> That is only true of the contents of the headers isn't explicitly part of
> the program. The effect of including a header is well-defined if either
> the header is a standard one or its text is specified as part of the program.
My point is that standards only go so far; this is probably an inevitable
situation. Different standards choose different boundaries.
(From your description, #include itself does not produce undefined behavior;
at compilation, the header file is either a standard one, or it is found by
the implementation (and thus its text is specified as part of the program),
or it's not found (in which case a diagnostic is required, and that effect
seems reasonably well-defined).)
> >while this is unfortunate, you will find that some sort of non-standard
> >extension is needed for real work in C, too.
>
> You can do a lot of real work in C without resorting to non-standard
> extensions.
And you can do real work in Pascal without resorting to non-standard
extensions. Yes, you can even write a C compiler. The amount of
real work that can't be done in either C or Pascal without non-standard
extensions is many, many times larger than the amount that can be done.
> >Generally, comparable implementations of separate compilation I have
> >used have been faster with Pascal than C because the compiler doesn't
> >have to rescan header files again and again; the worst instances for C
> >are header files that are included repeatedly and surrounded by the usual
> >
> >#ifndef THIS_HEADER
> >#define THIS_HEADER
> >...
> >#endif
> >
> >where the preprocessor, at least, looks at every line of the header
> >each time.
>
> This isn't generally a big problem because lexing code is generally a
> pretty fast operation. A number of compilers also recognide the #ifndef
> contruct above and know to ignore the header if any further attempt is
> made to include it in the translation unit. The ability to split
> large programs into smaller parts with selective recompilation, edit code
> while compilation takes place and the general speed of todays processors
> has in most cases made this a non-issue. I've used a lot of different C
> compilers with widely varying compilation speeds (also with different
> optimisation options which can have a very big effect) and generally
> I don't find any real advantage is choosing the faster compilers since the
> slower ones are still fast enough not to impact development time noticably.
> Of couse there will be some counter-examples, like maybe proving
> full source builds for different environments but these aren't that
> common.
Lexical analysis is a significant overhead for compilation. I was extremely
careful to qualify this as _my_ experience of comparable implementations of
separate compilation, by way of explaining why a Pascal programmer
asserted that C compilers are slow and why #include contributes to this.
--
MJSR
Lawrence Kirby
mar...@aziraphale.demon.co.uk "Martin Harvey" writes:
>Lawrence Kirby wrote:
>> This isn't generally a big problem because lexing code is generally a
>> pretty fast operation.
>
>Oh really? I was always told that the two most costly operations in any
>compiler were lexing and linking ... all the parsing and optimising in
>between is not I/O limited.
Why not try it out sometime, most compilers have options to run the
preproecessor only or link only. On typical PCs or workstations these
are by far the fastest phases (certainly when you have any significant
degree of optimisation set) Syntax analysis may be fast too but that is
generally not visible separately. Even slow hard disks these days can
transfer data at many megabytes per second and most OSs have caches (which
will help with, e.g. repeated header inclusion). Try comparing the time it
takes to file copy a large source set to the time it takes to compile it.
Compilation is usually not an I/O bound process (A quick test here
rebuilding a project measured the system stalled about 15-20% of the time
waiting for the hard disk).
>> The main thing to watch out for is of you change a header which many source
>> files depend on. Possibly a bigger problem is very large system headers such
>> as <windows.h>.
>
>Agreed. The great thing about most structured implementations of pascal
>is that every header part iof a file is compiled once and only once...
>everything else is then done by manipuating the intermediate
>representation in the "compiled units" (see my previous post about the
>difference between them and objs).
C allows the use of precompiled headers, although not all compilers support
this.
--
-----------------------------------------
Lawrence Kirby | fr...@genesis.demon.co.uk
Wilts, England | 7073...@compuserve.com
-----------------------------------------
Lawrence Kirby
dave...@home.com "David M. Cook" writes:
>On Mon, 14 Dec 98 21:08:24 GMT, Lawrence Kirby <fr...@genesis.demon.co.uk>
>wrote:
>
>>That is if anything a design flaw in Pascal. Certainly some mechanism is
>>required to affect more than one value in the caller but pass by reference
>>has some serious dawbacks. The problem is that you can't tell by looking
>>at the code in the caller which arguments can be modified by the called
>>function. C is much better in this respect (although not pefect).
>
>Neither can you tell with C. You know the value of a pointer isn't changed,
>but you don't know whether a memory location is being changed.
You know that the argument itself isn't changed and not all arguments are
pointers. It is also helpful to know that a pointer argument isn't
itself changed.
>Something like Ada's IN/OUT keywords would be better. I beleive Eiffel has
>something similar, but I haven't gotten to that part is Bertrand Meyer's
>massive OOSC2
The important thing is that the difference is visible in the code of the
*caller*. In C that is to some extent true because for a variable to be
modified elsewhere you have to have taken its address.
Igor Rafienko
<off-topic>
| Any Idiot can learn to program!
Wrong. Any idiot can learn how to output some source code. Writing
easy-to-read/maintain code requires quite a bit of intelligence. I'd
like to see some of your code that exceeds 5000 lines...
| Pick a place and start. Here at Minnesota State University I started with
| C++, and am now learning Java. I does not take a rocket scientist
| to do it.
... but I seriously doubt that "any idiot" could do it. Otherwise I'd
be concerned for the quality of education at Minnesota State University.
| In fact I started to teach my friends 14 year old brother C++ and after 2
| months he is right up to speed with the college students.
now I am concerned about the abovementioned quality, unless this guy
is a genious.
| Shure he may not
| understand the really hard stuff,
define 'really hard'
| but any moron can make a program to sort
| an array
No kidding? Can this friend of yours write a QSort from scratch and
sketch an analysis of why qsort is better than say ShellSort? This
*is* taught in introductory courses, you know...
| or output a fibinachi number sequence. That is what most
| introductory courses teach you anyway.
Wrong. Introductory courses teach you to structure your way of
thinking + _basic_ datastructures and algorithms. At least we do that at
the University of Oslo. Fibonacci programs come as a bonus, not as the
main goal.
| If you can already do this stuff
| when you get to college your life will be much easier.
Being able to write Fibonacci programs will not make your life
considerably easier.
| Once there, they
| will remove most of your bad habits and you will be on your way. If you
| want to know about programming ask students not teachers.
That depends on students and teachers.
| Most professors
| are teaching cis for a reason. Why else would somone turn down 100,000 for
| 50,000 to teach unless they are stupid.
Have you ever considered that some of them might want to contribute to
the quality of the next generation of programmers? And if you consider
your teaching stuff to be morons, what are you doing at the Minnesota
State Univ. anyway?
</off-topic>
igorR
--
"Ma belle demoiselle, repondit Candide, quand on est amoureux, jaloux
et fouette par l'Inquisition, on ne se connait plus"
Voltaire, "Candide"
George W. Bayles
> I know it's my prejudice, but I think there's just of much of the "I
> want to learn to program in C++" and not enough of "I want to learn to
> program." The whole purpose of a "teaching language" ought to be the
> the pursuit of the latter at the expense of the former.
You are not alone in that opinion.
Greg Comeau
>Greg Comeau wrote:
>> This is great, and I agree with you in part, however, is simply does not
>> address the issue that was raised. To wit, for instance, Delphi is
>> not pascal. It is just not pascal.
>
>No, I have to admit that it's 80% not pascal :-)
>
>> So, you cannot disagree, at least
>> with the argument you gave, because you haven't used pascal in the real work,
>> you've used delphi.
>
>Yep... I've relied very heavily on OO.... sure is nice tho. Perhaps I
>should have said that "Wirth style languages aren't dead!"
>
>> And it was Pascal that was the issue being raised
>> on the table. You may feel that Delphi has some nice extensions to it,
>> but that's a seperate issue.
>
>Pascal is a bit dysfunctional... However, I suspect if it allowed a bit
>of inline asm (perhaps in "unsafe units"), then there's no reason why it
>shouldn't be used instead of C.
I've always argued that from a high level, syntactically, that Pascal and C
are more or less similar. I mean yes, they both have if statement and loops
and functions. And from that "subset" they can more or less be used
interchangeably. Each has its pros and cons at this level (and BTW,
this is why I say that the classic argument that Pascal is/should be
the definitive teaching language is silly and archaic).
But there are other levels where they are not similar, including
syntactically, culturally, freedom of expression, no less with pointers,
techniques, etc. That is, from the previous paragraph, each _does_ branch off.
So, indeed, there can be reasons why Pascal shouldn't be used instead of C,
and vice versa.
Greg Comeau
>Lawrence Kirby wrote:
>> This isn't generally a big problem because lexing code is generally a
>> pretty fast operation.
>
>Oh really? I was always told that the two most costly operations in any
>compiler were lexing and linking ... all the parsing and optimising in
>between is not I/O limited.
I'm a compiler writer, and from my experience, the two most costly operations
in any compiler are the two that the compiler author "made" slowest,
whatever they happen to be.
>> The main thing to watch out for is of you change a header which many source
>> files depend on. Possibly a bigger problem is very large system headers such
>> as <windows.h>.
>
>Agreed. The great thing about most structured implementations of pascal
>is that every header part iof a file is compiled once and only once...
>everything else is then done by manipuating the intermediate
>representation in the "compiled units" (see my previous post about the
>difference between them and objs).
C and C++ compilers do this too, terming it ``pre-compiled headers''.
It is an implementation detail, as a quality of implementation issue.
The respective Standard's don't require this model but they do support it.
Greg Comeau
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
> >> Sigh. To say that Pascal is a good teaching language is
> >> mythology IMO. I say this because it just solemely rules out
> >> so many things. Note my saying this is NOT an attack on
> >> Pascal. But a statement that when one is armed with proper
> >> educations mediums, the sky is the only limit. IOWs, you're
> >> implying that say C or C++ are not good teaching languages and
> >> that is simply not so.
>
> Judson> Actually, in most places you look it says that Pascal is
> Judson> a good teaching language, or should we really start with
> Judson> <a href=http://www.cwi.nl/~steven/abc/">ABC</a>? And, i
> Judson> don't imply that C and C++ are not good teaching languages
>
>First, I must say that I have strong objections to the notion that the
>most important aspect of the introductory course is the language
>taught. If anything, a study of a language like Pascal serves a dual
Not just Pascal.
>purpose: it broadens the student's understanding of the "programming
>universe" and it forcibly prevents the students from focussing on the
>language itself, to the exclusion of concepts.
>
>I know it's my prejudice, but I think there's just of much of the "I
>want to learn to program in C++" and not enough of "I want to learn to
>program." The whole purpose of a "teaching language" ought to be the
>the pursuit of the latter at the expense of the former.
It's my prejudice too. That's why I say above that saying that
Pascal is the end all for newbies is ruling out so many things.
It misses the point, which you've articulated well.
Alex Yacoub
<754rkg$i5$1...@nitrogen.mankato.msus.edu>...
> Any Idiot can learn to program!
Yeah, any idiot can learn to write Fibonacci sequences. It just involves
writing an algorithm you already know in C and then outputting it. Real
programming requires hard work, experience, and high standards.
> Shure he may not understand the really hard stuff
The hard stuff is what counts. I don't think the people who write programs
you buy use just the "easy stuff". No offense. but it's people like you or
your friend who write all the code with bugs; good coders use good
practices which make debugging and keeping code updated much easier.
> Once there, they
> will remove most of your bad habits and you will be on your way.
>
Bad habits are almost impossible to get rid of. A quality computer science
program will teach you good habits from the beginning, so that when you
finally realize why they're so important, you won't have to worry about
changing your habits. That's one of the main reasons why Pascal is a good
teaching language, it pretty much forces you to program with good habits,
unless you want to do more work and make it look bad.
-Alex
Martin Ambuhl
>
> Is UCF University of Central Florida? If so, I didn't find it ranked at
> all (in Princeton's review, US News and World Report, or Money Mag).
You will find UCF in the US News survey in "Tier 4" schools. I admit
that I rarely pay much attention to even Tier 2 schools, so can have no
opinion about what some Tier 4 school has for a program. I know that
the Tier 1 schools that I attended and taught at are not quaking in fear
at the thought that UCF is on their heels.
> Not saying it doesn't have a good program, but I will have to say Rice
> is a great southern school, especially for engineering and comp sci. Of
> course, I'm biased in that respect :) Rice was ranked in all of the
> above mentioned mentioned reviews (#18 nation wide in Princeton, and #2
Your bias is well placed. Rice is an excellent school in most
disciplines.
Michael Powe
Hash: SHA1
>>>>> "Greg" == Greg Comeau <com...@panix.com> writes:
>> First, I must say that I have strong objections to the notion
>> that the most important aspect of the introductory course is
>> the language taught. If anything, a study of a language like
>> Pascal serves a dual
Greg> Not just Pascal.
No. Consider this:
"Programming should be from the beginning a creative and literate
endeavor. Our goal is to expose the reader to the exhilaration of
reading and creating beautiful programs. These programs should be
concise in their expression, general in their application, and easily
understood. These goals are achieved through creative use of
abstraction techniques that capture recurrent patterns of computation
and allow them to be simply used."
Huh. Well, that sure didn't come from a C or C++ "Introduction to CS"
textbook. Both of the C++ textbooks I have start right in with the
language itself. One of them actually starts out with "C and C++ are
seeing explosive growth, not only in industry but in academia, as
well." Yawn. So what?
No, the above quote comes from <Scheme and the Art of Programming>. A
fair number of schools use Scheme as the introductory language. I
spent a little time last night cruising around the web; as far as I
can see from a quick look, there's not a clear consensus. Rice
University uses Scheme for its intro, Oberlin uses Java, Princeton and
Stanford use C, Portland State University (alma mater) uses C++. At the
University of Washington, you get C first and then C++.
I do have to say, I'm glad I did not try to start learning to program
with Java. I'm sure that would have sent me screaming from the
classroom, never to return. It's bad enough now, when at least I have
some practical experience behind me.
mp
- --
Michael Powe
mic...@trollope.org http://www.trollope.org
Portland, Oregon USA
-----BEGIN PGP SIGNATURE-----
Version: PGP for Personal Privacy 5.0
Charset: noconv
iQA/AwUBNnbT5rajuNi/6Js3EQInvACgs7pwYWpdQC6TmvU2VwzfbKmgQM8An3H8
4uF2lJwaJmZTWMlAsJpnV1ud
=WM90
-----END PGP SIGNATURE-----
Daniel Barker
[snip]
> And Pascal is much more alive than Latin!
If you publish a new species of plant not previously described, a brief
Latin description of the species must be given. This rule is fixed by the
International Code of Botanical Nomenclature (ICBN).
So, in the botanical world, Latin is very much alive. Many of us can read
it. Those who can write it fluently are few, but much in demand. Those who
ignore it are unable to publish new species.
It looks like the rules will soon be relaxed to allow an English
description instead. However, reading the older descriptions will always
be necessary.
Daniel Barker,
Institute of Cell and Molecular Biology,
Swann Building,
King's Buildings,
Mayfield Road,
Edinburgh
EH9 3JR
UK
Daniel Barker
> I'm a senior in H.S. and I'm getting ready to go to college in the Fall. I
> am interested in learning a programming language but I'm not sure where to
> start. I thought perhaps C++? I don't know, any insight or advice is
> greatly appreciated.
[snip]
If I could begin again, I would begin with formal methods and only move on
to executable languages at some late stage. (As it happens, I began with
executable languages - BASIC then FORTRAN77 then C - and am now thinking
of moving on to formal methods.)
I suggest taking a look at a formal specification language like Z (see
news group comp.specification.z). This would give you a grounding in how
to design a correct program. This is much more important than knowing the
syntax of C, Scheme, Pascal or F. A Z specification uses the concepts of
mathematics and can be proved correct. Once you've done that, you can
translate the specification into whatever executable language you like.
Executable languages refer to details of the implementation (i.e.,
computer) and their behaviour is harder to infer from the text, especially
if you have a mathematical background.