Kids These Days! (not much Scheme content, but an appeal to Educators)
David Rush
<rant>
What is wrong with kids coning into this business these days?
OK, I'm just a line-grunt. In fact, I'm a University drop-out and my
aspirations to academia stem from an addiction to ACM TOPLAS (although
these days I don't have access to it anymore) and a fairly refined
coding aesthetic. But I've been programming professionally for 17
years now (and as an enthusiastic amateur for another few before
that), and I've got to say that the general quality of code that I'm
seeing deployed has dropped even as we've gotten better and better
tools for doing things right.
I can't even begin to tell you how frustrating this is.
I currently work for a *large* internet/media company whose
three-letter acronym shall remain unmentioned. As we've expanded into
international markets some of our core services have encountered
well-solved problems in the networking world. In this case, we're
delivering email from servers in the USA to clients in many timezones
around the world. The local portals want the delivery times to be
shown in their native TZ rather than EST.
This is a well-solved problem in the Unix world. /etc/zoneinfo (on
Linux, /usr/share/lib/zoneinfo on Solaris, &cet) has an extensive
database of precise timezone definitions. The basic library functions
exists on every Unix box (ctime, mktime, strptime &cet...), and the
library source is freely available at ftp://elsie.nci.nih.gov/pub
Over the last 4 workiung days I have watched a fairly sharp young guy
tell his manager that he can't implement code that correctly handles
DST changes *on Unix* because he doesn't have a *PERL* language
library to handle it. In two hours last night I hacked together a C
program (which you can bloody well call from PERL) that does the exact
job. When I demo'd it for the manager, said mgr gave me a FUD answer
about not trusting the zoneinfo DB (because he admittedly didn't know
Unix) and that I was only able to solve the problem because of my
*great expertise*. It's 62 lines of code! How hard can it be?!
Manager said that he couldn't expect guys fresh out of school to be
able to do what I did.
OK, so maybe it takes a day to read through all the manual pages if
you never dealt with that sort of turgid prose before, but what is
wrong with expecting new grads to be able to actually analyze a problem
and use the right tools to solve it? NOOOOOO! Programming has become a
matter of finding the right recipe and following it blindly! That
way you "de-risk the deliverable". As if the project wasn't already
late anyway, and it will actually *DO THE RIGHT THING*, which the
*CUSTOMER* actually *WANTS* (a rare enough thing in this business).
This problem cuts right to the heart of what's wrong with the web. In
another case I recently witnessed, an entertainment portal used
300+lines of Javascript to encode (things like town names into magic
numeric codes) *two* HTML form fields and send them off to back-end
server search processing. WTF? Aren't servers supposed to do the hard
work? And why are you doing this stupid encoding in the first place?
NOOOO! We've got some manager delegating to some half-assed new grad
who doesn't know enough to question the question, and thinks it's
really cool to solve something using the only technology he
knows^W^W^W^W latest buzzword^W web-standard technology! Nobody uses
those minority browsers anyway!
And then we have feckless^Wmecklund97and his amazing:
> > Bruce R Lewis <brl...@my-deja.com> wrote:
> > > mekl...@aol.com (Meklund97) writes:
> > > > If anyone out there likes to do this for fun please help me
> > > > because I am lost and this is the last homework before I am
> > > > done. Please help!!!!!!!!
> >
> > > But with eight exclamation points, it must be
> > > really important for us to do this for you. Please post your TA's
> > > address so that we can send the answer in directly and eliminate the
> > > middleman.
>
> my ta's address is sca...@cs.com if you were really serious. please
> send me a copy also if you could - mekl...@aol.com
The cynicism is just too much for words. This is *exactly* the
attitude that commits these coding/managerial atrocities. "I don't
care I just want it behind me" didn't build the internet. Or Linux. Or
Microsoft. Or Apple. Or *anything* worth having for that matter.
Profs! Please! Flunk 'em if they won't learn. You'll be doing the
world a favor. They're making what could be one of the most elegant of
engineering disciplines into a field I'm becoming increasingly ashamed
ot have my name associated with. How often do I have to say to
non-techies that "software sucks", and yes, I am a
co-perpetrator. Much more of this and I *will* just cash in and
exercise what we used to call the drywall option (although for me it
would be carpentry).
<ObScheme>
So anyway. Let's all push BRL, *and* SSAX, *and* PLT and make the
bastards learn how to program properly using good and apropriate
tools.
</ObScheme>
</rant>
Oh bloody hell. I'm running out of steam here. Sorry If I haven't got
Olin's style, but this is really annoying. The training of these
programming puppies *must* be more than teaching them not to piddle on
the floor. Those heads full of mush must learn to *think*, and a lot
of the people are responsible for that process. Fight for Scheme and
SICP. There's so much more to this craft than using the latest
libraries.
david rush
--
Finding a needle in a haystack is a lot easier if you burn down the
haystack and scan the ashes with a metal detector.
-- the Silicon Valley Tarot
Matthias Felleisen
> Profs! Please! Flunk 'em if they won't learn. You'll be doing the
> world a favor. They're making what could be one of the most elegant of
> engineering disciplines into a field I'm becoming increasingly ashamed
> ot have my name associated with.
Some of us feel like you do and some of us act and flunk such people.
You do need to understand, however, that grades are assigned to students
by the university _not_ the professors. If a student doesn't like the grade,
he sues, and the university president changes the grade. But at least some
of us have to have the guts to stand up and say what we think about these
non-learners.
-- Matthias
Ehud Lamm
> > world a favor. They're making what could be one of the most elegant of
> > engineering disciplines into a field I'm becoming increasingly ashamed
> > ot have my name associated with.
>
I always felt like you do. However, I also feel that most of these
non-learners will be found out after a ten seconds interview. So good
companies won't hire them, and good universities won't accept them for
graduate studies.
But perhaps I am just too optimistic by nature...
--
Ehud Lamm msl...@mscc.huji.ac.il
http://purl.oclc.org/NET/ehudlamm <== Me!
Eli Barzilay
> [...] You do need to understand, however, that grades are assigned
> to students by the university _not_ the professors. If a student
> doesn't like the grade, he sues, and the university president
> changes the grade. [...]
This is amazing! Is this really happening (students suing for better
grades)? Is there any text on the web about such cases?
"Ehud Lamm" <msl...@mscc.huji.ac.il> writes:
> I always felt like you do. However, I also feel that most of these
> non-learners will be found out after a ten seconds interview. So
> good companies won't hire them, and good universities won't accept
> them for graduate studies.
In addition, I think that such programming is an extreme case of a job
that the differences between real programmers and "programmers" are
*huge* -- which is (IMO) the real reason for that timezone incident --
in this was not a rare case (seeing people that one hour of a good
programmer is roughly equivalent) about two days of theirs. The
result of this is that (again, IMO) good programmers will always be a
precious thing to have, which makes me not worry much about graduating
into the dot-com breakdown...
> But perhaps I am just too optimistic by nature...
Me too. Must be an Israeli thing...
--
((lambda (x) (x x)) (lambda (x) (x x))) Eli Barzilay:
http://www.barzilay.org/ Maze is Life!
Max Hailperin
Like the optimistic Israelis, I don't really think this is a major
source of the problem. Yes, I too teach in the U.S., and so yes, I
too know that it occasionally happens. I just don't think very often,
by comparison with some of the other sources of problems.
One very fundamental problem with David Rush's wish is that it has an
underlying assumption: that we (the computer science professors) have
any opportunity to pass or fail those who go into industry. Actually,
only one quarter of those entering programming jobs have computer
science degrees. (See ``Are Too Many Programmers Too Narrowly
Trained'' by David Clark, Computer, June 2000, p. 13.)
Of course, even the quarter who do pass through our educational halls
are subject to grade inflation, etc. But there are much more
pernicious causes of that than the threat of suit. For example,
universities are increasingly making decisions which faculty to tenure
and promote based on on student ratings of their teaching. Students'
ratings are in turn greatly influenced by the grading. So there is a
feedback loop that works towards grade inflation. I'd say this is a
much bigger problem than the occasional post-facto intervention. -max
Joe Lisp
> Some of us feel like you do and some of us act and flunk such people.
> You do need to understand, however, that grades are assigned to students
> by the university _not_ the professors. If a student doesn't like the grade,
For one Harvard prof's solution to the problem of grades, check this out:
http://www.economist.com/world/na/displayStory.cfm?Story_ID=574187
Meklund97
about 10 minutes. goodness!! the one time i try an easy way out and everyone
starts freaking out...well if you are so unsatisfied with the programmers
coming out of college right now, maybe instead of pointing the finger at all of
us "little boys" for doing everything wrong you should take a look at who is
teaching us all this stuff - your generation of programmers - so if you want us
to do things right, then by all means mr. wizard show us how it's done and
communicate the material to us effectively so we can be as perfect as you are.
if you are all going to critique and write big editorials on the downfall of
the computer science field then maybe you picked the wrong profession anyhow -
you should go back to school and get an english major so you can critique and
cut everyone down for a living.
Matthias Felleisen
> Matthias Felleisen <matt...@rice.edu> writes:
>
> > [...] You do need to understand, however, that grades are assigned
> > to students by the university _not_ the professors. If a student
> > doesn't like the grade, he sues, and the university president
> > changes the grade. [...]
>
> This is amazing! Is this really happening (students suing for better
> grades)? Is there any text on the web about such cases?
About two or three weeks ago an appelate court handed down the
decision, and the media reported on this. I read it in the free part of
CHE, but I can't find the URL. You may be able to find it with a
Google search.
As Max correctly points out, this isnt the real problem.
The real problem: industry hires unqualified people, because we
can't produce enough qualified people.
Another real problem: colleges "qualify" many students who aren't
qualified. So, "hey, but if these guys are programmers, we are programmers,
too. They can do X. They make those mistakes. I am no worse. Right?"
Throw in a few students who complain loud enough to get a better grade,
and it doesn't make a real difference.
-- Matthias
David M Einstein
: ok, ok everyone settle down, i did my homework all by myself and it
I doubt that anyone is really surprised by the amount of effort needed.
: the one time i try an easy way out and everyone starts freaking out...
: well if you are so unsatisfied with the programmers coming out of college
: right now, maybe instead of pointing the finger at all of us "little boys"
: for doing everything wrong you should take a look at who is teaching us
: all this stuff - your generation of programmers - so if you want us
: to do things right, then by all means mr. wizard show us how it's done and
: communicate the material to us effectively so we can be as perfect as you
: are.
I believe that many of the people that you deride, especially
Mathias Felleisen,along with his cohorts and minions are working very
hard at that, and seem to be succeeding, though I doubt that their
students would agree with them until two or three years after they
have finished the course.
: if you are all going to critique and write big editorials on the downfall of
: the computer science field then maybe you picked the wrong profession anyhow -
: you should go back to school and get an english major so you can critique and
: cut everyone down for a living.
Don't knock english majors. It is a damn sight easier to
teach an english major to program than it is to teach a CS major to
write. Unfortunately the english major can usually get through life without
programming.
Robert S.
"Ehud Lamm" <msl...@mscc.huji.ac.il> wrote in message
news:9cpb0j$rmq$1...@news.huji.ac.il...
> However, I also feel that most of these
> non-learners will be found out after a ten seconds interview. So good
> companies won't hire them, and good universities won't accept them for
> graduate studies.
> But perhaps I am just too optimistic by nature...
In my 25+ years, (and about 10 different jobs), I have NEVER been asked a
technical question on a job interview!
In my current position, at an R&D group for a large media company, I'm known
for my tough interviewing style...often I'm the only interviewer who bothers
asking questions!
I've found that, for example, about 80% of folks who put "C++" on their
resumes can't answer FUNDAMENTAL questions about the language, stuff that's
in the first few chapters of Bjarne's book. About half the time they're just
lying, the rest of them beleive that since they ran some app builder that
made dialog boxes and some C++ glue code to tie a form to a database, that
they're C++ experts.
It's also funny how many people can't answer basic math questions, (I'm
talking algebra and trigonometry here).
Our group hires pretty well, but we do that mainly by picking top students
from top schools.
Robert S.
> teach an english major to program than it is to teach a CS major to
> write.
That's why you shouldn't hire CS majors, you should hire MATH MAJORS, who
tend to be good in just about any subject.
Eli Barzilay
> About two or three weeks ago an appelate court handed down the
> decision, and the media reported on this. I read it in the free part
> of CHE, but I can't find the URL. You may be able to find it with a
> Google search.
I tried but found nothing - what's CHE?
> As Max correctly points out, this isnt the real problem.
I know, I was just curious...
> The real problem: industry hires unqualified people, because we
> can't produce enough qualified people. [...]
Yeah, that's all true... So as someone who teaches programmers, I can
only be sad about the situation (and try to keep *myself* from going
down (which would probably make it tough for me to teach in America)),
but as a programmer who is concerned about future income I can just
sit back and enjoy the fact that chances are that I'll always be
valuable enough (up to the point where I'll get senile...)
The same goes for language wars -- you can either get frustrated at
people who think that C++ is the best thing that could ever happen to
mankind, or you can smile at the fact that it took you 4 days to make
a web-server that does things that C++ programmers would have to spit
blood to get...
Hartmann Schaffer
>
>> [...] You do need to understand, however, that grades are assigned
>> to students by the university _not_ the professors. If a student
>> doesn't like the grade, he sues, and the university president
>> changes the grade. [...]
>
>This is amazing! Is this really happening (students suing for better
>grades)? Is there any text on the web about such cases?
actually, this is an old thing. i remember a case in the mid 1960s where
we had to fail a guy who somehow had managed to get the qualifications
for our class on paper (easy to guess how). he complained to some advisor
and we had to give him a retest. his preparation was to write algol60
programs like
begin
real pi;
read(pi);
print(pi)
end;
even though we simplified the test (he was a foreign student and needed to
pass something to assure continued financing) he failed miserably again and
tried to get yet another retest. we stood our ground, but it wasn't easy
(no courts involved, though)
> ...
--
hs
----------------------------------------------------------------
"The cheapest pride is national pride. I demonstrates the lack of
characteristics and achievements you can be proud of. The worst loser
can have national pride" - Schopenhauer
Biep @ http://www.biep.org
seconds interview.
> So good companies won't hire them, and good universities won't accept them
for graduate studies.
> But perhaps I am just too optimistic by nature...
I am afraid many companies will accept a degree as proof of *technical*
competence, and spend their interviews finding out other qualities (team
work, customer-directedness, ...). Specific technical *knowledge* (the
stuff that's in manuals) will be picked up along the way.
So maybe companies should be educated not to trust degrees.. Or
universities not to issue them indiscriminately..
--
Biep
Reply via http://www.biep.org
Ehud Lamm
> from top schools.
I always thought that it is much safer to assume that top students are
indeed pretty clever, than to assume that passing grades mean anything.
Bruce Hoult
> seeing people that one hour of a good
> programmer is roughly equivalent about two days of theirs
That's conservative. I've seen plenty of examples where an hour of a
good programmer is worth a month of a poor one. And that's not even a
month of a totally incompetent one -- *those* actually produce negative
results, as often as not.
-- Bruce
Bruce Hoult
http://www.biep.org" <repl...@my.webpage.com> wrote:
> I am afraid many companies will accept a degree as proof of *technical*
> competence, and spend their interviews finding out other qualities (team
> work, customer-directedness, ...). Specific technical *knowledge* (the
> stuff that's in manuals) will be picked up along the way.
And that make good sense, as everyone here should appreciate. So you
don't know anything about some wacky programming language they want you
to use, but you know SICP inside-out. No problem -- you really *will*
learn it.
The problem is how to sort out who can learn the new stuff and who
can't. I don't know how to do that, but asking "puzzle" type questions
is probably reasonable. Maybe dropping the candidate a manual for
something an hour before the interview and then asking open-book
questions about it is a good way too.
I've only once formally interviewed someone (the local university wanted
an outside interviewer for confirmation about a particular candidate)
but I came up with a question that I thought was quite a good one:
What is the total mass of the Earth's atmosphere?
I've found from trying it on my friends and colleagues that this
question is actually a pretty good discriminator and there are many
others like it. I find the good people come up with a workable method
of attack within 10 or 15 seconds and the really good get one decimal
place numerical result within a minute or so.
-- Bruce
brl...@my-deja.com
ten minutes once you applied yourself, it seems to me that the material
*has* been communicated to you effectively.
Don't be surprised that "everyone starts freaking out" when you ask us
to do all the work for you and mail it to your TA. You didn't exactly
make a stellar first impression. Hopefully this kind of behavior is
atypical for you. It touched a nerve because there are students for
whom such behavior is typical.
Will Clinger - Sun Microsystems
If anyone out there likes to do this for fun please help me because I am lost
and this is the last homework before I am done. Please help!!!!!!!!
On 02 May 2001 at 01:26:10 GMT he wrote:
CAN ANYONE DO THESE?? PLEASE?? IT'S MY LAST PROJECT AND I AM DONE WITH SCHEME
FOREVER. IF ANYONE ENJOYS DOING THESE PROBLEMS THEN HAVE FUN AND LET ME KNOW
WHAT YOU COME UP WITH
On 02 May 2001 at 23:12:41 GMT he wrote:
ok, ok everyone settle down, i did my homework all by myself and it took all of
about 10 minutes. goodness!! the one time i try an easy way out and everyone
starts freaking out...
Sorry to change the subject, but it seems to me that the real
problem here may lie not with the young, with the old, or with
the educational system, but with the hardware vendors. When a
computer costs $1000 or more, its keyboard should come with
a reliable shift key.
> well if you are so unsatisfied with the programmers
> coming out of college right now,
My personal opinion is that the programmers coming out of college
right now are no worse than the programmers who came out of college
25 or 40 years ago. In fact, today's programmers may be better,
because even the best of the old programmers were awful in ways
that today's programmers can only dream about. You can read about
one of those old guys, Mel, at
http://www.ccs.neu.edu/home/will/com1205/mel.html
But I'll say this for Mel: He was honest.
> maybe instead of pointing the finger at all of
> us "little boys" for doing everything wrong you should take a look at who is
> teaching us all this stuff - your generation of programmers
Methinks Meklund97 has a point here.
> - so if you want us
> to do things right, then by all means mr. wizard show us how it's done and
> communicate the material to us effectively so we can be as perfect as you are.
Oh dear me. Pardon my laughter. I can think of no response to
this more appropriate than the words of Robert Hunter and Jerry
Garcia (Meklund97: These were holy men of antiquity):
If I knew the way, I would take you home.
> if you are all going to critique and write big editorials on the downfall of
> the computer science field then maybe you picked the wrong profession anyhow -
> you should go back to school and get an english major so you can critique and
> cut everyone down for a living.
This must be Meklund97's editorial comment on the field of
English majoring, but I doubt whether he has the proper
academic credentials to be cutting people down like that.
Will
Evan Prodromou
DME> Don't knock english majors. It is a damn sight easier
DME> to teach an english major to program than it is to teach a CS
DME> major to write.
Actually, I've found that good writers very often make better
programmers than good math or engineering folks. The process of
creating a program is much more analogous to writing a clear, concise
essay than it is to making a bridge or proving a theorem.
Skills that liberal arts majors develop, and especially literature or
English majors, are very applicable to programming. For example:
- using stages in development to aid the writing process
(notes -> outline -> draft 1, etc.)
- composing and decomposing complex ideas
- editing for clarity
- re-writing for efficiency
- keeping in mind a goal, and not straying from that goal
- making different parts of an idea work together as a whole
Most of all, I think that people with writing skills usually remember
that code is written not only to be _run_, but also to be _read_.
~ESP
--
Evan Prodromou
ev...@prodromou.san-francisco.ca.us
Eli Barzilay
at you, then please ignore.]
Evan Prodromou <ev...@prodromou.san-francisco.ca.us> writes:
> >>>>> "DME" == David M Einstein <Dei...@world.std.com> writes:
>
> DME> Don't knock english majors. It is a damn sight easier
> DME> to teach an english major to program than it is to teach a
> DME> CS major to write.
This is flame stuff, the results are expected.
> Actually, I've found that good writers very often make better
> programmers than good math or engineering folks. The process of
> creating a program is much more analogous to writing a clear,
> concise essay than it is to making a bridge or proving a theorem.
>
> Skills that liberal arts majors develop, and especially literature
> or English majors, are very applicable to programming. For example:
>
> - using stages in development to aid the writing process
> (notes -> outline -> draft 1, etc.)
> - composing and decomposing complex ideas
> - editing for clarity
> - re-writing for efficiency
> - keeping in mind a goal, and not straying from that goal
> - making different parts of an idea work together as a whole
>
> Most of all, I think that people with writing skills usually
> remember that code is written not only to be _run_, but also to be
> _read_.
So I'm going to have to go and disagree with you on this one...
You just wrote a list of things which are useful in almost any field I
can think about, so maybe English majors are the new super-humans? I
want to see the mathematician or the engineer that will agree that
development-stages, complex-ideas, editing, re-designing/writing,
goal-focusing and the parts thing are not really necessary qualities
for their field. Or that they are important, but not as important as
for the super-English majors. What is this kind of statement supposed
to say? "English majors are more goal oriented than engineers"?
"English majors are handling more complex ideas than programmers"? Or
that things with the different parts working together... All good
programmers that I have seen *know* that code needs to be read. Most
people here will agree that this is one strong feature of Scheme code.
If all of these things are important to programmers and to writers,
and you have two people with the same IQ, and both studied these
concepts but one from a CS dept and one from an English dept, who
would you take? Both of them have these things as important things,
except that the *specific* complex ideas the English major has been
dealing with for the last N years are pretty different than those
taught in the CS department.
Did you even see what you wrote? Would you hire plumbers for as
network consultants because they know about transferring stuff in a
pipe from point A to point B? How about gardeners who have total
respect for uniformity? What about that old man in the subway who
sits there every morning and solves crossword puzzles - those are
pretty complex things there, he'd be a good sysadmin. Hired killers
are very goal oriented. McDonald's employees know all about queueing
events.
You could talk about the *kind* of people who'd go to each field,
about the fact that the easy money chance draws a lot of
PLEASE-DO-MY-HOMEWORK people, that English majors have such low
chances of jobs in their field that they must be dedicated /
non-conformists / creative people, but your list of points above have
no value at all. As a matter of fact, all good programmers that I saw
were *really good* at writing, maybe I should reverse that list and
claim that good programmers are better writers?
Ji-Yong D. Chung
Meklund97 <mekl...@aol.com> wrote in message
news:20010502191241...@ng-mq1.aol.com...
> ok, ok everyone settle down, i did my homework all by myself and it took
all of
> about 10 minutes. goodness!! the one time i try an easy way out and
everyone
> starts freaking out...
It should have been fairly obvious that, after graduating,
people generally do not enjoy doing other people's homework.
If you, a student, do not like doing homework, what
in the world makes you think, other people here
who generally GIVE homework, would enjoy doing one?
Does that makes sense to you?
Personally, your homework is the last thing I'd want to do
today, or the rest of my life, and I am not even in
a teaching profession.
> well if you are so unsatisfied with the programmers
> coming out of college right now, maybe instead of pointing the finger at
all of
> us "little boys" for doing everything wrong you should take a look at who
is
> teaching us all this stuff - your generation of programmers - so if you
want us
> to do things right, then by all means mr. wizard show us how it's done and
> communicate the material to us effectively so we can be as perfect as you
are.
You sure don't know what you are asking for.--
God forbid an average programmer
of our generation successfully teach you how it is
done and make you as perfect as he is.
And with your attitude young man,
you have a good chance of becoming just like him.
Given that you have first tried to get others
to shoulder what is your responsibility,
you should be the last one on this planet to have
the attitude "... you should take a look at who
is teaching us all this stuff." Obviously, it wasn't
the educators who taught you to get others to do
your homework.
> if you are all going to critique and write big editorials on the downfall
of
> the computer science field then maybe you picked the wrong profession
anyhow -
> you should go back to school and get an english major so you can critique
and
> cut everyone down for a living.
No, son, people here were not writing editorials on the downfall
of the computer science field, but were expressing frustration at
their roles as educators/professionals when they saw your email.
They are dismayed at what you "represent"; their failings.
Good thing your email was, at least, chock full of amusing
and transparent mental gymnatics. Otherwise, it would
be just depressing.
Biep @ http://www.biep.org
news:skvgnho...@mojave.cs.cornell.edu...
> If all of these things are important to programmers and to writers,
> and you have two people with the same IQ, and both studied these concepts
> but one from a CS dept and one from an English dept, who would you take?
> Both of them have these things as important things,
> except that the *specific* complex ideas the English major has been
> dealing with for the last N years are pretty different than those
> taught in the CS department.
When grading CS work at university, I have always graded for natural
language as well as for the CS stuff, so ideally CS students should be
versed in both. (Reality was rather different, however..)
To what degree is your average student of English trained in programming?
ronald schroder
>language as well as for the CS stuff, so ideally CS students should be
>versed in both. (Reality was rather different, however..)
>To what degree is your average student of English trained in programming?
>
showed that training in natural language and training in maths are
equally important for CS professionals.
After all, a computer programming language is still a language and
should be treated as such.Then again, a computer program is nothing
more than a (long) mathematical formula, and (again) should be treated
as such.
In short, reading and writing skills are as important as maths skills.
Ronald Schröder
Ehud Lamm
>
Shorter version of this debate: smarts count.
How surprising.
Shriram Krishnamurthi
> What about that old man in the subway who
> sits there every morning and solves crossword puzzles - those are
> pretty complex things there, he'd be a good sysadmin. Hired killers
> are very goal oriented. McDonald's employees know all about queueing
> events.
I suspect that hired killers would make excellent computer scientists.
I pity the fool who fails to answer a codewalk question from one!
Perhaps the current labor shortage in IT/CIS stems from the fact that
companies are advertising in all the wrong places: /CACM/ and /IEEE
Computer/ instead of /Soldier of Fortune/.
Shriram
Evan Prodromou
EB> [I'm just freaking out here, if you feel nice and the sun is
EB> smiling at you, then please ignore.]
The sun is absolutely beautiful out here, but how could I ever pass up
an opportunity to practice the dialectic with a worthy opponent and
thus improve my own ideas?
Me> Actually, I've found that good writers very often make better
Me> programmers than good math or engineering folks.
This is the thesis, right here.
EB> So I'm going to have to go and disagree with you on this
EB> one...
But of course! I'm not sure what your point exactly is, though. Would
you say that writers are not -better- at programming than, say,
engineers, but equal? Or worse? Or totally unqualified whatsoever?
Or were you just saying that I'm a big dork?
EB> You just wrote a list of things which are useful in almost any
EB> field I can think about, so maybe English majors are the new
EB> super-humans?
I can't comment on how important editing for clarity is to, say,
lumberjacks. However, I can say pretty clearly that learning to
prepare a text can help someone who is going to be a programmer.
I don't think that that was a particularly comprehensive list of all
skills ever needed to survive, or even to be a good programmer. In
particular, there wasn't any mention of interpersonal skills or time
management.
EB> I want to see the mathematician or the engineer that will
EB> agree that development-stages, complex-ideas,
EB> editing, re-designing/writing, goal-focusing and the parts
EB> thing are not really necessary qualities for their field.
Yes, but are they the primary field of study? I'd say that they pick
up these skills as a means of surviving, and not as part of the formal
curriculum. For an English major, creating your own texts, or studying
how someone else created texts, are the main point of study.
Other educational backgrounds give different sets of skills that
translate into different styles of programming. An industrial
engineer, for example, learns some skills that are applicable to a
particular style of programming -- making little machines that do
stuff efficiently. I just happen to think that this style of
programming ends up with less useful programs in the long term than
those that are made with the goal of preparing a text.
EB> "English majors are more goal oriented than engineers"?
?? I don't think I said that. I said that English majors learn to
tailor a text towards a goal, and to edit out stuff that doesn't help
towards reaching that goal. This is an important skill for a
programmer, too.
EB> "English majors are handling more complex ideas than
EB> programmers"?
Why are you contrasting "English majors" with "programmers?" I was
talking about English majors who become programmers. Surely you can't
be suggesting that only people who did CS majors in college become
programmers, or that they are the only ones who are any good?
EB> All good programmers that I have seen *know* that code needs
EB> to be read.
Yes, indeed. And I would say that programmers with a liberal arts know
this better. And that programmers who don't realize this are bad
programmers.
EB> If all of these things are important to programmers and to
EB> writers, and you have two people with the same IQ, and both
EB> studied these concepts but one from a CS dept and one from an
EB> English dept, who would you take?
A couple of things: first, I was contrasting English majors and
engineering or math majors. So, I wasn't actually making the
comparison you made.
Second, have you ever hired someone right out of a CS program for a
professional programming job? They usually need as much on-the-job
training as any film major or art student. They have something of a
leg up in that they grok the basic premise of programming, and have
familiarity with the constructs used in common programming languages,
but that is usually about it. Knowing how to program and being a good
programmer are two very different things.
EB> Both of them have these things as important things, except
EB> that the *specific* complex ideas the English major has been
EB> dealing with for the last N years are pretty different than
EB> those taught in the CS department.
And, of course, both are different from any problems that I'll
actually need them to work on, or that they'll be dealing with in 5 or
10 years.
EB> Did you even see what you wrote? Would you hire plumbers for
EB> as network consultants because they know about transferring
EB> stuff in a pipe from point A to point B?
No, but I'd hire a network consultant who got a degree in plumbing. I
wouldn't demand that he had a degree in... whatever it is they give
people degrees in so they can become networking folks.
EB> What about that old man in the subway who sits there every
EB> morning and solves crossword puzzles - those are pretty
EB> complex things there, he'd be a good sysadmin.
Hrm. I dunno about that one, but I'd probably give him a shot at doing
database design work.
EB> Hired killers are very goal oriented.
They tend to be bad in team situations, though, and they don't do very
good reporting.
EB> McDonald's employees know all about queueing events.
I'd hire a good burger-flipper, sure. They have great team experience,
and McDonald's corporate culture puts a big emphasis on continuous
improvement of an efficient, repeatable process.
I'm only being semi-serious here, but I think my main point is that
considering programming as a strictly mathematical or engineering
discipline not only masks out a large amount of good input from other
fields of knowledge, and unfairly disqualifies people with
understanding in those fields, but also may be quite incorrect, and
lead to a style of programming that makes for bad software.
EB> You could talk about the *kind* of people who'd go to each
EB> field, about the fact that the easy money chance draws a lot
EB> of PLEASE-DO-MY-HOMEWORK people, that English majors have such
EB> low chances of jobs in their field that they must be dedicated
EB> / non-conformists / creative people, but your list of points
EB> above have no value at all.
Preparing a text is very much like creating a program, and the same
skills are needed for both. I think the list is entirely accurate.
EB> As a matter of fact, all good programmers that I saw were
EB> *really good* at writing, maybe I should reverse that list and
EB> claim that good programmers are better writers?
"Good programmers are better writers than people who aren't good
programmers."? I wouldn't say that's necessarily right, but I'd
probably say that that's it's a defensible claim.
thi
Other educational backgrounds give different sets of skills that
translate into different styles of programming. An industrial
engineer, for example, learns some skills that are applicable to a
particular style of programming -- making little machines that do
stuff efficiently. I just happen to think that this style of
programming ends up with less useful programs in the long term than
those that are made with the goal of preparing a text.
i've found liberal arts majors who become programmers to be heavy on the
metaphorical, which can be useful if the problem domain supports easily
maintained abstractions. however, such programmers often shirk the
gritty details and when it comes down to it, indulge mostly in hiding
their lack of grounding (w/ more metaphors). ymmv.
thi
Eli Barzilay
> But of course! I'm not sure what your point exactly is,
> though. Would you say that writers are not -better- at programming
> than, say, engineers, but equal? Or worse? Or totally unqualified
> whatsoever?
As someone just said, this NG is full of people who are related to
teaching in a university - being among these people (though on a much
smaller place than some), I find it, um, annoying when people come and
say that the kind of people I have been trying real hard to educate
are not better than English majors. I think that we might have a
misunderstanding about the definition of "engineer" - I assume talking
about CS majors (which are a subset of engineering in many places I've
seen around here) but you seem to be talking about engineers in
general - so I will restrict myself to "CS majors" (if you want to
compare non-CS engineers to English majors, that's something I really
couldn't care less about). Anyway, my point is that CS education
makes better programmers. You say that English majors learn property
X which is good for programmers and I say that if that's the case, not
only will CS majors learn X too, but they will learn it in a context
which is *by definition* closer to programming. I see no way that any
reasonable line of logic will make you disagree.
> I can't comment on how important editing for clarity is to, say,
> lumberjacks. However, I can say pretty clearly that learning to
> prepare a text can help someone who is going to be a programmer.
So maybe your claim is that preparing a text is something that English
majors study and CS majors don't study? I think that enough time has
passed having CS around for it to be mature enough to avoid such
obvious things as "studying X is good for programming and we don't
teach it. I have never yet seen anyone who thinks that students are
being taught too much CS and they should get more education in
literature or something instead of some CS material - all CS programs
I have seen were extremely intense, and all teachers were constantly
facing the problem of trying to teach more stuff while what they get
to is less than they wanted.
> EB> I want to see the mathematician or the engineer that will
> EB> agree that development-stages, complex-ideas,
> EB> editing, re-designing/writing, goal-focusing and the parts
> EB> thing are not really necessary qualities for their field.
>
> Yes, but are they the primary field of study? I'd say that they pick
> up these skills as a means of surviving, and not as part of the
> formal curriculum. For an English major, creating your own texts, or
> studying how someone else created texts, are the main point of
> study.
Take a few seconds to reverse the argument, and it makes perfect
sense: substitute "Math" for "English" and "proofs" for "texts". This
applies to other items on your list and I'd say that most of them are
better fitted to a mathematician or a CS major than to an English
major. Goal: when you program there is the very clear goal of a
working solution within resource constrains and also when you prove
something. Developement in stages: without this I would say that any
CS major is bound to get completely lost, and I'm sure that any CS or
Math majors would consider their domain's objects that they compose
and decompose on an order of magnitude more complex than the English
major's concepts. etc etc. As a demonstration of the last point, how
often do you see someone reading a single page for more than a week?
Compare the answer for CS and for English majors.
> Why are you contrasting "English majors" with "programmers?" I was
> talking about English majors who become programmers. Surely you
> can't be suggesting that only people who did CS majors in college
> become programmers, or that they are the only ones who are any good?
They are the *only* ones who were *educated* as programmers. If some
CS major starts a brilliant career making movies I wouldn't jump to
the nearest film newsgroup suggesting that CS education makes you a
better director. When English majors become programmers, they need to
fill in gaps that CS people spent lots of time over.
> A couple of things: first, I was contrasting English majors and
> engineering or math majors. So, I wasn't actually making the
> comparison you made.
Oh, so you do mean engineering as in the non-CS part... Then we go
into the I-don't-care zone.
> Second, have you ever hired someone right out of a CS program for a
> professional programming job? They usually need as much on-the-job
> training as any film major or art student. They have something of a
> leg up in that they grok the basic premise of programming, and have
> familiarity with the constructs used in common programming
> languages, but that is usually about it. Knowing how to program and
> being a good programmer are two very different things.
I have seen one two many uneducated (in CS) programmers who didn't
even realize the concept of runt-time (making their program fail
miserably when moving from toy test data to a customer). This is true
for many other things (not knowing about algorithms, not knowing about
control statements other than VB arrows, not understanding recursion
etc etc). (As funny as it may sound, I have personally witnessed an
implementation of a *three* level hierarchical tree - three meaning
that there was *no code* for handling deeper trees.)
> EB> Both of them have these things as important things, except
> EB> that the *specific* complex ideas the English major has been
> EB> dealing with for the last N years are pretty different than
> EB> those taught in the CS department.
>
> And, of course, both are different from any problems that I'll
> actually need them to work on, or that they'll be dealing with in 5
> or 10 years.
And this is the core of what I disagree on. While buzzwords (like
JavaJunk) will certainly change, CS programmers will change better
with them since they have good *general* knowledge - as hey say in the
beginning of most Scheme courses - it's not a Scheme course, but a
course about the principles of programming. English majors might
learn Perl as fast as CS majors, but not having the education, they
will still be shitty programmers (and Perl is not chosen at random
here, I think the main thing I don't like it is that every 4-year-old
writes Perl scripts that makes my hair fall).
> EB> Did you even see what you wrote? Would you hire plumbers for
> EB> as network consultants because they know about transferring
> EB> stuff in a pipe from point A to point B?
>
> No, but I'd hire a network consultant who got a degree in plumbing. I
> wouldn't demand that he had a degree in... whatever it is they give
> people degrees in so they can become networking folks.
If you were making a precondition to this that the plumber *looked*
more intelligent or whatever, I would at least see some vague logic.
Without it, I'm completely lost on this kind of reasoning.
> EB> You could talk about the *kind* of people who'd go to each
> EB> field, about the fact that the easy money chance draws a lot
> EB> of PLEASE-DO-MY-HOMEWORK people, that English majors have such
> EB> low chances of jobs in their field that they must be dedicated
> EB> / non-conformists / creative people, but your list of points
> EB> above have no value at all.
>
> Preparing a text is very much like creating a program, and the same
> skills are needed for both. I think the list is entirely accurate.
Either you stopped reading my reply a this point or you're at a
completely different wavelength. The list is completely irrelevant to
this discussion just because it is relevant to both fields (and many
others).
> EB> As a matter of fact, all good programmers that I saw were
> EB> *really good* at writing, maybe I should reverse that list and
> EB> claim that good programmers are better writers?
>
> "Good programmers are better writers than people who aren't good
> programmers."? I wouldn't say that's necessarily right, but I'd
> probably say that that's it's a defensible claim.
Good programmer => intelligent person => good writer.
The arrows are probabilistic implications.