🚀 go-pugleaf

Steve Franklin wrote:
> 
> In article <315ED94F.73E@airmail.net>, Brian Ebarb <ebarb@airmail.net> wrote:
> > Whether it's true or not in the artificial universe of academia is
> > beyond my experience since I'm self-taught. Nonetheless, 15 years of
> > professional programming has shown me that a) many bugs in PC-based
> > development (Windows, network, database...) are caused by too much code,
> > usually written by someone who isn't really sure what they're doing, and
> > b) most universities do a deplorable job of preparing their charges with
> > the skills necessary to hit the ground running and be even moderately
> > productive in a commercial programming environment.
> 
> Interestingly, I have found just the opposite. Of the programmers I've worked
> with (which amounts to many over about 7 years of varied employment), the
> worst code was usually written by the self-taught. Now, I should also add
> that some of the neatest and most creative code was written by the self-taught
> (often suitably given the name 'hackers').
> 
> Creative and interesting, however, is not always that useable in a
> productivity-oriented situation (e.g. industry) where unreadable, sloppy,
> obfuscated code is hardly useful. Getting a university education in
> computer science guarantees a certain standard of knowledge, an
> opportunity to learn disciplined code writing (if I see another person who
> doesn't indent their code I'm gonna...), and in most cases, a chance to
> get some good industry experience (co-op).
> 
> Now, I am not going to generalize that all self-taught programmers are
> poor, but I do feel that an education can greatly benefit one's
> programming talents avoiding the slow and tedious trial-and-error process
> that the self-taught go through.
> 
> Having said that, I do sympathize with your position Brian, being
> self-taught myself :)
> 
> Steve
> 
> --
> Steve Franklin                        franklsm@tuns.ca
> Halifax Nova Scotia                   http://www.tuns.ca/~franklsm/
> Technical University of Nova Scotia   Masters of Applied Science (EE)

Now, I'm not proposing that we pick teams of self-taught vs. academic programmers for a 
code-off! Both camps suffer from chronic mediocrity.

Stroustrup himself once stated that (paraphrasing) that "..C has the power to allow the 
programmer to shoot himself in the foot. In C++ the programmer can blow his whole foot 
off..". (That's not an exact quote, but it's as close as I can get to right now. You get 
the drift.) All that stuff about 'reusable code' and 'improved maintainability' of C++ 
code comapred to straight C looks good on paper, but most programmers don't follow that 
paper. Sometimes it's laziness, sometimes it's because of schedule constraints that 
force the programmer to emphasize function over elegance. For whatever reason, there's a 
lot of crappy code out there and a lot of blame to go around. A degreed programmer 
doesn't intrinsically have an advantage over a non-degreed programmer when faced with 
these hurdles.

I was a hacker 12-15 years when the term was endearing, but having learned all I felt 
I wanted by taking apart Mom's toaster (metaphor), these days I'm more of a fuddy-duddy. 
I remember how to research and debug, and I remember how to attack a problem from 
various fronts -- but I insist that my code, and the code produced by those I mentor, be 
well documented and conventional. Given the choice between using a solution provided by 
a third-party vendor and writing a solution using libraries provided by our compiler 
vendor, I tend to prefer that we roll our own. That way we aren't hhandcuffed to a 
vendor that may (or may not) not be there tomorrow when the thing has to be maintained. 
'Hackers' have the reputation of using anything and everything within reach to achieve a 
solution. SO -- I'm not a hacker now. 

-- 

- Brian Ebarb