Re: why did you chose compiler development?

Scott Stanchfield <scotts@metaware.com>
25 May 1997 13:41:25 -0400

          From comp.compilers

Related articles
why did you chose compiler development? ast@halcyon.com (1997-05-22)
Re: why did you chose compiler development? salomon@silver.cs.umanitoba.ca (1997-05-25)
Re: why did you chose compiler development? walter@bytecraft.com (Walter Banks) (1997-05-25)
Re: why did you chose compiler development? scotts@metaware.com (Scott Stanchfield) (1997-05-25)
Re: why did you chose compiler development? thetick@scruz.net (Scott Stanchfield) (1997-05-27)
Re: why did you chose compiler development? marssaxman@sprynet.com (1997-05-27)
Re: why did you chose compiler development? ian@five-d.com (1997-05-30)
Re: why did you chose compiler development? conway@mundook.cs.mu.OZ.AU (1997-05-30)
Re: why did you chose compiler development? dwight@pentasoft.com (Dwight VandenBerghe) (1997-05-31)
| List of all articles for this month |

From: Scott Stanchfield <scotts@metaware.com>
Newsgroups: comp.compilers
Date: 25 May 1997 13:41:25 -0400
Organization: MetaWare Incorporated
References: 97-05-242
Keywords: practice

Andrew Tucker wrote:
> I just quit my current job to take my first compiler development
> position. In doing so, lots of people have asked me why I want to get
> into compiler development and it has made me think about it. I have
> my standard reasons (listed below), but I wonder what makes you enjoy
> this area?
>
> My reasons:
>
> 1) It was the most interesting (and challenging) part of my undergrad
> work. I'll never forget the prof I had -- he influenced my career
> substantially.


Likewise for me. I took it as a grad course when I was an undergrad
at te University of Michigan. I enjoyed the programming but the
theory didn't seem to map to it, at least the way that prof taught it.
End result, "hated it"...


I took another compiler course during my Master's work at Johns
Hopkins. The only reason I took it was because a prof who had taught
two other classes (C++ and OOA&D) was teaching it and I really liked
his teaching method. He let us do the project with whatever tools we
wanted and I discovered PCCTS (1.10 at the time). The way he taught
it, the theory made sense and was interesting. It wasnt't just "the
mathematical definition of a context-free grammar is..." (which is the
way the Michigan guy taught it, yawn.) Sure, we went over that, but
instead of having the tests be regurgitations of definitions, it was
primarily applications of algorithms and writing grammars/REs. I
ended up really liking the stuff!


> 4) It has a rich history and some of the most famous computer
> scientists (ie Knuth, Grace Hopper, Turing) have made significant
> contributions.


I work for Frank DeRemer (father of LALR) and with Tom ("Very Fast
LALR Parsers") Pennello at MetaWare. Great guys, and it's fun to
argue LL vs. LALR with them... not that I'll change their minds,
though.


> 6) There are wide and deep mainstream applications. Aside from the
> obvious language compilers, there are macro interpreters, SQL,
> conversion utilities, configuration files, digital design...


BINGO! The biggest problem with schools is that the only time parsing
concepts come up are in courses with scary names and reputations like
"Compiler Construction" and "Parsing Theory" (you know how everyone
jumps at the chance to take a course with "theory" in its title.


These are really useful methods and tools that most programmer don't
know exist. I've used yacc (yuck) and PCCTS (smile) for several non
compiler projects. Someday I'll write a book on "Parsers ain't just
for compilers."


Finally, I've got a twisted mind, and you need that to enjoy this
stuff...


-- Scott


--
Scott Stanchfield
MetaWare Incorporated
Santa Cruz, CA
See my PCCTS Tutorial at http://www.scruz.net/~thetick/pcctstut
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.