Looking for balanced critique of Perl
Mark Lawrence
mark at DRD.Com
Wed Jun 27 05:03:09 AEST 1990
[I sent this via e-mail and then thought that the comments might be of
general interest]
Don,
I saw your post in comp.lang.perl and wanted to share our (admittedly
limited) experience with Perl.
Being fairly novice to UNIX (I'm the senior UNIX user in-house having used
it since 1986, others are much less comfortable with it), basic
capabilities that experienced folks might take for granted (effective use
of RegExps, awk, sed, sophisticated use of shell and so forth) has come very
hard to us. Perl sort of tied everything together in one place, gave all
these things a sense of cohesiveness, and now we understand a lot more
about the features we discover in awk, sed, shell and the like that Perl
obviously derived from. Incidently, we use Perl to write a lot of the code
that makes up the core of an application that I'm the project manager for.
It involves data management (because the application deals with a lot of
data from various sources) and generating code to model structures,
initialize maps and so forth is a very straightforward job with perl (as it
probably would be with a combination of shell, awk and sed, but as I say --
it took perl to put it all together for us).
The documentation ain't great, but I found that a single serious
read-throughof the notorious man page gave me enough to get going pretty
well. At present, I think it lacks heavily in the area of packages and how
to use them effectively. The reference cards that Vromans put together are
an invaluable help. Of course, Schwartz and Wall claim that a book is in
the works, and we'll probably purchase multiple copies when it comes available.
The text-oriented-ness of Perl seems really logical to us and having all the
capabilities in one tool seems like it should be a performance win.
Actually, the original reason I got interested in it was because awk didn't
have a debugger (except: bailing out at line n :-) and perl did.
In summary, our experience with Perl has been fairly positive. Obtuse code
*can* be written in Perl, but then, I've seen some obtuse shell/awk/sed
scripts, too. Certainly, Larry seems to be able to top anybody in terms of
reducing an algorithm to the tersest and most efficient set of statements,
but then, he wrote it. Doesn't bother me. I get done what needs to get done.
--
mark at DRD.Com uunet!apctrc!drd!mark$B!J%^!<%/!!!&%m!<%l%s%9!K(B(918)743-3013
More information about the Comp.unix.wizards
mailing list