On Tuesday, May 21, 2002, at 04:24 , Peter Scott wrote:
[..]
> This is somewhat religious, of course,
and the rest of software development is not a matter
of studying the medieval heresy trials and resolving
which side of satan your current project is on????
> and I'm not invalidating your approach, just exposing beginners to a
> different point of view.
points well made peter! my complements! I do not see us as
competing on the point - just trying to make sure that the
'full orthodoxy'[1] of perl is defended.... to be honest
until you raised the matter, and then I tried it free hand,
and got bit again on the same folly I had done before -
I really had forgotten why I didn't do it that way...
{ it's the miles kids, not the years... }
It would be wonderful if 'teaching/learning' were at least as exact a
science as perl ..... In the old days - things were a little
'hotter' in places like comp.lang.[perl|emacs|YourKultHere]
and we lost opportunities to 'teach and learn' - in
part because we were out trumpetting the greatness of a
new set of arcane symbols... So I am trying to avoid some
of that as I try to find a balance between 'text to describe'
and 'code to show'... { as well as trying to figure out what
is a 'homework' assignment question - my complements to those
who have helped in that space. }
To be honest, I am an 'old shell hack'[2] - so I am perchance a
little overly sensitive to the issues of folks who know how
to do x,y,z in sh/sed/awk - and have been working under the
cultural repression of not being taken seriously, because it
is 'just some scripting' - who are transitioning into perl,
so I do not wish them to feel 'frowned upon' - there is enough
of that in the RealCoder[tm] flag waving festivals as it is.
I think I got into the 'habit' of Getopts::Long because of
the latent gnuishNeff[3] of the beasties wandering around me and
it was simpler to write code that way to keep the beasties from
snuffelling and huffelling...
There are problems down that lane as well - what I refer to
as the 29 command line flags required to make thisPieceOfOpenNeff
actually do anything other than spit out error messages. Since it
was simpler to just 'cache' those commandline options in a
simpler set of shell scripts - and have short command line
typing to get the dope back from 'the real code'....
So no, I am not at all planning to support that level of madness
where we embark upon the SLOC - source lines of code - models of
failed analysis - to attempt to 'validate' this trick over that.
It's silly when management thinks that adding more SLOC adds
something to the process. The flip side is equally true -
the 'obfuscatory <yourLanguageHere>' contests are to allow
people the chance to do exactly what they know they should not do,
namely make totally unmaintainable code.
ciao
drieux
---
[1] - yes, It is important to me that I do things the
orthodox way - it's a lifeStyle Thing... In Perl, this
is not half as easy as it might look....
[2] cf:
http://www.wetware.com/drieux/CS/lang/shellScript/JustSomeScripting.html
The unpleasant part is that only a few of the decent shops understand
that 'perl' is not a scripting language and have what it takes to
list it in the section of 'coding languages'....
and trust me I am NOT KIDDING about 'shell libraries' -
http://www.wetware.com/drieux/CS/lang/shellScript/shlib.html
IF you find yourself creating these - it is time to get your
head around h2xs - and make the cut over to PERL NOW - no need
to bleed! Make Modules - the name space management solutions
alone are worth every farthing....
[3]
http://www.wetware.com/drieux/screeds/LiNox.html
honest, some of my best friends use emacs, I do not hold
that against them - just so long as they remember which
planet they are on....
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
