[Israel.pm] Is there a lack of (good) Perl programmers?

guy keren choo at actcom.co.il
Fri Mar 2 12:01:04 PST 2007


Gabor Szabo wrote:

> Regarding what Guy wrote finding non-perl programmers and teaching them
> Perl - that would be great, more work for me :-) - but as far as I can tell most
> companies prefer to get the programming knowledge with the programmer.

there is a way to partially circumvent that - the way of using 
underground perl secret agents ;) (see below)

> One less thing they need to invest in. Besides, I think that except of
> the really
> top programmers, for most Java or C programmer it would take a few months
> to be effective in Perl. Just like for me it would take months to be reasonably
> good in Java or C#.

i have to somewhat disagree with that - i saw people becoming effective 
in perl (read - 'productive', not 'got to their top performance') in 
less time than this.

> So in fact even if a good Perl programmer can produce application
> 3-4 times faster than a good Java programmer taking into account a
> learning period of (let's say) 3-6 months you might start to see the
> advantages only after 6-9 months. (with my totally mad up numbers)
> In most cases this is way too long to be in delay.
> 
> Besides, and here is my issue, how many Java or .NET programmer will
> be happy to learn Perl for a project? I think very few. The reason might
> be that Perl does not have high profile or that there are not many Perl
> jobs out there they can use their knowledge in or maybe other reasons.

[i answered this in my previous post - i won't repeat it here to reduce 
the noise]

> So I wonder what could be done by us and maybe by the companies
> to change this situation?

spread secret perl-agents into existing companies. ever since i worked 
in a "perl workshop" company, and came to see how far perl can go, i 
have pushed small perl projects everywhere i went to (except for one 
company who already had a more influential python+jython secret-agent).

in the current company, i developed an underground automatic-testing 
project. when it came into light, and the manager decided to make it 
official - the language was already chosen - unless someone wanted to 
re-write it - and no one had/has time to re-write it. so now when other 
programmers want or need to extend it - they do it in perl. it was 
surprising to see how many (young) programmers "once knew perl", and 
didn't find it too hard to get into speed with it.

it's still limiting with regards to which programmers can extend it - 
the older ones are more reluctant to learn perl, so it's mostly the 
young ones that are taking control of this project. they don't feel very 
bad about it (despite it being in the non-sexy language perl) for two 
reasons:

1. they know it's their side-prject, to aid in developing the real project.

2. it got enough visibility by the management by now, that they feel its 
importance.

so what you, gabor, can hope for, is to teach perl to as many people you 
can, spreading the sede. then find a way to plant a perl secret-agent in 
companies whose people you tought perl, and hope that the secret agent 
creates perl projects that will grow in time. the perl secret-agent must 
be profficient in the language of choice of the company (be it C++, java 
or C#) and have enough "huzpa" to create underground projects, and 
enough relentlessness to push them into view.

i assume there could also be other ways. after all, this is the first 
time i managed to push a perl project into a scale that requires other 
programmers to be involved with it. even thought i did manage to create 
2 new perl programmers in 2 previous jobs...

ofcourse, the real challenge will be whether the project survives after 
i leave the company. it already survived the fact that i hardly write 
code for it any longer. i truly wonder if it'll survive the arrival of a 
python secret agent if i'm not there ;)

--guy



More information about the Perl mailing list