Programmers with network engineering skills

Peter Kristolaitis alter3d at
Tue Feb 28 19:32:15 UTC 2012

Education in theoretical concepts is definitely a problem in general, 
but I've found it's particularly noticeable in the technology field 
which has become increasingly commercialized (or commoditized);  where 
once a "sysadmin" was a long-running process, grown from a person with 
the mindset of "how does this work?" and "let's build cool stuff!", 
eventually culminating in someone who knew operating systems, 
networking, and programming inside out (see: "UNIX greybeard" types), 
nowadays people haul off to a 5 day course in "Advanced System 
Administration(TM)", write a certificate exam or 15, and call themselves 
a "sysadmin".

There still are people with the "greybeard" mentality -- all of the very 
best sysadmins I know are of this type, and many know real programming 
(that is, data structures, algorithms, etc) better than the 
mass-produced developers our there -- but unfortunately, they're 
relatively rare, and usually very hard to recruit.   You have to have a 
very interesting project first and foremost -- salary is important, but 
building new, cool stuff is usually the #1 factor for top-notch 

- Pete

On 12-02-28 01:51 PM, Owen DeLong wrote:
> This problem is not limited to programming.
> Education in general has moved from teaching conceptual knowledge
> re-inforced by practical examples and exercises to teaching methodological
> and/or procedural knowledge without any effort to convey concepts.
> It's much like the difference between teaching a man to buy a fish using
> cash vs. teaching a man more generalized economic skills and money
> management.
> In the former case, you get a man who can eat fish as long as he still
> has some cash. In the latter case, you get a man who can keep cash
> coming in and use it to obtain a varied diet and other things he may
> want.
> Today, the indoctrination mills (hard to call them education centers
> at this point) churn out people who are good at repeating the same
> process and solving the same problems over and over.
> Unfortunately, when faced with a problem that doesn't look like something
> from their text book, they either become completely lost or they take
> the hammer approach (when the only tool you have is a hammer, every
> problem looks like a nail).
> I'm not sure how to solve this. Teaching methodologically is much much
> faster than teaching conceptually and the endemic lack of patience makes it
> hard to get people to sit still long enough to learn conceptually.
> Owen
> On Feb 28, 2012, at 6:03 AM, John Mitchell wrote:
>> <rant>
>> I would wholeheartedly agree with this, but I believe its worse than
>> just that. I used to categorize myself as a full developer, now I'm
>> slightly ashamed to be tainted with that brush since there's so many
>> people using the term who don't know the first thing about programming.
>> It used to be that when you were taught programming, you were taught
>> concepts (when to use a for loop, while loop, Boolean algebra), then
>> you built on the foundations by learning advanced concepts  (data
>> structures, how to program concurrently using semaphores etc etc), you
>> would then pick some optional classes to make up for some non
>> programming specific knowledge (networking, linux admin, etc etc).
>> I now have a lot of friends who work in academia and they are worried
>> by a decline (as am I when trying to hire new talent). Currently the
>> teaching process is one of learning to program like a monkey, monkey
>> see monkey do. People are no longer taught to think for themselves, but
>> instead taught to program in a specific language (PHP, Java, rarely C
>> or C++ any more, C#, or VB) and that is all they know. I don't believe
>> this is a failing with the lecturers but with the fundamental change in
>> attitudes to programming.
>> One of the tests I give all interviewees is write a very short program
>> in a language they have never ever used before ( personally I recommend
>> ) since this gives people a
>> chance to show they can program rather than being able to tell me "I
>> know PHP" or "I know C", suprisingly very few newer programmers can
>> make it through, or even try it, because the concept of thinking for
>> themselves is so last year.
>> </rant>
>> On 27 February 2012 20:02:13, Brandt, Ralph wrote:
>>> Generalists are hard to come by these days. They are people who learn
>>> less and less about more and more till they know nothing about
>>> everything. People today are specializing in the left and right halves
>>> of the bytes....  They learn more and more about less and less till they
>>> know everything about nothing.  And BTW, they are worthless unless you
>>> have five of them working on a problem because none of them know enough
>>> to fix it.  Worse, you can replace the word five with fifty and it may
>>> be still true.
>>> I know of three of these, all gainfully employed at this time and could
>>> each find at least a couple jobs if they wanted.  I am one, my son is
>>> two and a guy we worked with is the third.
>>> At one time (40 years ago) the mantra in IS was train for expertise, now
>>> it is hire for it.  Somewhere there has to be a happy medium.  I suggest
>>> this, find a good coder, not a mediocre who writes shit code but a good
>>> one who can think and learn and when you talk about branching out with
>>> his skill set he or she lights up.  His first thing on site is take the
>>> A+ networking course.
>>> No, I do not sell the courses.  But I have seen this kind of approach
>>> work when nothing else was.
>>> Ralph Brandt
>>> Communications Engineer
>>> HP Enterprise Services
>>> Telephone +1 717.506.0802
>>> FAX +1 717.506.4358
>>> Email Ralph.Brandt at
>>> 5095 Ritter Rd
>>> Mechanicsburg PA 17055
>>> -----Original Message-----
>>> From: A. Pishdadi [mailto:apishdadi at]
>>> Sent: Sunday, February 26, 2012 8:27 PM
>>> To: NANOG
>>> Subject: Programmers with network engineering skills
>>> Hello All,
>>> i have been looking for quite some time now a descent coder (c,php) who
>>> has
>>> a descent amount of system admin / netadmin experience. Doesn't
>>> necessarily
>>> need to be an expert at network engineering but being acclimated in
>>> understanding the basic fundamentals of networking. Understanding basic
>>> routing concepts, how to diagnose using tcpdump / pcap, understanding
>>> subnetting and how bgp works (not necessarily setting up bgp). I've
>>> posted
>>> job listings on the likes of dice and monster and have not found any
>>> good
>>> canidates, most of them ASP / Java guys.
>>> If anyone can point me to a site they might recommend for job postings
>>> or
>>> know of any consulting firms that might provide these services that
>>> would
>>> be greatly appreciated.

More information about the NANOG mailing list