Software Engineering - Is there a future in software Engineering for US?

Author

Is there a future in software Engineering for US?

ashia

2006-01-26, 3:59 am

Hi all:

I was reading the threads in response to the lady asking about career
change and I am quite surprised on the number of kind words and
encouragement that was given by this newsgroup.

I remember couple of years ago I went to some IEEE conference and there
was actually a topic on outsourcing and the future of US engineers. The
room was packed with upset looking engineers and defensive looking
management. I know this topic is so chewed on that everyone probably had
said they piece of it. But seriously, from a logical point of view, if
you have a 18 year old come to you today and ask you what major they
should choose, would you still recommend computer science? Is this a
waste of their time as they would have to switch career later on?
Is there such a thing as "US software engineer" anymore??

Bradley K. Sherman

2006-01-26, 3:59 am

In article <FsudnUN48uSLt0XenZ2dnUVZ_s-dnZ2d@comcast.com>,
ashia <the7of9@comcast.net> wrote:
>Hi all:
>
>I was reading the threads in response to the lady asking about career
>change and I am quite surprised on the number of kind words and
>encouragement that was given by this newsgroup.
>
>I remember couple of years ago I went to some IEEE conference and there
>was actually a topic on outsourcing and the future of US engineers. The
>room was packed with upset looking engineers and defensive looking
>management. I know this topic is so chewed on that everyone probably had
>said they piece of it. But seriously, from a logical point of view, if
>you have a 18 year old come to you today and ask you what major they
>should choose, would you still recommend computer science?
>

What is this idiocy about 'recommending' majors? If you need a major
recommended, go to trade school and learn TV Camera Operation or
Welding or something. If an 18-year-old has a flair for logic
and arithmetic and likes the idea of mastering a computer, then
Computer Science is a great major. If he/she likes reading history,
writing poetry and can't stand to be alone, it probably sucks.

--bks

Peter K.

2006-01-26, 3:59 am

bks@panix.com (Bradley K. Sherman) writes:

> What is this idiocy about 'recommending' majors? If you need a major
> recommended, go to trade school and learn TV Camera Operation or
> Welding or something. If an 18-year-old has a flair for logic
> and arithmetic and likes the idea of mastering a computer, then
> Computer Science is a great major.

Logic and arithmetic could imply mathematics, philosophy, accountancy,
engineering or any number of other disciplines. "Mastering a computer"
is about as fraught with peril as you seem to think TV Camera
Operation or Welding are.

The reason people ask for recommendations is to try to scope out their
options. Doing anything less would be just plain dumb.

Ciao,

Peter K.

Bradley K. Sherman

2006-01-26, 3:59 am

In article <u64o764v1.fsf@remove.ieee.org>,
Peter K. <p.kootsookos@remove.ieee.org> wrote:
>bks@panix.com (Bradley K. Sherman) writes:
>
>
>Logic and arithmetic could imply mathematics, philosophy, accountancy,
>engineering or any number of other disciplines. "Mastering a computer"
>is about as fraught with peril as you seem to think TV Camera
>Operation or Welding are.
>
>The reason people ask for recommendations is to try to scope out their
>options. Doing anything less would be just plain dumb.

All the options for majors are listed in the catalogue.

And all those other disciplines are just fine too.

As an undergraduate it's your job to choose. Recommending
a major is like recommending a wife.

One might well describe what the major is like, and answer
specific questions about it, but recommend? No.

--bks

Phlip

2006-01-26, 3:59 am

ashia wrote:

> I know this topic is so chewed on that everyone probably had said they
> piece of it. But seriously, from a logical point of view, if you have a 18
> year old come to you today and ask you what major they

Fear is a poor counsel.

I would tell them that one of the biggest trends in SE logistics is
co-locating teams with their customers.

The customers are generally in the USA, folks. And where they are not, there
are programmers for those customers, too.

The closer a team works with its customer, the leaner and more efficient
their activities. Any salary discrepancies between various populations
become less important. (Put another way, fewer programmers, closer to their
customers, can collect more of the profits from their code.)

--
Phlip
[url]http://www.greencheese.org/Z

Land[/url] <-- NOT a blog!!!

JXStern

2006-01-26, 3:59 am

On Wed, 25 Jan 2006 17:47:02 -0800, ashia <the7of9@comcast.net> wrote:
>said they piece of it. But seriously, from a logical point of view, if
>you have a 18 year old come to you today and ask you what major they
>should choose, would you still recommend computer science?

I would certainly recommend they take some CS classes, if they have
any interest or aptitude, because there will be computers in the
future!

That is a very different question from whether I would recommend they
pursue computer science as a career. Entrepreneurship in the computer
field, absolutely. Employee status? Probably not. Certainly not.

>Is there such a thing as "US software engineer" anymore??

No. *

But then, there hardly ever has been, and much of the current
negativity is just what you expect of an industry as it matures.

J.

* That's not an entire "no", but so much work is outsourced, so many
H-1Bs are here, that even the work that remains is increasingly
low-paying, structured for minimum skill levels, and isolated from the
main corporate opportunities, and increasingly low-status. "No" is
close enough for me.

Martin Brown

2006-01-26, 3:59 am

Bradley K. Sherman wrote:
> In article <u64o764v1.fsf@remove.ieee.org>,
> Peter K. <p.kootsookos@remove.ieee.org> wrote:
>
>
> All the options for majors are listed in the catalogue.
>
> And all those other disciplines are just fine too.

And some of them may be a better preparation for life in software
development than the more obvious computing courses. One company I knew
liked to recruit top grade classicists and then train them in house.
>
> As an undergraduate it's your job to choose. Recommending
> a major is like recommending a wife.
>
> One might well describe what the major is like, and answer
> specific questions about it, but recommend? No.

I am with you on this one. It is fair enough to warn people if there is
fierce competition and not many good jobs available at the end (eg
astronomy), and also on the relative merits and strengths of various
cities, universities and departments.

But *they* should decide what interests THEM and make their own decision
about what course to study and where. Another option is go to a
university that is flexible fairly laid back about changing subject.

The difference in some subjects is sufficient that some people at
university don't like the subject they excelled at in school.

However, there is still a shortage of good software engineers (and
probably always will be).

Incidentally those nervous about being outsourced as software engineers
could try trades like "plumbing" or "plastering" it is very difficult to
outsource those and they make as good a living as average software
engineers (or at least did until everyone jumped on the bandwagon). UK
is set to have a glut of plumbers shortly...

Regards,
Martin Brown

gds@best.cut.here.com

2006-01-26, 7:02 pm

ashia <the7of9@comcast.net> wrote:
>Hi all:
>
>I was reading the threads in response to the lady asking about career
>change and I am quite surprised on the number of kind words and
>encouragement that was given by this newsgroup.
>
>I remember couple of years ago I went to some IEEE conference and there
>was actually a topic on outsourcing and the future of US engineers. The
>room was packed with upset looking engineers and defensive looking
>management. I know this topic is so chewed on that everyone probably had
>said they piece of it. But seriously, from a logical point of view, if
>you have a 18 year old come to you today and ask you what major they
>should choose, would you still recommend computer science? Is this a
>waste of their time as they would have to switch career later on?
>Is there such a thing as "US software engineer" anymore??

I wouldn't recommend it to anyone unless he or she felt so
passionately about CS that he or she couldn't see themself in any
other field. IMO, there's just too much uncertainty. As far as
places to which work is outsourced goes, the risk is more acceptable
because the wages that are being offered are a substantial increase
over what they might otherwise be making.

--gregbo
gds at best dot com

H. S. Lahman

2006-01-26, 7:02 pm

Responding to Ashia...

THis happens to be a Major Hot Button for me...

> But seriously, from a logical point of view, if
> you have a 18 year old come to you today and ask you what major they
> should choose, would you still recommend computer science? Is this a
> waste of their time as they would have to switch career later on?
> Is there such a thing as "US software engineer" anymore??

Competent people will be able to get jobs in their chosen field so long
as the field is still relevant. The tricky part is understanding
competence.

The software industry is facing a paradigm shift today that is very much
like the paradigm shift that faced Manufacturing in the '80s. In the
'80s the conventional view of product quality became outmoded as the
PacRim nations demonstrated that much better quality was achievable when
one viewed testing as a process monitoring tool rather than a product
quality tool. IOW, quality was a process issue, not a test issue. That
paradigm shift caused very large and painful dislocations in the US and
Europe as markets shifted. The result was that the US and Europe had to
play catch-up for a decade because they were not prepared for product
quality to be a major competitive issue.

The same thing is facing the software industry today. Customer's are
beginning to notice that software is the only thing in their lives now
that breaks. Managers are noticing that /all/ software projects have
the kind of schedule and budget problems problems that only
unprecedented mega-projects have in other engineering areas. That
creates a situation where anyone who rocks the boat by providing better
reliability, better schedule adherence, and better budget containment
will trigger exactly the same sort of competitive advantage that the
PacRim did in the '80s.

The US and Europe got very complacent (though Europe less than the US)
with preeminence in their software industries. Like the PacRim built
manufacturing industries in the 60s, emerging nations have been building
software industries from scratch since the '80s. In doing so they have
taken to heart the /existing/ SE principles that the Western
Establishment ignored (just like the Western Establishment ignored the
preaching of Deming and Jurand in the '50s). So they largely did it
Right when they laid their foundation. The rapid growth of outsourcing
is just a manifestation of getting it right as an industry paradigm.

Cheap labor is really a red herring in the outsourcing debate. Any such
cost benefits are easily outweighed by the hidden costs of managing
arm's length, distributed projects. The real reason software is being
outsourced today is because offshore shops can usually do it better from
the perspective of reliability, schedule, and budget. Until the Western
Establishment figures out that software development is -- first and
foremost -- a team process rather than a collection of creative
individuals Doing Their Thing, outsourcing will just accelerate.

However, the demand for software is not going away. If the West wants
to maintain a position in the industry it is going to have to get its
act together and recognize that testing is about monitoring processes,
not product quality, while schedule and budget stability lies in
repeatable processes, not guesswork. Eventually I think it will, just
as it played catch-up to the PacRim in the '90s. That' simply because
the market is too big to let go and the basic ideas have been around for
decades. So all one needs is a change in leadership.

So, to answer the original question, I think an 18-year-old won't have a
problem in pursuing software development. They just need to recognize
the paradigm shift and focus on the SE process side of things so they
can be a leader rather than a follower.

*************
There is nothing wrong with me that could
not be cured by a capful of Drano.

H. S. Lahman
hsl@pathfindermda.com
Pathfinder Solutions -- Put MDA to Work
http://www.pathfindermda.com
blog: http://pathfinderpeople.blogs.com/hslahman
(888)OOA-PATH

JXStern

2006-01-27, 9:57 pm

On Thu, 26 Jan 2006 21:24:46 GMT, "H. S. Lahman"
<h.lahman@verizon.net> wrote:
>Cheap labor is really a red herring in the outsourcing debate.

Agreed.

My take on it is that US management likes the idea the work is done
far, far away. Sort of like sewage treatment.

>The real reason software is being
>outsourced today is because offshore shops can usually do it better from
>the perspective of reliability, schedule, and budget.

Very dubious.

I would be surprised if there is any major difference in quality
between projects done here or there - especially when most projects
here are heavily staffed with H-1Bs anyway.

> Until the Western
>Establishment figures out that software development is -- first and
>foremost -- a team process rather than a collection of creative
>individuals Doing Their Thing, outsourcing will just accelerate.

Can we get some people in Banaglore to respond to this?

I agree with your implicit criticism of technical management, but
whether the US cowboys are worse than outsource cowboys, or the US
flunkies are worse than the outsource flunkeys, or the US
pointy-haired-bosses are worse than the outsource pointy-haired-boses,
I very much doubt.

J.

H. S. Lahman

2006-01-27, 9:57 pm

Responding to JXStern...

>
>
> Very dubious.

I don't recall the exact numbers but a few years ago the proportion of
CMM, ISO, or Baldridge certified shops to the total was much higher
offshore than in the US. That alone is a pretty convincing indicator of
process orientation, which I believe to be the core of the paradigm
shift for both the manufacturing ('80s) and software realms ('00s).
Since such shops are well documented to have fewer and less serious
schedule, reliability, and budget problems, that proportionality is
indicative of an overall average competitive advantage. IOW, I'm sure
they have their zoos as well but they have fewer.

On a more informal basis, I was doing the rounds of process conferences
before retiring. One thing that struck me was the offshore
representation among both speakers and attendees. It was far higher
that at other software conferences.

Another indicator is the contributing authors to refereed SE venues like
the IEEE and ACM rags. Names from emerging nations now pretty much
dominate those publications. IOW, there is a changing of the guard
among SE leadership. Not all of them went home after getting schooled
in SE in the West, but quite a few did and those are the people who have
been defining the industry in those countries. That sort of strategic
positioning takes time, just like it took the PacRim three decades to
emerge as a force. But the offshore shops have been avidly involved in
that technology transfer since at least the '80s.

*************
There is nothing wrong with me that could
not be cured by a capful of Drano.

H. S. Lahman
hsl@pathfindermda.com
Pathfinder Solutions -- Put MDA to Work
http://www.pathfindermda.com
blog: http://pathfinderpeople.blogs.com/hslahman
(888)OOA-PATH

JXStern

2006-01-27, 9:57 pm

On Fri, 27 Jan 2006 18:33:46 GMT, "H. S. Lahman"
<h.lahman@verizon.net> wrote:
>
>I don't recall the exact numbers but a few years ago the proportion of
>CMM, ISO, or Baldridge certified shops to the total was much higher
>offshore than in the US.

Your evidence is secondary, mine is anecdotal but primary. Take the
system I was just working on, done offshore (via a chain of
outsourcing entities). I have no idea what their CMM status or other
certifications were supposed to be, but it was all big-name stuff.
And the product absolutely sucks: design gaffes galore, putrid
quality, incomplete implementations - even the project plan sucked,
but I concede that was done here (and accepted by client management!).

What did they deliver? Ooodles of documents, maybe twenty pages of
word docs and pictures per web page, seven months worth of effort, I'm
told. Absolutely non-agile. Which raises an interesting point, who
would say that a high CMM rating in a non-agile process, is even worth
having?

I have no great respect for most US shops, BUT I have yet to be
impressed by most offshore shops, either. And many direct and
indirect indicators of the shops that I've seen over the last few
years, suggest that even if there WERE quality work being done 10,000
miles away, the local connections would louse it up good, in the very
process of frittering away any putative cost benefits.

J.

H. S. Lahman

2006-01-29, 6:59 pm

Responding to JXStern...

>
>
> Your evidence is secondary, mine is anecdotal but primary. Take the
> system I was just working on, done offshore (via a chain of
> outsourcing entities). I have no idea what their CMM status or other
> certifications were supposed to be, but it was all big-name stuff.
> And the product absolutely sucks: design gaffes galore, putrid
> quality, incomplete implementations - even the project plan sucked,
> but I concede that was done here (and accepted by client management!).

As you say, this just suggests your anecdote context was one of the
unlucky ones. B-)

> What did they deliver? Ooodles of documents, maybe twenty pages of
> word docs and pictures per web page, seven months worth of effort, I'm
> told. Absolutely non-agile. Which raises an interesting point, who
> would say that a high CMM rating in a non-agile process, is even worth
> having?

The only way an outsourcing project can be agile is if the project is
subdivided into IID increments. But agility at the level of, say, XP
increments is simply not possible because the customer is not available
on a daily basis. So one of the big hidden costs of outsourcing is
defining requirements properly up front for large increments. That
sounds like a classic problem of your situation: great documentation of
the solution to the wrong problem.

BTW, I have a counter anecdote. I was involved to two coast-to-coast
projects between the same two divisions in the company separated by
nearly a decade. Basically the same technologies and pretty much the
same people on both projects. The first was a dismal failure and was
canceled after a year (ca '91?). The second worked like a dream (ca
'99) For example, total hardware and software integration was completed
and validated in three days on the other coast with only phone calls to
our coast. (For scale this was a megabuck hardware system with ~1MLOC
software.) The difference lay in the processes used to define things
like application partitioning and requirements allocation. Essentially
the disaster occurred before we had our act together and could even
define the requirements to be allocated.

*************
There is nothing wrong with me that could
not be cured by a capful of Drano.

H. S. Lahman
hsl@pathfindermda.com
Pathfinder Solutions -- Put MDA to Work
http://www.pathfindermda.com
blog: http://pathfinderpeople.blogs.com/hslahman
(888)OOA-PATH

JXStern

2006-01-29, 6:59 pm

On Sun, 29 Jan 2006 19:10:29 GMT, "H. S. Lahman"
<h.lahman@verizon.net> wrote:
>BTW, I have a counter anecdote.

I don't say that outsourcing *has* to fail.

But y'know, the difference between theory and practice, is greater in
practice than it is in theory.

J.

Bradley K. Sherman

2006-01-29, 9:56 pm

In article <Fq8Df.5898$Ix.3989@trnddc07>,
H. S. Lahman <hsl@pathfindermda.com> wrote:
>
>BTW, I have a counter anecdote. I was involved to two coast-to-coast
>projects between the same two divisions in the company separated by
>nearly a decade. Basically the same technologies and pretty much the
>same people on both projects. The first was a dismal failure and was
>canceled after a year (ca '91?). The second worked like a dream (ca
>'99) For example, total hardware and software integration was completed
>and validated in three days on the other coast with only phone calls to
>our coast. (For scale this was a megabuck hardware system with ~1MLOC
>software.) The difference lay in the processes used to define things
>like application partitioning and requirements allocation. Essentially
>the disaster occurred before we had our act together and could even
>define the requirements to be allocated.
>

Always plan to throw one away! (_Mythical Man Month_)

--bks

JXStern

2006-01-30, 3:57 am

On Mon, 30 Jan 2006 01:28:07 +0000 (UTC), bks@panix.com (Bradley K.
Sherman) wrote:
>Always plan to throw one away! (_Mythical Man Month_)

I dunno, these days we don't even toss the prototypes, we just
refactor the devil out of them!

I think the social anthropology of software projects has changed a lot
in thirty years, maybe as much or more than the technology or the
economics.

You throw away anything much recently?

J.

Bradley K. Sherman

2006-01-30, 8:00 am

In article <49irt1tua4ebgesh3lktkqe0bjtqoalapk@4ax.com>,
JXStern <JXSternChangeX2R@gte.net> wrote:
>On Mon, 30 Jan 2006 01:28:07 +0000 (UTC), bks@panix.com (Bradley K.
>Sherman) wrote:
>
>I dunno, these days we don't even toss the prototypes, we just
>refactor the devil out of them!

Amounts to the same thing. The important thing is that the
prototype must be re-worked.

--bks

JXStern

2006-01-30, 7:01 pm

On Mon, 30 Jan 2006 12:27:51 +0000 (UTC), bks@panix.com (Bradley K.
Sherman) wrote:
>
>Amounts to the same thing. The important thing is that the
>prototype must be re-worked.

Still makes a change from MMM days, I think.

Here in Los Angeles, people will knock down an old house, save about
one pebble of the original foundation, and build an entire new
structure three times as large, and it falls under the heading of
"renovation".

Our tools are a lot better, and, frankly, the systems much more
complex, than OS/360 was. We *can* refactor, and there seems more
sense in it.

Not sure what my point is, actually.

:)

J.

JXStern

2006-01-30, 7:01 pm

>> And all those other disciplines are just fine too.
>
>And some of them may be a better preparation for life in software
>development than the more obvious computing courses. One company I knew
>liked to recruit top grade classicists and then train them in house.

Seriously? I mean, this is since, say, 1990?

In olden days when there were no CS majors, there was a jones for
physics majors, "had the brains for it", and the like, but that was
mostly a long, long time ago.

Well, mostly. I did just run into one place that hired some kind of
biology PhD as their CIO. And yes, the place does run like a petri
dish, so I hope that's what they wanted.

J.

Sponsored Links