Re: CoBOL and Contracting (new thread)

On Wed, 6 Aug 2008 10:45:44 -0700 (PDT), "klshafer@xxxxxxx" <klshafer@xxxxxxx> wrote:

On Aug 5, 11:57 pm, Robert <n...@xxxxxx> wrote:
On Tue, 5 Aug 2008 16:29:31 -0700 (PDT), "klsha...@xxxxxxx" <klsha...@xxxxxxx> wrote:

Be wary of jobs in small cities, because the client company is likely to be a technical
backwater having employees terrified of losing the only programming job in town. They see
contractors as threats. In such places, an in person interview with *workers*, not
managers, is a must. You especially want to meet the guy who's been there 20+ years. If he
won't talk to you, pass.

We've come to call that phenomenon "contractor envy", and it is *very*
real. Caused me a lot of trouble on a couple of projects. One way the
fearful try to bait you is leading you into a discussion about money.
(_Never_ discuss money details on a project - it is right up there on
the same forbidden list of subjects as politics and religion.) Some
are quite brazen: "How much money could *I* expect to make as a
contractor?" or even more blatant: "Just how *much* money are they
paying you?"

The diplomatic way I've found to deflect this, while answering the
question, is to provide them a formula: "Well, it's competitive with
permanent salaries, calulated in the following way. I receive no
benefits, no social security, no 401K match, no vacation, no training,
none of that, and altogether, that's worth about 30-35% above your
base salary. And then there's the risk premium. Generally, that's
about 15%, which I have found to be *barely* adequate to cover my
downtime when I am not billable at all. So here's what you do - take
*your* salary, which I am sure is competitive, multiply by 1.5, divide
by 2000 hours a year, and that is your hourly rate."

I say, "The contract says I'm not allowed to discuss salaries." If they can't find the
answer online, they're clueless.

In private, I see nothing wrong with discussing salary with *other contractors*, not
company employees. We're all in the same exploitation boat. The pimp knows how much we all
make. We're entitled to know whether he's paying one of us $20/hr less than another.

And I agree wholeheartedly with your observation about interviewing
with the old-timers. Now that I think about it, the times they were
included in the vetting process, I had much better luck getting the
assignment.

The more important issue is keeping the assignment. Managers make hiring decisions;
old-timers make firing decisions. In some shops, being TOO competent will get you fired
faster than being incompetent. I wish I had a surefire question to identify such shops.
I've gooten good results with, "What development methodology do you use?" Bad shops
don't have any, but won't say so during an interview. They'll give a bull*** answer,
which is evident in their inability to answer a simple followup question. Good shops
describe in detail what they do have, even if informall.

3. Do you think that the relative proportion of permanent vs.
contractor status has changed significantly in the last ten years or
so? Does the future of CoBOL exist more in contracting now than in
permanent employment?

The most evident trend is toward outsourcing to India, predominantly maintenance, support
and testing. I estimate 10-30% of Cobol jobs go to contractors; the largest percentage are
still held by employees.

I wonder - is there a niche market for a "near shore" :-) maintenance
service bureau?

No. You can't compete with third world salaries.

I do think at least *some* of the bloom is off the
offshoring rose. My evidence for this is sparse, and only anecdotal,
but from sources I consider very reliable. Now that rates have been
"capped" domestically, I think it is more feasible than several years
ago.

I'm in the middle of it and I see the trend accelerating. First the client outsources
maintenance at a fixed price, then the outsourcer moves it offshore. Performance is easy
to measure, and glitches come out of the outsourcer's pocket.

Development is another story. The concern there is quality, not price. Clients and
outsourcers tend to fear cheap labor won't get it right. Hell, our own people can't always
get it right.

You overlooked asking about platform. It's 60-70% mainframe, 30-40% Unix, near zero on
Windows. I may be biased because my specialty is Unix. The mainframe market seems to have
a surplus of supply over demand. There is less competition on Unix.

Yes, I did overlook OS and "environment". Interesting what you say
about Unix, and the relative comparision of supply/demand with
mainframe. I wonder if CoBOL, being an ANSI standard language, and
Unix, being open-source, are being forced into a shotgun marriage of
convenience to compete with the proprietary likes of MS and C#?

Java and gcc compete with C# for the mainsteam. Cobol is a specialty language in the Unix
world.

Micro Focus owns the Unix Cobol market. Its compilers and runtimes are relatively
expensive. Others here talk about using Fujitsu Cobol; I've not seen it in any big US
company.

OK -- as I was typing this the idea occurred to me to do an informal
poll with DICE, state by state, and see what the job posts are for
CoBOL, with no other keywords. A rough measure, but it will give us an
idea. Can also get counts for employee/contractor, though some posts
will contain references to both.

CoBOL - no restrictions - 968
CoBOL - fulltime employee - 492
CoBOL - contractor (all kinds, including contract-to-hire) - 552

That's a reflection of DICE, not the market.

Are you saying that DICE has a "bias" toward contracting? That would
be good to keep in mind, if it is true.

Yes, DICE is a place for hiring contractors. Employers use more traditional channels to
look for employees. Indeed.com is a better measure of the total market because it
consolidates many channels, including newspapers.

By state - (without regard to permanent fulltime vs. contractor)
Alabama - 2
Alaska - 1

Indeed.com has better analytical tools (and better listings). Select Job Trends, search on
Cobol,  then combinations of keywords such as "cobol mainframe" and "cobol chicago". Use
the relative graph. To compare two trends, put a comma between the keywords (undocumented
feature). Try "cobol, plsql" for a shocker. Even worse "cobol, c#'. Note that Java shows
signs of aging.

Wow. I took a look at indeed.com and try the Trends. Very, very
interesting. Also, quite curious that I stumbled upon a significant
CoBOL listing, just browsing, at the University of Chicago Medical
Center. That's quite a good school, and *that* is an understatement.
That they are still using CoBOL must say *something*.

It's gone now. I'm nearing end of contract in Chicagoland


.