LT 99.5 percentile need not apply

[Homer Phillips]

I hear this from almost every software company. "We hire the top 1% or less," they all say.

[Jeroen Wenting]

Nice piece, and his description of the top people never applying for jobs because they're always asked may be partially true.
But more likely those people will have to apply for jobs just like the rest of us, they're just less likely to be turned down and therefore apply to less openings than most.

[peter wooster]

Originally posted by Homer Phillips:
I hear this from almost every software company. "We hire the top 1% or less," they all say.

As usual, there's a lot of truth in what Joel says.

[Eric Lemaitre]

Hi all !

This post, depite quite useless on a purely pragmactical point of view (interesting but conclusion is that nothing concrete can be drawn of it), is anyway refreshing and an occasion to wonder for recruiters "have I the right process for recruiting the right guys ?" and for candidates "am I relying on right criterias to be hired ?".

For recruiters, the only relevant process to be sure to hire the right guy is to make him work on the project and checking whether he copes with it or not, not only in a consequent amount of time but compared to fellows working along on the very same thing. Only past experience will prove the right guy was hired, it will always be impossible to foresee by advance it will be the case. So all recruiters' claims about "we hire the best" are vain, this can NEVER be proved by any way.

For candidates, there is no relevant process at all to be sure to be hired, despite job's only aspect is technical all will rely on appearances and speculation. In fact on an interview the very first minute will be critical, relying on gesture, clothes, behaviour, answer matches compared to recruiter's expectations, ... but totally ignoring years of technical knowledge for lack of time. So perhaps candidates should relay on pure relationship management on interviews, trying to gather by advance sensible answers most likely to be expected by recruiter. Considering this aspect their technical knowledge is vain too for getting the job.

In a word, present "must match criterias" behaviour of recruiters leads to nothing, it can only help to make a decision when piles of similar resumes are available, but "must match" won't ever mean for sure "best match" when someone is hired. In fact these last years companies recruiters forgot to keep sensible and trust their employees. The present (seen as such at least) attitude of recruiters who don't care who to hire because crowds are available, or don't care to fire much because outsourcing is so easy, is neither reliable nor sensible. Every employee has a real business value, and he should be trusted for his company's success drives his own success and vice versa. This is being forgotten or neglected. Perhaps this is why statically more than half US workers nowadays don't care any longer about their company. US companies cannot carry on (apparently) to lower costs at any price because ther is always a price at the end.

I particulary liked the signature of Tim Holloway under such considerations (recruiter for junior position in florida 2 forums above) : "If you extend the currently popular business philosophy to its ultimate end, we'll end up with apps produced in no time, for no cost that do nothing. And from the way some major corporate sites work, it won't be much longer".

Something should be done for changing US companies mentality, their former internet bubble mentality "trees will grow to the sky" was logically doomed by advance, but their present opposite one "costs can be decreased to nothing" is too. When will a sensible average reappear ?

Beat regards.

[Tim Holloway]

In the old days, we had Goldilocks, the Three Little Pigs, and Snow White and the 7 Dwarfs. Now we have "people with 20 years J2EE experience", "Economies of Scale", and "Leveraging our Synergies". That last one is even scarier than when the wicked witch tried to shove Hanel and Gretel in the oven.

Obviously it's statistically impossible for every company to be employing the top 1%, and it's been pretty well determined that about 80% of the work gets done by 20% of the people anyway. Efficiency is highly overrated, anyway. People are horrifically prone to measuring the wrong things and then committing resources based on the measurements. Worse yet, the usual response is to make more wrong measurements and thus compound the issue.

Anyway, all of the above mythologies aside, the most deadly fairy tale of them all in IT starts not with "Once upon a time", but with "All you have to do is...". Nothing, alas, is as quick or as easy as it seems, and it really is true that while 90% of the work takes 90% of the time, the last 10% of the work - the part that differentiates a model made of cardboard from an actual skyscraper - really does take 90% of the time. I can think of no other industry where the difference between the kind of work that amateurs do and a professional job are so hard to distinguish. Until the cardboard starts to buckle, anyway.

Certifications have never helped. It seems like almost anyone can get the knowledge needed to pass a cert. Various other techniques have also failed. I kind of wonder if maybe we shouldn't simply revert to the old guild approach, where in order to be recognized in the craft, an applicant shouldn't have to submit a "master piece" of work for peer review. That, however, has its own problems. Politics, for one. And if rumour is correct, the same person who produces a gem one time may produce swill the next or vice versa.

For my own self, I'd simply like to beg that people realize that contrary to intuition, the production of software is not a mechanical process. It can't be done on an assembly line, and that no amount of pressure can speed it up any more than you can make corn grow faster by going out and shouting at it or printers print faster by tugging on the paper as it comes out.

The current business climate places a premium on "first delivery" and "lowest price". But you don't get something for nothing. As consumers, we need to realize the difference between what's a good value and what's merely cheap.

Eventually, the cost-of-living imbalances that allow Western countries to buy goods at an unnatural discount will normalize themselves. Not only that, but one of the prime precepts of business for the entire history of the human race is expected to end within 45 years: the idea that the market (e.g. human population) is open-ended (expanding). For about 2050 the population of the planet is expected to stabilize. It's already doing so (or even shrinking) in some countries.

In short, changes are coming. Markets will move to a zero-sum game. I hope that this might change some of the fundamental precepts of business, and I'd like to see a world where everyone can afford a good set of the basics and would therefore see the value in searching for superior merchandise.

Admittedly, the world 5 years from now isn't a big concern in an economy where we reward executives from quarter to quarter (even when they tank the company). But FWIW, I myself will continue to agitate for quality. I hate cheap junk. I never found the thrill of opening new packages sufficient to offset the nuisance of discarding the old broken stuff. I just can't quite make myself believe that it simply disappears once I've thrown it out.

Wow. Did I really write all that drivel?

[Homer Phillips]

This post, depite quite useless on a purely pragmactical point of view

You did not notice that it correlates well with Matloff's assertion?

Industry employers claim a ``desperate'' labor shortage, but in fact their extreme pickiness in hiring shows they are not desperate at all. They are flooded with re'sume's but hire only a small percentage of the applicants.

Two or three times in the last few months I've gotten emails/postings from local employers claiming they just can't find any good local java people. One of them even went on to say that if they open the search up to nation wide they are flooded with resumes. But they did not want to waste their time going through the flood.

Soon there will be mounting pressure on Congress to increase the H1-B quota again because of the desperate labor shortage. Maybe Martin Fowler needs the top 1%, but I doubt it.

What the industry wants is not employees. The industry wants to buy silver bullets and winning lottery tickets. IMO, great software developers are made and not born. The theory of labor identified in the report commisioned for congress indicates that employers do not want to or refuse to invest in the formation of human capital because they are afraid it will move to a new high bidder.

There's a few contradictions here.

a) Hire good people but you can't invest in them because they bail. You can't hire good people because they won't or are afraid to move.

b) Labor cannot invest in formation of intellectual capital because only experience on the job is meaningful. Industry cannot invest in the formation of human capital because they are afraid of losing their investment.

c) People claim they can't get jobs because the industry won't hire them. They industry claims they can't find anybody.

Joel, if he has the credibility, needs to introduce the industry to Pogo. I forget how Pogo says it, perhaps even how to Pogo spells his name. As Pogo says - I've seen the enemy and I are it.

I am reasserting my claim since you missed it Eric that US Congress needs to quit meddling with the supply of high tech or college educated labor in the United States. IT'S COMMUNISM.

There is no shortage of good, profitable or available talent. Industry needs to buck up or go out of business. Government needs to enforce the age discrimmination act of 1969.

[Homer Phillips]

Certifications have never helped.

Maybe master's degrees and the U have never helped.

Certs created by industry and well proctored can distinguish knowledge people from non-knowledgable. Certs provide industry world-wide recognition. Certs provide a curriculum, self study pace, and economic efficiency.

If people believe in life-long learning certs meet the challenge.

Did you ever notice that they with validate a resume with brain-bench or teck-check? So if you have the stuff on your resume and you can pass a test you have credibility. If you just can pass the test you have no credibility.

Interviews are important but I been on interviews where I was sure my technology knowledge was higher than the interviewer.

Now I have my issues with a cert having the same 45 questions for two years. But if Sun, Oracle, IBM says a person is something like a certified bean developer and you need a bean developer... Does he show up for the interview and to work? Woody says 90% of success is showing up.

[Tim Holloway]

Whoa. When I say that "certifications have never helped", I'm referring to the fact that there's a major difference between knowing things and being able to do things. The complaint about wet-behind-the-ears johnies comping fresh out of college full of "book-larnin'" but no common sense is downright stereotypical, but in the IT profession there's also the perennial complaint that people fresh out of school may be able to determine whether an algorithm's NP-complete but are totally lost when directed to put together something that extracts data from 8 different legacy systems and feed them all to the General Ledger.

The original SCJP exam, for example, was a joke. It spent more time attempting to trip up people based on what they knew about C++ that was different in Java than on solving real problems. More recent example questions I've seen weren't much better - addressing little-used options best left in the reference section or a requirement for deciphering code that - while legal code - is probably grounds for assault on its author. Something, alas, often encountered in real life, but I'd gladly vote for DEcertification for anyone who produces such monstrosities.

I'm not inherently down on certs. It indicates a willingness to invest time and energy into memorization for at least as long as it takes to regurgitate it back on the exam. It's just that software certs don't accurately reflect what software people have to do. The RHCE exam, for example, gives you about half a day to take a broken machine and produce a functional server. That very closely matches both the amount of time and the tasks involved in the real world - I used to have to do the OS/2 equivalent, but the only paper I ever received for it was a paycheck.

Software development, on the other hand, isn't a 4-8 hour endeavor. For a meaningful real-world equivalent, you're talking weeks or months of labor. Not merely having the knowledge crammed into you head, but keeping that knowledge (or at least being able to retrieve it efficiently) and using it in many different ways to produce a useful result.

Medieval craft guilds probably would have scorned the very idea that an applicant could aspire to the rank of master by hacking out something in a day or less, just as today and advanced degree requires not merely an exam, but a submission of original work.

But returning to the topic of "know" vs. "can", One of my best friends is widely abused around town as someone who, despite 20+ years in the trade, including running his own IT service bureau is full of lots ok knowledge but little ability to deliver. Of course on bad days, I even get accused of that myself.

"Brownie points" inside the trade are one thing. When you go to [name of Fortune 50 corporation removed]'s website and have to try 2 different OS's and 3 different browsers before you can login (and the winner WASN'T Windows or IE), that obviously is incompetence that even laymen can understand. I have no idea whether the various lackwits who designed, implemented, managed, tested, and approved it for release had any certs, diplomas or other formal credentials at all. Presumably, they did, because Fortune 50 corporations tend to be really insistent about such things. Regardless, the site is garbage. It's almost completely nonfunctional. And it's not an isolated incident either for that company, or on the Internet as a whole.

Ideally the consuming public would get sick of this kind of thing and demand a minimal level of quality. After all, if public-facing stuff like this is done so poorly, is critical data being handled any better?

Oh wait. This is the planet where people will sit and listen to how important their call is until their brains start leaking out their ears...

[Kripal Singh]

I need more details about the magical screening process that found the "top 0.5 to 1%" and how accurate it is .

[Mike Gershman]

Tim:

I agree with you about the SCJP. Because there are a quite few people who have core Java down pat, Sun produces grade spread by using trick questions with errors no programmer would make and obscure typos which would show up instantly during compilation. Test constructors like Kathy Sierra, who is a quite a self-made expert in cognitive theory and would make a professional magician look like an amateur at misdirection, excels at this.

What about the SCJD, which involves programming a small but complete system that would take a pro two weeks to complete, and which is judged on quality, robustness, maintainability, etc. Do you think the SCJD bears little relationship to job effectiveness?

How would you compare the SCJD to an hour interview someone who has three years in a j2ee group on his resume? He may have three years of solid experience or he may have spent most of that time covering meetings, because he is so well spoken, and providing night support for other people's code, because his own code is correct but unmaintainable.
[ March 03, 2005: Message edited by: Mike Gershman ]

[Henry Wong]

While I agree with the article, it doesn't mention one major point. It is also very unlikely that a company will get a top 1% person via an general resume search.

Great people are remembered, and woo'ed from company to company. If by chance, a company gets one of these people, it is probably done via a quick search by a manager who knew the person -- not by a extensive search, thru hotjobs or a headhunter.

Henry

[Mark Herschberg]

Originally posted by Mike Gershman:
What about the SCJD, which involves programming a small but complete system that would take a pro two weeks to complete, and which is judged on quality, robustness, maintainability, etc. Do you think the SCJD bears little relationship to job effectiveness?

If I ever have a small project which is done by one person (as opposed to the teams I usually run), lasts two weeks (not months), dosn't intgrate to legacy systems and has no future requiring maintainance, maybe I'll hire someone with a good SCJD score. For the work my teams do, I'll hire people who have done it before, or who seem like a smart kid out of college.

Originally posted by Mike Gershman:
How would you compare the SCJD to an hour interview someone who has three years in a j2ee group on his resume. He may have three years of solid experience or he may have spent most of that time covering meetings, because he is so well spoken, and providing night support for other people's code, because his own code is correct but unmaintainable.

For me, if I thought the SCJD test was better, I would simply ask for those scores and dispense with the interviews. As it is, I learn much about their relevant skills through the interview. Most competant interviewers can determine who is talk and who isn't. And if they can't, that business unit is in trouble.

--Mark

[Sonny Gill]

On the topic of certifications, I dont think certifications prove any kind of competency, but in certain situations can be useful as a lowest benchmark.

If someone is SCJP, it does not say anything about how good they are at software development, but it does give me a fairly good idea of how well they know Java basics, which is something. But I would apply this only to people fresh out of a university, or with very little work experience. Asking somebody with years of experience to get an SCJP to prove they know Java would be ridiculous.

In fact, I am considering recommending to my boss that any junior programmers hired be given something like 6 months to get Sun certified (at company expense), as a way of making them responsible for their own learning. As a result I know that they wont make basic mistakes, such as overriding a class and not calling the wrong superclass constructor etc., and would be more capable when finding cause of bugs, or fixing them.

I will be happy to hear any comments on this.

[Mike Gershman]

From Mark Herschberg:

For the work my teams do, I'll hire people who have done it before, or who seem like a smart kid out of college.

So what is your prescription for COBOL/CICS programmers who learned Java on their own? Kool-Aid?

Are legacy programmers who worked on three generations of software suddenly not even worth interviewing to work on the fourth?

BTW, are there no age discrimination laws in Massachusetts?

[Mark Herschberg]

Originally posted by Mike Gershman:
From Mark Herschberg:

So what is your prescription for COBOL/CICS programmers who learned Java on their own? Kool-Aid?

Are legacy programmers who worked on three generations of software suddenly not even worth interviewing to work on the fourth?

If they've don it before, and seem competant I'll hire them. It's the small minded managers that only hire people who have done the exact same project before. I have often hired people who have never written a commercial line of Java, but know C++ and OO principles quite well.

Of course, if you only know and have worked on COBOL, you're going to have a hard time convincing me that you understand the issues in a larg OO project. On the other hand, if you come to the interview and show me some code that you've written on a mid-sized OO project and can answer my OO questions well, of course I'd be interested.

The short answer Mike, I can evaluate someone in much more detail than a prwritten exam--any half decent manager can.

--Mark

[Homer Phillips]

On the other hand, if you come to the interview and show me some code that you've written on a mid-sized OO project and can answer my OO questions well, of course I'd be interested.

You can be interpreted as the industry.

If you will only hire the top 1/2 percent, the odds are slim you'll interview them. You've made it clear to the contracting companies you don't want to see them. They watch what you hire and they listen to what you say.

The industry only wants to hire old-timers as on a try-before-you buy system. So very few get hired.

The hard data is there. The anecdotal evidence is there. It's McJobs for the 40+ crowd.

[peter wooster]

Originally posted by Homer Phillips:
...The hard data is there. The anecdotal evidence is there. It's McJobs for the 40+ crowd...

That's far too pessimistic. From what I've seen, few go into McJobs, most become Project Managers.

[Mike Gershman]

From Mark Herschberg:

if you come to the interview and show me some code that you've written on a mid-sized OO project and can answer my OO questions well, of course I'd be interested.

Do you apply the same standards, "show me some code that you've written on a mid-sized OO project", to "a smart kid out of college"?

[Tim Holloway]

Mark says that "any half-decent manager" can determine an applicant's competencies. Considering the amount of incompetent work floating out on the Internet right now, I'll not comment on any implication of the ratio of decent to non-decent managers.

My evidence isn't anecdotal. In addition to the previously-mentioned Fortune 50 website, two businesses I deal with lost several hundred dollars worth of sales from me in the last 3 months because their sites were malfunctioning and I went somewhere else. Another company had their order get lost for over 2 months before I finally realized that the shipment hadn't arrived because their website didn't finalize the transaction - and, in fact, couldn't because their credit-card verification agency wasn't working right (I had to switch to a different credit card). Allowing for litigation, I can name names, places, times and circumstances, turn you loose and let you scream at this nonsense. But I'll wager I don't need to. I don't think my experiences are all that unique.

Yes, the closer a certification process comes to modelling the circumstances under which real-world work is done, the more weight I'll give to it.

I believe someone brought up the point that in the real world, software development isn't a single-person effort. Good point. Although truthfully, over the years, my observation has been that often the software that actually gets the job done isn't something that was meticulously planned out and worked on in a formal process by a group of people - it's more often something that was hacked out in a fairly short period by one or two people and is often ugly as all get-out, but since it does the job, no one's inclined to to much about it. At least until the user requirements have built up to the point where the old system is no longer capable of doing the job. That's when the committee efforts begin, large amounts of time and money are wasted, and 2 times out of 3, the whole rewrite ends in the dumpster. YMMV.

Oddly enough, I hadn't correlated what's been said here with what we're doing where I work. Thanks to all who responded to the job posting, BTW. We didn't care about certs, diplomas, or other stuff. We were looking for someone who demonstrably had passion about working with IT. We figured that if you have the fire, you'll get the skills.

And, I guess what I'm whining about is that as a consumer I don't really care about the credentials of the people who design, implement, and maintain the applications I interact with. What I care about is the commitment to quality. Or lack thereof. Obviously if you start out with good talent, you have the potential to do better, but potential is useless if the attitude isn't there. And right now the attitude - not just in IT, but almost everywhere - is that quality doesn't count. Faster and cheaper is all that matters.

[Mark Herschberg]

Originally posted by Homer Phillips:

You can be interpreted as the industry.

This is news to me.

Originally posted by Homer Phillips:

You've made it clear to the contracting companies you don't want to see them.QB]

I have? Someone better alert the recruiters I use to send me candidates. I guess they didn't get the memo.

Originally posted by Homer Phillips:
[QB]
The industry only wants to hire old-timers as on a try-before-you buy system. So very few get hired.

...and given that I'm the industry I better tell those guys more than 5 years out of college that I was supposed to try them out first and not simply hire them after two rounds of interviews. Clearly my mistake.

Originally posted by Mike Gershman:

Do you apply the same standards, "show me some code that you've written on a mid-sized OO project", to "a smart kid out of college"?

Nope, I don't expect a kid in college to have the same skill set as someone with 10 years experience. As I am looking for different skills, I test them in different ways. Of course, I don't have many hard and fast rules about how interviewing/hiring should be done and I do what is appropriate for the particular candidate. (What I said above are only general guidelines in light of no other information about the candidate.)


--Mark

[Mike Gershman]

Mark:

As has been made abundantly clear on this board, US legacy programmers who have taught themselves j2ee have no realistic path into industry. By demanding to see OOD code from a substantial project, you create an almost insurmountable barrier to that first j2ee job. Of course, that's better than the usual requirement of 3-5 years of paid commercial experience, enforced by automated resume filters, semi-automated HR people, and hard-nosed head-hunters.

By allowing "a smart kid out of college" but not a 15-year-experienced COBOL/CICS programmer to be evaluated without this requirement is both unfair and illegal. You could suggest volunteering to contribute to an open-source project, but those smart kids can do the same.

I expect that reviewing a 3500 line program, written for an SCJD assignment which was designed by experts to assess key defined areas, gives your group a better basis for evaluating programming skills than examining a randomly assigned slice of someone's open code project. For assessing teamwork, there is no substitute for a personal interview, but the experienced programmer has much more to discuss in this and many other areas.

Obviously, anyone's first significant j2ee assignment involves a lot of learning, but why are you reserving that precious opportunity for young graduates?

[Mark Herschberg]

Originally posted by Mike Gershman:
Mark:

As has been made abundantly clear on this board, US legacy programmers who have taught themselves j2ee have no realistic path into industry.

I disagree completely both that there is no path and that it's been demonstrated on this board, but I'm tired of debating the point with you. (See my final point below for how this can easily be settled.)

Originally posted by Mike Gershman:

By allowing "a smart kid out of college" but not a 15-year-experienced COBOL/CICS programmer to be evaluated without this requirement is both unfair and illegal.

If you think I'm doing something illegal, you should probably report me to the appropriate authorities. At the very least it would seem JavaRanch should have me removed from running this discussion group.

Originally posted by Mike Gershman:

I expect that reviewing a 3500 line program, written for an SCJD assignment which was designed by experts to assess key defined areas, gives your group a better basis for evaluating programming skills than examining a randomly assigned slice of someone's open code project.

Feel free to interview your candidates any way you like. I'll interview them as I see fit.



What I love about free markets is that they're the great equalizer. If you think I'm being stupid/lazy/inefficent/obtuse you can start your own company and capitalize on my mistakes. If I'm too stupid to hire the best programmers and filter them out of the candidate pool, then you can create a better process, hire them, and produce better software. I wish you the best of luck.

--Mark

[Mike Gershman]

Mark Herschberg said:

What I love about free markets is that they're the great equalizer. If you think I'm being stupid/lazy/inefficent/obtuse you can start your own company and capitalize on my mistakes. If I'm too stupid to hire the best programmers and filter them out of the candidate pool, then you can create a better process, hire them, and produce better software. I wish you the best of luck.

I never said that age discrimination was stupid, inefficient, irrational, uneconomic, or counterproductive. I said it was unfair and illegal.

I was around during the civil rights battles of the 1960's and I can assure you that the manager of that Woolworths in North Carolina was right when he said that if black people ate at his lunch counter, white people would not. That didn't make his position moral.

I have no doubt that you are saving your company considerable money by hiring young kids, visa holders, etc., in place of middle aged American workers trying to continue their careers. The role of government includes regulating the free market when it goes too far.

The problem is that when the government does step in, it tends to overregulate and to create permanent solutions for temporary problems. When that occurs, if you want someone to blame, check your mirror.

Mike

[Jeroen Wenting]

There indeed is no career path for a 40+ (maybe 45) year old programmer who looses his job.
When the company I worked for (OK, this in Europe but the situation is similar. In fact it's better here) went bankrupt in 2003 5 programmers lost their jobs (including me), the rest (15 or so) having been laid off earlier).
All of them who were under 45 found jobs within a few months. Of the older ones only 1 found a job and that was as a PL/1 programmer, when he'd just spent over a year learning Java and OO skills and applying those to small projects. Noone wanted him for those skills, not even on a junior salary...

[Homer Phillips]

Even Harris Miller admits to the skills gap -

See page 105

The current slowdown in the information technology (IT) industry due to the overall American economic downturn has resulted in decreased
demand for IT workers this year. IT and non-IT companies have slowed
hiring and increased dismissals in 2001 and 2002. Despite this downturn,
a skills gap persists for IT workers in the U.S. Over the last five
years, employers of IT workers from both IT and non-IT organizations
have consistently told ITAA that there is a lack of properly skilled technology workers. ITAA original research suggests that even as demand
falls for IT workers, the skills gap remains largely unchanged, presenting
employers with limited pools of qualified applicants. Of deep concern
is the long-term ability to maintain and train an adequate supply of
technology workers with requisite math and science skills.

Up market or down market, they just can't find workers.

[Mike Gershman]

even as demand falls for IT workers, the skills gap remains largely unchanged, presenting employers with limited pools of qualified applicants.

That is because these employers will not consider interviewing older workers for current technology jobs except those workers forsighted enough to jump to an eCommerce group in the 1990's.

The programmers who loyally supported legacy applications until they (workers and systems) were phased out have no options - retraining, taking salary cuts, volunteer work, etc. - to get back in the game. That is why we need a temporary freeze on IT work visas until the overhang of unemployed legacy programmers finds jobs.

[Tim Holloway]

Originally posted by Mike Gershman:
Mark:

As has been made abundantly clear on this board, US legacy programmers who have taught themselves j2ee have no realistic path into industry. By demanding to see OOD code from a substantial project, you create an almost insurmountable barrier to that first j2ee job.

Hey, WAIT A MINUTE!!! I resemble that remark! My first language was FORTRAN IV. When I started out, they were still arguing the merits of Structured Programming.

I founded one of the first OO technology companies - Lattice C++ for the Amiga was my company's port of the original AT&T C++ translator done back when IBM PC's weren't up to the job. OO has been around for 20 yaers now, and anyone who can't do OOD/OOP by now has more serious issues than just their age.

Likewise, I made it my business to learn Java shortly after it hit the streets, and I've watched J2EE grow up from its beginnings, back before amenities like JSPs existed. In a few cases, it may even be that I helped shape the technology in some small way.

Does that mean that I think age discrimination doesn't exist? Not hardly. No more than I think that possession of an extensive and highly polished skillset - or a proven record of continous growth in the profession - is going to guarantee me a job.

I know that any hour of any workday I can be called in and told that my job is being replaced with offshore labor and/or a technology that I'm not practiced in and that they don't wish to invest the time and effort it takes me to switch over. I know - from experience - that no matter how highly my peers applaud my knowledge and skills, that it could take (has taken) literally years to get another job.

I know that I'm not operating in a totally free environment. Some restrictions are self-imposed, such as my unwillingness to move out of state for a job. Some restrictions are not, such as the Minimum Wage laws.

Nonetheless, I'm entirely too ornery to just sit down and cry. I may be victimized, but I refuse to be a professional victim. For those many places that so arrogantly think that I don't/can't have what it takes, they can just . My ultimate revenge comes when I can offer my services to their competitors. I don't like working for stupid people anyway, even if there are more of them than there are smart ones.

[Mike Gershman]

Tim:

What about those programmers who loyally supported, at their employers' request, the legacy systems that kept us in business while others went off to try to make their first billion in the dot com world?

It's no harder to self-learn j2ee/UNIX than it was to self-learn COBOL/CICS/MVS, but one still needs that first Java job. Are these people to just go off to McJobs or will the employers give them a chance? Is there an ethical duty on the part of the IT industry to provide a decent career path for the keepers of the legacy flame? Or do the employers stick to the safer choice of the "smart kid out of college", the L1/H1B, or someone like you who made the move to OOD back when it was possible?

Mike

[Homer Phillips]

Or do the employers stick to the safer choice of the "smart kid out of college", the L1/H1B, or someone like you who made the move to OOD back when it was possible?

There is no evidence the college kid is safer.

You are taking a long shot if you are thinking that getting certified without the actual experience is going to increase your marketability. The industry gives no credit for carrying a legacy torch. The industry recognizes the value of the value of legacy experience but does not want to pay for it.

There are few paths to that first java job. 1) get transferred onto the Java path while working in the company 2) knowing somebody that can get you hired.

IIRC SCJD is based on Swing and RMI. IMO, there is very little market for either of these technologies.

If you are over thirty-five and think that you would like to fill one of the thousands of unfilled java positions out there, think again.

[Jay Ashar]

Originally posted by Homer Phillips:

There is no evidence the college kid is safer.

You are taking a long shot if you are thinking that getting certified without the actual experience is going to increase your marketability. The industry gives no credit for carrying a legacy torch. The industry recognizes the value of the value of legacy experience but does not want to pay for it.

There are few paths to that first java job. 1) get transferred onto the Java path while working in the company 2) knowing somebody that can get you hired.

IIRC SCJD is based on Swing and RMI. IMO, there is very little market for either of these technologies.

If you are over thirty-five and think that you would like to fill one of the thousands of unfilled java positions out there, think again.

and to add more to this, one more hurdle to cross for getting a java job.

USCIS Link

USCIS had a press release, 20,000 new H1s are open to everyone now and not just US advance degree holder, as initially proposed. With this change I am sure this quota will be filled in couple of days as soon as these visas are available.

[Homer Phillips]

Originally posted by Mark Herschberg:

If you think I'm doing something illegal, you should probably report me to the appropriate authorities.

It's unusual a person would seek a confrontation with the authorities. Is that because you don't think the authorities can make a case against you, will make a case against you or you are in compliance with the law?

[Homer Phillips]

USCIS had a press release, 20,000 new H1s are open to everyone now and not just US advance degree holder, as initially proposed.

I am not suprised. IMO, the US government has launched an assualt on the domestic labor force. In view of the Rand Corp report of 2003 which indicates the market for science and technology people shows signs of being in surplus, what other conclusion can one make? I suspect the US government commisioned the Rand Corp, a respectected think tank, to conduct the study.

[Tim Holloway]

Originally posted by Mike Gershman:
Tim:

What about those programmers who loyally supported, at their employers' request, the legacy systems that kept us in business while others went off to try to make their first billion in the dot com world?

It's no harder to self-learn j2ee/UNIX than it was to self-learn COBOL/CICS/MVS, but one still needs that first Java job. Are these people to just go off to McJobs or will the employers give them a chance? Is there an ethical duty on the part of the IT industry to provide a decent career path for the keepers of the legacy flame? Or do the employers stick to the safer choice of the "smart kid out of college", the L1/H1B, or someone like you who made the move to OOD back when it was possible?

Mike

Actually, I am one of those faithful corporate drones. I'd racked up 13 years experience and never even considered jumping ship for a dot-com.

Then the last recession came and they canned me without a second thought. It took 28 months to get another job.

One thing I never did, however, was sit idle. Like I said, I actively investigated up-and-coming technologies and acted as an early adopter where warranted. It's how I went from mainframe assembly language to C, from C to C++, and from C++ to Java. Not without a few false starts, like doing systems programming in Pascal (actually IBM's compiler was pretty darned good) and learning, but never using Ada.

I'd already been working in J2EE before I got laid off, so at least I didn't have that hurdle to surmount, but even way back when IT jobs were "DP" jobs, the industry was allegedly recession-proof, and I was much younger/cheaper, I darned near had to pull a gun to get seriously considered/hired. Age hasn't really made that much difference for me so far.

[Jesse Torres]

Originally posted by Tim Holloway:

but even way back when IT jobs were "DP" jobs, the industry was allegedly recession-proof, and I was much younger/cheaper, I darned near had to pull a gun to get seriously considered/hired. Age hasn't really made that much difference for me so far.

What is "DP" jobs?

Thanks,

[Homer Phillips]

Information technology is sexier than data processing.

[Jason Cox]

This is tiring.

As a veteran software developer, I expect to be held to a higher standard than a kid fresh out of college. That has nothing to do with age, it has everything to do with experience.

I hear this "legacy developer" complaint a lot. I moved into IT about the same time Java was starting to take off as a language. That's back in 1999 and there were plenty of Cobol jobs to be had, but it was already in decline. The question was already being asked as to what would happen to a lot of the Cobol work once Y2K came and went. Java was not "new" in technology sense by that point. Even the IBM IDE's, notoriously behind Java releases back then, were up to JDK 1.1.7 and I took the Java 2 certification exam. It was already a proven technology.

Since then, Cobol and other legacy technologies have been in exponential decline. It was pretty apparent that a lot of these jobs were going away or at least being drastically reduced in demand. According to some of the people here, you're telling me in the past 6 years there has been NO opportunities to learn a more up-to-date technology? You don't have to be an early adopter, but you do have to move on when the writing is on the wall. When Java starts to decline, you can bet I will be looking at what the adopted current technologies are and I will move to them first chance I get. I don't plan to ever be a "legacy programmer".

It's not a lack of compassion either. I did a 7 month stint of unemployment and the job market is frustrating as hell. That doesn't justify the paranoia, xenophobia, and persecution complexes that seem to have followed. I'm not saying that prejuidice and bias doesn't enter the picture sometimes, but I don't think it's under every rock either.

I like working in the technology field. There are easier professions with more job security that pay better. I don't have to be doing this kind of work, I want to. As a result, I put up with the idiosyncracies, the misconceptions about technology and the professionals who work with it, the unstable job market, the deadlines, and the learning curve. All of these things are part of the game. If you can't tolerate it, why do it? Money? See comments above. If nothing else, I know people who have left the technology field, earn less money, but are far happier. Of course, they generally trade salary for security, so it's not a total net loss for them either.

[peter wooster]

Originally posted by Rob Aught:
This is tiring.

I agree this is tiring.

... When Java starts to decline, you can bet I will be looking at what the adopted current technologies are and I will move to them first chance I get. I don't plan to ever be a "legacy programmer"...

Just a hint of the writing on the wall. No Web App Guys. I believe half the problem is that far too many people have jumped on the J2EE bandwagon, to the exclusion of other opportunities and uses of Java technology. I see many posts where freshers are wanting to become J2EE architects, by the time they get there it will be obsolete.

[Mike Gershman]

Rob Aught said:

You don't have to be an early adopter, but you do have to move on when the writing is on the wall. When Java starts to decline, you can bet I will be looking at what the adopted current technologies are and I will move to them first chance I get. I don't plan to ever be a "legacy programmer".

Many of us continued to support legacy systems because they were still live and our companies and customers depended on them.

Except for independent consultants, this has never been a high risk career strategy. The transitions from second gen to DOS/360 to OS/360 to OS/MVS to CICS to structured programming, etc., were a routine part of the job and the jump to Java is no harder.

The clinker is the flood of underpaid, overworked, and powerless H1B/L1 programmers that gives employers a powerful financial incentive to declare all legacy programmers "15 years out of date" and totally useless regardless of their current training and certifications. This is complete nonsense and the employers know it is nonsense.

If you predicted this situation 6 years ago, you are truly a genius and should advise the rest of us lagards on how to recover from the consequences of our misguided company loyalty.

As for myself, I have given up on mere persuasion and become active in the movement to return the H1B program to its original purpose - as a source of extra help when no US talent can be found - not as competition for US workers. More on this later.

Homer Phillips said:

SCJD is based on Swing and RMI. IMO, there is very little market for either of these technologies.

The SCJD certification is not an assessment of knowledge of a specific Java technology. It is a totally untimed, open book exercise, and the choice of Swing and RMI is meant to stick to some of the most widely known technologies supporting GUI and multi-tier programming.

The SCJD measures robustness, readability, maintainability, standards compliance, OOD compliance, meeting the program spec, making good decisions when the spec is ambiguous, and many other things that add up to professional quality work. In addition to checking the SCJD scores, employers can choose to review the specs and the corresponding 3500+ lines of code against their own standards.

Combined with extensive legacy programming experience and the more technical Sun/IBM/Oracle certifications, the SCJD is a powerful predictor of Java success if the programmer is only given a chance.

Homer also said:

If you are over thirty-five and think that you would like to fill one of the thousands of unfilled java positions out there, think again.

Your comment reflects today's reality. It's up to us to change that reality. I spent 30 years fixing problems, not living with them. Now I'm going to try to fix this one.
[ March 09, 2005: Message edited by: Mike Gershman ]

[Homer Phillips]

This is tiring.

The sleepless nights are not only tiring they are debilitating.

According to some of the people here, you're telling me in the past 6 years there has been NO opportunities to learn a more up-to-date technology?

There have been opportunities to learn but not to sell.

If nothing else, I know people who have left the technology field, earn less money, but are far happier. Of course, they generally trade salary for security, so it's not a total net loss for them either.

IMO, it is not left the profession, it is has been driven out. I am sure what you say is tue about those people being happier is true but it's a limited survey of the population.

Are you asserting McJobs provides job security? Or perhaps you could expound on what sort of new careers the 45+ crowd has taken up.

It's not about money, it's about life and death.