Battle of the Ages; Stereotypes Collide 319
JCOTTON writes "A CIO.com article By Phil Murphy explains that "The hype around the shortage of qualified legacy technologists grows each day. Pundits would have us believe that 1.5 million COBOL programmers will suddenly disappear one day, leaving any company with legacy technology in dire straits. The truth is that there are far more programmers with legacy skills looking for work than there are jobs for them, as evidenced by organizations like Legacy Reserves, which functions as a training and job matching service for unemployed or underemployed programmers wishing to modernize their skills."
This article explains many of the issues facing "the upper half" of Information Technology workers."
Upper half? Unfair connotation (Score:0, Insightful)
Why do people tie themselves like this? (Score:3, Insightful)
Sure I have my favorite languages, but I treat each language I come across equally; hell, I tolerated and become proficient in Scheme of all things. This way, if the flavor of the day goes away, I can simply pick up a book on the new flavor, figure out how it does business, and get to work.
Good principles and techniques transcend language boundaries.
Other side of the coin... (Score:5, Insightful)
but...
There are too many companies that refuse to move out of the computing Bronze-Age; bite the bullet and upgrade.
The town that I work in (Blue-collar auto-industry) is filled with tool & die shops. Typical scenerio: The owner left the assembly line of Ford/GM/whatever 20 years ago and created his own company. He bought a DOS app to run his business on a 286-server/workstation, and he is surprised and insulted to find out that XP won't run on it.
I have seen shops that Net revenue >$10 million/year, and they depend on a app written in BASIC!!!! as their life-blood.
Holy shit people, it might be time to upgrade!
There is a reason we don't (all) still use Horse & buggys. There is still a market for companies to make horse-shoes and buggy whips, (and I bet that company has a monopoly) but there are valid reasons to upgrade.
There will always be a need for Legacy-based skills, but for the love of $deity don't hold onto old tech that you think "Well it used to be good enough!"
A thinking man's barriers. (Score:1, Insightful)
Yes they have things in common, but they also have different ways of approaching the same problem. For example the imperative way of programming, verses say the functional way, or the procedural way. Sometimes the barrier isn't the language, so much as it is "the way of thinking" that goes with it.
Windows Arrogance and Stereotyping (Score:5, Insightful)
My business concentrates on Mac OS X systems used in a publishing environment. They work much like their Windows counterparts and could even be integrated with the larger domain for more efficiency. But when I speak of this to others they look at me with confusion and, maybe, heresy?
These people act as if Macs are toys or inferior in some way. Of course, this is far from the case, but their training has changed how they see technology. This really isn't the old Mac/PC debate. (Apple lost the first war. But they still found an important place in today's computing world.)
No computer technology is perfect, of course. But the mistaken ubiquity that IT is Microsoft and Microsoft is IT makes all other non-MS technicial initiatives and products harder to sell in concept or through a store.
Re:Other side of the coin... (Score:3, Insightful)
But if it is still good enough, why change? Rewriting large apps will introduce new bugs and problems. I work at a company that writes programs in COBOL. It might be nice to my resume to redo everything whatever the flavor of the month language is, but why? Our apps work great and our customers really like them.
It usually isn't voluntary. (Score:3, Insightful)
A person with both good knowledge of C and good knowledge of COBOL is usually seen as being a "COBOL programmer" if their last work experience was mainly writing COBOL code.
It sounds silly, I know, but that's what I've seen (and what many others I know have also seen) in the current job market.
Microsoft Confirms it... (Score:3, Insightful)
Seriously, I think I remember reading that MS said that end-of-life for VB6 is coming up in 2006 or so, but can't find the article where I read that. If it exists, it's likely buried deep within MS's site.
The best advice I could give to someone who's been buried under a pile of MS technology for most of his/her education/career would be to go out and pick up some non-MS languages. That way, if Redmond (or its language of the month) disappears tomorrow, there's a chance that you'll still be employable, and you'll gain a perspective on programming that you might not otherwise have. That's just my opinion, though, and I'm sure there are thousands of MS flamers who would say that once you've gone down that path, you're damaged goods anyway. Take this sort of rambling in either direction with a tumbler of salt.
Upgrade to what? And why? (Score:4, Insightful)
Fortran was (and is) a perfect language for the type of problem being solved, since a lot of it actually does involve semi-complex calculations.
The mainframe platform is also ideal, as the system is designed as a centralized software app running on a large-scale server and being used by folks all over the world on remote terminals (be they "green screens" or web clients).
Sometimes the older languages and platforms in use really *are* a good fit. Or is it change for changes sake that you're asking for?
Multiple Languages, Anyone? (Score:4, Insightful)
A carpenter is not a hammer-er, or a saw-er, or a drill-er. He is expected to be able to quickly learn and use any of those tools, as needed for the project. A new project can use a new tool (language, os, whatever) as needed for the application. When an old program needs maintenence, it may require some re-learning of the old tool, but that should not be difficult.
I suspect the harder problem is preserving the old development systems and tools. If the compiler (or some other tool) hasn't been used in several years, there is a good chance that it won't work. Or, that we can't find it at all because it didn't get loaded onto the new host before the old host was scrapped. Or, that the old hard-copy manuals (how to use the tools) have rotted and/or been discarded in the trash.
The learning curve isn't COBOL itself... (Score:5, Insightful)
An IBM CICS programmer familiar with DB2 would have a tough time coming into a Unisys A-series shop that uses COMS and DMSII, not to mention the culture shock when his JCL-conditioned mind runs into a job control language like WFL.
Re:So much for keeping up on VB6... (Score:2, Insightful)
Re:Other side of the coin... (Score:5, Insightful)
Seriously. A 30-year-old custom COBOL app has, in all probability, had all of it's bugs resolved 20 years ago. It works. Replacing a legacy system with a million lines of tested and proven code is going to be an expensive and dangerous proposition.
If it works reliably and satisfies the business requirements, what does it matter what language it's written in? The answer is: it doesn't. If the bugs have been squashed and the requirements have not changed, there is NO reason whatsoever to monkey with a working, stable system. "BASIC is for n00bs; Python is l33t" is not an adequate justification to replace a proven system.There are plenty of applications that work perfectly with a curses-based interface runing on dumb green-screen terminals -- just because the technology used isn't "cool" does not mean that there's any benefit in replacing it with a GUI or web-based interface or whatever else is "cool" this year.
Holy shit people, it might be time to develop some professionalism. It's not about who has the coolest toys -- it's about satisfying the business requirements in the most cost-effective manner. The question isn't "did it used to be good enough?", the questions are "is it currently good enough?" and "can we justify the expense and risk of re-implementing it?"."Labor Shortage" yellow alert (Score:5, Insightful)
I am not saying that this is necessarily what the article's author has heard, but it would not surprise me. Organizations like ITAA are shrewd and tenacious. They recently managed to influence many small-city newspapers to publish articles about the dangers of tech labor shortages by quoting companies who allegedly will go under unless they import Indians or move to India. Their leader, Harris Miller, lobbied for more agricultural migrants (fruit pickers) from Mexico in his previous job, according to some sources.
The excuse is the same for tech as it was for agriculture: "Americans don't want fruit-picking jobs". At $3-per-hour, who would? They want to do to tech what they did for agriculture. Different career, same plan.
They should be on the same "geek enemy list" as SCO.
Re:Why do people tie themselves like this? (Score:1, Insightful)
I can't quite tell whether you're serious or this is extremely subtle satire. In view of my 20 years experience in the industry, I'm voting for satire.
This is bad news. (Score:4, Insightful)
"Welcome to Legacy Reserves, the largest U.S. databank of Legacy Professionals over the age of 35"
I think that is a new low in setting the threshold for being "over the hill". This means I was old 20 years ago...god, somebody see if I still have a pulse!
Re:Multiple Languages, Anyone? (Score:2, Insightful)
Nothing a US$30 UPS couldn't solve. :-) (Score:3, Insightful)
Sounds to me like he's got a sane idea.
A Lot of Them Can't (Score:4, Insightful)
To implement the software on modern gear would require a tremendous amount of time just sorting out what everone does and why. It's a much larger problem than just sitting down and hacking it out, even if you have the original source and want to blindly follow the last guy's design.
And then sometimes they just can't match the performance of the old system. IBM's been trying to do away with their RETAIN system since I first started working for them back in the mid '90's. At the time they thought they'd go to a Lotus Notes app on their 486 servers. After all, the 486 was designed to give you the same performance on your desktop as a mainframe, right? Sure, for a single user! They never could figure out how to match RETAIN's performance. To this very day they're still maintaining it. I don't think anyone understands it anymore, really. It's millions of lines of mainframe assembler code from what I hear. It's like this ancient evil that lurks under the surface of the apparently peaceful company, just waiting to consume the souls of young programmers. With Tentacles.
Re:So much for keeping up on VB6... (Score:2, Insightful)
One of the things that keeps COBOL alive is that it has been around for 45 years. Most of the newer languages are likely to disappear in a flash. What is more likely to be fairly common 10 years from now: COBOL or Ruby? Ruby may take off, or more likely die when the next Foobar Oriented Programming fad comes along and makes Ruby look legacy. Most "in" languages right now will likely be firewood in 10 years based on past patterns. Since COBOL is already known to be legacy, it is less likely to be affected by that perception. Nobody knows if Ruby can survive a legacy stigma. COBOL has proven to weather fad storms. Not that I defend COBOL as a language, but from a manager's standpoint, it takes a licking and keeps on ticking.
Re:It appears that they're hiring again (Score:3, Insightful)
How many successful politicians can you name who have also run successful companies? Very few, I bet. It takes a remarkably different temperament and management style, being a businessman versus being a politician. Ross Perot was successful in business but was a complete disaster in politics.
Why do people assume that if someone is good in business, they'll be good in politics (or vice-versa)? That's like saying if someone is a good football player, they should also be able to play the trombone well. (mmm..heh heh, yeah, he said 'bone', heh heh)