Congratulations, sir, you win a prize! You have discovered a sentence of meaningless bullshit in a mainstream technical publication that purports to understand technology. This is an incredibly rare event and you deserve recognition!
At least it's not the awesomeness on cobol.com which boils down to "If you don't use COBOL you will have to REWRITE software which is a RISKY and COSTLY endeavor. COBOL makes everyone's life easier because you won't get FIRED for suggesting alternatives."
It might also be referring to COBOL's built-in support for binary-coded decimal arithmetic. Which would still be a pretty silly statement, since other languages can also do BCD with the appropriate library.
Why should I trust what is basically a Micro Focus press release?
I actually had to use Micro Focus COBOL in a project once. It was awful: COBOL the worst language ever (awful syntax, no local variables!) But what's worse is that the Micro Focus product is nearly unusable (at least on unix platforms). It's loaded with copy-protection garbage, and the installer is deliberately left in a broken state so that you must call Micro Focus support (just guess what a support contract costs: it's not cheap.) This should not happen when you drop three grand on a compiler.
Summary: avoid COBOL, unless you want to rely on a company that should have been out of business long ago (and it probably will be; just wait a few years), and a language that should never have existed.
(user@pasta) /home/user $ psearch -l cobol
lang/open-cobol
OpenCOBOL is an open-source COBOL compiler,
which translates COBOL programs
to C code and compiles it using GCC.
WWW: http://www.opencobol.org/
2001 isn't nearly recent enough. For at least those 10 years, it has been very hard to find COBOL programmers. Businesses don't take long to get sick of paying a premium for antiquated goods.
I wouldn't be surprised to hear that that number is now more like 20%, or even less.
As sick of paying a premium for antiquated goods as businesses may be, they aren't going to make a decision that has costs far in excess of it's benefits.
The benefit of abandoning COBOL is greatly increased programmer productivity. The cost of abandoning COBOL is completely rebuilding your system from scratch.
I could go on for days about why it's dumb for an organization to undertake a massive project for which it is horribly ill-equipped in order to replace something that works. Instead I'll just note that I used to work (not as a programmer) at an organization that tried this. They ended up spending 8 figures for nothing more than the realization that they should hire more COBOL programmers.
I agree that the percentage has dropped (due to organizations that use COBOL dying and new ones avoiding it for obvious reasons), but I doubt it has dropped that much.
I took a COBOL class in college and I actually applied for a summer internship at US Steel working on COBOL.
Luckily, I didn't not get that job.
Looking back, I'm so glad I didn't, my life would have been so different.
The hardest part of learning COBOL for me was the spacing.
Comments had to start with a '*' in column 7 and are the rest of the line.
Labels had to start in column 8.
Statements have to start in column 12 of later.
It was not a fun language to program in.
If they were really desperate for COBOL programmers we would see job posts
offering $200k for COBOL programmers.
In many ways this makes sense. Many of the people running these COBOL companies are going to be retiring soon, creating a huge demand for COBOL programmers.
But on the other hand I haven't seen any job advertisements for COBOL programmers in years. Where is all this demand for COBOL engineers?
... because nothing beats using a 3270 font with ISPF...
There are tons of COBOL code out there, with tons more being written every day (because the tons already there won't be ported but still have to do new stuff). You may get a job, even a steady one, but you won't have much fun doing it. There are lots of programmers who write software not because they like doing it, but because they feel it's a nice job.
But yes, you should learn COBOL. Only after you do it you'll be able to bash it. There is nothing worse than a person who bashes a programming language they aren't fluent in.
I am one of those programmers that looks for companies that work with technologies I love, because that makes sure I will be happy and they will be happy. Jumping into a position using something I absolutely hate just because they'll throw in another $50K/year or so at me won't work.
Perhaps others have a different phillosophy -- "anything that brings in more money is fine", it is sort of a personal choice.
So my advice to kids is "Don't learn COBOL, learn something that is more relevant today and something you are absolutely passionate about. (If you just happen to be passionate about COBOL you are probably a very strange person ;-)"
I worked on a project once. It was COBOL but since COBOL wasn't quite good enough they introduced another custom programming layer on top of it...in COBOL.
I'm ashamed to admit that I actually have a qualification in COBOL, obtained many many years ago. Even at the time it was a dead/obsolete language.
Some COBOL programmers were employed in the last few years of the 20th century in order to combat the largely fictitious Y2K menace. There was a point at which I could have joined that particular bandwagon, and I probably would have had considerable earnings for a couple of years. But I knew enough about COBOL that I didn't want to be anywhere within a barge pole prodding distance of it.
COBOL is a language which was probably incredibly appropriate and highly useful within the culture and technological infrastructure of the 1960s/70s. If a company advertises that it's still using COBOL or Visual Basic 6 then this says to me that they're not interested in innovation, and that for someone like me working there would be like doing the coding equivalent of a jail sentence.
27 comments
[ 237 ms ] story [ 927 ms ] threadWhat does that mean? How is COBOL better at representing business data than another enterprise platform - say Java, or .NET?
http://www.cobol.com/competition
Processing fixed format records (preferably one at a time) is the only thing COBOL is somewhat skilled at, though.
I actually had to use Micro Focus COBOL in a project once. It was awful: COBOL the worst language ever (awful syntax, no local variables!) But what's worse is that the Micro Focus product is nearly unusable (at least on unix platforms). It's loaded with copy-protection garbage, and the installer is deliberately left in a broken state so that you must call Micro Focus support (just guess what a support contract costs: it's not cheap.) This should not happen when you drop three grand on a compiler.
Summary: avoid COBOL, unless you want to rely on a company that should have been out of business long ago (and it probably will be; just wait a few years), and a language that should never have existed.
It's still COBOL, but, at least, it doesn't come with Micro Focus.
I wouldn't be surprised to hear that that number is now more like 20%, or even less.
The benefit of abandoning COBOL is greatly increased programmer productivity. The cost of abandoning COBOL is completely rebuilding your system from scratch.
I could go on for days about why it's dumb for an organization to undertake a massive project for which it is horribly ill-equipped in order to replace something that works. Instead I'll just note that I used to work (not as a programmer) at an organization that tried this. They ended up spending 8 figures for nothing more than the realization that they should hire more COBOL programmers.
I agree that the percentage has dropped (due to organizations that use COBOL dying and new ones avoiding it for obvious reasons), but I doubt it has dropped that much.
If they were really desperate for COBOL programmers we would see job posts offering $200k for COBOL programmers.
But on the other hand I haven't seen any job advertisements for COBOL programmers in years. Where is all this demand for COBOL engineers?
Infosys / Patni / TCS etc work on quite a few COBOL projects.
There are tons of COBOL code out there, with tons more being written every day (because the tons already there won't be ported but still have to do new stuff). You may get a job, even a steady one, but you won't have much fun doing it. There are lots of programmers who write software not because they like doing it, but because they feel it's a nice job.
But yes, you should learn COBOL. Only after you do it you'll be able to bash it. There is nothing worse than a person who bashes a programming language they aren't fluent in.
I am one of those programmers that looks for companies that work with technologies I love, because that makes sure I will be happy and they will be happy. Jumping into a position using something I absolutely hate just because they'll throw in another $50K/year or so at me won't work.
Perhaps others have a different phillosophy -- "anything that brings in more money is fine", it is sort of a personal choice.
So my advice to kids is "Don't learn COBOL, learn something that is more relevant today and something you are absolutely passionate about. (If you just happen to be passionate about COBOL you are probably a very strange person ;-)"
Some COBOL programmers were employed in the last few years of the 20th century in order to combat the largely fictitious Y2K menace. There was a point at which I could have joined that particular bandwagon, and I probably would have had considerable earnings for a couple of years. But I knew enough about COBOL that I didn't want to be anywhere within a barge pole prodding distance of it.
COBOL is a language which was probably incredibly appropriate and highly useful within the culture and technological infrastructure of the 1960s/70s. If a company advertises that it's still using COBOL or Visual Basic 6 then this says to me that they're not interested in innovation, and that for someone like me working there would be like doing the coding equivalent of a jail sentence.