You make some good counterpoints about COBOL. My thought is, how much of COBOL's lack of continued popularity is due to the language's deficits (there are quite a few, I'll readily admit), and how many are because the majority of programmers are infected with shiny-language syndrome?
In other words, could the lack of COBOL developers be in part due to the regular statement that "COBOL is dead?"
Part of the difficulty in eliminating COBOL from the financial sector is its stability. Given how unstable code is, and how many projects fail dramatically, the financial sector is probably flighty about the idea of porting. They'd be trading speed, stability, and proven mathematical accuracy for what exactly? Maintainability is important, but in mission-critical code, it isn't usually worth the trade-off. Thus, it becomes more practical to find/train COBOL developers to patch the existing systems, rather than build new ones.
Besides that, according to some expert articles I read, so far no languages have really proven themselves equal to COBOL in the financial role. So, we have a language with no apparent successor, and a lot of already-stable code that we don't want to break.
So, I'll ask again: is the problem COBOL, or that we've been discouraging developers from learning it?
P.S. Yes, I'm well aware the language has a painful syntax. That's one of the drawbacks.
In other words, could the lack of COBOL developers be in part due to the regular statement that "COBOL is dead?"
I'm too young to know what happened to COBOL but I guess it's both :-) The article I linked talks about the shortcomings of the language itself. I have no idea how recent Cobol dialects are but the GO TO must be hell if used in programs with millions of lines of code
Thus, it becomes more practical to find/train COBOL developers to patch the existing systems, rather than build new ones.
Yep, but they can't keep up with the amount of devs they need in rapport to the amount of code there is.
So, I'll ask again: is the problem COBOL, or that we've been discouraging developers from learning it?
I think it's both. Devs like shiny things, COBOL isn't anymore :-) It's probably also because the landscape of development isn't the same as in the 60s. Web development didn't exist back then, now web development is a huge chunk of the software development industry.
I'll give you another example. I learned object oriented programming with Object Pascal/Delphi. I loved it, it was a good language BUT they were so concentrated on creating GUI when the world was clearly more interested in developing software for the web. Delphi mostly lacked support over there, people jumped ship. Is Delphi dead? Not at all. But it's not just the fault of people saying "don't bother", of that I'm sure.
Python might be dead in 30 years but its "general purposiness" from day 0 is one the reasons why is still one of the languages that people learn every day, even if it will be 30 years old shortly ;-)
Glad you liked the article.
You make some good counterpoints about COBOL. My thought is, how much of COBOL's lack of continued popularity is due to the language's deficits (there are quite a few, I'll readily admit), and how many are because the majority of programmers are infected with shiny-language syndrome?
In other words, could the lack of COBOL developers be in part due to the regular statement that "COBOL is dead?"
Part of the difficulty in eliminating COBOL from the financial sector is its stability. Given how unstable code is, and how many projects fail dramatically, the financial sector is probably flighty about the idea of porting. They'd be trading speed, stability, and proven mathematical accuracy for what exactly? Maintainability is important, but in mission-critical code, it isn't usually worth the trade-off. Thus, it becomes more practical to find/train COBOL developers to patch the existing systems, rather than build new ones.
Besides that, according to some expert articles I read, so far no languages have really proven themselves equal to COBOL in the financial role. So, we have a language with no apparent successor, and a lot of already-stable code that we don't want to break.
So, I'll ask again: is the problem COBOL, or that we've been discouraging developers from learning it?
P.S. Yes, I'm well aware the language has a painful syntax. That's one of the drawbacks.
I'm too young to know what happened to COBOL but I guess it's both :-) The article I linked talks about the shortcomings of the language itself. I have no idea how recent Cobol dialects are but the GO TO must be hell if used in programs with millions of lines of code
Yep, but they can't keep up with the amount of devs they need in rapport to the amount of code there is.
I think it's both. Devs like shiny things, COBOL isn't anymore :-) It's probably also because the landscape of development isn't the same as in the 60s. Web development didn't exist back then, now web development is a huge chunk of the software development industry.
I'll give you another example. I learned object oriented programming with Object Pascal/Delphi. I loved it, it was a good language BUT they were so concentrated on creating GUI when the world was clearly more interested in developing software for the web. Delphi mostly lacked support over there, people jumped ship. Is Delphi dead? Not at all. But it's not just the fault of people saying "don't bother", of that I'm sure.
Python might be dead in 30 years but its "general purposiness" from day 0 is one the reasons why is still one of the languages that people learn every day, even if it will be 30 years old shortly ;-)
A good, balanced rebuttal (well, half-rebuttal). Thanks for the added insight!