The art of software development - commonly referred to as programming - is often compared to the discipline of architecture. Both involve the creation of structures intended for use by humans; both use basic materials to carve something new; and both rely on function to dictate form, and imagination combined with experience and knowledge to deliver on that form in a unique way.
Cobol went out of fashion quite close to the time that grey, drab buildings fell into disfavour with the public.
Jason Norwood-Young, technology editor, ITWeb
The recent ITWeb round table on Cobol brought into question the basic building materials that developers should be reliant upon. While some agreed that Cobol was a strong basis for business development, it was out-dated and, in itself, not a good career path for graduates.
Cobol is a bit like concrete - it forms great foundations, you can build high-capacity constructions from it, but man, is it ugly! Users just don`t appreciate green-screen terminals anymore, just like people don`t want to live in a non-plastered, grey-walled concrete monolith. Cobol went out of fashion quite close to the time that grey, drab buildings fell into disfavour with the public. As one friend commented: "Cobol is so 20th century!"
Still useful
The utilitarian look does not appeal to many people, but (like Cobol) it is not cost-effective to tear the buildings down just to replace them with something more modern. Concrete is, despite public opinion, still a pretty useful material to build from.
Slap a veneer of plaster, paint, or even that ugly white self-cleaning bathroom tile on it and it gets a whole new lease on life. Cobol has had much the same done to it. People still use it, but not without a lovely veneer of a modern development environment adding a pretty Web-enabled front-end to keep the GUI-savvy end-user happy with the "look-and-feel".
So although the modern enterprise might not look like Cobol, and might not even act like Cobol for the end-user, its foundations are still steeped in that fourth-generation development language that the elder-geek cut his or her teeth on.
This still leaves the apprentice propeller-head with a quandary - study Cobol and be unemployable, or study something more modern and be under-skilled in the real world? Unfortunately for the Cobol industry, most newbies to the programming game choose the latter - they would rather get a salary and not know what`s going on than sit at home scrubbing up their CV and sending it out to disinterested employment agencies ("You`ve got a degree in what?"). Sending a Cobol-only graduate into the market is like training an architect to only design with concrete - it is career limiting in the extreme.
A dearth of fresh Cobol skills in the marketplace - combined with the drain of existing Cobol developers into management positions or to other technologies - makes this option even less attractive to enterprises looking for a language for a green-field project. So what other choice is there?
Taking a beating
Most topical at the moment is the pre-cast wall of development - Java. It looks good, it`s portable, but if you push it too hard it could just fall over. It is a great choice for a facade for a real business application, but it doesn`t have the strength alone to stand the beating that a banking institution would give it.
Then there are the steel girders of C++. Also utilitarian, strong, and having stood the test of time, it can be used to knock up a pretty stable environment, if you have the time and the skills to work with it.
How about the wooden structure of Visual Basic or one of the other visual languages? It looks lovely, and can be carved for great user-interface effect, but you can hardly build a skyscraper out of it.
The brick approach of object-driven languages, like Delphi, C++ (again) or Smalltalk, can be used to build pretty high, can look okay, and comes closest to concrete of all the materials. It still generally needs a face-lift, and its strength is predominantly reliant on how well you lay your bricks, and the quality of the grout that holds the structure together.
And so we see that, although we may not like concrete much, it can fulfil a purpose that no other material can easily replace. A number of competing architectures can be strung together to create a pretty good facsimile, but nothing beats the cold, hard reality of concrete.
Share