Interesting blog entry by Martin Fowler

[Argo]

http://martinfowler.com/bliki/...rTalentHypothesis.html

 

[kamper]

"Smarter programmers are better." Thanks Captain Obvious!
"We should try to pay smarter programmers more." Thanks Captain Obvious!
"We have no way of judging who is the smarter programmer." Then what was the point of this blog?

There's a lot of people blogging about how the way to build better software is just to have smarter developers. I have not seen many of them actually propose any practical advice.

 

[nordloewelabs]

Originally posted by: kamper
There's a lot of people blogging about how the way to build better software is just to have smarter developers. I have not seen many of them actually propose any practical advice.

there are too many ppl blogging nowadays.

 

[Markbnj]

Martin is a very smart guy, and his basic idea isn't wrong, or very novel. It's something we've known about for a long time. It's absolutely true that some indefinable talent makes some people much better at writing software than others. The word "smarter" is not even close to an adequate description of it. Creating software is still very much closer to painting or sculpting than to mechanical engineering or electronics. We'd like to get a handle on it. We keep trying, and occasionally when some entity that can afford it is willing to spend many millions and possibly decades to build something in a very tightly controlled manner (i.e. milspec), they succeed. But sometimes even in those situations someone uses meters and some else uses feet and your interplanetary probe digs a multi-hundred million dollar hole in the Martian dust, or your global telecomm network goes black.

The real gap isn't between talented and untalented programmers, but between what software really is, and what we would like to think it is. Sure, the average business person thinks anyone who can type and ostensibly knows a language is as good as anyone else with the same resume entries. But then those same managers assume they can build an entirely new and complex piece of virtual machinery, within a certain amount of time, and for a certain amount of money, but without actually thinking about what it is supposed to do at any level of detail.

The real gap is between how much we rely on software, and the aspirations we have for it, and the level of quality that can actually be expected from such stupid processes. That's a gap worth blogging about.

 

[Argo]

I know it seems like he's stating the obvious, but I think one of the main points he's making is that a lot of companies think they're better off hiring 2 average developers to do the work of one very good one, and Fowler disagrees with that approach.

 

[eleison]

Originally posted by: Markbnj
Martin is a very smart guy, and his basic idea isn't wrong, or very novel. It's something we've known about for a long time......

Martin Fowler has sold out. Ever since starting thoughtworks, he hasn't been focusing on development productivity. He's been trying to straddling between making $$$ and espousing good advice. The good advice will cost his company billing time... Hence his bias towards $$$$.


I read one of his bloggs on Spring (basically Inversion of control).. what a bunch of cow dung!!!! His early blogs are better -- the ones in which they weren't really about the $$$ but about developer productivity.

This current blog isn't about how companies are sensitive to price.. its basically a whinny diatribe about why companies are picking his competitors who are charging less over his consulting firm. Obviously, he is losing contracts because of this, else why the blog??????

He's also been the consulting field for a while... he should have know about this already.. after all, he was initially a one man consulting shop.. now that he has to increase revenue and he has other developers under his wings.. a lot of his time is now about the $$$$$$... For freaks sake, how long has the "mythical man month" book be out?? Years.. decades?? For those who don't know, this book talks about quality programmers and cost..

Its sad really... now the people who I can consistently look up to are the old timers .. Dennis Ritchie, Ken Thompson, etc.... these people where about getting things done... how to save development time.. etc... not about increasing develpment time to increase billing time.. Sadly these gray beards... while I look up to them for programming advice (via articles, books, etc); in the current IT field, I wonder if there would be able to survive in the BUZZ-word laden field.. Sorry Dennis, you are NOT using the latest and greatest XYX technology.. we are going to have to lay you off...