Tim Bray, co-editor of the XML specification:

“The Web These Days · It’s like this: The time between having an idea and its public launch is measured in days not months, weeks not years. Same for each subsequent release cycle. Teams are small. Progress is iterative. No oceans are boiled, no monster requirements documents written.

“And what do you get? Facebook. Google. Twitter. Ravelry. Basecamp. TripIt. GitHub. And on and on and on.

“Obviously, the technology matters… More important is the culture: iterative development, continuous refactoring, ubiquitous unit testing, starting small, gathering user experience before it seems reasonable.”

Italics mark my emphasis. Read the rest of Tim Bray’s Doing It Wrong. And my favorite book on iterative software development is Extreme Programming Explained, which you can find in our bookstore.

Topics Lean

5 comments · Show

  • Brian Dunbar

    He might be right about web development for entrepreneurs. He’s dead wrong about ERP.

    ERP is complicated because the business is. The business is not one bit more complex than it needs to be: complexity costs money; we know this.

  • dbv

    Having built large systems for enterprises and in the past few years consumer facing web services, I can’t help but think that Tim Bray has lost his marbles on this one. Enterprise systems have to consider so many different aspects about a business that a focused web service doesn’t.

    To avoid failure with enterprise systems is to start small, get something working and then build-out further. A huge problem in the enterprise are the systems integrators who believe the opposite: start big and if it fails make sure that contractually it is the customers fault. There is usually an immense amount of corporate politics involved too with large enterprise systems.

    A review of the financial size of the systems integration businesses will provide most of the answers to the high cost and failures of enterprise systems.

  • Krishna Muppaneni

    Facebook. Google. Twitter. Ravelry. Basecamp. TripIt. GitHub. And on and on and on are not developed over night. They evolved over a period of time and continue to grow with increased features and functions. It is hard to believe that all these were result of some over night invention without going through the whole 9 yards of SDLC processes. The success behind these are the take one bite at a time, take cutomer feedback and improve upon it. Whereas real life business systems are different. They need due deligence in documenting the requirements. I do not want to risk killing someone by not doing the proper requirements gathering for a critical healthcare system!

    IT Projects fail due to lack of understanding of the requirements, improper processes, lack of excutive sponsership and even individual egos and an attitude of “My way or highway”.

  • Brian Dunbar

    I think it is also important to note something that may be overlooked if you read horror stories on The Reg: we’re not out here in the wild endlessly re-creating the final scene to the movie ‘Titantic’.

    Stuff gets built, systems are improved. Much product is being shipped and a lot of business users are if not happy then content.

    If we were doing it completely wrong, the biz guys would demand that we do something to make it right.

    I’m not impartial: I work for a mid-sized electronics manufacturer that, over the past decade, has put a lot of capital into our ERP system. But I’m also not an integrator depending on ERP for a living. I have not always been in ERP-land and I will not be here forever.

    We’ve had our successes and a few failures. Overall, we’re a net positive for the company: we ship more, faster, than we did before.

    Bray is right — things can be changed for the better. But the result will not look like Web 2.0 but more like what has come before, but faster, more Agile, and in tune with our customers in business.