all bits considered data to information to knowledge

13Apr/1014

Agile Calculus

One of Agile tenets is cycle shortening : development cycle, communication cycle, integration cycleโ€ฆ It had occurred to me that takingย these to the logical conclusion results in the cycles becoming continuous.

Continuous development, continuous integration, continuous peer review (think XP)...All discrete activities meld into continuous processes.

Comments (14) Trackbacks (0)
  1. “It had occurred to me that Belgium, or, say, Netherlands are countries of diminutive size, so the philosophical thought is propagating through those faster.”
    Continuous development (of the same project) – is an ultimate result of the Agile (XP) approach: do first, think later. This created a generation of young software engineers for whom typing speed is the biggest asset.

  2. Every idea could be taken ad absurdum ๐Ÿ™‚
    Yet there is a rational core to it: shortening development cycles leads to faster feedback. If your code breaks the build you know about it immediately, and can fix it ASAP.
    A positive aspect of XP pair programming is continuous code and design review on a rather low level – class, function… Coupled with unit testing, model-driven development, sound software engineering practices (e.g. static/dynamic code analysis) this can lead to a solid robust implementation.

  3. My point exactly! Whether it’s “ad absurdum” or “a logical conclusion” is in the eye of the beholder ๐Ÿ™‚

  4. Ultimately, efficacy of a process will be judged by the results. Will the system be delivered on time? On budget? Did it implement all features requred? What is the systems defect rate? How many support tickets were opened after go-live? etc.
    So, there are relatively objective metrics attached to the development process…

  5. The metrics might be objective, but one can’t get into the same water twice. Those metrics are usually applied to different teams working on different projects using different processes. “Oh, the patient would have definitely survived if we used Panaceallin…”

  6. Developers are not interchangeable robots, granted; and the best metrics would be collected tracing evolution of a team. Nevertheless, given enough time, individual differences level out.

    While there are wide variations across the teams using different techniques in the long run the market settles the dispute – the best process wins. You cannot rely on super-stars as your business model. In fact, the studies has shown that the best developers are up to 10 times more productive than the worst ones, but the latter outnumber the best by a similar margin [Frederick P. Books The Mythical Man-Month] …

  7. Since it’s your blog I’ll grant you the last word ๐Ÿ˜€

  8. As Michael Bulgakov once noticed, “the point is not that man is mortal but rather that he’s mortal suddenly”…

    The keyword here is consistency of results. Any team can pull a success once; the probability of consistently solid results drops precipitously as more projects are added to the record.

    A validation of any methodology should be a solid track record over lengthy periods of time across multiple teams.

  9. Alex,

    The Mythical Man-Month hypothesis breaks down when you don’t have a profit motive driving your business but rather a policy motive, like the State does. You have seen the resources wasted, er thrown, at projects that would be killed in a “normal” business. We had a solid track record before we changed horses in mid-stream – now we haven’t had one for almost two years.

  10. There are many variables to the equation… changing technologies in mid-flight is one of them. While the State definitely has its specifics I have witnessed similar situations in private sector as well… ๐Ÿ™

  11. “Normal” business is a unicorn occasionally seen around in the form of companies of roughly 10 employees or less. Once the business grows sufficiently, the “profit motive” get diluted to the point of extinction. Do you have any doubt that a manager in one of those intels or microsofts would do everything to prevent a wasteful and useless project from being killed if his career is on the line? Even the stock options donโ€™t help, because the effect of individual projects on company’s performance becomes negligible. Collective farm mentality, if you will, is a function of size more than anything else.

  12. Sadly, I’ll have to second this… ๐Ÿ™

  13. OMG, did you just agree with me??? I must have said something I didn’t mean to ๐Ÿ˜€ ๐Ÿ˜€ ๐Ÿ˜€

  14. Even a stopped clock shows correct time twice a day ๐Ÿ™‚


Leave a comment

No trackbacks yet.