Daniel Lemire's blog

, 6 min read

Does time fix all?

6 thoughts on “Does time fix all?”

  1. jld says:

    “The argument that if enough time passes, the problem will be solved mostly make sense if you plan to live forever.”

    Uh! No! When you’re dead you don’t have any problem either, or may be you believe in after-life, and a one which sucks?

  2. I’ve been thinking about this as well[1]. The durability of a group often boils down to 4 words: Be Welcoming To Outsiders.

    A common response to software being broken is to say, “We rewrite everything in a few years anyway.” It seeks to turn a liability (code) into an asset (dynamism), which si good for morale in a large organization. However, it’s a terrible argument for two reasons:

    a) The code you have today affects what you write tomorrow, even if you replace all of it.

    b) You never replace all of it. A very good org might keep 5% of it. But which 5%?

    [1] http://article.gmane.org/gmane.lisp.readable-lisp/327

  3. It seems to be a corollary to the The Shirky Principle (http://www.kk.org/thetechnium/archives/2010/04/the_shirky_prin.php):

    “Institutions will try to preserve the problem to which they are the solution.” — Clay Shirky

  4. From above: “… What the librarians had come up with were terrible time-consuming systems. It took an outsider (Berners-Lee) to invent the Web….”


    You mean the librarians who had been working on digital retrieval since the late 1940’s and subject retrieval longer than that? Those librarians?

    With the web, every user repeats the search effort of others. Why isn’t repeating the effort of others a “terrible time-consuming system?”

    BTW, Berners-Lee invented allowing 404s for hyperlinks. Significant because it lowered the overhead of hyperlinking enough to be practical. It was other CS types with high overhead hyperlinking. Not librarians.

    Berners-Lee fixed hyperlinking maintenance, failed and continues to fail on IR. Or have you not noticed?


  5. Charlie says:

    When I was at University, all of our research was done the “old-fashioned way”- and, quite frankly it was much easier than it is today. Back when I first gained access to the DARPAnet, I was really excited about a new way to do research- until I found out that, when researching a new topic about which one had limited knowledge of the subject jargon, would result in thousands of citations, many of which were repeats of information published elsewhere, or rehashes of what others had published, or papers that had been rescinded or superseded by newer work…but with time, I developed a private filtering method (based on the inverse of the frequency of publication) that could get me to the core information quickly.
    Then the Internet came along. Great access to a much broader universe of information…finally evolving into a totally useless system for doing basic research because the marketers have gamed the system so thoroughly that the really good information is generally buried ten pages into the search (if it is picked up at all). Valuable information is harder to sort from the chaff than ever before…
    The conclusion I draw from my own experiences is that, no matter how good the system is, someone is going to come along and complicate it to the point where it is no longer functional. NO system is going to be self-correcting- it is going to go down ultimately to over-complication (or, if you want to think in terms of social systems, societies come to inglorious ends because they become overly complex- unable, in the end, to respond in a timely fashion to some crisis- unless they welcome outsiders with a different perspective).
    So, ” People too often believe that the systems they are in are self-correcting. Yet corrections, when they happen, are often actually disruptions brought forth by outsiders. Trust that systems, when left alone, will do the right thing, is overly optimist.

    We should celebrate outsiders and protect them from the wrath of the insiders.”

    I agree 100%, as long as you don’t just pile the contributions of the outsiders on top of the old, non-functional system…

  6. Charlie, it’s hard to understate how much I agree. http://news.ycombinator.com/item?id=4361596