Making the world safe for infrastructure

While I haven’t been blogging much in the last few weeks, and I haven’t also been in the hospital or otherwise slowed down, I’ve been writing for Linux Journal and the Berkman Center. Some of the Linux Journal stuff is what I write every month for issues that appear three months in the future. For example, yesterday I finished being late for the July issue deadline. Some of what I write is for LJ goes straight to the website, which has been improving steadily lately (quite aside from my own contributions). The latest of those is called Understanding Infrastructure. It’s close to four thousand words, and has been in the works for about a month now.

Infrastructure has been an interest of mine for some time. I think it’s one of the world’s most overused yet least understood concepts. And I have an immodest fantasy about correcting that.

Not alone, though. A bunch of other people I hang and talk with have similar ideas. This latest piece is grist for our collective mill. Maybe yours too.

It’s not perfect. In the words of Mark Twain, if I’d had more time I could have made it much shorter. But I needed to get it out there, so there it is.

The Berkman Center pieces are also kinda big, and will show up between now and Berkman@10.

4 comments

  1. Mike Warot’s avatar

    Doc, you’re right, a conversation about what we have, and what we want to end up with is a good idea right now.
    I’m trying to get my thoughts about this into a coherent whole, but it’s taking a while. You might think that it’s ok to post “half baked” thoughts in a stream of consciousness, but that’s because you’re a writer and editor, and quality of your first draft is much better than the second or third draft the rest of us might come up with.
    I think we’re not quite there yet… the infrastructure doesn’t support mobile code, and is paranoid about avoiding it, for good reason… the side-effects can be catastrophic, and there isn’t any way to sandbox things effectively.
    I know I’ve got an important point to make, several, in fact, but getting them out effectively is really much harder than it seems like it should be, if blogs worked as well as the hype says.
    Oh well… I’ll keep plugging away, and ping you back when it looks like it’s reasonably complete and readable.
    –Mike–

  2. Mike Warot’s avatar

    Programs are currently the most discriminated against users of the internet, followed closely by the data itself. We’re constantly told not to trust them, to treat any random email or web page as a potential serial computer serial killer, and this approach is based on common sense. Any random program can kill your computer before you even know it.
    The mechanisms to allow us to end this discrimination are starting to fall into place, with the java sandbox, and virtual machines for surfing, but they don’t go far enough. We need to be able to let our guests into our computers, feed them a good meal of data, and enjoy the results of the interesting conversation afterwards.
    Mobile code is one of the best uses of the internet, we need to figure out how to make it safe to use. I strongly suspect that never trusting it, and limiting its actions to those defined by “capabilities” is the best way to go. Perhaps there is a better scheme, but I don’t see it.
    –Mike–

  3. Peter Bodifée’s avatar

    Like Mike Warot I am also trying to get my thoughts in writing. And it isn’t easy because of fear for being misunderstood. Many organizations are still led by the idea that open and public domain based infrastructure is not for them. Organizations (communities if you will) should make different decisions in order to make it much easier for them to implement and in particular maintain applications. That could be the driver to achieve the necessary ubiquity which will make the world a better place to live.
    When I have my thoughts ready for publication I will let you know. Let’s keep the conversation alive!
    Peter

Comments are now closed.