Of course, if you‘ve ever stopped here before, you‘ll know my answer: trailing. Here are the main questions that are still percolating for me: does search always have to be trailing edge? What would leading edge search look like? What are the long term effects of living off of a trailing edge stream?
Almost everyone lives in a world where they think the whole web was just a huge pit of nothingness before Google. I'm not sure whether G rented crop dusters and aerosolized barbiturates, or what, but it's getting to be seriously stupid. For those of us who were not living under that lily pad, we sounded like kooks, then Bing came along and suddenly the memo went out: whatever Google is doing, other people can do it too, and they might have as or more valid approaches. Such a memo does not a paradigm shift make. The paradigm shift memo is when you wake up into a world where you not only don‘t see search in the center of your solar system, you find yourself hardly ever using it anymore. That happened to me recently, when I started doing iPhone programming.
Why? Simple: because the SDK comes with documentation that is excellent, and the thing works in ways that are predictable and contained. Is O-C the best language I have ever used in terms of its testability? it‘s transparency when things go wrong? No. BUT, what has started to really grate on me after years of programming is not that stuff. It‘s the ‘hey, the oil pan fell off again, stop programming and throw the jalopy environment/tool/API up on blocks and become a mechanic on someone else‘s crap‘ stuff. That literally never happens using Xcode/O-C/iPhone OS. Not only that, the new analysis tools make it utterly trivial to make sure that you are not unknowingly slitting your own throat.
People have been talking about how the iPad is going to save publishing. I have argued on here that publishing in the IT world is going down the rathole. That is, traditional publishing. The deluge of iPhone books were a treatise in this: most of them late, useless, and very quickly outdated. Today, I found out that there was a new edition of iPhone in Action, from Manning. I got the ebook version because the one thing that I think is sorely lacking in the SDK docs is coverage of the topic of styling and integrating design into apps (tables, using images, etc.). There is some stuff, but nothing that shows you a really slick interface and how it was created. (Apple could have done this. Clearly, they are believers in leaving a crack for the priests of their app establishment to stamp their imprimatur on their products, which is too bad: if there is one area where the iPhone has left itself open it is in design: the default styles/controls are already starting to reek of cheap.)
Anyway, to say the least, I was not impressed with the new edition. It‘s really a grab bag of odds and ends. Ironically, right in the beginning, the authors make some statement about where they see themselves fitting in, saying something like ‘the Apple docs are not very tutorially-oriented.‘ Um, gee, Guys, time to put the cap back on the model glue. The docs have a ton of working example projects. You can open them and run them in a few minutes. You can try changing them. You can write tests for them. Meanwhile, your book is ricocheting from ‘ok, if you‘re from Python or Ruby, this is what you need to know about Objective C, in a page and a half‘ to ‘this cool little toolkit can mimic native apps through CSS-based web components.‘
This is how these threads come together: search and this kind of literature are the same thing: trailing edge opportunism trying to sprout in the margins of someone else‘s success. I don‘t bring it up to scoff at the concept. There are plenty of cases of trailing edge technology being immensely useful. But at some point, it makes sense to just say ‘hey, maybe if we fixed the roof, we could get this garbage can out of the middle of the living room.‘