Hacker News Reader

Hacker News

Comment by icey | Hacker News Reader

Full thread

icey

TL;DR:

Prediction #1: XML databases will surpass relational databases in popularity by 2011.

Prediction #2: Someone will make a lot of money by hosting open-source web applications.

Prediction #3: Multi-threaded programming will fall out of favor by 2012.

Prediction #4: Java's "market share" on the JVM will drop below 50% by 2010.

Prediction #5: Lisp will be in the top 10 most popular programming languages by 2010.

Prediction #6: A new internet community-hangout will appear. One that you and I will frequent.

Prediction #7: The mobile/wireless/handheld market is still at least 5 years out.

Prediction #8: Someday I will voluntarily pay Google for one of their services.

Prediction #9: Apple's laptop sales will exceed those of HP/Compaq, IBM, Dell and Gateway combined by 2010.

Prediction #10: In five years' time, most programmers will still be average.

View on HN

#1 is kinda sorta like the NoSQL movement (hierarchical versus relational modeling, do away with much of the O-R impedance mismatch).

#2 Github?

#3 The evils of threads, and how to replace them with something else to use multi-cores, is very high on the list of blogged about programming topics. Several new languages have staked their claim to fame largely on how they handle concurrency. So, yes, this one was understated, if anything.

#4 Not there yet, but might be on that trajectory if there is some "tipping point" of Java programmers en masse getting fed up with Java's limitations. There are certainly enough alternative JVM languages to choose from.

#5 Does Clojure make the top 10? Certainly not in number of users, but maybe in most popular to blog about.

#6 Facebook.

#7 iPhone was released June, 2007, beating his prediction by a couple of years.

#8 Maybe if you buy an Android phone from them?

#9 Apple laptops are immensely popular, but the scale of this prediction was a little ridiculous. Between iPhone and iPod, it's possible that there are more people who own an Apple product than own a product from all those companies combined.

#10 Now that's just insulting. Who's he calling average?

#10 Now that's just insulting. Who's he calling average?

The average programmer.

As if that weren't bad enough, at least half of them are still below average.

Only if it's a standard distribution (eg bell curve). It's quite possible with a non-standard distribution to have much more than half above average as well. Now median on the other hand, by definition, separates the distribution in half... </pedantic>

Or perhaps much more than half below average.

Given that there’s no reasonable way to quantify programmers, and we could arbitrarily change the scale however we like even if there were, usage like “average programmer” is generally understood to mean the same thing that “median programmer” would mean, if we had such a rubric (since we don’t, “average” is a fine term). Your pedantry is misplaced.

Emacs Lisp seems to be the 11th most popular language on github. Common lisp, clojure and scheme seem to be 19,20 and 21 respectively. If we count them together, they might make it to top 10.

It depends on what else we count together.

#2 Heroku

Yes, #10 is a bit offensive, implying that he is in the top but you probably aren't. Also, it is rather unlikely that you can attain a huge productivity edge over others in everyday tasks like "take an old buggy codebase and add a simple feature here and there". This might work in more research-oriented programming, where some can create things that others didn't even think possible; but I don't think that's what he means.

I didn't find it offensive. It just means you have to think twice before using fancy stuff in code other people will be maintaining.

#8 Google Apps

The evils of threads, and how to replace them with something else to use multi-cores, is very high on the list of blogged about programming topics.

They may be high on the list of blogged about topics, but in places where performance really matters threads are used more and more. If you for example look at the article http://lwn.net/Articles/357658/ you'll see that Google uses thousands of threads on 16-32 core machines and in fact is making changes to the Linux kernel to make it support a larger number of threads.

OK, I'll take a stab at them. (Yes, I did read his Important Note, but why not?) Status:

1: No

2: No (I think). I don't know if Google App Engine is making money for them, and Github doesn't host apps, and Amazon Web Services doesn't (only) host open-source web apps.

3: Hopefully.

4: Doubtful, but one can dream =)

5: No.

6: HN? Is Steve Yegge here?

7: Yes. I think it's valid to put the iphone's success in 2009 (rather than earlier) due to the large number of apps and growing number of competitors in that year.

8: Well, considering he's working there, I think he's being paid instead =)

9: Probably.

10: Yes, although it's not a tautology. (You could have asymmetric heavy tails which would make most programmers below or above the mean.)

If you take prediction 1 as key-value stores (which I think is more-or-less where stevey was going) - he's spot on.

"surpass relational databases in popularity by 2011"

Are you really sure about that? AFAIK the use case for key-value stores isn't one that a lot of people are going to suffer (seriously high-volume).

I think there was a lot of excitement about XML databases around that time, because they would let you store more complex structures in the DB, not less complex (like key-value). Somehow XML did not take off, not sure why.

"Prediction #10: In five years' time, most programmers will still be average."

In order to sleep through the night, I'll just assume that he's not talking about me.

Reply on news.ycombinator.com