O, then mark that link as no-cache, and then it's unstable, but whatever, that's the price you pay.

Also, the partial quoting as inlined content that he suggests would probably fall under fair use if it's small enough.

If everything was sourced this way, we'd be better able to see if multiple sources were all in agreement, or if a single source was being disproportionately referenced.

If your document relies on another document, then your server could cache that other document, and serve it up. That way it's fault tolerant, and version preserving. Then the browser could provide a way to link to the other document's real URL, and show if it's still available, and how it's changed since then. It could summarise the deltas to the quoted portion of the document, and the document a a whole.

So you need to link to a version of the document. That way if/when it changes, your original intent is preserved, and then the user can follow the version history to what's current, to also see what has changed. And if the links are bidirectional, then there's a mechanism for you to know that what you quoted has been updated, so you can then know if you have to revise your own document.

So, for example, if your document was about treating ulcers, and you linked to a document about the causes of ulcers, then when they figured out that ulcers were caused by bacteria, and updated their document, that bidirectional link could provide a means for you to be notified to update your treatment document. Version 1 of your treatment document would be consistent with version 1 of the causes document, and your version 2 could then be consistent with their version 2.

This would then reflect the changing nature of our understanding of the world, and facilitate rippling new information from one area across all other related areas.

With DRAM, you have to power the memory, or it loses the data. You would have to use a completely different type of memory to actually power it down, regardless of whether it's currently being read from or written to.

Actually, they are. Basically, they're taking what's ready, or close to ready, and releasing that as Java 7, and then releasing the stuff that won't be ready for more than a year as Java 8.

These are Oracle's plans for Java:
http://www.oracle.com/technetwork/articles/java/java-platform-2010-174690.html?msgid=3-2517886426txt

These are Oracle's plans for Java:
http://www.oracle.com/technetwork/articles/java/java-platform-2010-174690.html?msgid=3-2517886426txt

These are Oracle's plans for Java:
http://www.oracle.com/technetwork/articles/java/java-platform-2010-174690.html?msgid=3-2517886426txt

I think we're actually agreeing on how open source software models work. I'm just acknowledging that what balances out the risk of giving up control is that there's an inherent inertia towards customers staying with you. If the inherent momentum was away, then no one would do it. And it's not so much a problem with facing competition, the problem is when someone takes the open product and gives it away for free, since their efforts are subsidised by some other revenue stream (like Google's advertising). In effect, they're leveraging one business into another. And that's a game that not everyone can play, especially not small players.

I've commented elsewhere that Sun should have known better. But they were probably busy fighting the last war, like everyone does. They were fighting Microsoft on the server and the desktop, and didn't realise the battle had shifted to the cellphone. They ignored Moore's Law, and lost.

I'm just pointing out that the legalities are less relevant. Sun's fate is less relevant. We should be concerned about the future of the open source business model, which I'm concerned that Google has fractured.

The cell phone market is now completely embroiled in litigation. Just about every company in that space is now involved in several patent lawsuits. It's going to be interesting to see what the outcome will be. Personally I believe in patents, but only when they follow the non-obvious, no prior art rule. But I can see why people are against them, when the are de facto not following the rules.

The First Google phone was released around a year ago, and so it itself must have been in development for a while, and Android must have been in development for at least a year or two before that. With Java, things were looking bright with OpenJDK, and all the JSRs that were shaping Java 1.5 and 1.6 were very community driven. And then Apache started complaining about getting locked out of the tests. So, maybe it was related to Android, and maybe it was just the pandora's box that Apache was opening.