Symphonious

I found it rather amusing that while filling out my census online it asked me:

Question 59: Can the internet be accessed at this dwelling?

To be fair, the options provided to answer included a description of what type of connection you had so the question couldn't just be assumed, but it's funny none the less.

I haven't paid a great deal of attention to the WWDC keynote details - just sampled the various discussions going on. I was however interested in a comment by Ted Leug that Apple were including DTrace in Apple's performance tools.

I wonder what Sun think about this. DTrace was a key Solaris feature and now it's coming out for OS X and I seem to recall mention of projects that are porting it to Linux. As a developer it's nice to see your code become popular so the DTrace team are probably thrilled. As a company though, it's hard to leverage the benefit of your investment when everyone else is reaping the benefits. Even if Sun get improvements back from Apple, where's the benefit for them? Apple and everyone else have those improvements too. The ubiquity argument of Java doesn't seem to apply here, you don't build on top of DTrace, you use it as a debugging tool. The support business model probably doesn't pan out either - Apple will be providing support for it themselves and they're the experts on DTrace on OS X, not Sun.

Overall, I can see benefit to Sun in open sourcing Solaris - the ubiquity and support business models have potential for the OS as a whole, but the ability to pinch key advantages of that OS and port them to other OS's seems to be a boon for Solaris competitors. Then again, is Sun really competing in the OS market or are they just using Solaris and Linux to sell their server hardware? Most likely, they don't stand to lose a lot of Solaris gradually dies off, as long as their hardware runs the OS that takes over from it.

That said, clearly this is a boon for OS X developers so you won't hear me complaining about it.

Mitch Tulloch raises a concern over how large enterprises would react to Microsoft moving to more regular, iterative releases. The answer for large enterprise who can't handle releases coming out more often than once every six years or so is to only update when they are ready instead of every time there's a new release.

With Apple, this isn't a great option because Apple don't have very long support periods for older OS's, but that's not the case with Microsoft - a legacy of the fact that they deal with large enterprise and Apple in general does not. Companies that were running Windows 95 would only have reached the point of having to upgrade, what, a year or two ago? That's easily more than the required six years between upgrades.

So if enterprises won't upgrade anyway, why would Microsoft want to release more regularly? Well for a start it avoids the embarrassment of huge schedule slippage like with Vista and more importantly, it would help keep Microsoft competitive in the consumer market - something that is very slowly moving away from Microsoft. That's not to suggest that Microsoft will lose it's commanding lead in the OS market or that you should expect Apple's market share to suddenly jump. However, Microsoft is clearly under increasing competitive pressure that would be relieved with more frequent releases.

It seems that the world of Java on Mac is always full of drama and gnashing of teeth and Chris Adamson has just the article to really kick it into full gear: Mustang for Mac PPC… any point now?

Sadly, I'm inclined to believe that Chris may well be right - the Java 1.6 release for PPC is likely to be a less than wonderful release as Apple focuses it's efforts on the Intel release. It's a given that the Intel release will be better than the PPC release just due to the fact that they now get all Sun's optimizations for free.

The good news for Mac Java developers is that this means that Java on Intel Mac is going to be pretty awesome. Apple can now devote most of their development time to polishing the Mac specific features and optimizing the OS specific bottle-necks. The Java 1.5 release on Intel is already a lot more comparable to performance of Java on Windows - partly from the optimizations they get from Sun and partly just because of the faster chips. Overall, I'm really looking forward to seeing what Apple's Mustang release is like - I expect it to be pretty impressive, and that makes a nice change when talking about Apple's Java releases.

Jason Calacanis' entry on Wikipedia considering adding a WYSIWYG editor to make it easier for people to contribute strikes a chord close to my heart. The argument that a WYSIWYG editor will cause more work for administrators is quite valid - making it easier for people to contribute will mean more contributions that need to be reviewed and checked. On the other hand though, the benefit of a WYSIWYG editor isn't just that more people will contribute, but that domain experts in fields other than computing will be able to and be more inclined to contribute. For Wikipedia, that's a pretty huge benefit - the people who know most about a subject will be more likely to actually be writing the Wikipedia article on that subject.

There is a proposal to do a "semi-WYSI" that is what will probably come to Wikipedia. A "semi-WYSI" would have only a handful of features and it would translate the edited text into Wiki markup (as opposed to HTML) for users.

A semi-WYSI, or rather a WYSIWYG editor configured to allow just the operations that add value and not those that make articles overly complex, is an excellent compromise between the two viewpoints. Even better, it actually makes it easier for people to learn to use the WYSIWYG editor and get to contributing. Less time is wasted on formatting and more on writing great content. That said, there's no reason this couldn't be done with HTML instead of wiki syntax, and the HTML editors are generally better than wiki syntax editors. It's not like it's hard to convert wiki syntax to HTML…