Saturday, January 27, 2007

Restaurant Recommendation

Last night, on a whim, Natascha and I ate at Ponti Seafood Grill. I have not eaten there in quite some time, but remember liking the food, and it is close to where I live. Either my memory failed me, or I just had an different experience last time. The experience I had last night was amazing! I had the pepper seared albacore, which was absolutely delicious (although too peppery for Natascha's tastes). Aside from the main course tasting so darn good, the overall pacing of the meal was excellent, and the staff was great. The sommelier had excellent recommendations, but more-over, I didn't get the normal, I-know-wine snobbery that I'm used to at upscale restaurants. The staff was very friendly and quite efficient. The final piece was their calamari... done right! The oddest bit of the evening, was that the owner was wandering around, chatting with tables. Often that would make me feel uncomfortable, but he really felt more like your neighbor.

Both Natascha and I had seafood, so I can't really speak to the rest of the menu. I can say with certainty, that I intend to visit more often.

[spelling update...]


Friday, January 19, 2007

For my birthday... someday

Someday, for my birthday, I want one of these amazing sculptures. As a math dork, these are incredible. My desk wants one of them....


Thursday, January 18, 2007

VisualStudio vs Eclipse

I met up with a friend from Microsoft this last weekend and was challenged to explain why I prefer Eclipse over VisualStudio. The answer wasn't obvious. It isn't really that one has features that the other doesn't have. Both are slow memory hogs, and in fact, Eclipse is usually worse than VS on this front. Both provide good intellisense and refactoring.

So why do I prefer Eclipse?

In the course of talking about it, I came up with my answer. Eclipse if built by top developers for themselves. VisualStudio is built by top developers for some other developers. Both IDEs were built by teams of some very smart developers, but Eclipse developers seem more focused on building a product that they themselves would want to use. Microsoft has a strong internal culture that says that their primary target developer isn't as good a developer as most Microsoft developers. Which do you think will result in a 'better' product?

There are truths to Microsoft's delema. A huge majority of their target customers are not code gurus who think in boolean logic and can read hexadecimal upside-down. The majority of VS's customers are still VB developers who are more interested in drag'n'drop forms layout and easy database integration. I think Microsoft got it backwards though. They should be building a product that they love to use first, and then worry about 'dumbing it down'. Why? Because I think it will surprise them how similar these app-devs are to themselves. I think they would find that rather than 'dumbing it down', the would find that the app-devs have some excellent ideas for improving work-flow for common operations, that benefit everyone.

As an aside, this also explains why I liked Visual C++ 6 and put off 'upgrading' as long as possible. VisualStudio 7 was when they merged VB and VisualInterDev with VC++. Before that, VC++ was built targeting their own use-cases. With the grand union of all dev products, it was a compromise for everyone. I know as a VC++ user, I felt like I lost a lot, and I bet VB users felt similarly.


Friday, January 12, 2007

Minor Upgrade

I just move to the 'New' Blogger. I'm not using any of it's features yet, but I hope to start using the tagging. If you use a aggregator, you notice the change because all my posts show up as new again...


MacFUSE make'n me jealous!

I've been watching the FUSE project for a while, and am sincerely jealous. For a variety of reasons (mostly hardware and software compatibility) I spend most of my time in Windows. I keep holding out the hope that someone will implement something like FUSE for Windows, or even better re-implement FUSE on Windows, but no.

And now Google comes along and releases MacFUSE!!! Arrggg!

Monday, January 08, 2007

W3C announces EXI standard to be based on Efficient XML

The W3C Efficient Xml Interchange Working Group just updated their published status to reflect that the group voted to base the forthcoming standard on our Efficient XML!  This is big news for my company! I have learned a great deal, and my opinions have changed significantly, of the course of my working with different groups building 'binary' XML solutions.

'Binary' XML is a complicated and controversial beast.  It needs to balance the desires for read/write performance and wire data-size.  I have been involved in multiple 'binary' XML efforts, and am exciting to see the W3C working to build a standard.

There is a lot of confusion about 'binary' XML.  In my mind, 'binary' XML is all about enabling XML for scenarios that were not viable before.  The main problem that Efficient XML solves, is how to leverage XML over low-bandwidth links.  Too many people are focused on using XML to improve XML read/write performance.  While XML can provide some gains there, improving performance will not significantly enable XML where it was not usable before, but improves existing scenarios.  By focusing on the limited-bandwidth scenarios, Efficient XML is already enabling existing XML services to integrate with clients that did not have the bandwidth to support those services as raw XML.  Computing processing power is still improving dramatically, but bandwidth growth has been much more limited.

None of this is to say that read/write performance is not important as well.  I've always been an evangelist for efficient implementations and the last year has been exciting for me.  I am amazed at how much performance can be achieved while still prioritizing wire efficiency.  I have been repeatedly surprised by how fast we were able to make out Efficient XML implementation.  I am very proud of our work and have no doubt that it blows the socks off any other 'binary' XML implementation with which I have been involved.

I look forward to seeing further progress from the EXI group.    As happened with XML, I expect a standard 'binary' XML will enable entirely new classes of applications to leverage XML.