In yet more SVG-related news, Microsoft demonstrated tools it's building for its SVG marginalizer - Sparkle. You can see more on Channel 9. I came across this on Slashdot. It appears that they're talking about Microsoft Expression, which I linked in the SVG discussion.

After reinstalling Visual Studio 2003 (while I could try diagnosing every single possibility, I just needed the machine up and running as soon as possible), the problem persisted - "Could not create automation server" whenever I added a project to a solution, failing the operation. Somehow something was broken during the process of uninstall a prior CTP of SQL Server 2005 (and the .NET 2.0 runtime), and installing the newer version.

In any case, I eventually tracked it down to the scripting engine being broken, so I downloaded the 5.6 scripting engine. Of course the MSI was blissfully unaware, and thus belligerent towards, Windows Server 2003 (which I use as a development/test workstation), so I had to run it in an XP compatibility mode, but that solved the problem. This is entered in the archives just in case someone in the future has this problem.

I've unintentionally come across this website several times (today I had a need to reference a double-slit experiment), and each time it has sucked me into meandering through its excellent interactive demonstrations. While this is high-school level physics, if you haven't brushed up in a while it can be eye-opening and educational. The site appears to have been designed some time back, and aesthetically it is subjectively rather dated, but it is pure genius the way it uses interactive elements for education - This is the sort of thing that demonstrates computer and web learning at its finest.
 
Irony (or is it?) - Mental confusion over how to spell genius ("Genious? Hrmmm...")

Brad excited by what he saw at PDC.
Brad Feld, venture capitalist: 2006 Will Be The Year of Microsoft.
[Scobleizer: Microsoft Geek Blogger]

A surefire way of getting linked on Robert Scoble's highly rated and ranked blog is to say something promising about Microsoft, or to say something negative about Microsoft in a way that Mr. Scoble can easily defuse while pretending to agree with you (such as http://radio.weblogs.com/0001011/2005/09/16.html#a11174, where Scoble subtly claims that Balmer was being misinterpreted). In this case, a VC was drunk on the PDC, and like many before him presumed that everything Microsoft spins is gold. Many foolish predictions have come out of such a situation.

In any case, while Microsoft will continue to do very well by any definition, it's quite a stretch to call 2006 the year of Microsoft.

• SQL Server 2005 and Visual Studio 2005 will barely be out of the gates, having come out (hopefully) at the end of this year - I believe early November. It will be quite some time before legacy shops (which are of course the bulk of SQL Server customers. The RDBMS market isn't a heavily expanding market, and is largely selling to existing customers or trying to steal customers from your competitors) will upgrade their SQL Server installs, especially given the large schism to 2005. Visual Studio 2005 is a great product, but it's largely an upgrade/maintenance release for the existing Visual Studio base, a large percentage of which are on the MSDN program anyways and thus will upgrade at no revenue advantage to Microsoft.
• Vista will see adoption among the leading edge, and of course will replace XP on new PCs (to no gain to Microsoft), but it will see negligible adoption in corporations. Many corporations haven't even modernized to XP at this point, 4 years after its release and despite it being a minor difference from 2000. Vista represents a pretty large transition.
• Office 2006 is similar to Vista - it'll take years to see heavy adoption in the corporate space, and otherwise will largely be sales to new buyers (who will just be getting 2006 instead of its closest competitor, Office 2003).
  
• Many of Microsoft's new technology platforms will see negligible adoption in the first, and even second, year. This is just experience talking, but there is a long list of examples of highly lauded Microsoft technologies that struggled or saw little adoption for years (.NET being an excellent case. By now we were supposed to be awash in .NET applications. While it is a superlative web platform, and was a critical upgrade to classic ASP, the desktop world has been barely impacted by .NET at all, against most predictions)

Undoubtedly Microsoft has some great products, but we're talking about a company that needs to maintain a revenue of $10 billion a quarter, and with programs like software assurance and the MSDN program, these upgrades are long overdue. If anything, these releases are about avoiding a loss of revenue stream rather than an increase of revenue stream.

This technology demo is a fantastic watch, and is truly a revolution in the way data access will occur in our .NET applications. A LINQ technology preview can be downloaded for the just-released Visual Studio 2005 Release Candidate here. The LINQ project page can be found here. I'm trying out the LINQ technology preview and will post more thoughts shortly.

Another part of the LINQ project to watch closely will be the Object Relational technology in DLINQ. Both of these will be hugely useful in dissolving the tremendous disconnect that has traditionally existed between relational persistent storage and our code.

Tough to listen to the hype machine kicking up for C# 3.0, though, given that C# 2.0 isn't even to production yet. It's great to be prepared for new technologies, and to get your input in at this stage so they can make it the best that it can be, but still it's really difficult to find the time to learn currently impractical technologies when there are so many practical technologies we can learn to help us today.

If you're in the Cogeco service area (in Ontario and Quebec), you owe it to yourself to look at their broadband solutions. I've been a cable-modem customer for about 4 years now, in different parts of the city, and have been extremely pleased. Prior to that I was a Rogers cable-modem customer, and my experience then was much less satisfactory.

This comes to mind watching the download task for the newest Visual Studio 2005 Release Candidate progress - 2.5GB in, my average download speed is a constant 650KB/second. Not Kb/second, but real, bonafide KB/s. Amazing.

While I don't actively target the Linux platform, I have always been comforted by the Mono Project. That cross-platform-.NET project gives me a feeling that my .NET work could at least partially be ported to other targets if a need arose, removing the vendor lock-in to Windows when using the .NET platform (even though I have never had the need to use it, I pursue the same target when developing in C++). Every couple of months I install the latest incarnation of Mono on a Linux virtual machine, copy over an assembly (ensuring that it uses only features that exist in Mono), and run it, and it really is a little bit of magic when it runs successfully (and quickly I should say - the performance on Mono is spectacular).

The thing that I don't get, though, is the bubbling animosity Microsoft displays towards the Mono project, and the legal uncertainty some within Microsoft continually try to create around Mono. Don't you realize, Microsoft, that Mono is one of your greatest allies? That .NET has been sold within countless shops based upon the argument that it was "cross platform", courtesy of Mono, and that the worries of the executive that Linux needs to be considered are all taken care of, again courtesy of Mono? Microsoft really should be sending Manuel a big, fat commission check.

I can appreciate that Microsoft doesn't want anyone usurping their creation, but really - in virtually every rational situation Mono is a winner for Microsoft.

• It helps sell .NET as "cross-platform"
• It will always be one step behind, ensuring that Microsoft has the upper hand. Sure, it'll implement some technologies (such as vector graphics) more quickly than the Microsoft's .NET framework, but to most developers that is non-standard, and instead it'll "really" come out when Microsoft releases it (and Mono will then have to play catch up)
  
• It's training loads of Microsoft-hating OSS lovers in the ways of .NET, drawing them away from other superb environs like Python. A very large percentage of the people targeting Mono would have targeted Python or J2EE or C++ otherwise.
  
• It helps Microsoft compete on their "home turf"

Microsoft really needs to put down the gauntlet, embrace Mono, and remove the legal uncertainty around it.