The State of Massachusetts has decided that they will continue using Microsoft Office but will make OpenDocument format (ODF) plug-in part of the standard install.

The irony is uncanny: in an attempt to be politically correct in support of open file formats, the Commonwealth's CIO has decided to tell people to use the world's de facto office software suite, but optionally (?) save their files in a format that may not maintain the full fidelity of the documents themselves.  The irony doesn't stop there: it turns out that Microsoft Office, rather than OpenOffice, is in fact the politically correct choice, given its richer accessibility features (i.e. features for people with vision, hearing and other disabilities).

That OpenOffice lacks these accessibility features cuts to the core failing of open source software: without a regular, paid staff of developers, some of the nitty-gritty features, the ones that are less "sexy" and which appeal to a minority of users, just don't get implemented.  With commercially developed software, people get paid to do what others might consider grunt work, and they get reviewed based on their performance on this work.  So it gets done.

The failing of OpenOffice will become more apparent once Office 2007 is released.  And maybe OpenOffice will respond, within two years or so, with something representing feature parity.  But playing catch-up isn't a software development strategy; it's a grudge match.

I just got my Treo 700w to sync over Bluetooth.  And I am very proud of myself.

But, in a technologically sensible world, I shouldn’t be…that is to say, setting up an ActiveSync partnership over Bluetooth shouldn’t be that hard.  In fact, there shouldn’t even be a concept of setting up a “partnership.”  I should just be able to tell ActiveSync to allow Bluetooth connections and go.

Instead, I have to go into the Bluetooth stack software on my laptop, enable the serial port profile, determine its COM port in the property sheet, go into the ActiveSync Connection Settings dialog, allow connections to that COM port, pair the phone with the laptop’s ActiveSync profile, fire up ActiveSync on the device, and then, finally, tell it to connect over Bluetooth.

Bluetooth was initially designed to replace the cables used to connect PCs and peripherals.  But until Bluetooth is as easy to use as a USB cable, it will be relegated to a specialized protocol for cell phone cordless headsets.  People will plug and play with cables until playing unplugged doesn’t require voodoo or passing a certification exam.

Something’s probably a little screwy with me.  I’ve chosen tonight, my first night of vacation, to try and get back to posting here.  No matter, whatever it takes, it takes.

Honestly, I am still convalescing.  About 2 months ago I finally finished work on Programming Microsoft SQL Server 2005, the book I’ve been working on with Stephen Forte, and nine other people too.  If you include the index, the book comes in at a hefty 900+ pages, and it still feels like we just scratched the surface of the product.

Our book was really, really late.  We should have released it in conjunction with the product itself.  Shame on us; I won’t bore you with the reasons why it took so long.  Except one: the product itself was late.  Want a fun factoid that illustrates this well?  We signed the contract for the book with Microsoft Press right around the time of the Yukon Technical Preview conference in Bellevue, in February 2003.  Yes, you read that right.  From the Tech Preview Conference, when the first alpha/preview bits were released in the form of VMWare images (Microsoft hadn’t yet acquired Virtual PC from Connectix!), to product release was almost three years.  Our book merely took another six months to finish and an additional month to hit availability on Amazon.

It’s not just SQL Server 2005 of course; Vista and Office 2007 are massively delayed too.  And what about “Longhorn” Server?  What about IE7, which went into Beta a year ago?  And when will Exchange 2007 really ship?

What’s going on in Redmond?  Clearly, I am not the first to observe this problem or ask this question.  But it bears asking nonetheless.  Forget all this talk about “innovation.”  I prefer to focus on simple “organization” and “dedication.”  What will it take for folks at Microsoft to get some hustle in them?  To feel the panic of a deadline?  To pull an occasional all-nighter?

Maybe I’m just bitter.  During the last three months of working on the book, I worked every day of the week, save for maybe four days off.  I closed deals at the office, I spoke at conferences, I wrote, I edited, and I answered hundreds email of messages a day.  I’m guessing many other software book authors go through the same thing.

As shareholders, as business partners, as platform advocates and as colleagues, we need everyone at Microsoft to do the same.  If the Zune is to succeed, if the XBox platform is to continue its upward climb, and if the Windows Server stack is to complete it ascension not just to serious contender, but to a ubiquitous no-excuses platform for the corporate IT world, Microsoft needs to bear down and push.

I'm not pushing again until next week though.  Now, back to vacation.

Materials from my Tech*Ed 2006 session "Developing SQL Server 2005 OLAP Applications with ADO MD.NET" (BIN319) are available right here.

 

While the value of BI has been well-demonstrated for quite some time, that value has seemed hypothetical to many, because of the difficulty in keeping OLAP cubes up-to-date.  Compared to an OLTP database, OLAP cubes have, for many, projected an image of inferiority because their data has for the most part been reflective of historical facts rather than of current transactions. 

 

Historical data, and analysis of it, is very important, but appreciating that importance takes a certain vision and leap-of-faith.  In most over-stressed IT organizations, dealing with historical data feels less urgent than dealing with current data.  This has pushed BI down on the list of priorities…far enough down, in fact, that it has never gotten off the backlog list in many organizations.  And while techniques have always existed for keeping cubes reasonably up–to-date, doing can involve a lot of work and require a lot of resources (of both the hardware and human variety), pushing BI projects further onto the back burner

 

SQL Server 2005’s Integration Services and Analysis Services 2005 change the landscape of the BI market by making real-time BI easy.  New advances in these components allow you to load data directly into your cubes bypassing your data warehouse and/or allow Analysis Services to process cubes automatically in the background, in response to changes in your data warehouse.

 

I believe these advances will eventually change the whole dynamic of the BI market.  People at all levels of management and even non-management will now be able to perform sophisticated analysis on current data, and they’ll be able to do so much faster than they could with a relational database.

 

As groundbreaking as these advances are, for people new to SQL 2005, it will not be immediately obvious how to take advantage of them and thus it will not be obvious that they even exist: the direct data push capabilities of SSIS are stashed away in the Visual Studio toolbox and the proactive caching features of Analysis Services are buried deep inside a wizard and some dialog boxes.

 

Take a look.  You can adjust proactive caching settings directly from Management Studio or from Analysis Services projects in Visual Studio:

 

 

 

You can also do so from the Aggregation Design Wizard (available in both Visual Studio and Management Studio):

 

 

And the SSIS data push tasks are highlighted below:

 

 

I am posting these screen captures to make it clear how easy these tools are to get at, if you know where to look.  In many cases, using them requires only a modicum of effort.  I will try to coverage some basic usage scenarios in another post.