Thursday, July 03, 2008

SOA Brouhaha

A recent guest editorial by ZapThink’s Ron Schmelzer on TechTarget.com has gone out on a limb and panned Microsoft’s messaging around Oslo and its greater SOA strategy.  Schmelzer and ZapThink really believe that Microsoft has SOA all wrong.  The most prominent objection in the editorial seems to be that Microsoft is way too nuts and bolts about SOA — seeing it as more about application integration and less about the architectural paradigm shift that many SOA proponents feel is crucial, not only to understanding SOA, but to understanding software today.  Quoting Schmelzer: “Microsoft has made a critical (if not fatal) mistake of turning SOA into a developer initiative focused on standards-based interoperability.”

At first blush, this critique sounds spot on, because Microsoft has made this alleged error before.  In fact, when the first version of BizTalk launched, it was hampered by Microsoft’s selling it as a developer tool.  If Steve Ballmer’s infamous Developers-Developers-Developers “monkey dance” is the company’s hammer, then we shouldn’t be surprised if Microsoft turns everything into developer nails.

But is that what’s really going on here?  For ZapThink, what is the crux of the philosophical/paradigmatic misstep on Microsoft’s part?  Quoting the editorial again: "The real power of SOA is not simply in standards-based integration (didn't XML and EDI provide that, too?), but in the power of composing heterogeneous services in environments of continual change."

In other words, SOA is all about loose coupling, and building (composite) apps around and for it, rather than point-to-point integration over Web services.  OK.  What in the Oslo messaging (and technology) somehow precludes this?  Isn't this what things like BizTalk Services, ADO.NET Data Services, SQL Server Data Services, support of RESTful services in WCF, and the WebClient programming model in Silverlight are all about?

I’m worried that there’s a failure of rhetoric (to quote The Police, backwards) here.  Microsoft does speak to the "connected business” in their messaging, and they absolutely speak to the necessity of helping developers build Web Services at scale, and with less plumbing engineering for each implementation.  Perhaps that’s not organized well for those that feel strongly that the priority should be on getting developers to build things that are solidly factored and designed for-reuse.  If Microsoft concentrates too heavily on helping developers hook stuff together, and doesn’t place a premium on changing the way developers think and plan how they build software, isn’t that selling SOA short?

At the very least, that criticism is doctrinaire.  It’s doesn’t speak to the substance of what .NET 3.0+ and Oslo are about (or not about), but is instead an oblique objection to what Microsoft is not saying, and not being fervent about.  Even if that sort of policy-oriented protectionism were reasonable, is the architectural principle of re-use-before-all-else providing the results to merit such an emotional defense?  If SOA is mostly about software re-use in the enterprise, how’s that mission working out, anyway?

To quote one analyst in a piece entitled Core Value of a SOA is the Ability to Reuse Services? Not a Chance.:  “[the value of] reuse is …much less than we expected, or the ‘SOA hype’ has been stating. The true value of SOA is the ability to create enterprise architectures that provide much better agility than the overly complex, static, and fragile architectures we have around today.”

Whoever said that seems in sync with some of the Oslo ideas around modeling, developer productivity and building scalable services.  Whoever said that must be in Microsoft’s pocket, right?  Ironically, it was ZapThink’s own David Linthicum.  If you’d like, read the full piece, so you can have all the context.

Is it really sensible for ZapThink to make re-use politics the backbone of its critique of Microsoft’s SOA strategy?  Granted re-use was not the phrase ZapThink used, but it’s the concept they seemed to invoke.  Meanwhile, ZapThink seems to lack re-use even amongst its own analysts.  Small wonder they dislike the phrase “connected business.”

#    |
 Monday, June 23, 2008

Analysis Services 2008 Wizards: Is the Magic Gone?

I’m updating chapters in my book on SQL Server 2005 for the impending release of SQL Server 2008. Right now, I’m focusing on the chapters that cover Analysis Services (AS), the BI component of SQL Server.  I’ve come across an “improvement” in the new version of AS, and I’m not sure I like it: the wizards do less.

In AS 2005, the Cube Wizard could create a time dimension for you, and the Dimension Wizard could create a parent-child dimension/hierarchy for you. 

Cube Wizard AS2005 Time Periods Cube Wizard Dimension Design AS2005 Parent-Child Dimension AS2005

In AS 2008, only the Dimension wizard will create a time dimension, but won’t do so very easily when you supply your own dimension table for it.  And, as best as I can tell, under AS 2008, there is no wizard interface at all for creating parent-child dimensions.  Even for regular dimensions, the Cube Wizard creates only a key attribute for each generated dimension.  If you want multiple attributes, you’ll need to add them manually or use the Dimension Wizard.  The cube wizard actually gives you no insight into attributes at all:

Cube Wizard AS2008 Cube Wizard AS2008 Finish page

This is most perplexing.  So let’s play devil’s advocate.  The AS 2005 wizards had a feature called Auto build (it was called IntelliCube when AS2005 was still in Beta).  It didn’t work that well, and I usually disabled it. 

Auto Build

I can understand that giving rise to a point of view that the AS 2005 Wizards were over-engineered.  Fine.  I can therefore see streamlining the wizards.  A little bit.

But why take away all that good, helpful functionality?  Has thin-client computing led to thin Wizard computing?  Frankly, I think this is absurd.  Under the guise of having the wizards build leaner cubes (and dimensions) in fewer steps, we end up with a product that simply does less.  It doesn’t affect me personally. I know how to use the Analysis Services designers and the Properties window to do on my own what the Wizards would do for me.  But beginners don’t know how to do that.  And aren’t Wizards for beginners?

#    |
 Friday, June 13, 2008

You Always Have Other Options?

With yesterday’s announcement that Yahoo and Microsoft have broken off all discussion of a merger (partial or otherwise) and that Yahoo has closed an agreement with Google to carry the latter’s paid search ads on the former’s search engine, it certainly seems like MicroHoo has bitten the dust.  Perhaps both companies have decided that the expansion of Yahoo’s fabled acronym (referenced in the title of this post) is the best philosophy.  I still believe there’s a lot of posturing going on, and I still believe a deal is necessary.  I also think a deal is possible, but I don’t think it’s guaranteed.

I once owned my own firm and I remember how strongly I felt that I would never sell it.  My own pride was wrapped up in the company and its success, and I took any subjugation of the company’s identity to equate to that of my own.  I suppose Jerry Yang may feel likewise about Yahoo.  But, with all due respect (whatever that means), that kind of internalized, emotional approach to business is more appropriate to running a candy store than to being the CEO of a top-tier Internet company.  I sold my firm in 2004 so that I, and my employees, could move on to bigger projects and new technologies.  I suppose the analogy's a weak one, but why won’t Jerry Yang do the same for his team?

Jerry may get over himself or Carl Icahn may succeed in ousting him and force Yahoo to pursue a Microsoft merger deal with vigor (instead of with passive aggression).  But the fact that so much time has elapsed since the beginning of this round of deal-making truly complicates things.  Whatever Microsoft may have felt was a fair price for Yahoo several weeks ago, the reality is that the passage of time and events have changed Yahoo’s intrinsic value and its value to Microsoft specifically.  Consider:

  • The Yahoo brand is now tarnished.  And a shareholder rebellion led by Icahn will tarnish it further.  This would ostensibly have impact on the reach of Yahoo’s network of sites and the enthusiasm and satisfaction of its customers while surfing it, which could impact click-through rates.
  • Microsoft has already lost valuable time and momentum pursuing this deal.  With each passing day, the potential boost to its Internet advertising prowess that a Yahoo acquisition could provide diminishes, and the lost potential revenue piles up.
  • The caché for Microsoft on the Web that would have come from a swiftly and amicably executed deal with Yahoo is now beyond reach.

All of the above are Yahoo’s losses, first and foremost, but they hurt Microsoft too.  There’s no other deal that Microsoft can do that will bring them the audience share, and the demographic diversity of it, that Yahoo can offer.  And whether you believe that Microsoft should go into the Web advertising business or not, it’s important to realize that Microsoft is absolutely convinced that it must do so.  Under these circumstances, it becomes important that it do so as shrewdly as possible. 

At this point, doing a deal depends on (a) Carl Icahn’s maneuvers, and their success, (b) Microsoft’s willingness to accept acquiring a distressed property, rather than a trophy prize and, quite possibly, (c) Yahoo willing to accept a significantly lower price than the one it has already turned down.  That’s a lot of intrigue and compromise, and it’s far from certain that either company knows how to endure each or both, let alone do so quickly.  There’s a good business school case study in here somewhere.  It remains to be seen if it will be a study of success or failure.

#    |
 Friday, June 06, 2008

An Unexpected Lunch

On Tuesday morning (June 3rd), at Microsoft’s Tech*Ed Developer conference, Bill Gates made his last scheduled public presentation as a full-time Microsoft employee.  About 90 minutes later, 15 tech “influencers,” myself included, gathered around a square table in a small room at the north end of the convention center, and were joined by Mr. Gates for lunch.

We were invited to this luncheon just a few weeks ago, and very unexpectedly.  A couple of weeks later, we were told that we’d each have the chance to ask “BillG” (as he is known on Microsoft’s email system) a question.  So I prepared one or two clever questions in advance.  I had planned to ask Mr. Gates about the future of the BASIC programming language…given its historical importance to Microsoft’s business, and considering its apparent demise of late, at least in the world of .NET development.  I also had a non-technical question prepared: whether Gates has ever considered running for public office.  I expected the latter question would be ill-suited for a luncheon at which 15 top evangelizing geeks were in attendance, but I was hoping the conversation might swing that way.

To my surprise (and, from what I can tell, the surprise of everyone there), the questions asked of Bill at lunch were almost entirely focused on education, policy and issues concerning the world’s poor.  The questions from our group were all quite astute, and I, for one, enjoyed very much being able to talk about matters beyond the keyboard.  Throwing out my prepared questions, I instead asked Gates of his views on the United Nations.  Gates answered, and I am paraphrasing, that the United Nations is important, necessary and broken.  In other words, if I may infer, it needs reform, and it necessarily suffers from bureaucratic inefficiencies, but it’s an irreplaceable institution, and that trashing it (or de-funding it) is naive and ill-advised.  That’s a pretty nuanced answer compared to the more polarized positions expressed by many politicians.  It bucked a trend.  There was more of that to come.

From others’ questions, I learned that Bill Gates is extremely focused on issues of secondary and higher education.  He is quite critical of teachers unions, as he believes they institutionalize a system where under-performing teachers become ensconced in the educational system; in effect, he believes the union discourages any notion of competition or real meritocracy.  Perhaps that sounds reactionary…but I have to tell you that in the context of his discussion, even to my own liberal sensibilities, it seemed, unexpectedly, quite reasonable.  Gates also feels that school systems controlled by mayors (including in my own home town of New York City) are more more investment-worthy than are schools controlled by school boards (which Gates feels lead to contract negotiations between teachers and teachers).  What is really hard to convey here is that Gates states these things more as empirical observations than partisan positions.  And as he does so, he eliminates the component of controversy that you might expect (and I would have expected) in such proclamations.

And the surprises didn’t end there.  I learned that Gates feels Junior/Community Colleges and Associates Degree programs are essential to helping develop the labor pool (including skilled labor) in this country.  I also learned, courtesy of a question from Stephen Forte, that Gates is somewhat critical of micro-financing and bullish on “micro savings.”  He feels micro-financed loans often have unreasonably high interest rates and can therefore be ineffective.  Meanwhile, the lack of access by the world’s poor to simple retail banking services requires them to do things like invest their money in livestock (often resulting in losses of up to 30% when they need to liquidate the investment) or in jewelry (which has inefficiencies of its own), or to posses the cash and suffer inflationary depreciation, and risk of theft. This can lead to problems more grave than those solved by micro-finance loans.  And between the use of small retail locations for deposits, ATMs and cell phones (for checking balances, and so forth), even remote, poor areas could have access to banking services.

There’s one thing about a person having financial and political independence: he can avoid politically correct rhetoric with impunity.  For Bill Gates, one example of this was his reference to poor countries, not as the “third world” or the “developing world” but as the “poor world.”  Trust me here: this phrase was not one he used in condescension.  Rather, I think Gates likes to use directly descriptive names for things, because it saves time, and gets right to the point.  It is also, I found, more respectful of the problems faced by the poor, and more respectful of the need to help them.  And to Gates, at least as I read him, helping more people emerge from poverty isn’t just about charity, ethics, or responsibility.  It’s about an opportunity, not just for them, but for the people who will employ them.  Lifting people out of indigence can actually mean economic benefit for the rich.  That was inspiring, because it made the problem of poverty seem approachable to me. Another unexpected outcome of the conversation.

Gates’ keynote at Tech*Ed was, unfortunately, anti-climactic: there was little new announced and the presentations lacked punch and enthusiasm.  Lunch, on the other hand, was inspired.  And inspiring.  It’s pretty clear then where Gates’ real passion now lies.  And with him (and his foundation’s money), focused full-time on a pragmatic, mostly a-political approach to solving some of the world’s worst problems, I am hopeful that he can be an agent for real and lasting improvements.  For Gates (to borrow a phrase attributed usually to the Mormons), solving world problems is not just about doing good, but also about doing well.  Each of these motivations is quite powerful; together they could be transformational, if not unexpected.

 

#    |
 Tuesday, May 20, 2008

NYC Franchise for FiOS TV?

This is a big deal folks.  Look what I just read in my inbox (it actually arrived yesterday):

Tomorrow the New York City Franchise and Review Commission (FCRC) will hold a public hearing regarding Verizon’s entrance into the New York City Cable Television market. On April 29, 2008 Deputy Mayor Lieber and Commissioner Cosgrave of the Department of Information and Technologies (DoITT) announced an agreement with Verizon to offer cable service throughout the City of New York. Currently, an overwhelming majority of New York City residents have only a single choice in cable television providers. If approved, the agreement would require Verizon to offer cable service to all residences in New York City, potentially establishing a competitive marketplace in an industry that has been dominated by single providers and a lack of competition since its creation. We encourage you to attend tomorrow’s public hearing taking place at NYC College of Technology, 285 Jay Street, Brooklyn, NY. The hearing begins at 3:00pm and is anticipated to run until at least 6:00pm. As with all public hearings, feel free to attend and testify regarding your concerns.

I’m no fan of Verizon the company, but I think the FiOS product is a good one, and the rigor of a full cable franchise agreement would hold Verizon to the same standard as today’s cable providers in NYC (Time Warner, RCN and Cablevision).  This would force Verizon to serve all NYC neighborhoods, not just the more lucrative ones, and would ostensibly also compel them to provide public access programming carriage.

I’d love to have the better HD picture and faster broadband speeds that FiOS provides.  Perhaps more importantly, I’d love to see Time Warner Cable match those offerings, thus allowing me to stay with them.

I imagine a full city-wide franchise will anger Time Warner and Cablevision, who now serve mutually exclusive territory (Cablevision in the Bronx and Southern/Eastern Brooklyn; Time Warner everywhere else) because the previous franchise areas and awards prevented them from wiring the whole city.  And they’d have a valid point.  Let’s see what happens.  Maybe they’ll each get the opportunity to serve a city-wide area as well.  More competition.  Something the cable industry needs.

#    |