Most of the time

A more rational discussion of OS choice

Dino Chiesa at Microsoft sent around this blog entry from Paul Murphy at ZDNet. Here's an excerpt:

I got involved in an email exchange recently with the director of a government office in Ottawa, whose IT troops maintain multiple racks of IBM Xeon servers running Red Hat Linux. In Canabucks (currently about $0.86 US) he's paying Red Hat about $120,000 a year in licensing, has contracted client side support from an outsourcer, and has five of his IT staff committed full time to babysitting the applications, mostly trivial (from a complexity and load viewpoint), running on these machines. Now the outsourcing contract is coming up for renewal in the next budget year, the hardware is almost three years old and overdue for replacement, his IT head is insistent that going back to Windows Servers will reduce costs and downtime, and he's feeling real pressure from above to fund a study on switching to Linux on an IBM mainframe.

So what should he do? I told him to go all Windows. Why? Because the social and political support he needs to run Linux isn't there, his people are clueless, and the vendors - including his own IT staff - have their own agendas.

Obviously, the sentiments in the blog are music to the ears of the folks at Microsoft (me included), but it did reinforce a fatigue I've felt over the endless debates over Linux and Windows and, indeed, the long-term Linux paranoia at Microsoft (that has now been replaced by paranoia over Google. Google has now teamed up with Sun, which suggests Google has stopped focusing on its users and started fixating on Microsoft -- that should slow them down).

Getting back to operating systems: for most customers, Linux is just Unix from Redhat or some other distributor. Customers will end up paying more or less what they pay for Windows, and if they want support from the supplier they won't take advantage of the open source aspects. That's pretty much what Paul's blog concludes. I think there's a pretty good case to be made that Windows will cost less to operate than Linux, and is as or more secure, but what the hell -- we're not talking about massive differences.

The real distinction for Linux is the open source aspect, and that matters to a small but important group: companies building Software as a Service applications that they intend to host themselves. I've blogged on this recently, so I won't go through the points, but the net is that Microsoft needs to deliver a Windows offering that offers these companies the control and flexibility they get with LAMP (LInux, Apache, MySQL, PHP).

I belioeve the best SaaS offerng will have a scalable database (SQL blows Oracle or DB2 away on value there) and a rich and varied client set (PDAs, phones, laptops, tablets, TVs etc). That's where Microsft delivers the most obvious value to SaaS. Perhaps we should give away Windows Server to that community, also provide them with special support, and get paid for the areas in which (for that community) we provide distinctive value?

The industry fights back: the Secure IT Alliance

Announced today, a community response to security. From the press release ...

... tackling the security challenge demands more than just unilateral development of technology. Microsoft is also focused on providing prescriptive guidance to customers to help provide timely information, tools and training. Making progress on the security challenge also requires partnership: working with law enforcement and policymakers collaboratively across the industry. The SecureIT Alliance will enable security ISVs such as VeriSign, Trend Micro, Symantec and the 30 founding members (see Gathering in the Security Space, below) to work closer with Microsoft and each other in order to more effectively and efficiently build and integrate their products for the Microsoft platform.

There are 29 companies already in the alliance, including McAfee, Verisign, and Symantec. Notably there are also *eight venture-backed start-up companies, five of which were sponsored by Don Canning and Sam Ramji of the Microsoft Emerging Business Team:

• e-Security
• Forum Systems
• Network Intelligence Corp
• Voltage Security
• Vormetric

This underpins yesterday's article from Always On, about how Microsoft invests in startups:

"At the core, we are a software platform  company. We do well when others build on our platform. Successful startups spur disruptive innovation and create unique end customer value propositions. We see our role as guiding entrepreneurs and venture capitalists about where our platform assets can be best applied, and helping them plug into our partner ecosystem—it’s all about leveraging our tools and platforms, and  our marketing channels."

[*minor edit: I added the number of companies who are venture-funded startups]

Check out ISV Connect ...

ISV Connect is a portal to Microsoft's Developer and Platform Evangelism strategy. Theer you can find technologies and resources for ISV partners, architects, IT pros and developers. ISV Connect provides firsthand knowledge from Microsoft executives, key Microsoft program and product managers as well as other industry influencers innovating on the Microsoft platform

   

Kari Martin runs ISV connect and also blogs about it.

Advice to Google ... from my scoutmaster

From Reuters yesterday ...

Sun and Google, a leading Web search company, said they would jointly promote Java Desktop and Open Office, Sun's operating system and office productivity software systems.

"Working with Google will make our technologies available more broadly, increase options for user, lower barriers, and expand participation worldwide," Scott McNealy, chief executive of Sun, said at a joint news conference in Silicon Valley.

The comment from Scott McNealy made me want to mention to Google something my old scoutmaster told me: "don't get too close to a drowning man, unless you plan to save him."

Newsgator acquires NetNewsWire

More news from Newsgator. These guys are on a roll, with:

• a great web-based service for consumers
• a Mobile Edition, for PDAs
• FeedDemon Desktop client
• FeedStation for Podcasts and media
• an email Edition, for Eudora, Outlook Express, etc
• an Outlook Edition,
• a server for enterprises who want to do their own aggregation

and now ...

• a newly-acquired Mac OS X client -- NetNewsWire. See Brad Feld's thoughts on this (Brad is at Mobius, a principal investor in Newsgator).

I think, when we look back at all these Web 2.0 predictions, they are going to look a little unbalanced. Newsgator's approach will stand out as rational: a great web service, complemented by products for businesses, and a rich and broad client set. Google, too, is on this path with its recently-announced support for Open Office  :-)

Web 2.0, and another "Tale of Two Posts"

Post 1: Tim O'Reilly answers his own question: What is Web 2.0?

Post 2: in response, and more cryptically, Rick Segal observes Web 2.0 != a check

Some observers saw these as opposing views. Here are a couple:

Web 2.0 --  Don't Believe the Hype (yet)

Web 2.0 is just hot air

Both posts are excellent. Sure, Tim focuses on newness, while Rick observes that:

Having said that, enough already with this Web 2.0 nonsense. We are doing the same thing we always do when “new” has “newer” come along. We hype the snot out of it and crap all over the ‘old stuff’.

...

I’m not going to Web 2.0 conferences and am going to take a very dim view of start ups who think a Web 2.0 badge is a waiver for actually doing the fundamentals of a business, those being: Solving a problem with a solution customers love and will pay for. Don’t look for the buzz words to get you into the game or get you a check.

All that said, I saw lots in Tim's long post to think about:

The value of key databases, including identity

The race is on to own certain classes of core data: location, identity, calendaring of public events, product identifiers and namespaces. In many cases, where there is significant cost to create the data, there may be an opportunity for an Intel Inside style play, with a single source for the data. In others, the winner will be the company that first reaches critical mass via user aggregation, and turns that aggregated data into a system service.

For example, in the area of identity, PayPal, Amazon's 1-click, and the millions of users of communications systems, may all be legitimate contenders to build a network-wide identity database. (In this regard, Google's recent attempt to use cell phone numbers as an identifier for Gmail accounts may be a step towards embracing and extending the phone system.) Meanwhile, startups like Sxip are exploring the potential of federated identity, in quest of a kind of "distributed 1-click" that will provide a seamless Web 2.0 identity subsystem.

Microsoft (more precisely, MSN) made a clumsy run at this in 2001, with Passport and the ill-fated Hailstorm. It seems to me that one premise of Hailstorm/Passport -- a managed identity and authentication service -- is crying out for a provider. I think ADFS (see this post) will go a long way to build out this vision technically and pragmatically, maybe in partnership with Verisign, Bank of America, USPS, ....

Tim also -- of course -- discusses Software as a Service (SaaS), along with a warning to Microsoft:

one of the defining characteristics of internet era software is that it is delivered as a service, not as a product. This fact leads to a number of fundamental changes in the business model of such a company ...

...

While Microsoft has demonstrated enormous ability to learn from and ultimately best its competition, there's no question that this time, the competition will require Microsoft (and by extension, every other existing software company) to become a deeply different kind of company. Native Web 2.0 companies enjoy a natural advantage, as they don't have old patterns (and corresponding business models and revenue sources) to shed.

One point we at Microsoft have to deal with is that if you are planning to host your own service (like Salesforce.com) then Open Source may have a number of advantages over Windows:

1)       Licensing cost … zero

2)       Licensing simplicity … compared to various Windows licensing options which distinguish betwem authenticated/anonymous users

3)       Ability to customize and tune the environment -- for performance etc … while Windows is clearly tunable, scriptable, and so on (and may in fact have a lower TCO, even be more secure than, say, Linux) the person who deploys an open source technology may place more value on control/discretion.

4)       Ability to fix bugs yourself or get them from the community … rather than depend on MSFT to agree it’s a bug, ship a fix, etc

Microsoft has some advantages (beyond TCO and security) in this narrow situation:

1)       If you plan to offer an enterprise-hosted server product, then some of the disadvantages listed above become advantages (eg. Enterprises may not want to rely in the community for support, and may prefer Microsoft)

2)       If you plan a Windowsibased client (standalone, plug-in to Outlook, etc) then Windows may bea better server (built-in authentication, etc)

3)       As your server scales, then MySQL will run out of gas. SQL Server plus Windows is a lot less expensive than Oracle plus Linux.

4)       .... there must (will) be more :-)

As Rick Segal put it, we at Microsoft have to deliver "... a solution customers love and will pay for."

Now we're talking ...

OK, sort of off-topic, but this is intriguing ... from Dow Jones VentureWire Alert today (subscription required, but free) ...

Ageia, which picked up $27 million in Series C financing, is looking to bring its flagship PhysX processor to the gaming market with the capital. The processor, which promises more realistic gaming graphics with by using physics calculations, is targeted at the PC industry, while its software development kit can work on both consoles and PCs. Providing a greater depth of physical interactivity than software-only physics, the chip powers the calculation of real-time interactive environments. The technology enables game developers to create graphics that obey the laws of classical physics, or create their own physical laws for more dynamic and creative gaming experiences.

Maybe now we'll see Snow Crash become (virtual) reality, or maybe Down and Out in the Magic Kingdom

The end of the beginning: SaaS starts walking upright

Don't you just love it when another hype cycle settles down? Is SaaS evolving from pure web-based access to client-server? Is it evolving from purely hosted environments to include enterprise server deployments? Will we finally stop hearing:

• "we are implementing everything on the web with thin clients. No desktop deployments!"
• "It's all On Demand, baby, no servers required!"

It appears so, and the evolution appears inevitable, driven by three forces:

1. Revenue, because users matter, and browser access penalizes power users
2. Lower training costs and higher adoption rates when using familar tools
3. Legacy Integration and customization

I think we'll all be better off for acknowledging this change, but let's start with a definition of SaaS. The Wikipedia lists the key characteristics of Software as a Service (SaaS) citing IDC:

• Network-based access to, and management of, commercially available (e.g., not custom) software
• Activities that are managed from central locations rather than at each customer's site, enabling customers to access applications remotely via the Web
• Application delivery that typically is closer to a one-to-many model than to a one-to-one model, including architecture, pricing, partnering, and management characteristics

The client-server evolution

The advantage that Saas companies claim (and deliver)is low or zero cost deployment via instant web access. The wikipedia definition also names some SaaS companies:

24SevenOffice
NetSuite
Salesforce.com
Siebel CRM OnDemand

What's interesting about the short list of examples is that the two big  players -- Salesforce.com and Siebel -- have both added client features:

Salesforce.com Office and Outlook Edition With Office Edition and Outlook Edition, workers no longer need to toggle back and forth between Salesforce and Microsoft's popular productivity and email tools. Our tightly integrated and easy-to-use Microsoft solutions allow users to work more productively in the critical applications that they use most.

CRM On-Demand Outlook Integration. Siebel CRM OnDemand allows users to associate inbound and outbound emails—including attachments—with contacts, leads, accounts, opportunities and service requests. Capture new email addresses, search for addressee information, and edit and create new records in Siebel CRM OnDemand with a single click -all while working directly within Outlook.

User Matter

The availability of processing power and storage is rising so much faster than bandwidth that it stands to reason that both data and function will inevitably move to the device. Local processing and data access lead logically to improvements in usability.

Usability has value, and Saas companies adopt the rich client because customers will pay for it. Look, for example at Newsgator, which provides a range of priced plug-ins for Outlook, and AtHoc which provides a number of client-side options (including a nifty toolbar).

At first glance you'd ask, though: doesn't the client element contradict zero-cost deployment? Of course, it does, but it's a pretty good tradeoff. The good news about browser access is that it's widely available. The bad news is that for any kind of power user interaction (the kind you do in a CRM system for example) the user experience is awful. As Sam Ramji says in this post: "... at the end of the day, we use browsers because we have to - not because we want to."

The value of familiar tools

A story from home: at Microsoft we spent millions of dollars customizing  Siebel for our sales teams and we still faced very poor adoption rates and low user satisfaction with the tool. Part of the problem was that we also had a number of other independent tools for our sales teams: billing, critical situation management, and so on. So, we put a single integreted web interface on top of web services that integrated Siebel and the other apps. Usage and satisfaction increased a bit .. but not much.

Then we decided to fully integrate Siebel and the other apps into Outlook. Essentially , we replaced the web interface with Outlook.  The familiar folder metaphor and offline support made the big difference and adoption rose dramatically. In use it's not dissimilar to Newsgator's Outlook plug-in for it's server-based aggregator.

The enterprise server evolution

While a hosted service may be exactly right for trial and rapid adoption, there are a number of good reasons customers want local versions: control (the NSA for example), legacy system integration, customization, ... Legacy system integration because a hoster cannot integrate -- economically at least -- with the huge range of line of businessand other applications which comprise enterprise IT shops.

Enterprise deployment also offers a different cost (and profit) model. A hosted environment's capital and operation costs will increase with the user base, while customers bear most of those costs in an enterprise deployment. These aren't the only factors in determining profitability and growth (suuport cots and development costs may have different curves too) but having both offers the ISV faster adoption, and flexibility over time as customer requirements change and new opportunities appera.

So, many Saas companies -- while they start out with a purely hosted service -- later add an enterprise Server. Here are a few examples from the collaboration family: Groove, AtHoc, Adesso, Newsgator, ...

I think we are starting to see, at last, the real value of SaaS (rapid trial and adoption; and broad accessibilty of applications) balanced with the usability and flexibility of rich clients and enterprise deployments. To borrow a quotation from Winston Churchill: Now this is not the end (of the SaaS hype cycle); it is not even the beginning of the end; but it is perhaps the end of the beginning.

Four legs good; two legs bad -- the WSA Panel today

Following up from an earlier post ... today at the WSA Investment Forum 2005, I sat on a panel that discussed: Platform Wars: A Civilized Conversation on the Best Solution

Moderator: Frank Catalano, Senior Vice President, Marketing, Pearson Assessments & Testing

Panelists:

- Stuart Cohen, CEO, Open Source Development Labs

- Cliff Reeves, General Manager, Microsoft .NET Platform Strategy, Microsoft Corporation

- Bill Vass, Chief Information Officer and Vice President, Information Technology, Sun Microsystems, Inc.

The panel was pretty tame -- in my experience they usually are --  but there was a little sloganeering here and there. with items like: "use open source software to avoid vendor lock-in," and "the Chinese government won't use proprietary software because they can't see the source."

Frank Catalano, the moderator, asked good questions and kept the discussion moving along.

Here's the list of questions, and a highly personal view of the highlights

- How would you define "open standards" from your perspective? ... I got to start and made a reasoned case that open implied "you can be confident that they are stable and long-lived," and that standards meant "widely available and widely supported." That enfranchises everything from IETF, through WS-I and de facto standards from Intel, Microsoft, others.
- Areas where hype has overtaken reality ... Stuart Cohen made the point that in many cases, open software (like Linux) was part of a commercial endeavour, and that there was a lot more proprietary software running on open software, than there actually is open software. He observed, a few times throughout, that open source does not mean "non-commercial."
- Sweet (and bad) spots for each platform ... Stuart noted that highly mobile platforms (like PDAs and laptops) would not yield easily or soon to open source platforms like Linux because of the deep function and additional applications that market demands. He was very bullish about Linux in the server space, though, and in some devices.

For my part, I suggested that an ISV building a hosted service might not see a Windows stack as optimal because of a need to be self-sufficient for tuning and service. Also, a licensed (even inexpensive) OS might be suboptimal in a low marging high-volume appliance or device. On the other hand, a platform like Windows is ideal in mass markets where add-on appkications are valued, and in critical business envrionmenst where strong support/service was necessary. Bll Vass didn't identify any areas where open source systems are not ideal :-)
- Realities of interoperability or homogeneity .. the panel pretty much agreed on heterogeneity, for a number of reasons. There are many reasons open source exists, not the least of which is that it's the obvious way to share technology that is funded by public money -- like government research.
- Possibility of current platforms being blindsided by something new ... no real insights here, although I called out Web 2.0 as something Microsoft hasn't yet come to grips with.

At one point, Bill Vass mentioned that Solaris was now Open Source. I suggested that was perhaps because it no longer had any value. This may have been mean of me.

Any way, fun panel, nice guys. Super job, Frank.

Watching the elephants dance

An interesting leader in the Economist this week drew heavily on eBay's proposed acquistion of Skype : How the internet killed the phone business.

"Watching the elephants dance

As is always the case with a disruptive technology, the incumbents it threatens are dividing into those who are trying to block the new technology in the hope that it will simply go away, and those who are moving to embrace it even though it undermines their existing businesses. Since VOIP will cause revenue from voice calls to wither away, the most vulnerable operators are those that are most dependent on such revenue.

In particular, that means mobile operators, which have been struggling for years to get their subscribers to spend more on data services, but are still hugely dependent on voice. Worse, the very “third generation” (3G) networks that are supposed to provide future growth for these firms could now undermine them, because such networks make mobile VOIP possible too. Least vulnerable, by contrast, are those fixed-line operators that are now building new networks based on internet technology, which will enable such firms to benefit from the greater efficiency and lower cost of VOIP compared with traditional telephony.

These operators are taking an “if you can't beat 'em, join 'em” approach and getting into the VOIP business. While their voice revenues will slowly evaporate, they will then be well placed to offer fee-based add-on services over their new networks. Again, this is a common pattern with disruptive technologies: forward-looking incumbents can end up giving upstart innovators a run for their money."

Now is the time to look at the companies who'll give  the fixed line operators a boots in this fight. Makes companies like Cicero Networks look even more attractive than I thought.