rwandering.net
The blogged wandering of Robert W. Anderson
May 23, 2006 at 7:11 am · Filed under .NET, Grid Computing
With WinHEC 2006 going on, this is a big news week for Microsoft. I blogged about FlexGo yesterday, but they also unveiled their virtualization road map (press release here). Their roadmap contains three main points:
- A commitment to release new Windows Server virtualization that is supported deep in the operating system. This is called hypervisor technology and will RTM 180 days after the release of Windows Server Longhorn. This is the most interesting part of the road map and something that I have been looking forward to for some time. The possible benefits of this to grid computing are enormous: the more the OS supports virtualization the easier it will become to instantiate virtualized sandboxes for secure and unintrusive distributed computing. Of course, my hope is that this technology will make it down into some Windows Vista SKU.
- The Microsoft Center Virtual Machine Manager. A way to manage the virtual machines from a centralized location. Not too much to say about this. Makes sense.
- The intent to acquire Softricity. Their SoftGrid software is a desktop provisioning product that eases deployment issues for large enterprises. This makes a lot of sense for Microsoft to be able to better support enterprise customers continuing to reduce the total cost of ownership for the Microsoft platform. It seems that there is potential synergy between the FlexGo provisioning server system and the Softricity SoftGrid software virtualization — though it doesn’t look like Softricity is currently part of the FlexGo technology.
Side note: Softricity’s product name embodies one of the reasons we at Digipede avoided the term grid for so long . . . there product looks cool, but I wouldn’t call it a grid.
May 22, 2006 at 8:26 am · Filed under Grid Computing
FlexGo is a new “pay-as-you-go” service from Microsoft. From their site:
Introducing Microsoft® FlexGo™ – an innovative technology that makes it easier for people with modest incomes in emerging markets to buy a full-featured PC for their families.
Nick Carr calls this a “full-fledged utility computing service for PCs”. While technically correct, calling this kind of service utility computing further muddles an already troubled term — rendering it useless.
Update: apparently Mr. Carr decided this isn’t really utility computing and has changed his post to read “a utility-like computing service.” Good. I feared he had gone around the bend.
Tags: FlexGo, grid, Microsoft, Utility-Computing, WinHEC
May 22, 2006 at 8:02 am · Filed under .NET, Miscellaneous
It is refreshing that Scheme (the variant of Lisp) is getting so much attention right now.
Check out www.schemers.org for Scheme resources (thanks to Steve Maine for the pointer).
It was used as a teaching language at Cal. Scheme was, without a doubt, the only mind-expanding programming language I’ve ever had the pleasure to ingest learn and use.
Tags: Programming, scheme, Software-Development
May 17, 2006 at 8:28 am · Filed under Miscellaneous
From Don Dodge: Microsoft has unveiled the Microsoft SaaS Lighthouse Program.
This program is designed to help SaaS startups with licensing, technical, and marketing assistance.
I have been talking about this hole in Microsoft Partner Programs for sometime — a little on my blog, but mostly in person to Robert Scoble, Sam Ramji, John deVadoss, and many others. Pretty much every Microsoft person I meet. Microsoft has a good program (called Empower) to help the ISV startup and has, up until now, really had nothing for the SaaS startup. Why is the SaaS startup important? Because Web 2.0 companies are essentially all SaaS startups. Why are Web 2.0 companies important? If you don’t know, well . . . stop reading here.
It is obvious that Microsoft is trailing in the Web 2.0 community — this is embodied in the negative perception of Microsoft I found at a TechCrunch party: “anti-Microsoft” is a buzz word?. Most Web 2.0 startups are making use of free tools to build their sites and view Microsoft’s products (primarily Server 2003 and SQL Server) as being too expensive. While the productivity gains attained using these tools trumps licensing costs, Microsoft still needs programs that will help the adoption of their technology in SaaS startups.
Unfortunately, this program isn’t it.
You see, to get into this program, your company must have venture funding. A few problems with this:
- A venture-backed company can afford the licenses and training. Period.
- A Web 2.0 startup doesn’t start with venture money. In fact, very few startups begin with venture. So, the Web 2.0 startup gets no help from Microsoft until after VC? Kind of obvious, I know, but the startup has probably already built a service and released it before they get serious interest from a VC.
- The venture clause is likely a way for Microsoft to get external validation that the startup isn’t wasting their time and money. OK, but that is no way to increase adoption.
Now, I cannot claim that the purpose of this program is to help Web 2.0 startups (or to increase adoption of Microsoft techologies).
Just to be clear: it doesn’t.
Tags: Microsoft, SaaS, Startup, Web2.0
May 11, 2006 at 7:22 am · Filed under .NET
C# 2 has anonymous delegates. I like them. Of course, they enable advanced constructs like closures and continuations, but one of I’m a pushover for language and environment features that improve readability and maintainability of code. Before C# 2, if you wanted to perform a task on another thread, you might do something like this:
/// Do something worker callback
private void DoSomethingImpl(object callerData) {
// implementation here
. . .
}
/// Do something on a threadpool thread
private void DoSomething() {
// invoke on threadpool thread
ThreadPool.QueueUserWorkItem(new WaitCallback(DemoJob));
}
With anonymous delegates, you can just do this:
/// Do something on a threadpool thread
private void DoSomething() {
// invoke on threadpool thread
ThreadPool.QueueUserWorkItem(delegate(object o) {
// implementation here
. . .
}) ;
}
Fewer lines of code. Easier to maintain. Logic easier to follow. The pattern gets even nicer when your worker method needs arguments. For example:
/// Do something on a threadpool thread
private void DoSomething(int a, double b, SomeClass c) {
// invoke on threadpool thread
ThreadPool.QueueUserWorkItem(delegate(object o) {
// implementation here can use a, b, and c
. . .
}) ;
}
To pass the three arguments a, b, and c to a named method would require a custom class or packaging in a collection or array. I won’t show you how to do it. It is tedious, after all.
Tags: .NET, c#
May 9, 2006 at 6:50 am · Filed under Web 2.0
Actually:
- I agree with Doc Searls: links are useful. I don’t think they are everything, and I bet Mr. Searls agrees with that too.
- If you haven’t tried to understand what Steve Gillmor is saying, I suggest you do.
- I believe a game-change is a foot. See #2.
- Notice that I find the arguments relevant and interesting and yet I still link. And I continue to use desktop Office.
- And Jim, not even Steve says you should stop linking. He still links when it is the right gesture.
Disagree all you like. But, if you haven’t taken the time to grok the actual argument beyond the low hanging fruit, then just move along. There is nothing to see here.
May 8, 2006 at 5:32 pm · Filed under Web 2.0
I’ve finally gotten around to listening to the discussion about links and the link economy between Steve Gillmor and Doc Searls (part I, II). Definitely worth a listen.
Good thing, too, since I stepped into an argument about this on Jim’s blog and found him taking issue and posting about it. Jim’s initial point was regarding a recent Seth Goldstein post, Attentrons on the cutting room floor, and specifically the statement:
Strong web bloggers no longer link.
Jim concludes that non-linkers are not blogging, that they are actually columnists, and says:
I [. . .] recognize that sometimes bloggers feel that it is no longer necessary to provide links to source information so that their readers can get context, further information, or view those that came up with the original idea credit.
He goes on to say (in response to me and, to some “elite”, of which I am not one):
The notion that links are merely to build readership is rather frightening to me. It says to me that conversation is important until you reach a point where you can pontificate. That blogging is a culture to grow TechnoPriests. What a drag that would be.
What I should have said is that as a new blogger, linking is required to get into the conversation. I do this not to build readership as I flipply suggested, but to be a part of a community — if readership was truly my goal, I would go on a snark hunt or something.
While there must be those who refrain from linking due to their own elitism and to withhold credit, suggesting this is the motive behind the whole “non-linking” crowd ignores the argument being made. To be fair, I don’t think Jim is ascribing these motives to all non-linkers; however, I cannot agree with the claim that not linking is not blogging. And if he turns out to be right (i.e., once the definition of blogging is finally complete), I suspect it won’t matter. Things are changing, blogs are evolving, tools are getting better. Coupling RSS, gestures, and discovery engines will bring more of the right information without relying on the sacred link.
In terms of the arguments against links, I boil this down to a few points:
- The link-model supports the page-view model. Steve is looking past this model.
- Decorating your text with links interrupts your writing (i.e., the reading). This isn’t about controlling the user, it is about composition.
- Linking is often excessive and can be insulting to your readers.
Steve Gillmor makes his points in his podcasts as well as on his blog — he doesn’t need me to make them for him. Read his latest, Back in the U.S.S.R, on this very topic.
May 6, 2006 at 4:47 pm · Filed under Grid Computing
The Digipede CEO, John T. Powers, has begun blogging at Powers Unfiltered. This will turn out to be an interesting blog. Aside from being really smart, John is funny and opinionated.
Our partner Nathan asks,
when do we get Powers filtered?
BTW: you may recognize his blog template . . . if any of you IE CSS wizards know how to fix the white gap between the header graphic and menu, let me know. John will buy you a drink. 😉
Update: Thanks to Shannon Whitley for the simple fix to the header. It still looks better in Firefox, but I’ll tweak it.
Tags: Digipede, grid
May 5, 2006 at 12:57 pm · Filed under .NET, Grid Computing
Just yesterday I compared the Digipede Network to Sun Grid. One point I made was that our product is very easy to use . . .
Of course, I’m biased, but: stay tuned for an upcoming independent review that contrasts the Digipede Network favorably against Sun Grid on this very point.
Little did I know that the review would be just a day away.
The review Gridlock Alert For Sun Grid? goes further than I thought. It slams Sun Grid and speaks well of our product. You can see it all in the article’s subheading:
Analysis finds hidden costs in Sun platform, while competing Digipede offering shines
The article also specifically supports my earlier point. The reviewing engineers gave up after spending five hours trying to get something to work on Sun Grid. On the Digipede Network, it took approximately 1/2 hour. In the reviewer’s words:
The difference in ease of use in Sun Grid vs. Digipede was enormous.
Thanks for the great review!
CRN has a more comprehensive review of the Digipede Network here: Grid Computing Turns .Net Into Enterprise Powerhouse.
Tags: .NET, Digipede, grid, SunGrid
May 4, 2006 at 10:05 am · Filed under .NET, Grid Computing
Jon Udell, lead analyst at InfoWorld, recently posted a screencast of the Sun Grid Compute Utility (Screening Room #4: The Sun grid compute utility).
I enjoy reading Mr. Udell on his blog and in InfoWorld as well as listening to him on the Gillmor Gang. He brings together both the broad view of an industry analyst with the deep dive on the technical only achieved by a developer. In fact, he is a developer (for example, he recently deployed metadata searching and exploring services for InfoWorld).
In the Sun Grid Utility screencast, Mr. Udell does a good job of quickly showing the major functionality of the Utility and the Compute Server project (i.e., Java APIs for communicating with the Utility). The screencast is not too different from the webcasts we do for the Digipede Network, our distributed computing solution for the Microsoft .NET platform. The quality of Mr. Udell’s questions in his screencast are on par with the best questions we get in our webcasts — he clearly groks distributed computing.
Sun Grid appears simple but is deceptively hard to use. Simplicity was a primary reason we developed the Digipede Network: we don’t think distributed computing needs to be complicated. I think we have been pretty successful at that. Of course, I’m biased, but: stay tuned for an upcoming independent review that contrasts the Digipede Network favorably against Sun Grid on this very point.
The Digipede Network is a Windows-focused product that shares some similarities to Sun Grid (specifically Sun Grid Engine). Here are some differences between the Digipede Network and Sun Grid:
- We do .NET 1.1 and 2.0, as well as command-line and other stand-alone applications.
- We have full support for COM on both the front-end and on the distributed program.
- .NET and COM developers can serialize assemblies, data and objects into messages do be distributed for execution.
- Our SDK and community site includes samples in many languages (C#, Python, PHP, VB6, VB Script). We plan to put together a Java example, though honestly the demand hasn’t been that great.
- The SDK integrates with VS2003 and VS2005, supplying developers with samples, a developer guide, API documentations, and XML schemas for Intellisense.
- We have a Workbench that allows you to design your jobs without writing any code.
- Our Agents are smart about data and applications. They can cache persistent applications and move data from and to file servers and web sites (in addition to the messaging described above).
- Applications can be configured externally.
- Jobs can be defined through XML or programmatically.
- We have a much richer user experience for control and monitoring of jobs through an ASP.NET Web site.
- You don’t need Solaris 😉
If you want to see this stuff in action, sign up for one of our webinars here or go to MSDN. They recently hosted two webinars on the Digipede Network: Object-Oriented Programming for Grid: Grid Computing for .NET and Advanced .NET Programming for Grid Computing.
And Mr. Udell, we would love to do a screencast with you, too.
Tags: .NET, Digipede, grid, SunGrid, Udell
« Previous entries ·
Next entries »
About me
Main feed
Email subscriptions