February 28, 2003
@ 08:07 PM

I've seen the future and it will be.

InfoPath may become for XML, what the Basic language was and is for binary stuff. 

Thinking that technical knowledge about all the odd stuff around angle brackets and infosets alone will remain to be a sufficient foundation for a career or even a whole business may be as fatal as thinking in the 1980's that your x86 assembly skills will be enough to support your family for the next 30 years. It works for a few, it won't work for most. Ladies and Gents, it seems like XML is growing up and is getting all dressed up for prime time -- get over it.


February 27, 2003
@ 02:17 PM

My dinner with Don.

I had a fantastic dinner with Don and Steve Swartz (Steve should start a blog so I can finally link to him and so that he can share his infinite wisdom; folks, you don't know what you're missing). Great Indian food and surprisingly good beer -- and we Germans have pretty high standards.

We were talking lots about (multidimensional!) XML, Web services moving forward and on the way home in the car, we've discussed a sufficient set of requirements for the metadata you need to bind to an XML Web service that is reliable, secure (and all the other WS-I goo). Result: You need a way to describe the message body and you need a way to describe the QoS and security features. (Mildly) annotated XSD covers the first, WS-Policy covers the second. Where does that leave WSDL?


February 22, 2003
@ 04:42 AM

Sun FUD against IBM: Now, here's an interesting quote from InfoWorld on Sun's upcoming Web services stack: "When you buy it from Sun, you'll be able to buy it all at one price, install it all [for] one price, and have blueprints and guidance to install it yourself," Bauhaus [Sun vice president of Java Web Services] said. This is unlike IBM's Web services products, which require purchasing integration services from IBM Global Services, Bauhaus contended.

100% pure comedy (too bad that "comedy" doesn't start with a "J") 


February 20, 2003
@ 06:13 AM
BusyBusyBusy (I am). For those with more time and some curiosity for things that mattered (mostly) in the past: check out Microsoft's newly released docs on NDR. ;)

February 9, 2003
@ 05:38 AM
Airport WLAN (no, not Apple). Greetings from Dubai airport. They've got a wireless LAN here; one more airport to add to the "we like that very much" list. Good connectivity and a freebie - at least in the lounge.

A picture named AOPinExcel.JPG

Ingo seems very happy about me announcing this, because he's been waiting impatiently for me to do that since we two had a little "backroom demo" session at a conference last November.

Status report: The GPF problem that had been bugging me (and the code) for three weeks has been solved and what I before suspected to be a GC bug turned out to be something else. All aspects are properly triggered and, more importantly, exceptions are properly relayed on the way back for all relevant scenarios in managed-to-managed (inproc, outproc, outproc/interop for SecureMethod).

Testing is underway for unmanaged-to-managed. The latter is actually the most significant difference to whatever can be done with the managed context. These attributes are actually working for unmanaged (read: COM) clients as can be seen in Excel's VBA stopping on an aspect-fired exception in this screenshot.

Still, stability is still an issue under some circumstances and there seem to be a couple of memory leaks deep down in the guts.

(By the way, Ingo asks over there, why I have a stateful variable in the code he's quoting. Answer, because I don't set the completion flag, state is held. SCs aren't uncontrollably stateless. Ah, and, if that variable was public, you could even put an aspect on it and I'll intercept that call as well.)


February 8, 2003
@ 01:36 AM

Crazy Travel Times. I am going to be traveling almost every weekday and some weekends until June (!). There are a few weekdays sprinkled across my schedule that I'll take off at home, but all in all, my schedule is entirely packed. There's even stretches of two or three weeks where I'll be home or near home only long enough to switch suitcases.

But not only for me, but also for Patricia, who's working at her cast iron engineering family business (GVA Krefeld) travel time starts tomorrow. They're in a serious hardware business. (The "slag pot" on the right is one of their products. Everything less than 10 metric tons of cast steel seems to be peanuts for them.)

She's going on once-around-the-world trip to South and North America, Oceania and SE Asia. And, of course, our trips in the next 3-4 months will overlap in a way that sometimes one of us will be just gone when the other comes back. Still, we shouldn't be complaining, it's still very exciting to travel!


February 7, 2003
@ 10:33 PM
Christian Weyer blogs. RSS subscribed.

February 5, 2003
@ 09:44 PM

Kartoo. I'm probably the last one to find out about Kartoo, but it's really cool.[Andres Aguiar's Weblog]

A link to Kartoo has been in my navigation bar (this page, on the left) ever since this blog exists ;-)  Also ever since then there has been a link to Corona Production's "Upcoming Attractions" site. Want to know what the current rumors are about Star Wars 3? Go here.


February 5, 2003
@ 05:25 PM

AOP ServicedComponents!

For those attending my talks at the Microsoft Strategic Architect's Forum in South Africa next week (that's an "by invite only" event), I am very happy to announce that I am going to be demonstrating, for the first time, a stable version of my brand-new AOP extensions for Enterprise Services. It's an extensible framework that allows writing custom, managed attributes, which are invoked from within the server-side context and allow interception and inspection at the class, interface, method, property, field and argument level. 

The example below illustrates the functionality using a set of validation attributes, but there's significantly more you could do. I will have to do quite a bit of documentation work before I can make this available, but there will be downloadable bits for everyone to play with within the next couple of weeks.

   public class Component : newtelligence.EnterpriseServices.AspectServicedComponent
      private string val = "";

      public Component()

      public string SetMeGetMe
         get { return val; }
         set { val = value; }

      public int TestField;

       [return: MinLength(1)]
      public string Hello( [MinLength(1),MaxLength(20)] string Name,
                          [Between(1,100)] int Age,
                          [MinLength(1)] out string retString,
                          [LaterThanToday] DateTime when)
           return "Test";


February 5, 2003
@ 04:45 AM
Classpath hell as seen from inside the great ball of fire.

J2EE 1.4 Waits for Web Services.


Sun Microsystems, which controls the widely used Java standard, said Tuesday that it will push out the delivery of the Java 2 Enterprise Edition (J2EE) 1.4 specification until this summer. The J2EE 1.4 specification, which gives Java licensees the blueprint for building Java programming tools and server software, was set to debut in the current quarter.

The forthcoming J2EE specification incorporates Web services protocols, a set of standards and a programming method for connecting disparate computing systems. Adoption of Web services is accelerating as companies look for ways to lower the cost of sharing information.

Sun representatives said the company chose to push back the finalized J2EE 1.4 specification in order to comply with interoperability guidelines set forth by the Web Services Interoperability Organization (WS-I).

[Matt Croydon::postneo]

And in three months they are expecting the profile to be stable and final? J2EE 1.3 may be the end of the road in terms of an agreed-upon all-vendor standard. With the ongoing delays of these specs, economic pressure on the middleware providers, and the volatility of all the "standards" we're currently dealing with... is anyone really waiting for this? Wouldn't J2EE 1.4 only be a snapshot that's valid only on the exact day the spec hits the street? The way and intensity in which the WS-* arena works is apparently very incompatible with how the JCP works.


February 5, 2003
@ 12:26 AM

newtelliblog. Seems like a lot of people are leaving Radio behind are are using other blogging tools. While I like Radio for what it does, it shouldn't take much longer and I'll be leaving as well.

I've got the core SQL store for a weblog centric "content system" almost feature complete; an extensible aggregation NT service that can read from RSS, RDF and Exchange web stores and which spits out the aggregated content via a pub/sub interface is already running; the rendering engine ("downwards compatible" with Radio design templates) is working and sits on top of ASP.NET and spits out RSS 2.0 feeds; the client tool looks a bit like MSN messenger and allows remote management via Web services as well as offline postings and already has the all-essential tiny HTTP server listening on port 5335 to catch clicks on Radio coffee-mugs.

And of course, it's all built on "dogfood". The Enterprise Services utilities, some of my Web services extensions, attribute-driven state management, JITA pooling and all that. So that noone can ever say we're not using our own stuff ;)


A picture named shuttle-breakup.jpg A picture named shuttle-breakup.jpg 

Space Shuttle Columbia  breaks up on descent after atmosphere reentry

Terrible, terrible news. (check Google News)

Transcribing from TV news: All 7 astronauts aboard space shuttle Columbia (OV-102) [mirrored here] on mission STS-107 were killed when the shuttle apparently broke up on descent after earth atmosphere reentry. Contact was lost at 1500CET, 0800CST over Texas, scheduled landing time was 0916EST at Kennedy Space Center. The space shuttle is presumed to have disintegrated at 200000 feet and at 12500mph, some 100 miles south-east of Dallas. People in Texas are advised to report any findings of debris to local authorities and not to go near them due to the toxic propellant used in the shuttle.

The videos seem to indicate the the descent was mostly nominal until the shuttle suddenly breaks up into two and shortly afterwards into multiple pieces after a large flash as shown on the second picture above.

NASA statement:

Entry Flight Director Leroy Cain declared a contingency for the shuttle Columbia at around 8:14 central time this morning (1414 GMT) as the shuttle and its seven astronauts headed for a landing at the Kennedy Space Center.

Columbia fired its braking rockets at 7:16 this morning (1316 GMT) and entered the Earth’s atmosphere with all of its systems functioning normally for a landing at the Florida spaceport at 8:16 a.m.(1416 GMT).

But communications were lost with Columbia around 8 o‘clock (1400 GMT) as the orbiter streaked over Texas.

NASA began to use all of its tracking facilities to look for Columbia, but communications were not restored by the time the shuttle had been scheduled to land.

Contingency procedures remain in effect and landing support officials are currently being dispatched near the Dallas-Forth Worth area to search for possible debris.

Witnesses in Palestine, TX report loud explosion sounds that shook their buildings. [16:30CET] Police in Nacogdoches reports finding debris.

A picture named palestinetx1.gif


STS-107 mission information at NASA: Crew, Timeline, Cargo, Press Kit (11MB PDF)
Better reachable: European Space Agency STS-107 information.

A picture named nasacrew.jpgA picture named nasacrew.jpg
Commander Rick Husband
Pilot William McCool
Mission Specialist Kalpana Chawla
Mission Specialist David Brown
Mission Specialist Michael Anderson
Mission Specialist Laurel Clark
Payload Specialist Ilan Ramon

[16:17 CET] Meanwhile, http://www.spaceflight.nasa.gov/shuttle is swamped and virtually unreachable


February 1, 2003
@ 04:15 PM
Huh! Someone really likes my utility set for Enterprise Services and correctly points out that they are in "preview" state, but very useful. The good news is: there's no timebomb. And, yes, there's going to be an update once Windows Server 2003 is "official". And, yes, there's going to be an option to get at the source code for all of it. These classes and a lot more code that only few folks have seen yet will be at the center of my developer-topics talks this year.