Matevž Gačnik's Weblog

There were a couple of sunsets lying around on Rožnik yesterday.

Thought I would make them available, because they make quite a desktop wallpaper.

Download: Here [Nikon D200, Nikkor VR 18-200 DX]

I just finished presenting my talk on the October 2006 Slovenian MSDN event.

Here are the deliverables:

Code: Download
PPT: Download [Slovenian]

Solution file includes:

1. WCF service using wsDualHttpBinding for duplex session communication scenario. WCF service hosts a workflow which computes stock tax based on the current (complex) Slovenian legislation.
2. WF workflow named ComputeTaxWorkflow. Workflow calls out to our public ASP .NET WebServices: StockQuotes and ExchangeRates.
3. WCF Client, which sends requests and receives responses when the workflow is done

Check it out. It's free.

Second incarnation of the .NET 3.0 Middleware Technologies day went through today.

Here are the deliverables:

Code: Download
PPT: Download [Slovenian]

Frequent answer still stands. This is the best playground.

Since there is so much interest, we are discussing a third repeat of the series. Anyone who wanted to attend, but did not get a seat, stay tuned.

We had a nice discussion throughout the day today, together with around 50 attendees. It's hard to even cover feature changes of .NET 3.0 in a day, but we managed to cover the important things of WCF and WF, spending a complete day inside the Visual Studio.

Here are the demos: Download
And the PPT slides: Download [Slovenian]

And the link to the Tom Archer's compatibility matrix for the .NET Framework 3.0 downloads.

Thanks to everyone who attended.

Microsoft has issued a statement that all (jointly) developed web services specifications have an "open use" policy attached to them.

Legal talk:

"Microsoft irrevocably promises not to assert any Microsoft Necessary Claims against you for making, using, selling, offering for sale, importing or distributing any implementation to the extent it conforms to a Covered Specification (“Covered Implementation”), subject to the following. This is a personal promise directly from Microsoft to you, and you acknowledge as a condition of benefiting from it that no Microsoft rights are received from suppliers, distributors, or otherwise in connection with this promise. If you file, maintain or voluntarily participate in a patent infringement lawsuit against a Microsoft implementation of such Covered Specification, then this personal promise does not apply with respect to any Covered Implementation of the same Covered Specification made or used by you. To clarify, “Microsoft Necessary Claims” are those claims of Microsoft-owned or Microsoft-controlled patents that are necessary to implement only the required portions of the Covered Specification that are described in detail and not merely referenced in such Specification. “Covered Specifications” are listed below."

The promise is given to the specifications themselves and, of course, not to implementations.

GoodMove^tm.

This post has been cooking for quite some time, quietly sitting on my desktop. Since Miha started the debate, I'm letting it go...

I've had a pleasure to work with the Acer Ferrari 4005 machine for a while. It was a great machine: AMD Turion 2.0 GHz, 64 bit, 2 GB RAM workhorse. Until I left it on the roof of my car and drove off...

Since then, I've been hammering on IBM Lenovo ThinkPad T60p, same specs, although x86 architecture. This is, in all terms, a great machine.

Having said that, I was running Windows XP x64 SP2 + Windows Vista x64, and Ferrari is actually one of the best machines to be on, when running x64. They have flawless driver support.

Let me get straight to the point.

Current prevailing architecture is x86. It's not going to stay that way for long. In the beginning of next year 99% of machines sold will have x64 support. Core 2 Duo is going to sweep the x86's dusty history.

The problem is, the majority of consumer base will decide by comparison, as always. It's just the magic of numbers, again. Imagine all the talking going on inside different computer stores and online forums, speculating how much better x64 is. In reality, x64 is currently (and for at least a couple of years) not going to be substantially faster - in the consumer space - than x86.

Nevertheless, a lot of people, who will now own the x64 chip, will want to run a x64-based edition of the OS. And here the problem lies.

Consumer Windows drivers have not been known for their robustness in the x86 world. There are devices that have real trouble running on Windows XP x86. Even though Vista will require signed x64 drivers, their availability is subject to questioning.

So the situation is this:

• You get the latest and greatest hardware, including a Core 2 Duo
• You get the latest and greatest software, including Windows Vista x64
• There are numerous well known problems with running apps in WoW, on x64 machines
• Currently, general device support is, well, flawed
• The drivers that exist have not been tested - for the consumer market.

Enterprise x64 market is quite different. There are a lot of production systems running Windows Server x64 successfully.

People are going to be pissed. It's Vista x64 and it is not going to launch successfully to the customer base.

Great tool was released today by the XML Team (Webdata), from Microsoft.

Find it here: http://www.microsoft.com/downloads/details.aspx?FamilyID=72D6AA49-787D-4118-BA5F-4F30FE913628&displaylang=en

It's a .NET Framework 2.0 application which can be used as a simple raw XML editor. It's got XSL support, XML differentiation, XML Schema validation, entity name intellisense, and, as the name suggests, it's as simple as notepad.exe. Superb performance on large documents, too.

Great. Tune it up, change the icons and layout then ship it with Vista, I say.

I find it quite attractive, since nowadays I don't spend as much time looking at angle brackets anymore.

There was a sincere and simple remark on Carl Franklin's blog, posted yesterday by Jonathan Parker, a CS student from down under.

If one recaps what he wanted to push through, it was this:

Microsoft's general public perception sucks.

He writes:

Earlier this year I had the privelage of attending a lecture from a Google employee and also one from a Microsoft employee. They were both open to all CS students and staff. The Google lecture was packed! My estimate is over 300 people (I sat on the floor most of the time). The Microsoft lecture had about 15 people. :(

He is right. Microsoft is trying to keep up with the coolness factor, all to no avail. Microsoft is not cool. Apple is cool. Google is cool. Myspace is cool. Microsoft should stop trying to be cool, this charade should stop. They should be doing what they do best in the world - write platform software. Zune, anyone?

In Jonathan's analysis on why Google is nowadays more popular than Microsoft, he writes:

I think the whole Web 2.0 thing is in Google's favour because they know how to run a business that is free for the consumers of their software.

Also I think that Google's software is much more discoverable than Microsoft's. If you want to learn more about Google you just go to their site and click on more. If you want to find out more about Microsoft you just go to their site and click your way around in circles for hours finding nothing.

True. Google's site is so much simpler, more intuitive and easier to navigate. But then, they only ship like a dozen products. Microsoft ships dozen squared.

His thoughts continue:

Open source is good for end user software. That's why Google is seen to be on the open source side of the fence even though Google is not open source in any sense of the word. All their code is tucked nicely away behind their servers. Yet people associate google with freedom of information. This is because there is not much use in installing a google search engine on your PC at home.

A general misconception is that open source equals free. Open source is far from free, especially in the enterprise space. And, as Jonathan states - I hope - Google managed to inject itself into this line of thought in general public.

There are two things that are wrong with this (general public, that is) way of thinking:

• Google/Apple is not open source
  Google took a lot from open source, but returned little. Read carefully - they returned little to Open Source, but gave a mountain to the public and the internet as a whole. Remember? Google is a good company.
  
  There is a difference between Apple and Google in this sense. Cupertino also took a lot from Open Source, but didn't contribute to it, nor release anything free to the public.
  
• Google is not percieved as a good company because of it being open source - its public perception is good, because everything they do is free
  If anybody does something really useful and keeps doing it for free, you just can't criticize it, right? Right.

If only there would be a simple formula to get Microsoft's public perception to even being average. I hate to say it, but, returning it back to good, is not possible. They were not there yet.

For the record: I do not percieve Microsoft as a bad, nor cool company. Microsoft makes the best fat client software in the world, period. They have the best development platform for any software, full stop. You can't fool the market over 20 years.

This has bothered me for two weeks. There was no way to install .NET Fx 1.1 on a Vista 5456 box.

Here's the solution: http://blogs.msdn.com/astebner/archive/2006/07/06/658484.aspx

After the permission fix, Framework 1.1 and SP1 for 1.1 install flawlessly.

Thanks Aaron Stebner!

There is Background Intelligent Transfer Service (BITS) present in every Windows XP/2003/Vista setup. BITS manages Windows Updates downloads, but is also capable of transfering other files.

More on BITS can be found here.

Since the infrastructure is there, I wrote a lightweight application, which manages the user queue of the BITS service.

Here are some screenshots:

Main features:

• Download files in the background
• Fire and forget
• Handles dropped connections
• Handles system downtime
• Handles bandwidth usage
• Can start download from IE (IE right click integration)
• Can autolaunch at system boot
• System tray notifications
• Harmless, small footprint
• Windows Vista support

You can download the installer or a ZIP version. If you grab the ZIP, you should change the installation path inside the .html file for the IE integration to work properly.

Download (Version 1.1.0):

• Installer
• ZIP

Requirements:

• .NET Framework 2.0

BITS Download Manager will quitely sit in your system tray and wait for you to give it something to download. When you initiate the download, it will progress in the background only if there is enough bandwidth available.

I use it to download large files over HTTP, being from my own server of those damn Windows SDK 1GB downloads which seem to break every now and then.

Update: Version 1.0.2 available [2006-07-04]

Minor bugs fixed regarding appropriate single instancing when launching a download from IE. Context menus fixed when no downloads are in progress. You do not need to uninstall version 1.0.0 before installing this one.

Update: Version 1.1.0 available [2007-02-20]

Minor bugs fixes, Windows Vista support.

A long running rumor has been made public a couple of days ago. WinFX name is dead and is superseded by Microsoft .NET Framework 3.0.

A part of me is grateful for being able to have a common name for the shipping technologies. Who wants to be acknowledged as a WinFX developer, when in reality she/he is a .NET developer? The decision to bundle WinFX technologies with Windows Vista and making it available down-level is awesome. Having it included with the .NET Framework is even better.

But there's another poor decision being made by the marketing folks at Microsoft. If this decision wouldn't be rushed, it would have made a wonderful release.

Microsoft .NET Framework 3.0 is ambiguous as a platform name. Since the press release states this explicitly, one can not complain as loudly as the reality demands. .NET Framework 3.0 is the first Microsoft .NET platform release that isn't really a platform - it's a marketing release. There's no CLR 3.0, no LINQ stuff, which was supposed to be in 3.0 release, no ASP .NET 3.0, no ADO .NET 3.0. In reality, there is current CLR 2.0 shipping, together with all the known friends: ADO .NET 2.0, ASP .NET 2.0. It even gets installed in the c:\windows\microsoft.net\framework\v2.0.50727 folder. At least, it's got its own folder inside the c:\windows\microsoft.net\framework directory. It is the 2.0 of the framework including WCF, WF and WPF. Goodness?

There exists a patronizing line of thought that nobody cares about the naming/install coherence and version matching. This is mostly heard from Microsoft, but the development community does not swallow it. The issue does not surface as this, actually.

Putting it mildly, this isn't serious. Suddenly separating CLR and framework version is a hack. A hack done to limit the damage that has already been done by establishing the WinFX bundle name. I'm happy that the LINQ team has not made a 'product name' breakthrough because suddenly we would run out of framework version numbers just to bundle the thing.

How is it possible, even in marketing terms, that .NET Framework 3.0 includes .NET Framework 2.0? What happens when LINQ gets shipped? Do we get CLR 3.0 and .NET Framework 3.5? .NET Framework 4.0? Do they get synchronized again and CLR skips versions? How do you distinguish the framework version? We will not be able to believe the compiler version, CLR version or BCL assembly version. They will all be 2.0.50727, but the framework is versioned as 3.0.

Even corflags.exe will still report a 2.5 CLR Header version, which is a mess even in current, shipping technology stack. It all follows the 'when the shit hits the fan' direction. Let's make it splatter, let's make even more mess.

Where is the compiler version change, runtime version change and metadata format version change? Where's the synchronization?

This is not serious.

Update: 2006-11-08, .NET Fx 3.0 is installed in a separate directory inside c:\windows\microsoft.net\framework.

If you happen to run Windows XP SP2 on a machine with more than a GB of RAM, you may likely see the following notification appear on the system tray:

This happens after hibernation is attempted. The error is: "Insufficient system resources exist to complete the API."

The solution is Q909095. There is a hotfix available, but you have to call Microsoft PSS to get it download the patch. It includes a new OS kernel which works flawlessly on my ThinkPad T60p with 2GB.

Knock knock. 

[Update 10/17/2006, Download available]

Having dropped all my current articles, this post is intended to sum it all up.

XML and XML Schema territory:

• Type Systems Compared - XML <> CLR (XML, XML Schema, CLR) [Preview, Download]
• The Importance of XML Typization (XML, XML Schema) [Preview, Download]
• XML Namespaces and PSVI Problems (XML, XML Schema) [Preview, Download]
• XML Schema - Specification Primer (I/II) (XML, XML Schema) [Preview, Download]
• XML Schema - Specification Primer (II/II) (XML, XML Schema) [Preview, Download]

WinFX territory:

• Concepts and Semantics of Service Contracts (WCF) [Preview, Download]
• Transactional Semantics in Loosely Coupled Distributed Systems (WCF) [Preview, Download]
• Cooperation Between Workflows and Services (WCF, ASMX, WF) [Preview, Download]

Article language: Slovenian

Last article discusses service-workflow cooperation options in WinFX and dives into communication scenarios for Windows Workflow Foundation.

Language: Slovenian

Naslov:

Sodelovanje storitev in delovnih tokov

Last article finished my XML series. This one focuses on transactional semantics in a service oriented universe.

Language: Slovenian

Naslov:

Transakcijska semantika v šibko sklopljenih, porazdeljenih sistemih

Next article in XML series is discussing XML Schema. This is a two part article.

Language: Slovenian

Naslov:

XML Schema (1/2)

XML Schema (2/2)

Next article in XML series is discussing XML Namespaces and PSVI problems.

Language: Slovenian

Naslov:

Imenski prostori XML in problemi v PSVI

This article is starting the XML series. First we dive into XML typization importance and XML Infoset.

Language: Slovenian

Naslov:

Pomembnost tipizacije XML

There's a ton of hot press stories going around today about Adobe not letting Microsoft implement the PDF export functionality in 2007 Office System.

While I can understand that Adobe is running scared by even the thought of everyone producing PDFs without a licensed Adobe Acrobat, I cannot understand the differentiation of source.

It's no secret that Apple uses PDF functionality in the base OS and its applications. You can export everything as a PDF in MacOS X. Hell, even the UI engine uses PDF technology to render those beautiful OS X windows.

See?

And this is not the only case. There are nearly 2000 products out there which can use PDF and this is a GoodThing^tm for a public specification.

WordPerfect Office has already implemented the same functionality Adobe is complaining to Microsoft about.

OpenOffice has already implemented the same functionality Adobe is complaining to Microsoft about.

While I am among the precious few, who condone IP ambivalence, this is not an IP issue, nor is it a licensing issue. Adobe can and will, as it seems, use their right to stop Microsoft in implementing their specification. And it is their specification, with all the rights the owner gets.

It seems that 'Save as...' functionality is indeed too much for Adobe to swallow when a 98% office productivity market share player gets real about it.

The second article is about concepts and semantics of service contracts. It deals with WCF (Windows Communication Foundation) contract definition and its behavioral aspects.

Language: Slovenian

Naslov:

Koncepti in semantike storitvenih pogodb