BizTalk Blog by Chris Han

Poka-yoke (mistake-proofing) is a technique for mistake-proofing in process design. It's trading the flexibility for less variance. There is a similar concept in software called defensive design or defensive programming.

Yesterday, I just experienced a good example of how important defensive design is. I was trying to plug my external hard-drive to my laptop via USB port. As you know the USB port actually has some sort of defensive design - the connectors in the metal enclosure is actually located only at the lower half. It supposes you can not plug the male connector into it at all if flips it upside down. The problem I see is the implementation of this defensive design on my laptop. The plastic connector is not strong enough to proof a determined idiot like myself. The consequence is serious – the laptop’s power unit is damaged. I guess I have to change the mother board or rather throw away the whole laptop.

Am I too 'Paka' to be a good user? If you use DFMEA method to calculate a Risk Priority Number (RPN) in this case, I'd say hell no. It gets a pretty high score on my calculation:

Severity: 10
Occurrence: 2
Detect and prevent: 5
RPN = S*O*D = 100 

Although it's highly unlikely happens, it is sure one of the failure modes both hard to detect and cause fatal damage. I like the IEEE 1394 (Firewire) socket better at this account. Like at it shape - reduced occurrence; and it's metal - hard to push through. 

 PS. I just realized that USB has a Mini version. It's much closer to 1394's design idea. It's better.

I was working a project schedule with MS project. I started by estimate the duration in 'months' using PERT, and later I decided to switch to 'days'. The I run into the duration format convertion trouble. I fund the set the 'Duration is entered in' drop down list in option menu didn't really do the job. See the full solution:

geekswithblogs.net/chrishan/archive/2008/07/18/123879.aspx

This is part of my school work at Steven Institute of Technology. The assignment is to come up with a definition of a system architecture by my own words. Here it is: 

A system architecture is the organization of a set of components forming up a system. It provides a plan that arranges the responsibilities of each internal component in the system and their interactions to each other as well as the interactions between the system and the external environment it resides and/or being operated in.

The last point about methodology is what I learned from my work experience, and inspired by the architecuture from civile engineering.

I haven't written any thing for a while. Part of the reason is because I'm now back to school - Stevens Institute of Technology.

As my career advancing, I'm more and more focusing on Enterprise Architecture and Business Process Management. Of course, as an EA, you'll never hand off on technologies. So as a Microsoft guy, I'm always curious about what Microsoft will offer to BPM.

I'm a little suprised when I talk to my professor Dr. Michael zur Muehlen (his blog: bpm-research.com). He sees Microsoft as just a vender of System Integration Solution. And it seems Gartner is agreeing with him. Check out this gartner BPM Magic Quadrant 2007(free with registered user). I also find that Sandy Kemsley had a chance to closely examine Microsoft's tech stack, and shared with on her blog column2.com.  

But I still remain curious on what are the commonly agreed criteria of a qualified BPM solution. I agree that BizTalk is not a complete business architecting to automatic execution solution, but can we at least say it's a workflow system than a simply EAI/ESB? I mean, it got BPEL engine, rule engine, BAM....

Is that because the lack of 'human-centric' features that BizTalk is not recognized as a BPM player? I'll find out, later.

It used to be a joke when there was some software design flaws caught at the test phase. It is amazing to see that the joke presents a truth by reading this new from MSNBC http://www.msnbc.msn.com/id/21112810/

If you put the hybrid car as an architect, one of the most important questions you should have asked yourself as an architect is 'how it will interact with the environment it resides?'

You thought the noise of the gasoline car is something unwanted? Think again. It is really depends on which stakeholder you are talking about. Blind people are in deed participating in the traffic - the architectural environment your system interacts with. To them, the noise indeed is a feature!

Now ask yourself, do you think vibration is a feature of the car or rather a technical constraint?

I'm running 64 bit Visa and trying to install my new phone toos software 4.5.1c. After get this error, I have to exchange the 4.5.1.c with a 4.5.  After install v4.5 and let it update itself to 4.5.1.c online, everything works.

Hope this will help ohters Moto owners.

I remember when the title 'architect' in IT became a buzz word around 2005, a fellow blogger comments that it is just a new way to say ‘I'm senior enough not to code any more’. It is a joke, a good one, if you don’t really understand what architecture is supposed to do with the system, and how you are supposed to make that happen.

Today, an even bigger title flowing around the job market – ‘Enterprise Architect’. What is that supposed to mean? A more senior-er position that you don’t have to code plus a c-level bonus?

Before you can call yourself an ‘Enterprise Architect’, you have to understand what an Enterprise Architecture is. The ‘architecture’ in system engineering perspective is a structure that defines its internal components and the relationship between the components, as well as the relationship to environment it resides. The ‘Enterprise Architecture’ comes from the theory that an organization (enterprise) is a system. Therefore it can be viewed as a structure that defines its internal components (functional departments or regional divisions) and the relationship between them (power and resource sharing, command chains), as well as relationship to its environment – political, environmental, competitive…

An ‘Enterprise Architect’ is a role who comes from engineering or science discipline with a good understanding of businesses and industries; who is responsible for over seeing the business as a system and design and govern it as such a system that can only gain its competitive edge by generating the output effectively and efficiently from its limited resources. It’s not a MBA, not an EMBA. It’s a system engineer at a larger, human-centered system.

Chris
http://www.stevens.edu/sse/

I got this error when testing my web service POC which is runing on .netFx 2.0:

System.Data.SqlClient.SqlException: Timeout expired.  The timeout period elapsed prior to completion of the operation or the server is not responding.

I find Angel's article gives the most accurate diagnosis. Check it out here:
http://blogs.msdn.com/angelsb/archive/2004/08/25/220333.aspx

But the solution he gives out doesn't solve the problem:
      public void DoesNotLeakConnections()

      {          

                  Using (SqlConnection sqlconnection1 = new SqlConnection("Server=.\\SQLEXPRESS ;Integrated security=sspi;connection timeout=5")) {

                              sqlconnection1.Open();

                              SqlCommand sqlcommand1 = sqlconnection1.CreateCommand();

                              sqlcommand1.CommandText = "raiserror ('This is a fake exception', 17,1)";

                              sqlcommand1.ExecuteNonQuery();  //this throws a SqlException every time it is called.

                              sqlconnection1.Close(); //Still never gets called.

                  } // Here sqlconnection1.Dispose is _guaranteed_

      }

What you should do which is the best practice for any resource related job is to close the resource in a Final block like this:

      public void DoesNotLeakConnections()  

      {          

                  Using (SqlConnection sqlconnection1 = new SqlConnection("Server=.\\SQLEXPRESS ;Integrated security=sspi;connection timeout=5")) {

try

{

                              sqlconnection1.Open();

                              SqlCommand sqlcommand1 = sqlconnection1.CreateCommand();

                              sqlcommand1.CommandText = "raiserror ('This is a fake exception', 17,1)";

                              sqlcommand1.ExecuteNonQuery();  //this throws a SqlException every time it is called.

}

catch  (Exception ex)

{

  //do something

}

finally

 

 

{

          sqlconnection1.Close(); //Always gets called.

}

                  } // Here sqlconnection1.Dispose is _guaranteed_

      }

My friend told me this morning, he's little confused with his UPS shipment tracking result.

We all know that UPS' using 18 bits tracking number to 'uniquely' identify a shipment. But after I just saw my friend's tracking result page, I'm not sure about the 'unique' part any more.

On the 'Package Progress' pane, UPS lists all the history of the shipment associated with the number:

See the two high-lighted lines? Apparently, the tracking number is reused by UPS! I don't what's the policy for reusing a tracking number at UPS. But the common sense would be unless you have used all other possible numbers, right? That means all 18 bits of numbers have been used in roughly half year (11/06/2006~06/25/2007)! That's 10^17*26= 2,600,000,000,000,000,000 shipments! Man, 2.6 trillion, what a business!

I guess it's a good news for sales department at UPS but headache for DBAs. Most our SQL server database are still using 'int' type as the unique id, which only supports 2^31-1 (2,147,483,647), even 'bigint' type which supporting up to 2^63-1 (9,223,372,036,854,775,807) - about 3.5 times of what UPS has used out in 6 months. It's about 1.7 yeas of OLTP data. After that, you have to move the data to a data warehouse! DBAs or database designer probably never think of such a big number in such a short time. Similar problem as Y2K, but this comes much quicker than any of us can ever imagine. Even Mr. Moore needs to adjust his law.

What a great time for all geeks like you who’s reading my blog. But be prepared! Love yo’ll.