·         .NET assemblies

·         Unified Modeling Language 2 (UML2)

·         Next generation identity

As we finalize the plan for our V1, we are trying to figure out what other schemas (and corresponding import/export tools) do we need to ship out-of-box for V1.  Our primary target customer for V1 is developer – in particular a developer who is an ISV/System Integrator, building database-based end-user solutions for people involved in software development lifecycle.   Our wish is that these developers will find it easy to leverage SQL Server Modeling Services as a platform to build metadata-oriented solutions for their customers (especially software developers and software architects). 

We will create some schemas, our internal Microsoft partners will create some and we expect our customers to create others.    One of our challenges is determining whether we take a breadth or depth approach to define metadata model for these artifacts.  So far, we have gone deep when modeling the CLR and UML domain. 

What top 5 model schemas would YOU like to see us ship as part of SQL Server Modeling Services V1?

Thinking broadly, we can foresee creating schemas for following type of assets (just a bucket list).

·         Conceptual Artifacts

o   Business Process Modeling Notation (BPMN)

o   Unified Modeling Language (UML)

o   Entity Relationship Models – Logical Data Model

·         Implementation Artifacts

o   .NET Assemblies (.dll/.exe)

o   Native Assemblies

o   Windows Communication Foundation (WCF)

o   Windows Workflow Foundation (WF)

o   WSDL/XSD

o   DAC Application Packages (.dacpac)

o   Web Deploy Application Packages (.zip)

o   Build Scripts (Build Configuration)

o   VS Project (*.*proj)

o   Relationship to Source Control Store / Team Foundation Server

o   Configuration Files (.config)

o   Windows Presentation Foundation (WPF)

o   ASMX Web Services

o   Windows Forms

o   ASP.NET

o   MSI Files

o   BizTalk Orchestrations (.odx)

o   Others (to be done by someone other than us)

§  Java Code (JAR)

§  Java Web Archives (WAR)

§  Eclipse Project

§  PHP

§  Ruby

o   …

·         Deployment, Configuration and Management Artifacts

o   Windows Server AppFabric

o   Windows IIS

o   AppFabric Azure

o   Microsoft Virtual Machine Manager – VHDs, etc.

o   System Center content – monitoring, configuration, health, operations

o   Machines, Data Center (CMDB)

o   …

If you are a developer what type of applications are you writing (Web, Rich Client, Midddle-Tier, Frameworks, Tools, Cloud, ….)? If we were to bring the metadata about your applications into an easy-to-query database, what metadata about those applications would be interesting to you for change and impact analysis?  Would you rather have a conceptual model that describes the application and not go deeper into individual technology implementations?

Thanks!

There are about ~20 Technical Fellows (all men) and ~48 Distinguished Engineers (all men) at Microsoft.  There are a handful of women who are in the executive ranks, including two Senior Vice Presidents.  Mary Jo Foley (author of "All About Microsoft") has started blogging a new series titled, Microsoft women worth watching, to profile 'techy' women who provide significant contribution to the company.  She starts the series off by profiling Julie Larson-Green, who is the Corporate Vice President for Program Management in Windows Client.  There is also a Microsoft Channel 9 series on Women in Microsoft and they also interviewed Julie in 2007. 

The percentage of women in product engineering groups (and especially key technical roles) varies from product to product.  The ratio also changes significantly across the higher levels in the company.  I wonder if the choice of product also determines the number of women in the technical role and their corresponding growth in that product area.  This article got me counting the number of women in our product group.  The SQL Server Modeling Services product engineering team, which is a small product team, has following ratios.

• The PM team ratio is 56% male and 44% female (which includes a Partner GPM Shoshanna Budzianowski and a PM lead)
• The SDE team ratio is 92% male and 8% female
• The SDET team ratio is 83% male and 17% female (which includes a test lead)

There isn’t a single woman architect in my product group. In fact, in my last 3 years at Microsoft, I have come across only one woman architect. 

Microsoft provides an environment where women are treated with respect.  We don't regularly even get to think about this gap, because we are so busy thinking and working on the product issues.  There are multiple reasons why there is such a divide, including availability and interest. 

There is a smaller pool of women out there to pick from (women participation in Computer Science has dropped significantly).  The situation is especially grim in the United States. Young people in many less-developed countries now outperform their American counterparts in both science and math.  I know this, because I, along with other volunteers, have been researching and collecting these statistics from various sources for a local non-profit organization called Cascades Science Center Foundation.  As part of this foundation, we are deeply interested in creating a physical environment locally for the children (boys and girls) to get interested in Science, Technology, Engineering and Mathematics (STEM).  Microsoft also has programs dedicated in this area – check this article titled STEM: A Foundation for the Future. Also, last year Microsoft launched a website as part of U.S. Computer Science Education Week to encourage youth to pursue careers in computer science.

Another factor in the decline for some women can be lack of interest.  I have experience working in both a technical role in high tech company (3 years) and a technology position in a company other than high-tech (11 years in financial industry).  I am a Computer Science graduate and I started out my career as a computer programmer.  I worked with many women developers and testers during my technology career in non-tech industry.  I was one of the few women, who was the technical lead, the IT architect and (for a small duration) the IT strategist.  I still remember attending meetings in IBM office at Poughkeepsie, NY where I was the only woman sent by my company amongst the total of 18 people (and yes, there is usually a line-up outside men’s washrooms at the high-tech conferences.J). During the course of my career, I saw many women developer and technical leads, either becoming generalist development managers or switching disciplines including Project Management and Business Analysis.  Despite the move away from core technical involvement, many are still working in the technology departments and didn’t move into the business groups.  For many of them, the reason for the switch was boredom and lack of interest. I personally believe (IMHO), women are wired to be ‘global’ and ‘big picture’ thinkers.  They want to connect many dots.  They want to connect to people (the whole IT-is-geeky argument).  They want to ‘see’ the impact of their work on others. Many times, one can get lost in mundane issues while working deep in one technical feature.  There are many other cool roles surrounding technology (Product Management, Technical Writing, UX Design, Release Management, etc.) which meet the needs and interests of certain women at different phases of their lives. 

Then there are always some other reasons! I From workplace ceiling to personal reasons. I have a very talented girlfriend, who used to work as a SDE at Microsoft.  She is super interested in technology.  She quit her job last year, after she gave birth to her second child.  She said that she found her job as a developer to be very stressful.  Now that her second child is over two years old, she is thinking of returning back to the workforce, but wants to switch the discipline, where stress level is low.  And then I have other friends, who find that working for a corporation (technology or otherwise) is not fulfilling enough for them and really yearn for working in a non-profit environment. 

Anyways, in general, Microsoft is hungry for talented technical women (and men)! If you fit the bill, do apply.  DMG is hiring (PM, Dev and Test positions are available)!

http://live.visitmix.com/videos

There are so many good sessions that I want to watch. I am especially interested in the following to begin with, purely because of personal interest.  (Unfortunately there is buffering issue which hopefully they will fix soon.)

• The Art, Technology and Science of Reading
• Total Experience Design
• The Laws of User Experience
• A Case Study: Rapid WordPress Design and Prototyping with Expression Web 3
• The Elephant in the Room
• An Hour With Bill Buxton

The session format was:

·         10 minute introduction (5 minute study intro and 5 minute product context)

·         1 hour and 30 minute of actual study

·         20 minute wrap up

Folks from product groups (PM, SDE, SDET) in the room observe the participants during the study, write the feedback on huge stickies and put them on the whiteboard.  We categorize the feedback during the session into high level categories and then we discuss the most relevant category in the post-study 20-minute wrap up.   I found this study format to be great as it was low overhead and makes it quite interactive.    We provided virtual machines with pre-loaded bits to the participants – all they had to do was bring their own laptops.  At recruitment, we had specifically asked for .NET developers with SQL experience for this study. 

In my 5 minute context (why) about SQL Server Modeling Services, I didn’t mention “M” or “Quadrant” at all.   Our usability study for System.Runtime domain had the following three primary evaluation goals:

1)      User comprehension of the System.Runtime database model (tables, views, table value functions)

2)      Usability of LoadAssembly.exe tool

3)      Usability of writing T-SQL queries over the model

The scenario, sample application and task list is available at: System.Runtime Model/Loader Usability Task List.

The sessions like these make us (folks from product groups) get out of our cocoon and see the ‘truth’. J 

I will do another post with analysis of the feedback once we have compiled it all, but here were my high level takeaways:

1)      Don’t make me write T-SQL queries, provide me a tool ( seems i am the only one who likes writing T-SQL queries ;-) )

2)      It is very difficult to navigate the structure of the schema in SQL Server Management Studio (SSMS)

3)      Provide an EDM / object model head over the schema so I can then write queries in code

It was apparent in this session, that unless our target customer is an ISV developer who is building metadata-aware tools on top of the SQL Server Modeling Services database, an enterprise developer or architect will not write T-SQL queries for analysis.   If our value proposition is ‘impact analysis over codebase’ then we need to provide a targeted tool for those users (and they need not worry about the database structure).  

What are your thoughts? Can you benefit by having your .NET applications metadata in a database?  How much metadata do you really need in there?  Are you interested in participating in our agile usability studies (i.e. if you are in the Seattle area)?

In this study, we will focus on the relational model and loader for .NET Assemblies.

The System.Runtime model in SQL Server Modeling Services contains extents, types and constraints that represent .NET Assembly metadata in a relational format.  The SQL Server Modeling CTP also contains a tool, LoadAssembly.exe, that can extract metadata from .NET assemblies and insert that metadata into the System.Runtime domain models in the SQL Server Modeling Services database. Once the metadata is loaded, the assembly metadata can be queried to obtain information about multiple applications across an entire enterprise.

Scenario

You are one of the developers working in a development group within an IT department in a financial company.  Your test team has just given you a sign-off on your .NET application which is now ready to be released to the customers.  As part of the IT asset management policy in your organization, besides releasing the application, you will also publish the application metadata for .NET assemblies into a shared SQL Server Modeling Services database (repository).  In this study, your objective is to publish .NET assemblies that make up your application to a given SSMoS database and query metadata about that application to obtain few analytics.

Your application is based on .NET Framework 4.0.  The high level architecture of your canonical application is shown below:

The attached Calculator.zip file includes the following file.  You can extract it in a directory in file system: “C:\SystemRuntime\SampleApp”.

·         Calculator/Calculator.exe

·         Calculator/CalculatorLibrary.dll

·         CalculatorService/CalculatorService.dll

·         Common/Utility.dll

·         Common/CalculatorLibrary.dll

Tasks

Task 1 – Load Calculator application

Use SQL Server Modeling Services LoadAssembly.exe tool for this exercise.

1.       Load Calculator.exe application into the folder “Lab/Calculator”.

Task 2 –Insight into the Calculator application using SSMoS

Use SQL Server Management Studio tool for this exercise. Can you write T-SQL queries for the following?

1.       Show list of assemblies in folder “Lab/Calculator”.   Hint: Use sproc [Repository.Item].PathsFolder.

2.       Count the number of public types in Calculator application. Hint: ‘Access = 6’

3.       Show all the type definitions and namespaces they belong to for Calculator application

4.       Show all the types being referenced by Calculator application.  Which assemblies do these types belong to? Result set in three columns “module name”, “type name”, “referenced assembly name”.

5.       Show all the assemblies that Calculator.exe depends upon

Task 3 – Load Calculator Service with its dependencies

Use SQL Server Modeling Services LoadAssembly.exe tool for this exercise. See if you can load the Calculator Service and its dependencies without explicitly specifying them in the command line.   Hint: LoadAssembly /?

1.       Load the Calculator Service with its known dependencies

Task 4 – Understand relationships across assemblies

Use SQL Server Management Studio tool to write T-SQL queries for this exercise.

1.       Show all assemblies that depend on CalculatorLibrary.dll

2.       Show all assemblies that depend on a type “ComplexNumber” in CalculatorLibrary.dll

3.       Show all callers (method name, assembly name) of method System.Console.WriteLine in the repository. How many times is System.Console.WriteLine being called?

Task 5 – Load all Calculator assemblies and add to binding context

1.       Load all assemblies related to Calculator application in a single load episode and add to binding context

2.       How would you get answers for following questions? You are welcome to write a T-SQL query or describe the flow to/from various tables. 

a.       The Calculator.exe depends on few other assemblies.  Which assembly references are already loaded in the repository?

b.      Show all types from assemblies that are calling any method defined in CalculatorLibrary.dll.

c.       Show all methods that CalculatorService.dll calls in other assemblies.  Note: Only show those method references, the definition for which is available in the database.

Calculator.zip

Context

Binding Context is a unique concept in SSMoS System.Runtime model and provides a resolution for the types and methods that are referenced between the assemblies AND loaded into the repository.  This allows you to navigate the references between assemblies faster than having to use a name-based approach for finding references.  

Let me walk you through a scenario:

·         FooAssembly.dll depends on BarAssembly.dll, because a type Foo in FooAssembly.dll references a type Bar in BarAssembly.dll

·         Both FooAssembly.dll and BarAssembly.dll have been loaded into the repository

How would you write this query – “Show all dependent assemblies for FooAssembly.dll that are also loaded into the repository”?  If you just wanted to find out the references, you can easily obtain that by querying the AssemblyReferences entity, but having a reference in that table doesn’t necessarily mean the metadata for that assembly is also available into the repository.    To try with some of our internal assemblies, I loaded Microsoft.M.LanguageServices.dll into the repository. 

The following T-SQL queries for all assembly references for Microsoft.M.LanguageServices that begin with “Microsoft.*”:

Select AR.* from [Repository].[System_Runtime].[AssemblyReferences] AS AR

INNER JOIN [Repository].[System_Runtime].[Assemblies] AS A ON A.Id = AR.Assembly

WHERE A.Name = 'Microsoft.M.LanguageServices' AND AR.Name LIKE 'Microsoft.%'

The result set looks like this on my machine:

But what this query doesn’t tell me yet is whether or not the dependent assemblies are also loaded into the repository.  To find that I need to look into Assemblies table for existence of Microsoft.M.dll assembly that matches the identity of assembly being referenced (i.e. Name + Version + PublicKeyToken + Culture).  Here is the revised query (this only matches based on the assembly name for now).         

Select AR.* from [Repository].[System_Runtime].[AssemblyReferences] AS AR

INNER JOIN [Repository].[System_Runtime].[Assemblies] AS A ON A.Id = AR.Assembly

INNER JOIN [Repository].[System_Runtime].[Assemblies] AS A2 ON ( A2.Name = AR.Name )

WHERE A.Name = 'Microsoft.M.LanguageServices' AND AR.Name LIKE 'Microsoft.%'

The result set of this query is only one row, because I only loaded Microsoft.M.dll into the database.

The Binding Context concept lets the system do the assembly name-matching for you and populate a bunch of tables which you can directly query to find answers.   The above query can be re-written like this using the BindingContext and yields the same result. 

SELECT AR.* FROM [Repository].[System_Runtime].[AssemblyReferences] AS AR

INNER JOIN [Repository].[System_Runtime].[Assemblies] AS A ON A.Id = AR.Assembly

WHERE AR.Id IN

   (SELECT BAR.SourceAssemblyReference FROM [Repository].[System_Runtime].[BoundAssemblyReferences] AS BAR)

Did you notice that we didn’t have to specify the name matching in above query? The system takes care of that.

Binding Context Overview

Binding Context makes it easier to write dependency queries to resolve type/method references to the type/method definition in the repository.  The following diagram shows the high level relationship between core entities related to the Binding Context.

n  BindingContexts

o   Holds individual binding context for a group of assemblies that make a particular application.  For example, you may want to create separate binding contexts for each individual tier in your application.  3^rd party  libraries may themselves be loaded into a binding context of their own.

n  BindingContextAssemblies

o   Holds links to each assemblies that are part of the binding context

n  BoundAssemblyReferences

o   Provides resolution between assembly references and assemblies

o   System populates this table when assemblies are inserted into BindingContextAssemblies table

n  BoundTypeReferences

o   Provides resolution between type references (TypeReferences) and type definitions (TypeDefinitions)

o   System populates this table when assemblies are inserted into BindingContextAssemblies table

n  BoundMethodReferences

o   Provides resolution between method references (MethodReferences) and method definitions (MethodDefinitions)

o   System populates this table when assemblies are inserted into BindingContextAssemblies table

Before the binding context can be used, you need to:

The insert of assemblies into the binding context invokes a trigger that resolves type and method references between assemblies and populate BoundAssemblyReferences, BoundTypeReferences and BoundMethodReferences entities.

Once the assemblies are loaded into the binding context, you can perform queries like these. 

For following queries, I have already oaded m.exe, Microsoft.M.dll and Microsoft.Oslo.Internal.dll into the repository and into a binding context 3. 

Show all methods that are bound to and being referenced in m.exe assembly

SELECT A.Name AS SourceAssembly, A2.Name AS ReferencedAssembly, M2.Name AS ReferencedMethod

FROM [Repository].System_Runtime.BoundMethodReferences AS BMR

INNER JOIN [Repository].System_Runtime.Methods AS M ON M.Id = BMR.SourceMethodReference

INNER JOIN [Repository].System_Runtime.TypeReferences AS TR ON TR.Id = M.DeclaringType

INNER JOIN [Repository].System_Runtime.AssemblyReferences AS AR ON TR.AssemblyReference = AR.Id

INNER JOIN [Repository].System_Runtime.Assemblies AS A ON AR.Assembly = A.Id

INNER JOIN [Repository].System_Runtime.Methods AS M2 ON M2.Id = BMR.TargetMethodDefinition

INNER JOIN [Repository].System_Runtime.Types AS T ON M2.DeclaringType = T.Id

INNER JOIN [Repository].System_Runtime.Modules AS MO ON T.Module = MO.Id

INNER JOIN [Repository].System_Runtime.Assemblies AS A2 ON MO.Assembly = A2.Id

WHERE BMR.Context = 3 AND A.Name = 'm'

Here is partial result for abovementioned query.  

Since this post has already gotten too long, I will write more binding context related queries in a future post.   Here is one query that I have been planning on writing and I think the binding context can be useful there.  

“Find all methods in your application that are not being called, across all the assemblies in the database (i.e. find all dead methods)”

Summary

Binding Context entities in the System.Runtime schema provide a quicker way to go from a reference to its definition in the repository.  What do you think of the Binding Context set of entities in the System.Runtime schema? Do you think they can be useful for your query scenarios? Does the name binding context resonate?   We appreciate your feedback!

Type definitions are made up of methods.  A method is a subroutine defined within a type that contains a sequence of statements to perform an action.  It is analogous to an “operation”.  It contains a set of input and output parameters and possibly a return value.   Some of the types of methods are: abstract method, constructors, destructors, accessor method and static method.   Methods can be overridden by another type to provide a specific implementation for the method.

The following ER diagram shows the relationship between different entities related to the definition of a CLR method. 

Method Definitions

The MethodDefinitions entity represents details for a method definition for a type in an assembly.  Let’s say you have a method “int add (int a, int b)” defined in Calculator type in Calculator assembly.   When the Calculator assembly is loaded into a SQL Server Modeling Services System.Runtime database, the MethodDefinitions table is populated with the add method details.  If a Calculator type was calling a method that’s defined in another assembly, that knowledge is stored in CalledMethods and MethodReferences tables.  Check out the MethodDefinitions.m and Methods.m for attribute details.

Note: You can further scope these queries by assemblies and folders.

Method Parameters

As the name suggests, the Parameters entity captures the parameters for the methods.  Check out Parameters.m, MethodDefinitions.m and Method.m for attribute details.

Note: You can further scope these queries by assemblies and folders.

Delegates and Events

Delegates and Events are how CLR implements the Observer design pattern.    The Observer pattern defines one-to-many dependency between objects so that when on object (publisher) changes state, all its dependents (subscribers) are notified and updated automatically.   This allows for loose coupling between the publishers and subscribers, so that the publishing code can change without breaking any of the subscriber code.  In .NET, the events are properties of the type publishing the event. 

To try these queries out, just load your .NET application's assemblies in the SQL Server Modeling Services database and happy analyzing!

Context

As a developer you may not be interested in leaving your favorite IDE to interact with another database.  Or if you are an IT manager, you may not be interested in using same tools as developers to perform business intelligence over the software application lifecycle data.  With SQL Server Modeling Services, we encourage ISVs to integrate with a SSMoS database from the tools targeted for the end users in software application lifecycle.  For example, as an ISV building a Model-Driven Development Toolset or a Business Process Management System, even if you want to leverage the common application lifecycle schemas provided my Microsoft for software assets, you may want to abstract the data access from your end users to give them a seamless experience.  ISVs can build application lifecycle tool specific plug-ins so their users interact with the database without having to leave the tool they love to use.  

Let’s take a hypothetical scenario in the context of a hypothetical Contoso Bank.  The IT development organization in Contoso Bank has a central architecture group that is responsible for defining common data and architectural models for the systems development projects in the enterprise.  This group uses IBM Rational Software Architect (RSA) to do their modeling.  In particular, many folks in this group like to use Unified Modeling Language (UML) as a software modeling language to design various aspects of a system.  At times, they also apply various specific UML Profiles (such as UML Profile for Business Process or UML Profile for Software Service) on the UML models, so certain tools can process these models to transform them to another implementation technology.    The software projects in Contoso use a mix of .NET and Java.   The java developers use IBM Rational Software Developer (RSD) and the .NET developers use Visual Studio (VS) to develop the software applications for Contoso’s business needs.   During the software analysis and design phase, the solution architects can publish the system models from IBM RSA to the SSMoS database for consumption by .NET developers.   The following picture depicts the high level use case.

In this scenario there are three plug-ins required:

1)      A VS plug-in that allows a VS user to publish a UML model developed in VS to the SSMoS database

2)      An IBM RSA plug-in that allows the user to publish a UML model developed in RSA to the SSMoS database

3)      A VS plug-in that allows a VS users to browse, search and use a UML model from the SSMoS database

VS Architect Tooling team at Microsoft can choose to develop plug-ins that provide good user experience for publishing and consuming UML artifacts from SSMoS database, if they find the scenario compelling.  VS 2010 has 5 UML diagram designers: UML Class Diagram, UML Use Case Diagram, UML Sequence Diagram, UML Activity Diagram and UML Component Diagram.  Dave Langer, who is the PM for UML model in SSMoS team, is working closely with VS Architect Tooling team to do next level of planning there.   Other tool vendors can also build SSMoS plug-ins for their tools.  (Hint: Is anyone interested in creating models for java-based assets and SSMoS plug-ins for Eclipse? J)

My experience building an Eclipse plug-in

So, I developed a bare bone Eclipse plug-in as a sample!!  I do not consider myself to be a very strong developer and it’s been many years since I have written any Java code.   But last year, after some encouragement by Doug Purdy, I decided to create a simple plug-in to publish a UML model from Eclipse-based tool to SSMoS UML database.   Fortunately, this task proved to be simpler than I thought.  Those open source Eclipse developers have done a very good job (Kudos!) in providing excellent wizards for creating Eclipse plug-ins.  The use of XMI by SSMoS LoadUml.exe utility combined with the Eclipse’s ability to serialize a UML object model as XMI, made it very easy to write this plug-in.   You can see this plug-in being used as part of the demo (at 37:10) for PDC 2009 session demo “SQL Server Modeling Services: Using Metadata to Drive Application Design, Development and Management”.   I can confidently say that other Eclipse plug-in developers can do a much better job than I did in creating Eclipse plug-ins for much richer scenarios over the SSMoS database.   Besides publishing data for a specific model, such as UML, one can imagine creating a new database from an Eclipse .ecore model (under the covers do .ecore -> “M” Schema transformation and then create a new database from the “M” schema).   I personally am a big fan of interoperability and tend to favor “better together” philosophy when it comes to software systems.

For those who have created Eclipse plug-ins, here are the extensions I created (captured in plugin.xml) for my plug-in:

1)      Custom preference page (“org.eclipse.ui.preferencePages”)  

a.       Get to it by clicking on Window -> Preferences -> Microsoft SQL Server Modeling Services

b.      This is where user can set database connection preferences.   User also sets the SSMoS folder name here, which is not ideal, because user would have to come to preferences page to change it for every new load.  If I had more time, I would have liked to build another view that allowed users to select an existing folder or create a new folder, before they publish. 

2)      Custom console view (“org.eclipse.ui.views”)

a.       Get to it by clicking on Window -> Show View -> Other … -> Microsoft SQL Server Modeling Services Console View

b.      The error messages from LoadUml.exe are piped to this view.

3)      Custom popup-menu (“org.eclipse.ui.popupMenus”)

a.       Get to it by selecting a file with extension “*.uml” or “*.xmi” in Eclipse workspace and right click Microsoft SQL Server Modeling Services -> Publish

b.      Uses couple of object contributions “org.eclipse.core.resources.IFile” and “org.eclipse.core.resources.IProject”. 

c.       Calls a custom action

4)      Custom action “PublishToRepository” (implements “org.eclipse.ui.IObjectActionDelegate)

a.       User uses the custom popup-menu to invoke this action

b.      Action uses the selected file(s) and values set in custom preference page to invoke the SSMoS LoadUml.exe utility

If you are interested in getting my plug-in source code, contact me and I can give it to you.  Just remember, this code is written by a PM (me) and not vetted by any Microsoft developer or tester. 

Thanks for reading!