1/2 Cup Dijon mustard
1/4 Cup Mayo
1/4 Cup Cheve (Softened in microwave)
2 green onions, chopped fine
1/2 tsp Celery Salt
Mix until smooth, eat with pretzels.
This week, I’ll be doing three neat security events, and you are invited!
Wednesday morning, I’ll be speaking at the Central Ohio ISSA about Windows Identity Foundation, OpenID and Claims Based Authentication. Details are here. This is the topic description:
“Escalation of privilege is based on a model of security that is driven by roles and groups for a given application. I am in the Administrator role, the Accounting group contains your username. What if instead you carried a token with a verifiable set of claims about your identity? One that is encrypted, requires no round trip to an authorization server, and can be coded against in a native API? Would that bring more security to our government and medical applications? Or is it just as full of holes as everything else? Join Bill in checking out Claims Based Security via Windows Identity Foundation, and see if it fixes problems or is the problem.”
That evening (wshew!) I’ll be giving a presentation on high-security locks at the Columbus Locksport International meeting at the Columbus Idea Foundry. You can sign up here. Please RSVP if you are coming, because we need to plan for a crowd if we have one. I’ll be covering security pins, and the idea behind sidebar locks.
Then, Friday, I’ll be at B-Sides Cleveland giving the WIF talk again. It’s at the House of Blues, and I’ll be talking at 10AM. The conference is sold out, though. Too bad - it sounds like an awesome lineup, and I am just floored to be among them. Freaking ReL1K is speaking – he built the Social Engineer’s Toolkit for crying out loud. I’m truly honored. I am looking forward to this.
Another CodeMash is in the books, and all kinds of new stuff was in the offing for me this year. But first I would be remiss if I didn’t thank Jason Gilmore, Brian Prince and especially Jim Holmes (along with the rest of the board) for uncompromising management of simply the best conference on this topic. Period. Not for the money, not for the constraints of space. It is simply the best code-centric conference on the planet.
I owe a lot of people a lot of links and information on a lot of topics.
First and foremost, I was delighted to be asked to speak again, and was pleased to have Matthew Groves join me for a discussion on Monodroid. We had 100 people join us for a look at how Monodroid came to be and what the future holds.
Then Matt took us for a tour of his excellent Stock Tracker application (shown left), converted from Windows Mobile. There were a number of good points made all around, and generally a good time was had by all.
The Monodroid documentation contains nearly everything that you need to know to get programming. The tutorials are the best starting point, and provide the templates for all of the major use cases. Matt’s application is on GitHub – please feel free to get it an mess around. It’s a good app. I’ll have BabyTrak up here in a couple of months.
The Locksport openspace was a rousing success. About 40 people were taught to pick, and about that many more stopped me in the halls and told me that they would like to have been there. I was frankly astonished by the turnout, and would have brought 5 times as many picks if I would have known about the interest – all 15 of the sets I brought were sold.
For those looking for more information:
The Locksport International site has a lot of good links to publications and whatnot. Deviant Ollem’s book, Practical Lock Picking, is excellent – he is the guy who wrote the presentation that I gave (twice). The best community is online at Lockpicking101, and they have an IRC channel too. If you need to order picks, look at LockPickShop – Red does an awesome job. The 14 piece set is on sale right now and is a great learners set!
Finally, if you are in the Columbus area please join us at the Columbus branch of Locksport International. We have a Meetup group – just go sign up and you’ll get the locations for each meeting. You can attend for free, but if you want a membership card and to participate in competitions, it’s $20 a year.
And last but not least, I got a ton of comments on the jam band. Lots of questions too. Yes, I was a professional musician for many years. I taught at a lot of area band camps, like Upper Arlington and Teays Valley. I played in a Dixieland band in London Ohio called the Lower London Street Dixieland Jazz Band and Chamber Music Society for nearly ten years. I haven’t played in quite a while, and I have to say it was a lot of fun. Hope to do it again next year.
All in all, an awesome conference. Again, I was a net producer of content rather than a consumer of content, and that’s OK. I still learned a ton just by chatting with friends old and new, and picked up information about the hip new technologies that the cool kids are using by osmosis.
Hope to see everyone at DevLink!
I recently was needful of adding some instrumentation to the Dot Net Nuke code base, and decided to use the new SQL Server Developer Tools to load in the database as a project and manage it there. I had written some internal documentation for the project while it was still under wraps, and was glad to see it so strong when it was released for public consumption. I thought seeing how I used the application might give someone out there a hand.
Since we have an existing project, we have an existing database. Fortunately, SSDT has an app for that. You can add a new SQL Server Database Project, which will effectively take a snapshot of the database and expose it to Visual Studio as T-SQL Scripts. These are the scripts that will eventually make up the development base for the software
We will start by creating a new project. Right click on the Solution file and select New Project … The Project Selection dialog appears, and if you click Data, you get the template for the SSTP. Name the project and move forward.
This project represents everything that is part of a database file in SQL Server. There is a properties folder in the project that will show you all of the database level properties – usually handled by SSMS. This is just one of many examples of SSDT bringing the DBA and developer closer together, as shown in the figure to the left. Operational properties such as the filegroup and transaction details are at least available for viewing by the developer and alterable locally. Permissions still hold, so you as a developer have to be set up to change these kinds of details to change the production system. At least you can alter them locally and see what works without a call to the DBA.
The original project is empty. In order to get the existing database into the project, an import needs to occur. Right click on the new project and then Import Objects and Settings. Select the local database and pass in the appropriate credentials. I selected the DotNetNuke database from my developer’s instance but you should select whatever you want to incorporate.
The Import Database Schema Wizard has all of the options that define how you will interact with the database once it is in the project. I like the default settings, which define a folder for each schema, and a folder within for each database item type.
There is an option to import the SQL Server permission structure, but I find that most projects don’t use that. My DotNetNuke project uses the SQL Membership Provider, however, so there is a mapping between the login structure of the database and the Users table of the membership provider. For that reason, I do turn on Import Permissions.
Once the values are set, just follow the steps:
- Make a new connection
- Click Start
- Watch the magic happen
- Click Finish
- Let’s see what we have
What we have here is everything that the database has to offer, in T-SQL Scripts. This is important. Every change that is made can be included in a DAC, because there is source control and an understood level of alteration. Changes are known by the system, and go into a pot to turn over to operations, or to be reviewed by the DBAs. Changes are not made by altering the database anymore.
Taking a look at the DotNetNuke database project, you’ll see the main schema (dbo) broken into the familiar Tables, Views, sprocs and functions. The project also has scripts for three other database members – Scripts, Security and Storage.
Storage is just the database definition. In the case of this project, it is simply:
ALTER DATABASE [$(DatabaseName)]
ADD FILE (NAME = [DotNetNuke],
FILENAME = '$(DefaultDataPath)
SIZE = 9216 KB,
FILEGROWTH = 1024 KB)
TO FILEGROUP [PRIMARY];
It’s part of the completeness, but not something that you will alter a lot. The Scripts directory is another that you’ll change once and forget about – it contains the pre and post build scripts for the deployment of the database. Every time you build and push to the actual database server, these scripts will be run.
The Security folder is fascinating. It contains all of the roles and related schemas, and the associated authorizations. If you have a project that secures database assets in the database management system, this could be awesome.
The meat is in the schema folder, called ‘dbo’ in the DNN example. This is where the scripts you would expect to see in a project like this are held, and where we will be doing the majority of our work. Each entity in the database is scripted separately here, and we can modify or add to our hearts content, and deploy separately.
Set up some new entities
The first ting needed for the instrumentation being added is a table for some timing data. Start by right clicking on the Tables folder and selecting ‘Add New …’. Notice the nice Visual Studio integration that shows asset types which can be added. Select a Table. Name the table ‘Instrumentation’ in the Add New Item dialog and click OK. There is a base template for a table; go ahead and change it for the new fields:
CREATE TABLE [dbo].[Instrumentation]
InstrumentationId int NOT NULL IDENTITY (1, 1) PRIMARY KEY,
ClassName varchar(64) NOT NULL,
MemberName varchar(64) NOT NULL,
ElapsedSeconds bigint NULL,
Exception text null
There is a Commit button that is so tempting at this point, but it isn’t the button you want right now. Commit peeks at the declarative information in the script – effectively doing a pre-parse – and try and make appropriate changes to the target database. It is more or less like the Table Designer in SSMS at this point, using the references to concepts in the database to make decisions rather than just running the T-SQL as coded.
Click the Execute SQL button in the text window’s toolbar to run the CREATE TABLE procedure and save the table to the database defined in the project properties. (You can read more about that in the Deployment section below.) In this way, the database is effectively disposable. At any time, you can get a copy of the scripts from Team Foundation Server, and generate a whole new copy of the database, minus reference data. For right now, though, we are just adding this one table.
Great, that gives us a place to put the data. Next step is a way to get it there. Right click on the Stored Procedures folder and select Stored Procedure from the context menu. Just like the table above, Visual Studio will give a base template. I changed it to the code below:
CREATE PROCEDURE [dbo].[AddInstrumentationEvent]
@ElapsedSeconds bigint = 0,
@Exception text = ''
INSERT INTO Instrumentation
I need to make a quick shout out here for the IntelliSense. It’s expected I suppose, that this should support full IntelliSense integration but I was just shocked at how comfortable to use I personally found it. I do not care to write SQL code because it is so cumbersome. Having that modern development experience talked about in the introduction makes a big difference.
That’s all we got for new features – it’s a short paper after all. Clearly any entity that SQL Server supports can easily be added to the project, stored in source control, and managed like the rest of the application. What’s more, it can be tested and refactored like any other part of the application.
Red, green, refactor
After further review, it was decided that Instrumentation wasn’t a good enough name for the table, since it doesn’t accurately represent what was actually put in the rows of the table. Instead, the name InstrumentationEvents is supposed to be used, so we need to rename the table.
Right click on the table name in the CREATE statement of Instrumentation.sql and select Refactor -> Rename. Change the name to InstrumentationEvents and click Next. Notice, as shown in the figure to the left, that SSDT got it right. The preview is very helpful. It finds all of the consuming database members, and lets you determine which of them to apply the change to.. Even in the stored procedure, the pluralization is correct, changing AddInstrumentation to AddInstrumentationEvent rather than AddInstrumentationEvents. That trailing s might not seem like much to some people but it can make a big difference in a convention over configuration based system.
Rename isn’t the only refactoring available in SSDT, there are also T-SQL specific features. If you are working in DotNetNuke, open up the GetUsersByUserName.sql script in the Stored Procedures folder. It’s a little overdeveloped and has too much UI logic in it, but it works for this example.
Line 31 has a SELECT * on it, and frankly this procedure is too slow as it is. We don’t need to add a table scan. The refactor menu has an Expand Wildcards option, and I recommend its use here. Right click on the asterisk and select Refactor -> Expand Wildcards, then click on SELECT * in the treeview. The Preview Changes dialog now will show us how the sproc will look with the wildcard expanded, just like the figure to the right. Click Apply to have the changes applied to the procedure.
Don’t overlook the various features of the Visual Studio IDE that now can be used to manage the T-SQL code. For instance, consider that GetUserByUserName.sql file. Right click on the vw_Users token on line 10 and select Go To Definition to be taken to the View definition in the database project. The view doesn’t help us much, because we want to see the table in question. Scroll down in the view to the FROM statement and right click on Users and select Go To Definition again to see the Users table.
As expected, you can then right click on the Users table name and Find All References to see the 44 places that the Users table is used in the database project. The usefulness of this in a legacy environment can’t be overestimated. Finding your way around a project this easily, digging from the core code to the database and navigating using the built-in features will significantly reduce time spent in getting up to speed on existing applications.
What’s more, code-level refactoring isn’t the only thing that the data modeling group is pushing for in SSDT. There is project-level refactoring available as well, which is a step in the right direction of whole-project management. Across-the-board changes to software code are something that Visual Studio already excels at, and SSDT is working toward providing the same kinds of features for database projects.
For instance, right click on the database project and select Refactor from that context menu. Aside from the wildcard expansion seen in the code-level refactoring, not the Rename Server/Database reference. It’s a whole-project way to change references that would be manages in configuration files for a C# project, but needs to be controlled with refactoring in T-SQL.
Refactoring is an important part of software development. Though it has been available in third party tools for a while having a standardized experience that is so tightly integrated with Visual Studio will make a big difference to the average developer. While unit testing integration with Visual Studio still isn’t in there for T-SQL it is still a step in the right direction toward that modern development experience we have been talking about.
Bastard child of the Screwdriver and the Shirley Temple.
- 1 shot Absolut Vodka
- 1 shot Mandarin Orange Juice (just pour it right out of the jar or oranges)
- 4 slices of Mandarin orange
- Fill the glass with Sprite
Preferably put it in a Pleasure Island Jazz Company 8oz shot glass.
And yes, no ice. Make sure the ingredients are cold.
Thanks to Matt Groves for the name.
There are a lot of questions about Visual Basic.NET 2005 For Dummies and Visual Basic.NET 2008 For Dummies and the use of Northwind for the samples. When I wrote the majority of that book in 2004, Northwind was still a common sample database for SQL Server. When I updated it in 2007, it will still provided as a sample. Since then, it has been totally replaced by AdventureWorks. Now, finding the sample is hard, and installing it is even harder.
I was going to write a large post on how to do the install, but Pinalkumar Dave did such an awesome job on his blog that I don't have to. Here is the link:
You can get the samples still, from this link:
I hope this helps. I will be updating the VB book series after I am done with Programming Data, and will change the data samples to use AdventureWorks, or the latest and greatest at that time, if it changes again.
I'll be speaking at the 2010 DogFood Conference that Microsoft puts on here in Columbus. Danilo Castilo runs it (largely) and it is pretty cool - a neat community event on a budget.
It's a cool collection of talks about the newest Microsoft products and how people are using them. Thus the name: 'DogFood' from the phrase 'eating your own dog food.'
I'll be speaking with Mario Fulan about using AFDS 2.0 to cross domain borders. If you don't already know Mario, he is a beast - one of like ten Certified Sharepoint Masters in the whole freakin universe or something. He has forgotten more about SharePoint than I will ever learn. I do know Windows Identity Foundation a little bit though, so that's what I'll be covering.
The conference is at www.dogfoodcon.com and is selling out really fast. If you are interested in the hot new stuff, check it out and get registered while you can. It's next month - November 4 and 5.
It's common knowledge that I have been following Oslo / SQL Server Modeling Services very closely. I am working on a book on the topic, and have posted a number of blog entries. The speaking circuit has been good to me too, and I have given my Software Modeling With ASCII talk five or six times already this year.
My focus has been on M, but today we are talking about Quadrant. Quadrant is part of a trio of tools that includes M (a language to build data and domain models) and Modeling Services (a set of common models and repository). Quadrant itself is tool to interact visually with SQL Server databases.
I've been watching Quadrant for over a year now, and I had a lot of questions about its viability in the marketplace. As a data management tool, it was underpowered, but as a data browsing tool, it was overpowered. When I eventually came to realize that it could be a domain model browser, my interest was piqued. Since you could define the quadrants using M, it would be effectively possible to build comprehensive data management 'dashboards' in Quadrant, and use them in a power user role.
Over time, however, I began to realize that this was an edge case. The business users and data managers that need these solutions will still find development in M too time-consuming, and the professional developers who would be asked to help them would just rather work in C#. It will end up that the business users will go back to Access and Excel, data managers will just use SQL Management Studio, and professional developers will use Windows Forms or XAML in C#.
Apparently, Microsoft saw this too.
I am sad to see Quadrant go. It was a beautiful application, and could have been a foundation to a number of very, very cool tools. Hopefully the Data team wil find another use for the technology.
It should be noted that SQL Server Modeling is not necessarily dead at all. Models can still be built with M, and exposed to the world in OData. Applications can still be built against this model using Visual Studio, and the data can still be managed using SQL Management Studio.
The loss of Quadrant doesn't impact this vision, and I hope that Microsoft realizes this and continues down the path toward an enterprise-class repository. It's the last piece of the puzzle that keeps large enterprises from deploying SQL Server in application-centric environments.
When I was a sophomore in high school, we had a unit in our World History class about the Holocaust. Fran LaBuda, a German Jew who escaped to the US through despite the Nazis, would stand at the door of our classroom and bark orders to us in German as we entered, using a pointer to tell us where to sit, and even push us around as necessary. A militant looking fellow (later I learned it was her son, and as gentle a guy as you could imagine in real life) escorted anyone who didn’t take it seriously out of the room, rather roughly.
The point was to show us in general how easily we could be cowed by a force we didn’t understand taking our power of independence. These are upper middle class high school students, and used to getting their way. Their parents bought them the cool clothes and looked the other way when the rules were transcended. They wore their ego on their shoulder like a badge of honor.
But when the going got rough, they folded like a bad hand at cards. Only one person tried to joke about the event with Mrs. LaBuda, and was taken from the room. He was the class clown, but was nearly in tears when pushed out of the door by the enforcer.
Fast forward to today. I was in line at the TSA’s security gate at SeaTac. Walking up and down the line was a rather militant looking fellow yelling out in plain, though loud, English:
“If you do not take your liquids and gels out of your carry-on luggage you will not be allowed to get on your plane. You will be escorted to enhanced screening, and there is a half day wait.”
Next to me stood a seventy year old woman, grey hair, a Russian Jew by her accent; tears were streaming down her face. She was frantically digging through her plain bag looking for the satchelof toiletries that was plainly sitting on the table in front of her, unnoticed.
“Your bottles are right here,” I showed her.
“Oh, thank you son,” she sighed in relief. “I’m just trying to get home to Florida to see my grandson. I’m so terrified that these people will lock me up.”
She was so terrified that those people would lock her up. People that were purportedly trying to keep us safe, but who were instead driving this woman, others, myself to tears with worry that one wrong move with the toothpaste could cost us time with loved ones, money, business, whatever.
The terrorists have won.
The goal of a ‘terrorist,’ and thus the name, is terror. They don’t really care, as a group, if they kill anyone. As long as the people they attack live in fear. They state that they want to kill Americans, and then, largely, don’t. They just want us to think that they will. (Remember, while 9/11 was a huge tragedy, it doesn’t make much of a mark in the numbers that have died in simple in-fighting in the Arab Alliance. The deaths weren’t the point. The after-effects were the point.)
We, as a country, as a people, as individuals, have folded. Just like that classroom of sophomores 20 years ago, we have turned in our independence to the authorities with our papers and our shampoo. Even the clowns in Washington, once a source of hope, are led crying from the classroom the moment the chips are down.
Please don’t think your humble author is putting himself above you, the reader. I had planned on traveling with a firearm this trip: because I can, then lock my luggage with my locks, and pretend that I am more secure than most. I did not, fearing hassle, fearing delay, or just fearing – I’m not sure which.
I don’t have a solution to suggest, dear reader. I simply needed to lament the passing of a once great country – the greatest of social experiments – into the waste bin of political history. I do not believe that is within any of us to turn the social tide now, unless Atlas truly does shrug and some number of us retreat to a contemporary Galt’s Gulch. The slope of our decline is too firmly now in place. We have lost.
For almost 200,000 years, humans were indistinguishable from animals.
For 5,000 years, we had only achieved the advancements of agriculture and prostitution. Nothing to be sneezed at for sure, but certainly not the pinnacle of potential.
In 100 years, we went from farming to the Industrial Revolution. There were a lot of reasons, but note the sudden easy availability of fiction. I know, I know, correlation doesn’t equal causation, but I can’t help but wonder how much the insurgence of fiction, and how it influenced the play of children, impacted the next generation and the ideas they worked from. Factories? Space travel? Computing?
Fast forward to Asimov, Clarke, and the other Science Fiction writers of the 50s. They pointed our eyes to the stars and our minds to the unimaginable. Is it a surprise that the generation that grew up reading their books and reenacting it in their play gave us the fathers of the Internet?
Please don’t dismiss child’s play as a waste of time. Please don’t assume that the introduction of fictional universes into a children’s playtime is an “overdose of media.” You don’t know what the availability of universes is doing for our children’s fertile minds. Wouldn’t you rather let them run with it and see what becomes of it, rather than shut it down, afraid of the future it might bring?