Search Results

Search found 9771 results on 391 pages for 'equivalence classes'.

Page 347/391 | < Previous Page | 343 344 345 346 347 348 349 350 351 352 353 354  | Next Page >

• A new SQL, a new Analysis Services, a new Workshop! #ssas #sql2012

  - by Marco Russo (SQLBI)

  One week ago Microsoft SQL Server 2012 finally debuted with a virtual launch event and you can find many intro sessions there (20 minutes each). There is a lot of new content available if you want to learn more about SQL 2012 and in this blog post I’d like to provide a few link to sessions, documents, bits and courses that are available now or very soon. First of all, the release of Analysis Services 2012 has finally released PowerPivot 2012 (many of us called it PowerPivot v2 before this official name) and also the new Data Mining Add-in for Microsoft Office 2010, now available also for Excel 64bit! And, of course, don’t miss the Microsoft SQL Server 2012 Feature Pack, there are a lot of upgrades for both DBAs and developers. I just discovered there is a new LocalDB version of SQL Express that can run in user mode without any setup. Is this the end of SQL CE? But now, back to Analysis Services: if you want some tutorial on Tabular, the Microsoft Virtual Academy has a whole track dedicated to Analysis Services 2012 but you will probably be interested also in the one about Reporting Services 2012. If you think that virtual is good but it’s not enough, there are plenty of conferences in the coming months – these are just those where I and Alberto will deliver some SSAS Tabular presentations: SQLBits X, London, March 29-31, 2012: if you are in London or want a good reason to go, this is the most important SQL Server event in Europe this year, no doubts about it. And not only because of the high number of attendees, but also because there is an impressive number of speakers (excluding me, of course) coming from all over the world. This is an event second only to PASS Summit in Seattle so there are no good reasons to not attend it. Microsoft SQL Server & Business Intelligence Conference 2012, Milan, March 28-29, 2012: this is an Italian conference so the language might be a barrier, but many of us also speak English and the food is good! Just a few seats still available. TechEd North America, Orlando, June 11-14, 2012: you know, this is a big event and it contains everything – if you want to spend a whole day learning the SSAS Tabular model with me and Alberto, don’t miss our pre-conference day “Using BISM Tabular in Microsoft SQL Server Analysis Services 2012” (be careful, it is on June 10, a nice study-Sunday!). TechEd Europe, Amsterdam, June 26-29, 2012: the European version of TechEd provides almost the same content and you don’t have to go overseas. We also run the same pre-conference day “Using BISM Tabular in Microsoft SQL Server Analysis Services 2012” (in this case, it is on June 25, that’s a regular Monday). I and Alberto will also speak at some user group meeting around Europe during… well, we’re going to travel a lot in the next months. In fact, if you want to get a complete training on SSAS Tabular, you should spend two days with us in one of our SSAS Tabular Workshop! We prepared a 2-day seminar, a very intense one, that start from the simple tabular modeling and cover architecture, DAX, query, advanced modeling, security, deployment, optimization, monitoring, relationships with PowerPivot and Multidimensional… Really, there are a lot of stuffs here! We announced the first dates in Europe and also an online edition optimized for America’s time zone: Apr 16-17, 2012 – Amsterdam, Netherlands Apr 26-27, 2012 – Copenhagen, Denmark May 7-8, 2012 – Online for America’s time zone May 14-15, 2012 – Brussels, Belgium May 21-22, 2012 – Oslo, Norway May 24-25, 2012 – Stockholm, Sweden May 28-29, 2012 – London, United Kingdom May 31-Jun 1, 2012 – Milan, Italy (Italian language) Also Chris Webb will join us in this workshop and in every date you can find who is the speaker on the web site. The course is based on our upcoming book, almost 600 pages (!) about SSAS Tabular, an incredible effort that will be available very soon in a preview (rough cuts from O’Reilly) and will be on the shelf in May. I will provide a link to order it as soon as we have one! And if you think that this is not enough… you’re right! Do you know what is the only thing you can do to optimize your Tabular model? Optimize your DAX code. Learning DAX is easy, mastering DAX requires some knowledge… and our DAX Advanced Workshop will provide exactly the required content. Public classes will be available later this year, by now we just deliver it on demand.

  Read the article

• Calculated Columns in Entity Framework Code First Migrations

  - by David Paquette

  I had a couple people ask me about calculated properties / columns in Entity Framework this week.  The question was, is there a way to specify a property in my C# class that is the result of some calculation involving 2 properties of the same class.  For example, in my database, I store a FirstName and a LastName column and I would like a FullName property that is computed from the FirstName and LastName columns.  My initial answer was: 1: public string FullName 2: { 3: get { return string.Format("{0} {1}", FirstName, LastName); } 4: } Of course, this works fine, but this does not give us the ability to write queries using the FullName property.  For example, this query: 1: var users = context.Users.Where(u => u.FullName.Contains("anan")); Would result in the following NotSupportedException: The specified type member 'FullName' is not supported in LINQ to Entities. Only initializers, entity members, and entity navigation properties are supported. It turns out there is a way to support this type of behavior with Entity Framework Code First Migrations by making use of Computed Columns in SQL Server.  While there is no native support for computed columns in Code First Migrations, we can manually configure our migration to use computed columns. Let’s start by defining our C# classes and DbContext: 1: public class UserProfile 2: { 3: public int Id { get; set; } 4: 5: public string FirstName { get; set; } 6: public string LastName { get; set; } 7: 8: [DatabaseGenerated(DatabaseGeneratedOption.Computed)] 9: public string FullName { get; private set; } 10: } 11: 12: public class UserContext : DbContext 13: { 14: public DbSet<UserProfile> Users { get; set; } 15: } The DatabaseGenerated attribute is needed on our FullName property.  This is a hint to let Entity Framework Code First know that the database will be computing this property for us. Next, we need to run 2 commands in the Package Manager Console.  First, run Enable-Migrations to enable Code First Migrations for the UserContext.  Next, run Add-Migration Initial to create an initial migration.  This will create a migration that creates the UserProfile table with 3 columns: FirstName, LastName, and FullName.  This is where we need to make a small change.  Instead of allowing Code First Migrations to create the FullName property, we will manually add that column as a computed column. 1: public partial class Initial : DbMigration 2: { 3: public override void Up() 4: { 5: CreateTable( 6: "dbo.UserProfiles", 7: c => new 8: { 9: Id = c.Int(nullable: false, identity: true), 10: FirstName = c.String(), 11: LastName = c.String(), 12: //FullName = c.String(), 13: }) 14: .PrimaryKey(t => t.Id); 15: Sql("ALTER TABLE dbo.UserProfiles ADD FullName AS FirstName + ' ' + LastName"); 16: } 17: 18: 19: public override void Down() 20: { 21: DropTable("dbo.UserProfiles"); 22: } 23: } Finally, run the Update-Database command.  Now we can query for Users using the FullName property and that query will be executed on the database server.  However, we encounter another potential problem. Since the FullName property is calculated by the database, it will get out of sync on the object side as soon as we make a change to the FirstName or LastName property.  Luckily, we can have the best of both worlds here by also adding the calculation back to the getter on the FullName property: 1: [DatabaseGenerated(DatabaseGeneratedOption.Computed)] 2: public string FullName 3: { 4: get { return FirstName + " " + LastName; } 5: private set 6: { 7: //Just need this here to trick EF 8: } 9: } Now we can both query for Users using the FullName property and we also won’t need to worry about the FullName property being out of sync with the FirstName and LastName properties.  When we run this code: 1: using(UserContext context = new UserContext()) 2: { 3: UserProfile userProfile = new UserProfile {FirstName = "Chanandler", LastName = "Bong"}; 4: 5: Console.WriteLine("Before saving: " + userProfile.FullName); 6: 7: context.Users.Add(userProfile); 8: context.SaveChanges(); 9:  10: Console.WriteLine("After saving: " + userProfile.FullName); 11:  12: UserProfile chanandler = context.Users.First(u => u.FullName == "Chanandler Bong"); 13: Console.WriteLine("After reading: " + chanandler.FullName); 14:  15: chanandler.FirstName = "Chandler"; 16: chanandler.LastName = "Bing"; 17:  18: Console.WriteLine("After changing: " + chanandler.FullName); 19:  20: } We get this output: It took a bit of work, but finally Chandler’s TV Guide can be delivered to the right person. The obvious downside to this implementation is that the FullName calculation is duplicated in the database and in the UserProfile class. This sample was written using Visual Studio 2012 and Entity Framework 5. Download the source code here.

  Read the article

• The illusion of Competence

  - by tony_lombardo

  Working as a contractor opened my eyes to the developer food chain.  Even though I had similar experiences earlier in my career, the challenges seemed much more vivid this time through.  I thought I’d share a couple of experiences with you, and the lessons that can be taken from them. Lesson 1: Beware of the “funnel” guy.  The funnel guy is the one who wants you to funnel all thoughts, ideas and code changes through him.  He may say it’s because he wants to avoid conflicts in source control, but the real reason is likely that he wants to hide your contributions.  Here’s an example.  When I finally got access to the code on one of my projects, I was told by the developer that I had to funnel all of my changes through him.  There were 4 of us coding on the project, but only 2 of us working on the UI.  The other 2 were working on a separate application, but part of the overall project.  So I figured, I’ll check it into SVN, he reviews and accepts then merges in.  Not even close.  I didn’t even have checkin rights to SVN, I had to email my changes to the developer so he could check those changes in.  Lesson 2: If you point out flaws in code to someone supposedly ‘higher’ than you in the developer chain, they’re going to get defensive.  My first task on this project was to review the code, familiarize myself with it.  So of course, that’s what I did.  And in familiarizing myself with it, I saw so many bad practices and code smells that I immediately started coming up with solutions to fix it.  Of course, when I reviewed these changes with the developer (guy who originally wrote the code), he smiled and nodded and said, we can’t make those changes now, it’s too destabilizing.  I recommended we create a new branch and start working on refactoring, but branching was a new concept for this guy and he was worried we would somehow break SVN. How about some concrete examples? I started out by recommending we remove NUnit dependency and tests from the application project, and create a separate Unit testing project.  This was met with a little bit of resistance because - “How do I access the private methods?”  As it turned out there weren’t really any private methods that weren’t exposed by public methods, so I quickly calmed this fear. Win 1 Loss 0 Next, I recommended that all of the File IO access be wrapped in Using clauses, or at least properly wrapped in try catch finally.  This recommendation was accepted.. but never implemented. Win 2  Loss 1 Next recommendation was to refactor the command pattern implementation.  The command pattern was implemented, but it wasn’t really necessary for the application.  More over, the fact that we had 100 different command classes, each with it’s own specific command parameters class, made maintenance a huge hassle.  The same code repeated over and over and over.  This recommendation was declined, the code was too fragile and this change would destabilize it.  I couldn’t disagree, though it was the commands themselves in many cases that were fragile. Win 2 Loss 2 Next recommendation was to aid performance (and responsiveness) of the application by using asynchronous service calls.  This on was accepted. Win 2 Loss 3 If you’re paying any attention, you’re wondering why the async service calls was scored as a loss.. Let me explain.  The service call was made using the async pattern.  Followed by a thread.sleep  <facepalm>. Now it’s easy to be harsh on this kind of code, especially if you’re an experienced developer.  But I understood how most of this happened.  One junior guy, working as hard as he can to build his first real world application, with little or no guidance from anyone else.  He had his pattern book and theory of programming to help him, but no real world experience.  He didn’t know how difficult it would be to trace the crashes to the coding issues above, but he will one day.  The part that amazed me was the management position that “this guy should be a team lead, because he’s worked so hard”.  I’m all for rewarding hard work, but when you reward someone by promoting them past the point of their competence, you’re setting yourself and them up for failure.  And that’s lesson 3.  Just because you’ve got a hard worker, doesn’t mean he should be leading a development project.  If you’re a junior guy busting your ass, keep at it.  I encourage you to try new things, but most importantly to learn from your mistakes.  And correct your mistakes.  And if someone else looks at your code and shows you a laundry list of things that should be done differently, don’t take it personally – they’re really trying to help you.  And if you’re a senior guy, working with a junior guy, it’s your duty to point out the flaws in the code.  Even if it does make you the bad guy.  And while I’ve used “guy” above, I mean both men and women.  And in some cases mutant dinosaurs. 

  Read the article

• Trace File Source Adapter

  The Trace File Source adapter is a useful addition to your SSIS toolbox.  It allows you to read 2005 and 2008 profiler traces stored as .trc files and read them into the Data Flow.  From there you can perform filtering and analysis using the power of SSIS. There is no need for a SQL Server connection this just uses the trace file. Example Usages Cache warming for SQL Server Analysis Services Reading the flight recorder Find out the longest running queries on a server Analyze statements for CPU, memory by user or some other criteria you choose Properties The Trace File Source adapter has two properties, both of which combine to control the source trace file that is read at runtime. SQL Server 2005 and SQL Server 2008 trace files are supported for both the Database Engine (SQL Server) and Analysis Services. The properties are managed by the Editor form or can be set directly from the Properties Grid in Visual Studio. Property Type Description AccessMode Enumeration This property determines how the Filename property is interpreted. The values available are: DirectInput Variable Filename String This property holds the path for trace file to load (*.trc). The value is either a full path, or the name of a variable which contains the full path to the trace file, depending on the AccessMode property. Trace Column Definition Hopefully the majority of you can skip this section entirely, but if you encounter some problems processing a trace file this may explain it and allow you to fix the problem. The component is built upon the trace management API provided by Microsoft. Unfortunately API methods that expose the schema of a trace file have known issues and are unreliable, put simply the data often differs from what was specified. To overcome these limitations the component uses  some simple XML files. These files enable the trace column data types and sizing attributes to be overridden. For example SQL Server Profiler or TMO generated structures define EventClass as an integer, but the real value is a string. TraceDataColumnsSQL.xml  - SQL Server Database Engine Trace Columns TraceDataColumnsAS.xml    - SQL Server Analysis Services Trace Columns The files can be found in the %ProgramFiles%\Microsoft SQL Server\100\DTS\PipelineComponents folder, e.g. "C:\Program Files\Microsoft SQL Server\100\DTS\PipelineComponents\TraceDataColumnsSQL.xml" "C:\Program Files\Microsoft SQL Server\100\DTS\PipelineComponents\TraceDataColumnsAS.xml" If at runtime the component encounters a type conversion or sizing error it is most likely due to a discrepancy between the column definition as reported by the API and the actual value encountered. Whilst most common issues have already been fixed through these files we have implemented specific exception traps to direct you to the files to enable you to fix any further issues due to different usage or data scenarios that we have not tested. An example error that you can fix through these files is shown below. Buffer exception writing value to column 'Column Name'. The string value is 999 characters in length, the column is only 111. Columns can be overridden by the TraceDataColumns XML files in "C:\Program Files\Microsoft SQL Server\100\DTS\PipelineComponents\TraceDataColumnsAS.xml". Installation The component is provided as an MSI file which you can download and run to install it. This simply places the files on disk in the correct locations and also installs the assemblies in the Global Assembly Cache as per Microsoft’s recommendations. You may need to restart the SQL Server Integration Services service, as this caches information about what components are installed, as well as restarting any open instances of Business Intelligence Development Studio (BIDS) / Visual Studio that you may be using to build your SSIS packages. Finally you will have to add the transformation to the Visual Studio toolbox manually. Right-click the toolbox, and select Choose Items.... Select the SSIS Data Flow Items tab, and then check the Trace File Source transformation in the Choose Toolbox Items window. This process has been described in detail in the related FAQ entry for How do I install a task or transform component? We recommend you follow best practice and apply the current Microsoft SQL Server Service pack to your SQL Server servers and workstations. Please note that the Microsoft Trace classes used in the component are not supported on 64-bit platforms. To use the Trace File Source on a 64-bit host you need to ensure you have the 32-bit (x86) tools available, and the way you execute your package is setup to use them, please see the help topic 64-bit Considerations for Integration Services for more details. Downloads Trace Sources for SQL Server 2005 -- Trace Sources for SQL Server 2008 Version History SQL Server 2008 Version 2.0.0.382 - SQL Sever 2008 public release. (9 Apr 2009) SQL Server 2005 Version 1.0.0.321 - SQL Server 2005 public release. (18 Nov 2008) -- Screenshots

  Read the article

• What should you bring to the table as a Software Architect?

  - by Ahmad Mageed

  There have been many questions with good answers about the role of a Software Architect (SA) on StackOverflow and Programmers SE. I am trying to ask a slightly more focused question than those. The very definition of a SA is broad so for the sake of this question let's define a SA as follows: A Software Architect guides the overall design of a project, gets involved with coding efforts, conducts code reviews, and selects the technologies to be used. In other words, I am not talking about managerial rest and vest at the crest (further rhyming words elided) types of SAs. If I were to pursue any type of SA position I don't want to be away from coding. I might sacrifice some time to interface with clients and Business Analysts etc., but I am still technically involved and I'm not just aware of what's going on through meetings. With these points in mind, what should a SA bring to the table? Should they come in with a mentality of "laying down the law" (so to speak) and enforcing the usage of certain tools to fit "their way," i.e., coding guidelines, source control, patterns, UML documentation, etc.? Or should they specify initial direction and strategy then be laid back and jump in as needed to correct the ship's direction? Depending on the organization this might not work. An SA who relies on TFS to enforce everything may struggle to implement their plan at an employer that only uses StarTeam. Similarly, an SA needs to be flexible depending on the stage of the project. If it's a fresh project they have more choices, whereas they might have less for existing projects. Here are some SA stories I have experienced as a way of sharing some background in hopes that answers to my questions might also shed some light on these issues: I've worked with an SA who code reviewed literally every single line of code of the team. The SA would do this for not just our project but other projects in the organization (imagine the time spent on this). At first it was useful to enforce certain standards, but later it became crippling. FxCop was how the SA would find issues. Don't get me wrong, it was a good way to teach junior developers and force them to think of the consequences of their chosen approach, but for senior developers it was seen as somewhat draconian. One particular SA was against the use of a certain library, claiming it was slow. This forced us to write tons of code to achieve things differently while the other library would've saved us a lot of time. Fast forward to the last month of the project and the clients were complaining about performance. The only solution was to change certain functionality to use the originally ignored approach despite early warnings from the devs. By that point a lot of code was thrown out and not reusable, leading to overtime and stress. Sadly the estimates used for the project were based on the old approach which my project was forbidden from using so it wasn't an appropriate indicator for estimation. I would hear the PM say "we've done this before," when in reality they had not since we were using a new library and the devs working on it were not the same devs used on the old project. The SA who would enforce the usage of DTOs, DOs, BOs, Service layers and so on for all projects. New devs had to learn this architecture and the SA adamantly enforced usage guidelines. Exceptions to usage guidelines were made when it was absolutely difficult to follow the guidelines. The SA was grounded in their approach. Classes for DTOs and all CRUD operations were generated via CodeSmith and database schemas were another similar ball of wax. However, having used this setup everywhere, the SA was not open to new technologies such as LINQ to SQL or Entity Framework. I am not using this post as a platform for venting. There were positive and negative aspects to my experiences with the SA stories mentioned above. My questions boil down to: What should an SA bring to the table? How can they strike a balance in their decision making? Should one approach an SA job (as defined earlier) with the mentality that they must enforce certain ground rules? Anything else to consider? Thanks! I'm sure these job tasks are easily extended to people who are senior devs or technical leads, so feel free to answer at that capacity as well.

  Read the article

• My Right-to-Left Foot (T-SQL Tuesday #13)

  - by smisner

  As a business intelligence consultant, I often encounter the situation described in this month's T-SQL Tuesday, hosted by Steve Jones ( Blog | Twitter) – “What the Business Says Is Not What the  Business Wants.” Steve posed the question, “What issues have you had in interacting with the business to get your job done?” My profession requires me to have one foot firmly planted in the technology world and the other foot planted in the business world. I learned long ago that the business never says exactly what the business wants because the business doesn't have the words to describe what the business wants accurately enough for IT. Not only do technological-savvy barriers exist, but there are also linguistic barriers between the two worlds. So how do I cope? The adage "a picture is worth a thousand words" is particularly helpful when I'm called in to help design a new business intelligence solution. Many of my students in BI classes have heard me explain ("rant") about left-to-right versus right-to-left design. To understand what I mean about these two design options, let's start with a picture: When we design a business intelligence solution that includes some sort of traditional data warehouse or data mart design, we typically place the data sources on the left, the new solution in the middle, and the users on the right. When I've been called in to help course-correct a failing BI project, I often find that IT has taken a left-to-right approach. They look at the data sources, decide how to model the BI solution as a _______ (fill in the blank with data warehouse, data mart, cube, etc.), and then build the new data structures and supporting infrastructure. (Sometimes, they actually do this without ever having talked to the business first.) Then, when they show what they've built to the business, the business says that is not what we want. Uh-oh. I prefer to take a right-to-left approach. Preferably at the beginning of a project. But even if the project starts left-to-right, I'll do my best to swing it around so that we’re back to a right-to-left approach. (When circumstances are beyond my control, I carry on, but it’s a painful project for everyone – not because of me, but because the approach just doesn’t get to what the business wants in the most effective way.) By using a right to left approach, I try to understand what it is the business is trying to accomplish. I do this by having them explain reports to me, and explaining the decision-making process that relates to these reports. Sometimes I have them explain to me their business processes, or better yet show me their business processes in action because I need pictures, too. I (unofficially) call this part of the project "getting inside the business's head." This is starting at the right side of the diagram above. My next step is to start moving leftward. I do this by preparing some type of prototype. Depending on the nature of the project, this might mean that I simply mock up some data in a relational database and build a prototype report in Reporting Services. If I'm lucky, I might be able to use real data in a relational database. I'll either use a subset of the data in the prototype report by creating a prototype database to hold the sample data, or select data directly from the source. It all depends on how much data there is, how complex the queries are, and how fast I need to get the prototype completed. If the solution will include Analysis Services, then I'll build a prototype cube. Analysis Services makes it incredibly easy to prototype. You can sit down with the business, show them the prototype, and have a meaningful conversation about what the BI solution should look like. I know I've done a good job on the prototype when I get knocked out of my chair so that the business user can explore the solution further independently. (That's really happened to me!) We can talk about dimensions, hierarchies, levels, members, measures, and so on with something tangible to look at and without using those terms. It's not helpful to use sample data like Adventure Works or to use BI terms that they don't really understand. But when I show them their data using the BI technology and talk to them in their language, then they truly have a picture worth a thousand words. From that, we can fine tune the prototype to move it closer to what they want. They have a better idea of what they're getting, and I have a better idea of what to build. So right to left design is not truly moving from the right to the left. But it starts from the right and moves towards the middle, and once I know what the middle needs to look like, I can then build from the left to meet in the middle. And that’s how I get past what the business says to what the business wants.

  Read the article

• A new SQL, a new Analysis Services, a new Workshop! #ssas #sql2012

  - by Marco Russo (SQLBI)

  One week ago Microsoft SQL Server 2012 finally debuted with a virtual launch event and you can find many intro sessions there (20 minutes each). There is a lot of new content available if you want to learn more about SQL 2012 and in this blog post I’d like to provide a few link to sessions, documents, bits and courses that are available now or very soon. First of all, the release of Analysis Services 2012 has finally released PowerPivot 2012 (many of us called it PowerPivot v2 before this official name) and also the new Data Mining Add-in for Microsoft Office 2010, now available also for Excel 64bit! And, of course, don’t miss the Microsoft SQL Server 2012 Feature Pack, there are a lot of upgrades for both DBAs and developers. I just discovered there is a new LocalDB version of SQL Express that can run in user mode without any setup. Is this the end of SQL CE? But now, back to Analysis Services: if you want some tutorial on Tabular, the Microsoft Virtual Academy has a whole track dedicated to Analysis Services 2012 but you will probably be interested also in the one about Reporting Services 2012. If you think that virtual is good but it’s not enough, there are plenty of conferences in the coming months – these are just those where I and Alberto will deliver some SSAS Tabular presentations: SQLBits X, London, March 29-31, 2012: if you are in London or want a good reason to go, this is the most important SQL Server event in Europe this year, no doubts about it. And not only because of the high number of attendees, but also because there is an impressive number of speakers (excluding me, of course) coming from all over the world. This is an event second only to PASS Summit in Seattle so there are no good reasons to not attend it. Microsoft SQL Server & Business Intelligence Conference 2012, Milan, March 28-29, 2012: this is an Italian conference so the language might be a barrier, but many of us also speak English and the food is good! Just a few seats still available. TechEd North America, Orlando, June 11-14, 2012: you know, this is a big event and it contains everything – if you want to spend a whole day learning the SSAS Tabular model with me and Alberto, don’t miss our pre-conference day “Using BISM Tabular in Microsoft SQL Server Analysis Services 2012” (be careful, it is on June 10, a nice study-Sunday!). TechEd Europe, Amsterdam, June 26-29, 2012: the European version of TechEd provides almost the same content and you don’t have to go overseas. We also run the same pre-conference day “Using BISM Tabular in Microsoft SQL Server Analysis Services 2012” (in this case, it is on June 25, that’s a regular Monday). I and Alberto will also speak at some user group meeting around Europe during… well, we’re going to travel a lot in the next months. In fact, if you want to get a complete training on SSAS Tabular, you should spend two days with us in one of our SSAS Tabular Workshop! We prepared a 2-day seminar, a very intense one, that start from the simple tabular modeling and cover architecture, DAX, query, advanced modeling, security, deployment, optimization, monitoring, relationships with PowerPivot and Multidimensional… Really, there are a lot of stuffs here! We announced the first dates in Europe and also an online edition optimized for America’s time zone: Apr 16-17, 2012 – Amsterdam, Netherlands Apr 26-27, 2012 – Copenhagen, Denmark May 7-8, 2012 – Online for America’s time zone May 14-15, 2012 – Brussels, Belgium May 21-22, 2012 – Oslo, Norway May 24-25, 2012 – Stockholm, Sweden May 28-29, 2012 – London, United Kingdom May 31-Jun 1, 2012 – Milan, Italy (Italian language) Also Chris Webb will join us in this workshop and in every date you can find who is the speaker on the web site. The course is based on our upcoming book, almost 600 pages (!) about SSAS Tabular, an incredible effort that will be available very soon in a preview (rough cuts from O’Reilly) and will be on the shelf in May. I will provide a link to order it as soon as we have one! And if you think that this is not enough… you’re right! Do you know what is the only thing you can do to optimize your Tabular model? Optimize your DAX code. Learning DAX is easy, mastering DAX requires some knowledge… and our DAX Advanced Workshop will provide exactly the required content. Public classes will be available later this year, by now we just deliver it on demand.

  Read the article

• NoSQL Java API for MySQL Cluster: Questions & Answers

  - by Mat Keep

  The MySQL Cluster engineering team recently ran a live webinar, available now on-demand demonstrating the ClusterJ and ClusterJPA NoSQL APIs for MySQL Cluster, and how these can be used in building real-time, high scale Java-based services that require continuous availability. Attendees asked a number of great questions during the webinar, and I thought it would be useful to share those here, so others are also able to learn more about the Java NoSQL APIs. First, a little bit about why we developed these APIs and why they are interesting to Java developers. ClusterJ and Cluster JPA ClusterJ is a Java interface to MySQL Cluster that provides either a static or dynamic domain object model, similar to the data model used by JDO, JPA, and Hibernate. A simple API gives users extremely high performance for common operations: insert, delete, update, and query. ClusterJPA works with ClusterJ to extend functionality, including - Persistent classes - Relationships - Joins in queries - Lazy loading - Table and index creation from object model By eliminating data transformations via SQL, users get lower data access latency and higher throughput. In addition, Java developers have a more natural programming method to directly manage their data, with a complete, feature-rich solution for Object/Relational Mapping. As a result, the development of Java applications is simplified with faster development cycles resulting in accelerated time to market for new services. MySQL Cluster offers multiple NoSQL APIs alongside Java: - Memcached for a persistent, high performance, write-scalable Key/Value store, - HTTP/REST via an Apache module - C++ via the NDB API for the lowest absolute latency. Developers can use SQL as well as NoSQL APIs for access to the same data set via multiple query patterns – from simple Primary Key lookups or inserts to complex cross-shard JOINs using Adaptive Query Localization Marrying NoSQL and SQL access to an ACID-compliant database offers developers a number of benefits. MySQL Cluster’s distributed, shared-nothing architecture with auto-sharding and real time performance makes it a great fit for workloads requiring high volume OLTP. Users also get the added flexibility of being able to run real-time analytics across the same OLTP data set for real-time business insight. OK – hopefully you now have a better idea of why ClusterJ and JPA are available. Now, for the Q&A. Q & A Q. Why would I use Connector/J vs. ClusterJ? A. Partly it's a question of whether you prefer to work with SQL (Connector/J) or objects (ClusterJ). Performance of ClusterJ will be better as there is no need to pass through the MySQL Server. A ClusterJ operation can only act on a single table (e.g. no joins) - ClusterJPA extends that capability Q. Can I mix different APIs (ie ClusterJ, Connector/J) in our application for different query types? A. Yes. You can mix and match all of the API types, SQL, JDBC, ODBC, ClusterJ, Memcached, REST, C++. They all access the exact same data in the data nodes. Update through one API and new data is instantly visible to all of the others. Q. How many TCP connections would a SessionFactory instance create for a cluster of 8 data nodes? A. SessionFactory has a connection to the mgmd (management node) but otherwise is just a vehicle to create Sessions. Without using connection pooling, a SessionFactory will have one connection open with each data node. Using optional connection pooling allows multiple connections from the SessionFactory to increase throughput. Q. Can you give details of how Cluster J optimizes sharding to enhance performance of distributed query processing? A. Each data node in a cluster runs a Transaction Coordinator (TC), which begins and ends the transaction, but also serves as a resource to operate on the result rows. While an API node (such as a ClusterJ process) can send queries to any TC/data node, there are performance gains if the TC is where most of the result data is stored. ClusterJ computes the shard (partition) key to choose the data node where the row resides as the TC. Q. What happens if we perform two primary key lookups within the same transaction? Are they sent to the data node in one transaction? A. ClusterJ will send identical PK lookups to the same data node. Q. How is distributed query processing handled by MySQL Cluster ? A. If the data is split between data nodes then all of the information will be transparently combined and passed back to the application. The session will connect to a data node - typically by hashing the primary key - which then interacts with its neighboring nodes to collect the data needed to fulfil the query. Q. Can I use Foreign Keys with MySQL Cluster A. Support for Foreign Keys is included in the MySQL Cluster 7.3 Early Access release Summary The NoSQL Java APIs are packaged with MySQL Cluster, available for download here so feel free to take them for a spin today! Key Resources MySQL Cluster on-line demo  MySQL ClusterJ and JPA On-demand webinar  MySQL ClusterJ and JPA documentation MySQL ClusterJ and JPA whitepaper and tutorial

  Read the article

• Mocking HttpContext with JustMock

  - by mehfuzh

  In post , i will show a test code that will mock the various elements needed to complete a HTTP page request and  assert the expected page cycle steps. To begin, i have a simple enumeration that has my predefined page steps: public enum PageStep {     PreInit,     Load,     PreRender,     UnLoad } Once doing so, i  first created the page object [not mocking]. Page page = new Page(); Here, our target is to fire up the page process though ProcessRequest call, now if we take a look inside method though reflector, we will find calls stack like : ProcessRequest –> ProcessRequestWithNoAssert –> SetInstrinsics –> Finallly ProcessRequest. Inside SetIntrinsics , where it requires calls from HttpRequest, HttpResponse and HttpBrowserCababilities. With this , we can easily know what are classes / calls  we need to mock in order to get though the expected call. Accordingly, for  HttpBrowserCapabilities our required test code will look like: Mock.Arrange(() => browser.PreferredRenderingMime).Returns("text/html"); Mock.Arrange(() => browser.PreferredResponseEncoding).Returns("UTF-8"); Mock.Arrange(() => browser.PreferredRequestEncoding).Returns("UTF-8"); Now, HttpBrowserCapabilities is get though [Instance]HttpRequest.Browser. Therefore, we create the HttpRequest mock: var request = Mock.Create<HttpRequest>(); Then , add the required get call : Mock.Arrange(() => request.Browser).Returns(browser); As, [instance]Browser.PerferrredResponseEncoding and [instance]Browser.PreferredResponseEncoding  are also set to the request object and to make that they are set properly, we can add the following lines as well [not required though]. bool requestContentEncodingSet = false; Mock.ArrangeSet(() => request.ContentEncoding = Encoding.GetEncoding("UTF-8")).DoInstead(() =>  requestContentEncodingSet = true); Similarly, for response we can write:  var response = Mock.Create<HttpResponse>();    bool responseContentEncodingSet = false;  Mock.ArrangeSet(() => response.ContentEncoding = Encoding.GetEncoding("UTF-8")).DoInstead(() => responseContentEncodingSet = true); Finally , I created a mock of HttpContext and set the Request and Response properties that will returns the mocked version. var context = Mock.Create<HttpContext>();   Mock.Arrange(() => context.Request).Returns(request); Mock.Arrange(() => context.Response).Returns(response); As, Page internally calls RenderControl method , we just need to replace that with our one and optionally we can check if  invoked properly: bool rendered = false; Mock.Arrange(() => page.RenderControl(Arg.Any<HtmlTextWriter>())).DoInstead(() => rendered = true); That’s  it, the rest of the code is simple,  where  i asserted the page cycle with the PageSteps that i defined earlier: var pageSteps = new Queue<PageStep>();    page.PreInit +=      delegate      {          pageSteps.Enqueue(PageStep.PreInit);      };  page.Load +=      delegate      {          pageSteps.Enqueue(PageStep.Load);      };    page.PreRender +=      delegate      {          pageSteps.Enqueue(PageStep.PreRender);      };    page.Unload +=      delegate      {          pageSteps.Enqueue(PageStep.UnLoad);      };    page.ProcessRequest(context);    Assert.True(requestContentEncodingSet);  Assert.True(responseContentEncodingSet);  Assert.True(rendered);    Assert.Equal(pageSteps.Dequeue(), PageStep.PreInit);  Assert.Equal(pageSteps.Dequeue(), PageStep.Load);  Assert.Equal(pageSteps.Dequeue(), PageStep.PreRender);  Assert.Equal(pageSteps.Dequeue(), PageStep.UnLoad);    Mock.Assert(request);  Mock.Assert(response);   You can get the test class shown in this post here to give a try by yourself with of course JustMock. Enjoy!!

  Read the article

• In the Mobile and Tablet World, How Much is Too Much?

  - by andrewbrust

  The week of April 26th was a huge one in the world of mobile and tablet devices,  There were so many individual developments, announcements and solidifications of strategy, it’s almost impossible to believe they occurred in the same month, let alone the same week. Things started with Apple and Gizmodo having a Law and Order moment over the latter’s procurement of what appears to be the former’s 4th gen iPhone prototype.  We found out on the 26th that Gizmodo blogger Jason Chen’s apartment was raided by police and, honestly, that was a bit much. But Apple didn’t stop there.  They also published Steve Job’s critique of Adobe Flash and his explanation of Cupertino’s embargo of Flash on iPhones, iPods and iPads.  If you ask me, this too, was a bit much. Apple finished up the week by releasing the 3G version of its iPad product to the US market. I like (iLike?) my WiFi iPad.  The idea of getting a version of it that required a second 3G service monthly subscription, is, well, a bit  much. Microsoft was in the news too.  It killed a project it hadn’t even acknowledged the existence of: the Courier tablet.  That’s a bit much too.  If a tree falls in the woods, and Microsoft says they can’t hear it anyway, could they really have chopped it down? Maybe Microsoft Research should have licensed some of Courier’s technology from other parts of Microsoft.  Then maybe they could have kept the product alive.  Ask HTC: they’re going to be licensing technology from Microsoft because Redmond insists that Google’s Android operating system infringes on certain of their patents.  And since HTC now builds a number of handsets on Android, instead of being beholden, as they once were, to Windows Mobile, that means they can keep making their products.  Why does HTC have to pay the royalties, and not Google?  Maybe Microsoft decided that going after GOOG would have been a bit much, even for them. The agreement came not a moment to soon: HTC released their “Droid Incredible” (that name’s a bit much), an Android 2.1 handset with amazing hardware and HTC’s own Sense UI, on April 30th (this past Friday). This phone is very well-reviewed.  Maybe that’s why Google basically decided to beg off introducing a version of its Nexus One phone (also manufactured by HTC) on the Verizon Wireless network.  Google backing down?  That’s incredible, if not also a bit much. And that brings us to HP.  Which this week announced its acquisition of Palm and its webOS mobile phone touch-oriented operating system.  HP also killed its own Slate initiative.  Apparently HP realized that Windows 7, even with a proprietary HP touch UI added on top, is no match for the iPad.  I’m guessing they think webOS might work a bit better,  And I’m wondering if HP even wants to use webOS for phone handsets, beyond the Pre and Pixi.  Using it just for slate devices would be a bit extreme, but maybe not too much. Honestly, this was not Microsoft’s best week.  It killed a project and a close partner did likewise.  Then that same partner bought a competing OS product, while another partner released their new product that uses yet another competing OS platform. What did Microsoft actually produce this past week? An update to its Windows Phone 7 developer tools that actually works with the version of Visual Studio 2010 released on April 12th, and the version of Silverlight released three days later. That took three weeks to get synced up, and that’s a bit much too. But at least it happened. Windows Phone 7 is Microsoft’s best hope for a comeback in the SmartPhone market and to offer a credible touch-based tablet device.  This week, two of Microsoft’s slate initiatives died, and its only mobile phone victory was around its competitor’s operating system.  I hope the new platform gets Redmond out of the PC ghetto and into the classes of device that get people really excited today.  If it can’t, that would be a bit much; probably too much.  And, as the signs at the Lonestar Cafe in NYC used to say, too much ain’t enough.

  Read the article

• problem with loading in .FBX meshes in DirectX 10

  - by N0xus

  I'm trying to load in meshes into DirectX 10. I've created a bunch of classes that handle it and allow me to call in a mesh with only a single line of code in my main game class. How ever, when I run the program this is what renders: In the debug output window the following errors keep appearing: D3D10: ERROR: ID3D10Device::DrawIndexed: Input Assembler - Vertex Shader linkage error: Signatures between stages are incompatible. The reason is that Semantic 'TEXCOORD' is defined for mismatched hardware registers between the output stage and input stage. [ EXECUTION ERROR #343: DEVICE_SHADER_LINKAGE_REGISTERINDEX ] D3D10: ERROR: ID3D10Device::DrawIndexed: Input Assembler - Vertex Shader linkage error: Signatures between stages are incompatible. The reason is that the input stage requires Semantic/Index (POSITION,0) as input, but it is not provided by the output stage. [ EXECUTION ERROR #342: DEVICE_SHADER_LINKAGE_SEMANTICNAME_NOT_FOUND ] The thing is, I've no idea how to fix this. The code I'm using does work and I've simply brought all of that code into a new project of mine. There are no build errors and this only appears when the game is running The .fx file is as follows: float4x4 matWorld; float4x4 matView; float4x4 matProjection; struct VS_INPUT { float4 Pos:POSITION; float2 TexCoord:TEXCOORD; }; struct PS_INPUT { float4 Pos:SV_POSITION; float2 TexCoord:TEXCOORD; }; Texture2D diffuseTexture; SamplerState diffuseSampler { Filter = MIN_MAG_MIP_POINT; AddressU = WRAP; AddressV = WRAP; }; // // Vertex Shader // PS_INPUT VS( VS_INPUT input ) { PS_INPUT output=(PS_INPUT)0; float4x4 viewProjection=mul(matView,matProjection); float4x4 worldViewProjection=mul(matWorld,viewProjection); output.Pos=mul(input.Pos,worldViewProjection); output.TexCoord=input.TexCoord; return output; } // // Pixel Shader // float4 PS(PS_INPUT input ) : SV_Target { return diffuseTexture.Sample(diffuseSampler,input.TexCoord); //return float4(1.0f,1.0f,1.0f,1.0f); } RasterizerState NoCulling { FILLMODE=SOLID; CULLMODE=NONE; }; technique10 Render { pass P0 { SetVertexShader( CompileShader( vs_4_0, VS() ) ); SetGeometryShader( NULL ); SetPixelShader( CompileShader( ps_4_0, PS() ) ); SetRasterizerState(NoCulling); } } In my game, the .fx file and model are called and set as follows: Loading in shader file //Set the shader flags - BMD DWORD dwShaderFlags = D3D10_SHADER_ENABLE_STRICTNESS; #if defined( DEBUG ) || defined( _DEBUG ) dwShaderFlags |= D3D10_SHADER_DEBUG; #endif ID3D10Blob * pErrorBuffer=NULL; if( FAILED( D3DX10CreateEffectFromFile( TEXT("TransformedTexture.fx" ), NULL, NULL, "fx_4_0", dwShaderFlags, 0, md3dDevice, NULL, NULL, &m_pEffect, &pErrorBuffer, NULL ) ) ) { char * pErrorStr = ( char* )pErrorBuffer->GetBufferPointer(); //If the creation of the Effect fails then a message box will be shown MessageBoxA( NULL, pErrorStr, "Error", MB_OK ); return false; } //Get the technique called Render from the effect, we need this for rendering later on m_pTechnique=m_pEffect->GetTechniqueByName("Render"); //Number of elements in the layout UINT numElements = TexturedLitVertex::layoutSize; //Get the Pass description, we need this to bind the vertex to the pipeline D3D10_PASS_DESC PassDesc; m_pTechnique->GetPassByIndex( 0 )->GetDesc( &PassDesc ); //Create Input layout to describe the incoming buffer to the input assembler if (FAILED(md3dDevice->CreateInputLayout( TexturedLitVertex::layout, numElements,PassDesc.pIAInputSignature, PassDesc.IAInputSignatureSize, &m_pVertexLayout ) ) ) { return false; } model loading: m_pTestRenderable=new CRenderable(); //m_pTestRenderable->create<TexturedVertex>(md3dDevice,8,6,vertices,indices); m_pModelLoader = new CModelLoader(); m_pTestRenderable = m_pModelLoader->loadModelFromFile( md3dDevice,"armoredrecon.fbx" ); m_pGameObjectTest = new CGameObject(); m_pGameObjectTest->setRenderable( m_pTestRenderable ); // Set primitive topology, how are we going to interpet the vertices in the vertex buffer md3dDevice->IASetPrimitiveTopology( D3D10_PRIMITIVE_TOPOLOGY_TRIANGLELIST ); if ( FAILED( D3DX10CreateShaderResourceViewFromFile( md3dDevice, TEXT( "armoredrecon_diff.png" ), NULL, NULL, &m_pTextureShaderResource, NULL ) ) ) { MessageBox( NULL, TEXT( "Can't load Texture" ), TEXT( "Error" ), MB_OK ); return false; } m_pDiffuseTextureVariable = m_pEffect->GetVariableByName( "diffuseTexture" )->AsShaderResource(); m_pDiffuseTextureVariable->SetResource( m_pTextureShaderResource ); Finally, the draw function code: //All drawing will occur between the clear and present m_pViewMatrixVariable->SetMatrix( ( float* )m_matView ); m_pWorldMatrixVariable->SetMatrix( ( float* )m_pGameObjectTest->getWorld() ); //Get the stride(size) of the a vertex, we need this to tell the pipeline the size of one vertex UINT stride = m_pTestRenderable->getStride(); //The offset from start of the buffer to where our vertices are located UINT offset = m_pTestRenderable->getOffset(); ID3D10Buffer * pVB=m_pTestRenderable->getVB(); //Bind the vertex buffer to input assembler stage - md3dDevice->IASetVertexBuffers( 0, 1, &pVB, &stride, &offset ); md3dDevice->IASetIndexBuffer( m_pTestRenderable->getIB(), DXGI_FORMAT_R32_UINT, 0 ); //Get the Description of the technique, we need this in order to loop through each pass in the technique D3D10_TECHNIQUE_DESC techDesc; m_pTechnique->GetDesc( &techDesc ); //Loop through the passes in the technique for( UINT p = 0; p < techDesc.Passes; ++p ) { //Get a pass at current index and apply it m_pTechnique->GetPassByIndex( p )->Apply( 0 ); //Draw call md3dDevice->DrawIndexed(m_pTestRenderable->getNumOfIndices(),0,0); //m_pD3D10Device->Draw(m_pTestRenderable->getNumOfVerts(),0); } Is there anything I've clearly done wrong or are missing? Spent 2 weeks trying to workout what on earth I've done wrong to no avail. Any insight a fresh pair eyes could give on this would be great.

  Read the article

• About Me

  - by Jeffrey West

  I’m new to blogging.  This is the second blog post that I have written, and before I go too much further I wanted the readers of my blog to know a bit more about me… Kid’s Stuff By trade, I am a programmer (or coder, developer, engineer, architect, etc).  I started programming when I was 12 years old.  When I was 7, we got our first ‘family’ computer – an Apple IIc.  It was great to play games on, and of course what else was a 7-year-old going to do with it.  I did have one problem with it, though.  When I put in my 5.25” floppy to play a game, sometimes, instead loading my game I would get a mysterious ‘]’ on the screen with a flashing cursor.  This, of course, was not my game.  Much like the standard ‘Microsoft fix’ is to reboot, back then you would take the floppy out, shake it, and restart the computer and pray for a different result. One day, I learned at school that I could topple my nemesis – the ‘]’ and flashing cursor – by typing ‘load’ and pressing enter.  Most of the time, this would load my game and then I would get to play.  Problem solved.  However, I began to wonder – what else can I make it do? When I was in 5th grade my dad got a bright idea to buy me a Tandy 1000HX.  He didn’t know what I was going to do with it, and neither did I.  Least of all, my mom wasn’t happy about buying a 5th grader a $1,000 computer.  Nonetheless, Over time, I learned how to write simple basic programs out of the back of my Math book: 10 x=5 20 y=6 30 PRINT x+y That was fun for all of about 5 minutes.  I needed more – more challenges, more things that I could make the computer do.  In order to quench this thirst my parents sent me to National Computer Camps in Connecticut.  It was one of the best experiences of my childhood, and I spent 3 weeks each summer after that learning BASIC, Pascal, Turbo C and some C++.  There weren’t many kids at the time who knew anything about computers, and lets just say my knowledge of and interest in computers didn’t score me many ‘cool’ points.  My experiences at NCC set me on the path that I find myself on now, and I am very thankful for the experience.  Real Life I have held various positions in the past at different levels within the IT layer cake.  I started out as a Software Developer for a startup in the Dallas, TX area building software for semiconductor testing statistical process control and sampling.  I was the second Java developer that was hired, and the ninth employee overall, so I got a great deal of experience developing software.  Since there weren’t that many people in the organization, I also got a lot of field experience which meant that if I screwed up the code, I got yelled at (figuratively) by both my boss AND the customer.  Fun Times!  What made it better was that I got to help run pilot programs in Taiwan, Singapore, Malaysia and Malta.  Getting yelled at in Taiwan is slightly less annoying that getting yelled at in Dallas… I spent the next 5 years at Accenture doing systems integration in the ‘SOA’ group.  I joined as a Consultant and left as a Senior Manager.  I started out writing code in WebLogic Integration and left after I wrapped up project where I led a team of 25 to develop the next generation of a digital media platform to deliver HD content in a digital format.  At Accenture, I had the pleasure of working with some truly amazing people – mentoring some and learning from many others – and on some incredible real-world IT projects.  Given my background with the BEA stack of products I was often called in to troubleshoot and tune WebLogic, ALBPM and ALSB installations and have logged many hours digging through thread dumps, running performance tests with SoapUI and decompiling Java classes we didn’t have the source for so I could see what was going on in the code. I am now a Senior Principal Product Manager at Oracle in the Application Grid practice.  The term ‘Application Grid’ refers to a collection of software and hardware products within Oracle that enables customers to build horizontally scalable systems.  This collection of products includes WebLogic, GlassFish, Coherence, Tuxedo and the JRockit/HotSpot JVMs (HotSprocket, maybe?).  Now, with the introduction of Exalogic it has grown to include hardware as well. Wrapping it up… I love technology and have a diverse background ranging from software development to HW and network architecture & tuning.  I have held certifications for being an Oracle Certified DBA, MSCE and Cisco Certified Network Professional (CCNP), among others and I have put those to great use over my career.  I am excited about programming & technology and I enjoy helping people learn and be successful.  If you are having challenges with WebLogic, BPM or Service Bus feel free to reach out to me and I’ll be happy to help as I have time. Thanks for stopping by!   --Jeff

  Read the article

• C# async and actors

  - by Alex.Davies

  If you read my last post about async, you might be wondering what drove me to write such odd code in the first place. The short answer is that .NET Demon is written using NAct Actors. Actors are an old idea, which I believe deserve a renaissance under C# 5. The idea is to isolate each stateful object so that only one thread has access to its state at any point in time. That much should be familiar, it's equivalent to traditional lock-based synchronization. The different part is that actors pass "messages" to each other rather than calling a method and waiting for it to return. By doing that, each thread can only ever be holding one lock. This completely eliminates deadlocks, my least favourite concurrency problem. Most people who use actors take this quite literally, and there are plenty of frameworks which help you to create message classes and loops which can receive the messages, inspect what type of message they are, and process them accordingly. But I write C# for a reason. Do I really have to choose between using actors and everything I love about object orientation in C#? Type safety Interfaces Inheritance Generics As it turns out, no. You don't need to choose between messages and method calls. A method call makes a perfectly good message, as long as you don't wait for it to return. This is where asynchonous methods come in. I have used NAct for a while to wrap my objects in a proxy layer. As long as I followed the rule that methods must always return void, NAct queued up the call for later, and immediately released my thread. When I needed to get information out of other actors, I could use EventHandlers and callbacks (continuation passing style, for any CS geeks reading), and NAct would call me back in my isolated thread without blocking the actor that raised the event. Using callbacks looks horrible though. To remind you: m_BuildControl.FilterEnabledForBuilding(    projects,    enabledProjects = m_OutOfDateProjectFinder.FilterNeedsBuilding(        enabledProjects,             newDirtyProjects =             {                 ....... Which is why I'm really happy that NAct now supports async methods. Now, methods are allowed to return Task rather than just void. I can await those methods, and C# 5 will turn the rest of my method into a continuation for me. NAct will run the other method in the other actor's context, but will make sure that when my method resumes, we're back in my context. Neither actor was ever blocked waiting for the other one. Apart from when they were actually busy doing something, they were responsive to concurrent messages from other sources. To be fair, you could use async methods with lock statements to achieve exactly the same thing, but it's ugly. Here's a realistic example of an object that has a queue of data that gets passed to another object to be processed: class QueueProcessor {    private readonly ItemProcessor m_ItemProcessor = ...     private readonly object m_Sync = new object();    private Queue<object> m_DataQueue = ...    private List<object> m_Results = ...     public async Task ProcessOne() {         object data = null;         lock (m_Sync)         {             data = m_DataQueue.Dequeue();         }         var processedData = await m_ItemProcessor.ProcessData(data); lock (m_Sync)         {             m_Results.Add(processedData);         }     } } We needed to write two lock blocks, one to get the data to process, one to store the result. The worrying part is how easily we could have forgotten one of the locks. Compare that to the version using NAct: class QueueProcessorActor : IActor { private readonly ItemProcessor m_ItemProcessor = ... private Queue<object> m_DataQueue = ... private List<object> m_Results = ... public async Task ProcessOne()     {         // We are an actor, it's always thread-safe to access our private fields         var data = m_DataQueue.Dequeue();         var processedData = await m_ItemProcessor.ProcessData(data);         m_Results.Add(processedData);     } } You don't have to explicitly lock anywhere, NAct ensures that your code will only ever run on one thread, because it's an actor. Either way, async is definitely better than traditional synchronous code. Here's a diagram of what a typical synchronous implementation might do: The left side shows what is running on the thread that has the lock required to access the QueueProcessor's data. The red section is where that lock is held, but doesn't need to be. Contrast that with the async version we wrote above: Here, the lock is released in the middle. The QueueProcessor is free to do something else. Most importantly, even if the ItemProcessor sometimes calls the QueueProcessor, they can never deadlock waiting for each other. So I thoroughly recommend you use async for all code that has to wait a while for things. And if you find yourself writing lots of lock statements, think about using actors as well. Using actors and async together really takes the misery out of concurrent programming.

  Read the article

• Goto for the Java Programming Language

  - by darcy

  Work on JDK 8 is well-underway, but we thought this late-breaking JEP for another language change for the platform couldn't wait another day before being published. Title: Goto for the Java Programming Language Author: Joseph D. Darcy Organization: Oracle. Created: 2012/04/01 Type: Feature State: Funded Exposure: Open Component: core/lang Scope: SE JSR: 901 MR Discussion: compiler dash dev at openjdk dot java dot net Start: 2012/Q2 Effort: XS Duration: S Template: 1.0 Reviewed-by: Duke Endorsed-by: Edsger Dijkstra Funded-by: Blue Sun Corporation Summary Provide the benefits of the time-testing goto control structure to Java programs. The Java language has a history of adding new control structures over time, the assert statement in 1.4, the enhanced for-loop in 1.5,and try-with-resources in 7. Having support for goto is long-overdue and simple to implement since the JVM already has goto instructions. Success Metrics The goto statement will allow inefficient and verbose recursive algorithms and explicit loops to be replaced with more compact code. The effort will be a success if at least twenty five percent of the JDK's explicit loops are replaced with goto's. Coordination with IDE vendors is expected to help facilitate this goal. Motivation The goto construct offers numerous benefits to the Java platform, from increased expressiveness, to more compact code, to providing new programming paradigms to appeal to a broader demographic. In JDK 8, there is a renewed focus on using the Java platform on embedded devices with more modest resources than desktop or server environments. In such contexts, static and dynamic memory footprint is a concern. One significant component of footprint is the code attribute of class files and certain classes of important algorithms can be expressed more compactly using goto than using other constructs, saving footprint. For example, to implement state machines recursively, some parties have asked for the JVM to support tail calls, that is, to perform a complex transformation with security implications to turn a method call into a goto. Such complicated machinery should not be assumed for an embedded context. A better solution is just to expose to the programmer the desired functionality, goto. The web has familiarized users with a model of traversing links among different HTML pages in a free-form fashion with some state being maintained on the side, such as login credentials, to effect behavior. This is exactly the programming model of goto and code. While in the past this has been derided as leading to "spaghetti code," spaghetti is a tasty and nutritious meal for programmers, unlike quiche. The invokedynamic instruction added by JSR 292 exposes the JVM's linkage operation to programmers. This is a low-level operation that can be leveraged by sophisticated programmers. Likewise, goto is a also a low-level operation that should not be hidden from programmers who can use more efficient idioms. Some may object that goto was consciously excluded from the original design of Java as one of the removed feature from C and C++. However, the designers of the Java programming languages have revisited these removals before. The enum construct was also left out only to be added in JDK 5 and multiple inheritance was left out, only to be added back by the virtual extension method methods of Project Lambda. As a living language, the needs of the growing Java community today should be used to judge what features are needed in the platform tomorrow; the language should not be forever bound by the decisions of the past. Description From its initial version, the JVM has had two instructions for unconditional transfer of control within a method, goto (0xa7) and goto_w (0xc8). The goto_w instruction is used for larger jumps. All versions of the Java language have supported labeled statements; however, only the break and continue statements were able to specify a particular label as a target with the onerous restriction that the label must be lexically enclosing. The grammar addition for the goto statement is: GotoStatement: goto Identifier ; The new goto statement similar to break except that the target label can be anywhere inside the method and the identifier is mandatory. The compiler simply translates the goto statement into one of the JVM goto instructions targeting the right offset in the method. Therefore, adding the goto statement to the platform is only a small effort since existing compiler and JVM functionality is reused. Other language changes to support goto include obvious updates to definite assignment analysis, reachability analysis, and exception analysis. Possible future extensions include a computed goto as found in gcc, which would replace the identifier in the goto statement with an expression having the type of a label. Testing Since goto will be implemented using largely existing facilities, only light levels of testing are needed. Impact Compatibility: Since goto is already a keyword, there are no source compatibility implications. Performance/scalability: Performance will improve with more compact code. JVMs already need to handle irreducible flow graphs since goto is a VM instruction.

  Read the article

• Playing with http page cycle using JustMock

  - by mehfuzh

  In this post , I will cover a test code that will mock the various elements needed to complete a HTTP page request and  assert the expected page cycle steps. To begin, i have a simple enumeration that has my predefined page steps: public enum PageStep {     PreInit,     Load,     PreRender,     UnLoad } Once doing so, i  first created the page object [not mocking]. Page page = new Page(); Here, our target is to fire up the page process through ProcessRequest call, now if we take a look inside the method with reflector.net,  the call trace will go like : ProcessRequest –> ProcessRequestWithNoAssert –> SetInstrinsics –> Finallly ProcessRequest. Inside SetInstrinsics ,  it requires calls from HttpRequest, HttpResponse and HttpBrowserCababilities. With this clue at hand, we can easily know the classes / calls  we need to mock in order to get through the expected call. Accordingly, for  HttpBrowserCapabilities our required test code will look like: Mock.Arrange(() => browser.PreferredRenderingMime).Returns("text/html"); Mock.Arrange(() => browser.PreferredResponseEncoding).Returns("UTF-8"); Mock.Arrange(() => browser.PreferredRequestEncoding).Returns("UTF-8"); Now, HttpBrowserCapabilities is get though [Instance]HttpRequest.Browser. Therefore, we create the HttpRequest mock: var request = Mock.Create<HttpRequest>(); Then , add the required get call : Mock.Arrange(() => request.Browser).Returns(browser); As, [instance]Browser.PerferrredResponseEncoding and [instance]Browser.PreferredResponseEncoding  are also set to the request object and to make that they are set properly, we can add the following lines as well [not required though]. bool requestContentEncodingSet = false; Mock.ArrangeSet(() => request.ContentEncoding = Encoding.GetEncoding("UTF-8")).DoInstead(() =>  requestContentEncodingSet = true); Similarly, for response we can write:  var response = Mock.Create<HttpResponse>();    bool responseContentEncodingSet = false;  Mock.ArrangeSet(() => response.ContentEncoding = Encoding.GetEncoding("UTF-8")).DoInstead(() => responseContentEncodingSet = true); Finally , I created a mock of HttpContext and set the Request and Response properties that will returns the mocked version. var context = Mock.Create<HttpContext>();   Mock.Arrange(() => context.Request).Returns(request); Mock.Arrange(() => context.Response).Returns(response); As, Page internally calls RenderControl method , we just need to replace that with our one and optionally we can check if  invoked properly: bool rendered = false; Mock.Arrange(() => page.RenderControl(Arg.Any<HtmlTextWriter>())).DoInstead(() => rendered = true); That’s  it, the rest of the code is simple,  where  i asserted the page cycle with the PageSteps that i defined earlier: var pageSteps = new Queue<PageStep>();   page.PreInit +=      delegate      {          pageSteps.Enqueue(PageStep.PreInit);      }; page.Load +=      delegate      {          pageSteps.Enqueue(PageStep.Load);      };   page.PreRender +=      delegate      {          pageSteps.Enqueue(PageStep.PreRender);      };   page.Unload +=      delegate      {          pageSteps.Enqueue(PageStep.UnLoad);      };   page.ProcessRequest(context);    Assert.True(requestContentEncodingSet);  Assert.True(responseContentEncodingSet);  Assert.True(rendered);    Assert.Equal(pageSteps.Dequeue(), PageStep.PreInit);  Assert.Equal(pageSteps.Dequeue(), PageStep.Load);  Assert.Equal(pageSteps.Dequeue(), PageStep.PreRender);  Assert.Equal(pageSteps.Dequeue(), PageStep.UnLoad);    Mock.Assert(request);  Mock.Assert(response);   You can get the test class shown in this post here to give a try by yourself with of course JustMock :-).   Enjoy!!

  Read the article

• Implementing synchronous MediaTypeFormatters in ASP.NET Web API

  - by cibrax

  One of main characteristics of MediaTypeFormatter’s in ASP.NET Web API is that they leverage the Task Parallel Library (TPL) for reading or writing an model into an stream. When you derive your class from the base class MediaTypeFormatter, you have to either implement the WriteToStreamAsync or ReadFromStreamAsync methods for writing or reading a model from a stream respectively. These two methods return a Task, which internally does all the serialization work, as it is illustrated bellow. public abstract class MediaTypeFormatter { public virtual Task WriteToStreamAsync(Type type, object value, Stream writeStream, HttpContent content, TransportContext transportContext); public virtual Task<object> ReadFromStreamAsync(Type type, Stream readStream, HttpContent content, IFormatterLogger formatterLogger); }   .csharpcode, .csharpcode pre { font-size: small; color: black; font-family: consolas, "Courier New", courier, monospace; background-color: #ffffff; /*white-space: pre;*/ } .csharpcode pre { margin: 0em; } .csharpcode .rem { color: #008000; } .csharpcode .kwrd { color: #0000ff; } .csharpcode .str { color: #006080; } .csharpcode .op { color: #0000c0; } .csharpcode .preproc { color: #cc6633; } .csharpcode .asp { background-color: #ffff00; } .csharpcode .html { color: #800000; } .csharpcode .attr { color: #ff0000; } .csharpcode .alt { background-color: #f4f4f4; width: 100%; margin: 0em; } .csharpcode .lnum { color: #606060; } .csharpcode, .csharpcode pre { font-size: small; color: black; font-family: consolas, "Courier New", courier, monospace; background-color: #ffffff; /*white-space: pre;*/ } .csharpcode pre { margin: 0em; } .csharpcode .rem { color: #008000; } .csharpcode .kwrd { color: #0000ff; } .csharpcode .str { color: #006080; } .csharpcode .op { color: #0000c0; } .csharpcode .preproc { color: #cc6633; } .csharpcode .asp { background-color: #ffff00; } .csharpcode .html { color: #800000; } .csharpcode .attr { color: #ff0000; } .csharpcode .alt { background-color: #f4f4f4; width: 100%; margin: 0em; } .csharpcode .lnum { color: #606060; } However, most of the times, serialization is a safe operation that can be done synchronously. In fact, many of the serializer classes you will find in the .NET framework only provide sync methods. So the question is, how you can transform that synchronous work into a Task ?. Creating a new task using the method Task.Factory.StartNew for doing all the serialization work would be probably the typical answer. That would work, as a new task is going to be scheduled. However, that might involve some unnecessary context switches, which are out of our control and might be affect performance on server code specially.   If you take a look at the source code of the MediaTypeFormatters shipped as part of the framework, you will notice that they actually using another pattern, which uses a TaskCompletionSource class. public Task WriteToStreamAsync(Type type, object value, Stream writeStream, HttpContent content, TransportContext transportContext) {   var tsc = new TaskCompletionSource<AsyncVoid>(); tsc.SetResult(default(AsyncVoid));   //Do all the serialization work here synchronously   return tsc.Task; }   /// <summary> /// Used as the T in a "conversion" of a Task into a Task{T} /// </summary> private struct AsyncVoid { } .csharpcode, .csharpcode pre { font-size: small; color: black; font-family: consolas, "Courier New", courier, monospace; background-color: #ffffff; /*white-space: pre;*/ } .csharpcode pre { margin: 0em; } .csharpcode .rem { color: #008000; } .csharpcode .kwrd { color: #0000ff; } .csharpcode .str { color: #006080; } .csharpcode .op { color: #0000c0; } .csharpcode .preproc { color: #cc6633; } .csharpcode .asp { background-color: #ffff00; } .csharpcode .html { color: #800000; } .csharpcode .attr { color: #ff0000; } .csharpcode .alt { background-color: #f4f4f4; width: 100%; margin: 0em; } .csharpcode .lnum { color: #606060; } They are basically doing all the serialization work synchronously and using a TaskCompletionSource for returning a task already done. To conclude this post, this is another approach you might want to consider when using serializers that are not compatible with an async model. Update: Henrik Nielsen from the ASP.NET team pointed out the existence of a built-in media type formatter for writing sync formatters. BufferedMediaTypeFormatter http://t.co/FxOfeI5x

  Read the article

• How to build a Singleton-like dependency injector replacement (Php)

  - by Erparom

  I know out there are a lot of excelent containers, even frameworks almost entirely DI based with good strong IoC classes. However, this doesn't help me to "define" a new pattern. (This is Php code but understandable to anyone) Supose we have: //Declares the singleton class bookSingleton { private $author; private static $bookInstance; private static $isLoaned = FALSE; //The private constructor private function __constructor() { $this->author = "Onecrappy Writer Ofcheap Novels"; } //Sets the global isLoaned state and also gets self instance public static function loanBook() { if (self::$isLoaned === FALSE) { //Book already taken, so return false return FALSE; } else { //Ok, not loaned, lets instantiate (if needed and loan) if (!isset(self::$bookInstance)) { self::$bookInstance = new BookSingleton(); } self::$isLoaned = TRUE; } } //Return loaned state to false, so another book reader can take the book public function returnBook() { $self::$isLoaned = FALSE; } public function getAuthor() { return $this->author; } } Then we get the singelton consumtion class: //Consumes the Singleton class BookBorrower() { private $borrowedBook; private $haveBookState; public function __construct() { this->haveBookState = FALSE; } //Use the singelton-pattern behavior public function borrowBook() { $this->borrowedBook = BookSingleton::loanBook(); //Check if was successfully borrowed if (!this->borrowedBook) { $this->haveBookState = FALSE; } else { $this->haveBookState = TRUE; } } public function returnBook() { $this->borrowedBook->returnBook(); $this->haveBookState = FALSE; } public function getBook() { if ($this->haveBookState) { return "The book is loaned, the author is" . $this->borrowedbook->getAuthor(); } else { return "I don't have the book, perhaps someone else took it"; } } } At last, we got a client, to test the behavior function __autoload($class) { require_once $class . '.php'; } function write ($whatever,$breaks) { for($break = 0;$break<$breaks;$break++) { $whatever .= "\n"; } echo nl2br($whatever); } write("Begin Singleton test", 2); $borrowerJuan = new BookBorrower(); $borrowerPedro = new BookBorrower(); write("Juan asks for the book", 1); $borrowerJuan->borrowBook(); write("Book Borrowed? ", 1); write($borrowerJuan->getAuthorAndTitle(),2); write("Pedro asks for the book", 1); $borrowerPedro->borrowBook(); write("Book Borrowed? ", 1); write($borrowerPedro->getAuthorAndTitle(),2); write("Juan returns the book", 1); $borrowerJuan->returnBook(); write("Returned Book Juan? ", 1); write($borrowerJuan->getAuthorAndTitle(),2); write("Pedro asks again for the book", 1); $borrowerPedro->borrowBook(); write("Book Borrowed? ", 1); write($borrowerPedro->getAuthorAndTitle(),2); This will end up in the expected behavior: Begin Singleton test Juan asks for the book Book Borrowed? The book is loaned, the author is = Onecrappy Writer Ofcheap Novels Pedro asks for the book Book Borrowed? I don't have the book, perhaps someone else took it Juan returns the book Returned Book Juan? I don't have the book, perhaps someone else took it Pedro asks again for the book Book Borrowed? The book is loaned, the author is = Onecrappy Writer Ofcheap Novels So I want to make a pattern based on the DI technique able to do exactly the same, but without singleton pattern. As far as I'm aware, I KNOW I must inject the book inside "borrowBook" function instead of taking a static instance: public function borrowBook(BookNonSingleton $book) { if (isset($this->borrowedBook) || $book->isLoaned()) { $this->haveBook = FALSE; return FALSE; } else { $this->borrowedBook = $book; $this->haveBook = TRUE; return TRUE; } } And at the client, just handle the book: $borrowerJuan = new BookBorrower(); $borrowerJuan-borrowBook(new NonSingletonBook()); Etc... and so far so good, BUT... Im taking the responsability of "single instance" to the borrower, instead of keeping that responsability inside the NonSingletonBook, that since it has not anymore a private constructor, can be instantiated as many times... making instances on each call. So, What does my NonSingletonBook class MUST be in order to never allow borrowers to have this same book twice? (aka) keep the single instance. Because the dependency injector part of the code (borrower) does not solve me this AT ALL. Is it needed the container with an "asShared" method builder with static behavior? No way to encapsulate this functionallity into the Book itself? "Hey Im a book and I shouldn't be instantiated more than once, I'm unique"

  Read the article

• Integrating Windows Form Click Once Application into SharePoint 2007 &ndash; Part 2 of 4

  - by Kelly Jones

  In my last post, I explained why we decided to use a Click Once application to solve our business problem. To quickly review, we needed a way for our business users to upload documents to a SharePoint 2007 document library in mass, set the meta data, set the permissions per document, and to do so easily. Let’s look at the pieces that make up our solution.  First, we have the Windows Form application.  This app is deployed using Click Once and calls SharePoint web services in order to upload files and then calls web services to set the meta data (SharePoint columns and permissions).  Second, we have a custom action.  The custom action is responsible for providing our users a link that will launch the Windows app, as well as passing values to it via the query string.  And lastly, we have the web services that the Windows Form application calls.  For our solution, we used both out of the box web services and a custom web service in order to set the column values in the document library as well as the permissions on the documents. Now, let’s look at the technical details of each of these pieces.  (All of the code is downloadable from here: )   Windows Form application deployed via Click Once The Windows Form application, called “Custom Upload”, has just a few classes in it: Custom Upload -- the form FileList.xsd -- the dataset used to track the names of the files and their meta data values SharePointUpload -- this class handles uploading the file SharePointUpload uses an HttpWebRequest to transfer the file to the web server. We had to change this code from a WebClient object to the HttpWebRequest object, because we needed to be able to set the time out value.  public bool UploadDocument(string localFilename, string remoteFilename) { bool result = true; //Need to use an HttpWebRequest object instead of a WebClient object // so we can set the timeout (WebClient doesn't allow you to set the timeout!) HttpWebRequest req = (HttpWebRequest)WebRequest.Create(remoteFilename); try { req.Method = "PUT"; req.Timeout = 60 * 1000; //convert seconds to milliseconds req.AllowWriteStreamBuffering = true; req.Credentials = System.Net.CredentialCache.DefaultCredentials; req.SendChunked = false; req.KeepAlive = true; Stream reqStream = req.GetRequestStream(); FileStream rdr = new FileStream(localFilename, FileMode.Open, FileAccess.Read); byte[] inData = new byte[4096]; int bytesRead = rdr.Read(inData, 0, inData.Length); while (bytesRead > 0) { reqStream.Write(inData, 0, bytesRead); bytesRead = rdr.Read(inData, 0, inData.Length); } reqStream.Close(); rdr.Close(); System.Net.HttpWebResponse response = (HttpWebResponse)req.GetResponse(); if (response.StatusCode != HttpStatusCode.OK && response.StatusCode != HttpStatusCode.Created) { String msg = String.Format("An error occurred while uploading this file: {0}\n\nError response code: {1}", System.IO.Path.GetFileName(localFilename), response.StatusCode.ToString()); LogWarning(msg, "2ACFFCCA-59BA-40c8-A9AB-05FA3331D223"); result = false; } } catch (Exception ex) { LogException(ex, "{E9D62A93-D298-470d-A6BA-19AAB237978A}"); result = false; } return result; } The class also contains the LogException() and LogWarning() methods. When the application is launched, it parses the query string for some initial values.  The query string looks like this: string queryString = "Srv=clickonce&Sec=N&Doc=DMI&SiteName=&Speed=128000&Max=50"; This Srv is the path to the server (my Virtual Machine is name “clickonce”), the Sec is short for security – meaning HTTPS or HTTP, the Doc is the shortcut for which document library to use, and SiteName is the name of the SharePoint site.  Speed is used to calculate an estimate for download speed for each file.  We added this so our users uploading documents would realize how long it might take for clients in remote locations (using slow WAN connections) to download the documents. The last value, Max, is the maximum size that the SharePoint site will allow documents to be.  This allowed us to give users a warning that a file is too large before we even attempt to upload it. Another critical piece is the meta data collection.  We organized our site using SharePoint content types, so when the app loads, it gets a list of the document library’s content types.  The user then select one of the content types from the drop down list, and then we query SharePoint to get a list of the fields that make up that content type.  We used both an out of the box web service, and one that we custom built, in order to get these values. Once we have the content type fields, we then add controls to the form.  Which type of control we add depends on the data type of the field.  (DateTime pickers for date/time fields, etc)  We didn’t write code to cover every data type, since we were working with a limited set of content types and field data types. Here’s a screen shot of the Form, before and after someone has selected the content types and our code has added the custom controls:     The other piece of meta data we collect is the in the upper right corner of the app, “Users with access”.  This box lists the different SharePoint Groups that we have set up and by checking the boxes, the user can set the permissions on the uploaded documents. All of this meta data is collected and submitted to our custom web service, which then sets the values on the documents on the list.  We’ll look at these web services in a future post. In the next post, we’ll walk through the Custom Action we built.

  Read the article

• Why enumerator structs are a really bad idea

  - by Simon Cooper

  If you've ever poked around the .NET class libraries in Reflector, I'm sure you would have noticed that the generic collection classes all have implementations of their IEnumerator as a struct rather than a class. As you will see, this design decision has some rather unfortunate side effects... As is generally known in the .NET world, mutable structs are a Very Bad Idea; and there are several other blogs around explaining this (Eric Lippert's blog post explains the problem quite well). In the BCL, the generic collection enumerators are all mutable structs, as they need to keep track of where they are in the collection. This bit me quite hard when I was coding a wrapper around a LinkedList<int>.Enumerator. It boils down to this code: sealed class EnumeratorWrapper : IEnumerator<int> { private readonly LinkedList<int>.Enumerator m_Enumerator; public EnumeratorWrapper(LinkedList<int> linkedList) { m_Enumerator = linkedList.GetEnumerator(); } public int Current { get { return m_Enumerator.Current; } } object System.Collections.IEnumerator.Current { get { return Current; } } public bool MoveNext() { return m_Enumerator.MoveNext(); } public void Reset() { ((System.Collections.IEnumerator)m_Enumerator).Reset(); } public void Dispose() { m_Enumerator.Dispose(); } } The key line here is the MoveNext method. When I initially coded this, I thought that the call to m_Enumerator.MoveNext() would alter the enumerator state in the m_Enumerator class variable and so the enumeration would proceed in an orderly fashion through the collection. However, when I ran this code it went into an infinite loop - the m_Enumerator.MoveNext() call wasn't actually changing the state in the m_Enumerator variable at all, and my code was looping forever on the first collection element. It was only after disassembling that method that I found out what was going on The MoveNext method above results in the following IL: .method public hidebysig newslot virtual final instance bool MoveNext() cil managed { .maxstack 1 .locals init ( [0] bool CS$1$0000, [1] valuetype [System]System.Collections.Generic.LinkedList`1/Enumerator CS$0$0001) L_0000: nop L_0001: ldarg.0 L_0002: ldfld valuetype [System]System.Collections.Generic.LinkedList`1/Enumerator EnumeratorWrapper::m_Enumerator L_0007: stloc.1 L_0008: ldloca.s CS$0$0001 L_000a: call instance bool [System]System.Collections.Generic.LinkedList`1/Enumerator::MoveNext() L_000f: stloc.0 L_0010: br.s L_0012 L_0012: ldloc.0 L_0013: ret } Here, the important line is 0002 - m_Enumerator is accessed using the ldfld operator, which does the following: Finds the value of a field in the object whose reference is currently on the evaluation stack. So, what the MoveNext method is doing is the following: public bool MoveNext() { LinkedList<int>.Enumerator CS$0$0001 = this.m_Enumerator; bool CS$1$0000 = CS$0$0001.MoveNext(); return CS$1$0000; } The enumerator instance being modified by the call to MoveNext is the one stored in the CS$0$0001 variable on the stack, and not the one in the EnumeratorWrapper class instance. Hence why the state of m_Enumerator wasn't getting updated. Hmm, ok. Well, why is it doing this? If you have a read of Eric Lippert's blog post about this issue, you'll notice he quotes a few sections of the C# spec. In particular, 7.5.4: ...if the field is readonly and the reference occurs outside an instance constructor of the class in which the field is declared, then the result is a value, namely the value of the field I in the object referenced by E. And my m_Enumerator field is readonly! Indeed, if I remove the readonly from the class variable then the problem goes away, and the code works as expected. The IL confirms this: .method public hidebysig newslot virtual final instance bool MoveNext() cil managed { .maxstack 1 .locals init ( [0] bool CS$1$0000) L_0000: nop L_0001: ldarg.0 L_0002: ldflda valuetype [System]System.Collections.Generic.LinkedList`1/Enumerator EnumeratorWrapper::m_Enumerator L_0007: call instance bool [System]System.Collections.Generic.LinkedList`1/Enumerator::MoveNext() L_000c: stloc.0 L_000d: br.s L_000f L_000f: ldloc.0 L_0010: ret } Notice on line 0002, instead of the ldfld we had before, we've got a ldflda, which does this: Finds the address of a field in the object whose reference is currently on the evaluation stack. Instead of loading the value, we're loading the address of the m_Enumerator field. So now the call to MoveNext modifies the enumerator stored in the class rather than on the stack, and everything works as expected. Previously, I had thought enumerator structs were an odd but interesting feature of the BCL that I had used in the past to do linked list slices. However, effects like this only underline how dangerous mutable structs are, and I'm at a loss to explain why the enumerators were implemented as structs in the first place. (interestingly, the SortedList<TKey, TValue> enumerator is a struct but is private, which makes it even more odd - the only way it can be accessed is as a boxed IEnumerator!). I would love to hear people's theories as to why the enumerators are implemented in such a fashion. And bonus points if you can explain why LinkedList<int>.Enumerator.Reset is an explicit implementation but Dispose is implicit... Note to self: never ever ever code a mutable struct.

  Read the article

• Some non-generic collections

  - by Simon Cooper

  Although the collections classes introduced in .NET 2, 3.5 and 4 cover most scenarios, there are still some .NET 1 collections that don't have generic counterparts. In this post, I'll be examining what they do, why you might use them, and some things you'll need to bear in mind when doing so. BitArray System.Collections.BitArray is conceptually the same as a List<bool>, but whereas List<bool> stores each boolean in a single byte (as that's what the backing bool[] does), BitArray uses a single bit to store each value, and uses various bitmasks to access each bit individually. This means that BitArray is eight times smaller than a List<bool>. Furthermore, BitArray has some useful functions for bitmasks, like And, Xor and Not, and it's not limited to 32 or 64 bits; a BitArray can hold as many bits as you need. However, it's not all roses and kittens. There are some fundamental limitations you have to bear in mind when using BitArray: It's a non-generic collection. The enumerator returns object (a boxed boolean), rather than an unboxed bool. This means that if you do this: foreach (bool b in bitArray) { ... } Every single boolean value will be boxed, then unboxed. And if you do this: foreach (var b in bitArray) { ... } you'll have to manually unbox b on every iteration, as it'll come out of the enumerator an object. Instead, you should manually iterate over the collection using a for loop: for (int i=0; i<bitArray.Length; i++) { bool b = bitArray[i]; ... } Following on from that, if you want to use BitArray in the context of an IEnumerable<bool>, ICollection<bool> or IList<bool>, you'll need to write a wrapper class, or use the Enumerable.Cast<bool> extension method (although Cast would box and unbox every value you get out of it). There is no Add or Remove method. You specify the number of bits you need in the constructor, and that's what you get. You can change the length yourself using the Length property setter though. It doesn't implement IList. Although not really important if you're writing a generic wrapper around it, it is something to bear in mind if you're using it with pre-generic code. However, if you use BitArray carefully, it can provide significant gains over a List<bool> for functionality and efficiency of space. OrderedDictionary System.Collections.Specialized.OrderedDictionary does exactly what you would expect - it's an IDictionary that maintains items in the order they are added. It does this by storing key/value pairs in a Hashtable (to get O(1) key lookup) and an ArrayList (to maintain the order). You can access values by key or index, and insert or remove items at a particular index. The enumerator returns items in index order. However, the Keys and Values properties return ICollection, not IList, as you might expect; CopyTo doesn't maintain the same ordering, as it copies from the backing Hashtable, not ArrayList; and any operations that insert or remove items from the middle of the collection are O(n), just like a normal list. In short; don't use this class. If you need some sort of ordered dictionary, it would be better to write your own generic dictionary combining a Dictionary<TKey, TValue> and List<KeyValuePair<TKey, TValue>> or List<TKey> for your specific situation. ListDictionary and HybridDictionary To look at why you might want to use ListDictionary or HybridDictionary, we need to examine the performance of these dictionaries compared to Hashtable and Dictionary<object, object>. For this test, I added n items to each collection, then randomly accessed n/2 items: So, what's going on here? Well, ListDictionary is implemented as a linked list of key/value pairs; all operations on the dictionary require an O(n) search through the list. However, for small n, the constant factor that big-o notation doesn't measure is much lower than the hashing overhead of Hashtable or Dictionary. HybridDictionary combines a Hashtable and ListDictionary; for small n, it uses a backing ListDictionary, but switches to a Hashtable when it gets to 9 items (you can see the point it switches from a ListDictionary to Hashtable in the graph). Apart from that, it's got very similar performance to Hashtable. So why would you want to use either of these? In short, you wouldn't. Any gain in performance by using ListDictionary over Dictionary<TKey, TValue> would be offset by the generic dictionary not having to cast or box the items you store, something the graphs above don't measure. Only if the performance of the dictionary is vital, the dictionary will hold less than 30 items, and you don't need type safety, would you use ListDictionary over the generic Dictionary. And even then, there's probably more useful performance gains you can make elsewhere.

  Read the article

• problems with my slotgame [delphi]

  - by Raiden2k

  hey guys im coding at the moment on a slotgame for the learning effect. here is the source code. my questions are below: unit Unit1; {$mode objfpc}{$H+} interface uses Classes, SysUtils, Windows, FileUtil, Forms, Controls, Graphics, Dialogs, StdCtrls, ExtCtrls, ComCtrls, Menus, ActnList, Spin, FileCtrl; type { TForm1 } TForm1 = class(TForm) FloatSpinEdit1: TFloatSpinEdit; Guthabenlb: TLabel; s4: TLabel; s5: TLabel; s6: TLabel; s7: TLabel; s8: TLabel; s9: TLabel; Timer3: TTimer; Winlb: TLabel; Loselb: TLabel; slotbn: TButton; s1: TLabel; s2: TLabel; s3: TLabel; Timer1: TTimer; Timer2: TTimer; procedure FormCreate(Sender: TObject); procedure slotbnClick(Sender: TObject); procedure Timer1Timer(Sender: TObject); procedure Timer2Timer(Sender: TObject); procedure Timer3Timer(Sender: TObject); private { private declarations } FRollen : array [0..2, 0..9] of String; public { public declarations } end; var Form1: TForm1; wins,loses : Integer; guthaben : Double = 10; implementation {$R *.lfm} { TForm1 } procedure TForm1.slotbnClick(Sender: TObject); begin Guthaben := Guthaben - 1.00; Guthabenlb.Caption := FloatToStr(guthaben) + (' €'); Timer1.Enabled := True; Timer2.Enabled := True; slotbn.Enabled := false; end; procedure TForm1.FormCreate(Sender: TObject); var i: integer; j: integer; n: integer; digits: TStringlist; begin Digits := TStringList.Create; try for i := low(FRollen) to high(FRollen) do begin for j := low(FRollen[i]) to high(FRollen[i]) do Digits.Add(IntToStr(j)); for j := low(FRollen[i]) to high(FRollen[i]) do begin n := Random(Digits.Count); FRollen[i, j] := Digits[n]; Digits.Delete(n); end; end finally Digits.Free; end; for i:=low(FRollen) to high(FRollen) do begin end; end; //==================================================================================================\\ // Drehen der Slots im Zufallsmodus //==================================================================================================// procedure TForm1.Timer1Timer(Sender: TObject); begin s1.Caption := IntToStr(Random(9)); s2.Caption := IntToStr(Random(9)); s3.Caption := IntToStr(Random(9)); s4.Caption := IntToStr(Random(9)); s5.Caption := IntToStr(Random(9)); s6.Caption := IntToStr(Random(9)); s7.Caption := IntToStr(Random(9)); s8.Caption := IntToStr(Random(9)); s9.Caption := IntToStr(Random(9)); end; //==================================================================================================// //===================================================================================================\\ // Gewonnen / Verloren abfrage //===================================================================================================// procedure TForm1.Timer2Timer(Sender: TObject); begin Timer1.Enabled := False; Timer2.Enabled := false; if (s1.Caption = s5.Caption) and (s1.Caption = s9.Caption) then begin Guthaben := Guthaben + 5.00; Inc(wins); end else if (s1.Caption = s4.Caption) and (s1.Caption = s7.Caption) then begin Guthaben := Guthaben + 5.00; Inc(wins); end else if (s2.Caption = s5.Caption) and (s2.Caption = s8.Caption) then begin Guthaben := Guthaben + 5.00; Inc(wins); end else if (s3.Caption = s6.Caption) and (s3.Caption = s9.Caption) then begin Guthaben := Guthaben + 5.00; Inc(wins); end else if (s3.Caption = s5.Caption) and (s3.Caption = s7.Caption) then begin Guthaben := Guthaben + 5.00; Inc(wins); end else Inc(loses); slotbn.Enabled := True; Loselb.Caption := 'Loses: ' + IntToStr(loses); Winlb.Caption := 'Wins: ' + IntTostr(Wins); end; procedure TForm1.Timer3Timer(Sender: TObject); begin if (guthaben = 0) or (guthaben < 0) then begin Timer3.Enabled := False; MessageBox(handle,'Du hast verloren!','Verlierer!',MB_OK); close(); end; end; //======================================================================================================\\ end. How can i replace the labels through icons 16 x 16 pixels? How can i adjust the winning sum according to the icons.(for example 3 crowns give you 40 € and 3 apples only 10 €) How can i adhust the winning sum with a sum for every round?

  Read the article

• Functional Adaptation

  - by Charles Courchaine

  In real life and OO programming we’re often faced with using adapters, DVI to VGA, 1/4” to 1/8” audio connections, 110V to 220V, wrapping an incompatible interface with a new one, and so on.  Where the adapter pattern is generally considered for interfaces and classes a similar technique can be applied to method signatures.  To be fair, this adaptation is generally used to reduce the number of parameters but I’m sure there are other clever possibilities to be had.  As Jan questioned in the last post, how can we use a common method to execute an action if the action has a differing number of parameters, going back to the greeting example it was suggested having an AddName method that takes a first and last name as parameters.  This is exactly what we’ll address in this post. Let’s set the stage with some review and some code changes.  First, our method that handles the setup/tear-down infrastructure for our WCF service: 1: private static TResult ExecuteGreetingFunc<TResult>(Func<IGreeting, TResult> theGreetingFunc) 2: { 3: IGreeting aGreetingService = null; 4: try 5: { 6: aGreetingService = GetGreetingChannel(); 7: return theGreetingFunc(aGreetingService); 8: } 9: finally 10: { 11: CloseWCFChannel((IChannel)aGreetingService); 12: } 13: } Our original AddName method: 1: private static string AddName(string theName) 2: { 3: return ExecuteGreetingFunc<string>(theGreetingService => theGreetingService.AddName(theName)); 4: } Our new AddName method: 1: private static int AddName(string firstName, string lastName) 2: { 3: return ExecuteGreetingFunc<int>(theGreetingService => theGreetingService.AddName(firstName, lastName)); 4: } Let’s change the AddName method, just a little bit more for this example and have it take the greeting service as a parameter. 1: private static int AddName(IGreeting greetingService, string firstName, string lastName) 2: { 3: return greetingService.AddName(firstName, lastName); 4: } The new signature of AddName using the Func delegate is now Func<IGreeting, string, string, int>, which can’t be used with ExecuteGreetingFunc as is because it expects Func<IGreeting, TResult>.  Somehow we have to eliminate the two string parameters before we can use this with our existing method.  This is where we need to adapt AddName to match what ExecuteGreetingFunc expects, and we’ll do so in the following progression. 1: Func<IGreeting, string, string, int> -> Func<IGreeting, string, int> 2: Func<IGreeting, string, int> -> Func<IGreeting, int>   For the first step, we’ll create a method using the lambda syntax that will “eliminate” the last name parameter: 1: string lastNameToAdd = "Smith"; 2: //Func<IGreeting, string, string, int> -> Func<IGreeting, string, int> 3: Func<IGreeting, string, int> addName = (greetingService, firstName) => AddName(greetingService, firstName, lastNameToAdd); The new addName method gets us one step close to the signature we need.  Let’s say we’re going to call this in a loop to add several names, we’ll take the final step from Func<IGreeting, string, int> -> Func<IGreeting, int> in line as a lambda passed to ExecuteGreetingFunc like so: 1: List<string> firstNames = new List<string>() { "Bob", "John" }; 2: int aID; 3: foreach (string firstName in firstNames) 4: { 5: //Func<IGreeting, string, int> -> Func<IGreeting, int> 6: aID = ExecuteGreetingFunc<int>(greetingService => addName(greetingService, firstName)); 7: Console.WriteLine(GetGreeting(aID)); 8: } If for some reason you needed to break out the lambda on line 6 you could replace it with 1: aID = ExecuteGreetingFunc<int>(ApplyAddName(addName, firstName)); and use this method: 1: private static Func<IGreeting, int> ApplyAddName(Func<IGreeting, string, int> addName, string lastName) 2: { 3: return greetingService => addName(greetingService, lastName); 4: } Splitting out a lambda into its own method is useful both in this style of coding as well as LINQ queries to improve the debugging experience.  It is not strictly necessary to break apart the steps & functions as was shown above; the lambda in line 6 (of the foreach example) could include both the last name and first name instead of being composed of two functions.  The process demonstrated above is one of partially applying functions, this could have also been done with Currying (also see Dustin Campbell’s excellent post on Currying for the canonical curried add example).  Matthew Podwysocki also has some good posts explaining both Currying and partial application and a follow up post that further clarifies the difference between Currying and partial application.  In either technique the ultimate goal is to reduce the number of parameters passed to a function.  Currying makes it a single parameter passed at each step, where partial application allows one to use multiple parameters at a time as we’ve done here.  This technique isn’t for everyone or every problem, but can be extremely handy when you need to adapt a call to something you don’t control.

  Read the article

• When does "proper" programming no longer matter?

  - by Kai Qing

  I've been a full time programmer for about 8 years now. Web based mostly, ranging in weird jobs for clients. Never anything I "want" to do. So my experience is limited to what I've been contracted to do, having no real incentive to master anything in particular. So here's my scenario and ultimately what I wonder about... I've been building an android game in my spare time. It's using the libgdx library so quite a bit of the heavy lifting is done for me. I don't read much of the docs cause unless it's in tutorial format I will just not care, and ultimately most of my questions have already been asked on stackoverflow. I get along fine and my game works as expected... Suspiciously well, even. So much so that I wonder why one should bother to be "proper" when coding if the end result is ultimately the same. To be more specific, I used a hashtable because I wanted something close to an associative array. Human readable key values. In other places to achieve similar things, I use a vector. I know libgdx has vector2 and vector3 classes, but I've never used them. When I come across weird problems and search stackoverflow for help, I see a lot of people just reaming the questions that use a certain datatype when another one is technically "proper." Like using an ArrayList because it does not require defined bounds versus re-defining an int[] with new known boundaries. Or even something trivial like this: for(int i = 0; i < items.length; i ++) { // do something } I know it evaluates item.length on every iteration. I just don't care. I know items will never be more than 15 to 20 items. So why bother caring if I evaluate items.length on every iteration? So I wonder - why does everyone get all up in arms over this? Who cares if I use a less efficient datatype to get the job done? I ran some tests to see how the app performs using the lazy, get it done fast and don't look back method I just described versus the proper, follow the tutorial and use the exact data types suggested by the community. The results: Same thing. Average 45 fps. I opened every app on the phone and galaxy tab. Same deal. No difference. My game is pretty graphic intensive. It's not like it's just a simple thing. I expected it to perform kind of badly since I don't care to optimize image assets or... well, you probably get the idea. I'm making the game for fun. As a joke, really. But in doing so I'm working outside the normal scope of my job, which is to always follow the rules and do it the right way. So to say, I am without bounds here and this has caused me to wonder why I ever really care to be "proper" So I guess my question to you is this: Is there a threshold when it no longer matters to be proper? Is there a lasting, longer term consequence to the lazy, get it done and don't look back route? Is it ok to say - "so long as it gets the job done, I don't care?" Disclaimer: When I program my game, I am almost always drunk. I do it to remember why I got into this stuff to begin with because the monotony of client based web work will make you hate being a programmer. I'm having a blast and my game is not crashing, tests well, performs well, looks good on all devices so far and has no noticeable negative impact on any of my testing devices. I expected failure because I was being so drunkenly careless with my code, but to my surprise, it had no noticeable impact. I am now starting to question the need to be careful. Help me regain the ability to care! ... or explain why it's not a bad thing to not care. Secondary disclaimer: I am aware of the benefits of maintainability. For myself and others. Agreed. But it's not like someone happening across my inefficient int[] loop won't know what it does. As an experienced programmer those kinds of things are just clear on sight. I document the complex stuff for myself knowing I was drunk and will probably need a reminder. Those notes would clarify any confusion for someone who might ever gaze upon my ridiculous game - though the reality is that either I maintain it myself or it fades into time. I'm ok with that. But if it doesn't slow the device down, or crash, then crossing the t's and dotting the i's might actually require more time than it's worth.

  Read the article

• C#/.NET Little Wonders: Static Char Methods

  - by James Michael Hare

  Once again, in this series of posts I look at the parts of the .NET Framework that may seem trivial, but can help improve your code by making it easier to write and maintain. The index of all my past little wonders posts can be found here. Often times in our code we deal with the bigger classes and types in the BCL, and occasionally forgot that there are some nice methods on the primitive types as well.  Today we will discuss some of the handy static methods that exist on the char (the C# alias of System.Char) type. The Background I was examining a piece of code this week where I saw the following: 1: // need to get the 5th (offset 4) character in upper case 2: var type = symbol.Substring(4, 1).ToUpper(); 3:  4: // test to see if the type is P 5: if (type == "P") 6: { 7: // ... do something with P type... 8: } Is there really any error in this code?  No, but it still struck me wrong because it is allocating two very short-lived throw-away strings, just to store and manipulate a single char: The call to Substring() generates a new string of length 1 The call to ToUpper() generates a new upper-case version of the string from Step 1. In my mind this is similar to using ToUpper() to do a case-insensitive compare: it isn’t wrong, it’s just much heavier than it needs to be (for more info on case-insensitive compares, see #2 in 5 More Little Wonders). One of my favorite books is the C++ Coding Standards: 101 Rules, Guidelines, and Best Practices by Sutter and Alexandrescu.  True, it’s about C++ standards, but there’s also some great general programming advice in there, including two rules I love:         8. Don’t Optimize Prematurely         9. Don’t Pessimize Prematurely We all know what #8 means: don’t optimize when there is no immediate need, especially at the expense of readability and maintainability.  I firmly believe this and in the axiom: it’s easier to make correct code fast than to make fast code correct.  Optimizing code to the point that it becomes difficult to maintain often gains little and often gives you little bang for the buck. But what about #9?  Well, for that they state: “All other things being equal, notably code complexity and readability, certain efficient design patterns and coding idioms should just flow naturally from your fingertips and are no harder to write then the pessimized alternatives. This is not premature optimization; it is avoiding gratuitous pessimization.” Or, if I may paraphrase: “where it doesn’t increase the code complexity and readability, prefer the more efficient option”. The example code above was one of those times I feel where we are violating a tacit C# coding idiom: avoid creating unnecessary temporary strings.  The code creates temporary strings to hold one char, which is just unnecessary.  I think the original coder thought he had to do this because ToUpper() is an instance method on string but not on char.  What he didn’t know, however, is that ToUpper() does exist on char, it’s just a static method instead (though you could write an extension method to make it look instance-ish). This leads me (in a long-winded way) to my Little Wonders for the day… Static Methods of System.Char So let’s look at some of these handy, and often overlooked, static methods on the char type: IsDigit(), IsLetter(), IsLetterOrDigit(), IsPunctuation(), IsWhiteSpace() Methods to tell you whether a char (or position in a string) belongs to a category of characters. IsLower(), IsUpper() Methods that check if a char (or position in a string) is lower or upper case ToLower(), ToUpper() Methods that convert a single char to the lower or upper equivalent. For example, if you wanted to see if a string contained any lower case characters, you could do the following: 1: if (symbol.Any(c => char.IsLower(c))) 2: { 3: // ... 4: } Which, incidentally, we could use a method group to shorten the expression to: 1: if (symbol.Any(char.IsLower)) 2: { 3: // ... 4: } Or, if you wanted to verify that all of the characters in a string are digits: 1: if (symbol.All(char.IsDigit)) 2: { 3: // ... 4: } Also, for the IsXxx() methods, there are overloads that take either a char, or a string and an index, this means that these two calls are logically identical: 1: // check given a character 2: if (char.IsUpper(symbol[0])) { ... } 3:  4: // check given a string and index 5: if (char.IsUpper(symbol, 0)) { ... } Obviously, if you just have a char, then you’d just use the first form.  But if you have a string you can use either form equally well. As a side note, care should be taken when examining all the available static methods on the System.Char type, as some seem to be redundant but actually have very different purposes.  For example, there are IsDigit() and IsNumeric() methods, which sound the same on the surface, but give you different results. IsDigit() returns true if it is a base-10 digit character (‘0’, ‘1’, … ‘9’) where IsNumeric() returns true if it’s any numeric character including the characters for ½, ¼, etc. Summary To come full circle back to our opening example, I would have preferred the code be written like this: 1: // grab 5th char and take upper case version of it 2: var type = char.ToUpper(symbol[4]); 3:  4: if (type == 'P') 5: { 6: // ... do something with P type... 7: } Not only is it just as readable (if not more so), but it performs over 3x faster on my machine:    1,000,000 iterations of char method took: 30 ms, 0.000050 ms/item.    1,000,000 iterations of string method took: 101 ms, 0.000101 ms/item. It’s not only immediately faster because we don’t allocate temporary strings, but as an added bonus there less garbage to collect later as well.  To me this qualifies as a case where we are using a common C# performance idiom (don’t create unnecessary temporary strings) to make our code better. Technorati Tags: C#,CSharp,.NET,Little Wonders,char,string

  Read the article

• Subterranean IL: Generics and array covariance

  - by Simon Cooper

  Arrays in .NET are curious beasts. They are the only built-in collection types in the CLR, and SZ-arrays (single dimension, zero-indexed) have their own commands and IL syntax. One of their stranger properties is they have a kind of built-in covariance long before generic variance was added in .NET 4. However, this causes a subtle but important problem with generics. First of all, we need to briefly recap on array covariance. SZ-array covariance To demonstrate, I'll tweak the classes I introduced in my previous posts: public class IncrementableClass { public int Value; public virtual void Increment(int incrementBy) { Value += incrementBy; } } public class IncrementableClassx2 : IncrementableClass { public override void Increment(int incrementBy) { base.Increment(incrementBy); base.Increment(incrementBy); } } In the CLR, SZ-arrays of reference types are implicitly convertible to arrays of the element's supertypes, all the way up to object (note that this does not apply to value types). That is, an instance of IncrementableClassx2[] can be used wherever a IncrementableClass[] or object[] is required. When an SZ-array could be used in this fashion, a run-time type check is performed when you try to insert an object into the array to make sure you're not trying to insert an instance of IncrementableClass into an IncrementableClassx2[]. This check means that the following code will compile fine but will fail at run-time: IncrementableClass[] array = new IncrementableClassx2[1]; array[0] = new IncrementableClass(); // throws ArrayTypeMismatchException These checks are enforced by the various stelem* and ldelem* il instructions in such a way as to ensure you can't insert a IncrementableClass into a IncrementableClassx2[]. For the rest of this post, however, I'm going to concentrate on the ldelema instruction. ldelema This instruction pops the array index (int32) and array reference (O) off the stack, and pushes a pointer (&) to the corresponding array element. However, unlike the ldelem instruction, the instruction's type argument must match the run-time array type exactly. This is because, once you've got a managed pointer, you can use that pointer to both load and store values in that array element using the ldind* and stind* (load/store indirect) instructions. As the same pointer can be used for both input and output to the array, the type argument to ldelema must be invariant. At the time, this was a perfectly reasonable restriction, and maintained array type-safety within managed code. However, along came generics, and with it the constrained callvirt instruction. So, what happens when we combine array covariance and constrained callvirt? .method public static void CallIncrementArrayValue() { // IncrementableClassx2[] arr = new IncrementableClassx2[1] ldc.i4.1 newarr IncrementableClassx2 // arr[0] = new IncrementableClassx2(); dup newobj instance void IncrementableClassx2::.ctor() ldc.i4.0 stelem.ref // IncrementArrayValue<IncrementableClass>(arr, 0) // here, we're treating an IncrementableClassx2[] as IncrementableClass[] dup ldc.i4.0 call void IncrementArrayValue<class IncrementableClass>(!!0[],int32) // ... ret } .method public static void IncrementArrayValue<(IncrementableClass) T>( !!T[] arr, int32 index) { // arr[index].Increment(1) ldarg.0 ldarg.1 ldelema !!T ldc.i4.1 constrained. !!T callvirt instance void IIncrementable::Increment(int32) ret } And the result: Unhandled Exception: System.ArrayTypeMismatchException: Attempted to access an element as a type incompatible with the array. at IncrementArrayValue[T](T[] arr, Int32 index) at CallIncrementArrayValue() Hmm. We're instantiating the generic method as IncrementArrayValue<IncrementableClass>, but passing in an IncrementableClassx2[], hence the ldelema instruction is failing as it's expecting an IncrementableClass[]. On features and feature conflicts What we've got here is a conflict between existing behaviour (ldelema ensuring type safety on covariant arrays) and new behaviour (managed pointers to object references used for every constrained callvirt on generic type instances). And, although this is an edge case, there is no general workaround. The generic method could be hidden behind several layers of assemblies, wrappers and interfaces that make it a requirement to use array covariance when calling the generic method. Furthermore, this will only fail at runtime, whereas compile-time safety is what generics were designed for! The solution is the readonly. prefix instruction. This modifies the ldelema instruction to ignore the exact type check for arrays of reference types, and so it lets us take the address of array elements using a covariant type to the actual run-time type of the array: .method public static void IncrementArrayValue<(IncrementableClass) T>( !!T[] arr, int32 index) { // arr[index].Increment(1) ldarg.0 ldarg.1 readonly. ldelema !!T ldc.i4.1 constrained. !!T callvirt instance void IIncrementable::Increment(int32) ret } But what about type safety? In return for ignoring the type check, the resulting controlled mutability pointer can only be used in the following situations: As the object parameter to ldfld, ldflda, stfld, call and constrained callvirt instructions As the pointer parameter to ldobj or ldind* As the source parameter to cpobj In other words, the only operations allowed are those that read from the pointer; stind* and similar that alter the pointer itself are banned. This ensures that the array element we're pointing to won't be changed to anything untoward, and so type safety within the array is maintained. This is a typical example of the maxim that whenever you add a feature to a program, you have to consider how that feature interacts with every single one of the existing features. Although an edge case, the readonly. prefix instruction ensures that generics and array covariance work together and that compile-time type safety is maintained. Tune in next time for a look at the .ctor generic type constraint, and what it means.

  Read the article

< Previous Page | 343 344 345 346 347 348 349 350 351 352 353 354  | Next Page >