Search Results

Search found 1506 results on 61 pages for 'ben racicot'.

Page 8/61 | < Previous Page | 4 5 6 7 8 9 10 11 12 13 14 15  | Next Page >

  • Writing a Book, and Moving my Blog

    - by Ben Nevarez
    I started blogging about SQL Server here at SQLblog back in July, 2009 and it was a lot of fun, I enjoyed it a lot. Then later, after a series of blog posts about the Query Optimizer, I was invited to write an entire book about that same topic. But after a few months I realized that it was going to be hard to continue both blogging and writing chapters for a book, this in addition to my regular day job, so I decided to stop blogging for a little while.   Now that I have finished the last chapter of the book and I am working on the final chapter reviews, I decided to start blogging again. This time I am moving my blog to   http://www.benjaminnevarez.com   Same as my previous posts I plan to write about my topics of interest, like the relational engine, and basically anything related to SQL Server. Hopefully you find my new blog interesting and useful.   Finally, I would like to thank Adam for allowing me to blog here. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • jQuery Context Menu Plugin and Capturing Right-Click

    - by Ben Griswold
    I was thrilled to find Cory LaViska’s jQuery Context Menu Plugin a few months ago. In very little time, I was able to integrate the context menu with the jQuery Treeview.  I quickly had a really pretty user interface which took full advantage of limited real estate.  And guess what.  As promised, the plugin worked in Chrome, Safari 3, IE 6/7/8, Firefox 2/3 and Opera 9.5.  Everything was perfect and I shipped to the Integration Environment. One thing kept bugging though – right clicks aren’t the standard in a web environment. Sure, when one hovers over the treeview node, the mouse changed from an arrow to a pointer, but without help text most users will certainly left-click rather than right. As I was already doubting the design decision, we did some Mac testing.  The context menu worked in Firefox but not Safari.  Damn.  That’s when I started digging into the Madness of Javascript Mouse Events.  Don’t tell, but it’s complicated.  About as close as one can get to capture the right-click mouse event on all major browsers on Windows and Mac is this: if (event.which == null) /* IE case */ button= (event.button < 2) ? "LEFT" : ((event.button == 4) ? "MIDDLE" : "RIGHT"); else /* All others */ button= (event.which < 2) ? "LEFT" : ((event.which == 2) ? "MIDDLE" : "RIGHT"); Yikes.  The content menu code was simply checking if event.button == 2.  No problem.  Cory offers a jQuery Right Click Plugin which I’m sure works for windows but probably not the Mac either.  (Please note I haven’t verified this.) Anyway, I decided to address my UI design concern and the Safari Mac issue in one swoop.  I decided to make the context menu respond to any mouse click event.  This didn’t take much – especially after seeing how Bill Beckelman updated the library to recognize the left click. First, I added an AnyClick option to the library defaults: // Any click may trigger the dropdown and that's okay // See Javascript Madness: Mouse Events – http: //unixpapa.com/js/mouse.html if (o.anyClick == undefined) o.anyClick = false; And then I trigger the context menu dropdown based on the following conditional: if (evt.button == 2 || o.anyClick) { Nothing tricky about that, right?  Finally, I updated my menu setup to include the AnyClick value, if true: $('.member').contextMenu({ menu: 'memberContextMenu', anyClick: true },             function (action, el, pos) {                 … Now the context menu works in “all” environments if you left, right or even middle click.  Download jQuery Context Menu Plugin for Any Click *Opera 9.5 has an option to allow scripts to detect right-clicks, but it is disabled by default. Furthermore, Opera still doesn’t allow JavaScript to disable the browser’s default context menu which causes a usability conflict.

    Read the article

  • Installing SubText with Web PI

    - by Ben Griswold
    SubText is the engine behind our company blog. With the goal of ensuring a smooth transition between the main website and the blogs, I spent some time tightening up the styles for the aggregate and individual blogs last week.  This required a custom SubText skin and lot of css tweaking. Though I’ve previously had the SubText source running on my machine, there was no need to update or rebuild the solution in my current case so just went ahead with a local installation using the Microsoft Web Platform Installer (Web PI).  I just checked the SubText box, provided answers to a few key setup questions (admin user credentials, SubText database, etc) and I was up and running in minutes.   Once the setup was complete, I was asked if I’d like to launch SubText.  The SubText Installation Wizard picked up where Web PI left off and the setup couldn’t have been easier.  Web PI provides quick and easy installs for lots of goodies.  Check it out.

    Read the article

  • Continuous Integration for SQL Server Part II – Integration Testing

    - by Ben Rees
    My previous post, on setting up Continuous Integration for SQL Server databases using GitHub, Bamboo and Red Gate’s tools, covered the first two parts of a simple Database Continuous Delivery process: Putting your database in to a source control system, and, Running a continuous integration process, each time changes are checked in. However there is, of course, a lot more to to Continuous Delivery than that. Specifically, in addition to the above: Putting some actual integration tests in to the CI process (otherwise, they don’t really do much, do they!?), Deploying the database changes with a managed, automated approach, Monitoring what you’ve just put live, to make sure you haven’t broken anything. This post will detail how to set up a very simple pipeline for implementing the first of these (continuous integration testing). NB: A lot of the setup in this post is built on top of the configuration from before, so it might be difficult to implement this post without running through part I first. There’ll then be a third post on automated database deployment followed by a final post dealing with the last item – monitoring changes on the live system. In the previous post, I used a mixture of Red Gate products and other 3rd party software – GitHub and Atlassian Bamboo specifically. This was partly because I believe most people work in an heterogeneous environment, using software from different vendors to suit their purposes and I wanted to show how this could work for this process. For example, you could easily substitute Atlassian’s BitBucket or Stash for GitHub, depending on your needs, or use an alternative CI server such as TeamCity, TFS or Jenkins. However, in this, post, I’ll be mostly using Red Gate products only (other than tSQLt). I would do this, firstly because I work for Red Gate. However, I also think that in the area of Database Delivery processes, nobody else has the offerings to implement this process fully – so I didn’t have any choice!   Background on Continuous Delivery For me, a great source of information on what makes a proper Continuous Delivery process is the Jez Humble and David Farley classic: Continuous Delivery – Reliable Software Releases through Build, Test, and Deployment Automation This book is not of course, primarily about databases, and the process I outline here and in the previous article is a gross simplification of what Jez and David describe (not least because it’s that much harder for databases!). However, a lot of the principles that they describe can be equally applied to database development and, I would argue, should be. As I say however, what I describe here is a very simple version of what would be required for a full production process. A couple of useful resources on handling some of these complexities can be found in the following two references: Refactoring Databases – Evolutionary Database Design, by Scott J Ambler and Pramod J. Sadalage Versioning Databases – Branching and Merging, by Scott Allen In particular, I don’t deal at all with the issues of multiple branches and merging of those branches, an issue made particularly acute by the use of GitHub. The other point worth making is that, in the words of Martin Fowler: Continuous Delivery is about keeping your application in a state where it is always able to deploy into production.   I.e. we are not talking about continuously delivery updates to the production database every time someone checks in an amendment to a stored procedure. That is possible (and what Martin calls Continuous Deployment). However, again, that’s more than I describe in this article. And I doubt I need to remind DBAs or Developers to Proceed with Caution!   Integration Testing Back to something practical. The next stage, building on our set up from the previous article, is to add in some integration tests to the process. As I say, the CI process, though interesting, isn’t enormously useful without some sort of test process running. For this we’ll use the tSQLt framework, an open source framework designed specifically for running SQL Server tests. tSQLt is part of Red Gate’s SQL Test found on http://www.red-gate.com/products/sql-development/sql-test/ or can be downloaded separately from www.tsqlt.org - though I’ll provide a step-by-step guide below for setting this up. Getting tSQLt set up via SQL Test Click on the link http://www.red-gate.com/products/sql-development/sql-test/ and click on the blue Download button to download the Red Gate SQL Test product, if not already installed. Follow the install process for SQL Test to install the SQL Server Management Studio (SSMS) plugin on to your machine, if not already installed. Open SSMS. You should now see SQL Test under the Tools menu:   Clicking this link will give you the basic SQL Test dialogue: As yet, though we’ve installed the SQL Test product we haven’t yet installed the tSQLt test framework on to any particular database. To do this, we need to add our RedGateApp database using this dialogue, by clicking on the + Add Database to SQL Test… link, selecting the RedGateApp database and clicking the Add Database link:   In the next screen, SQL Test describes what will be installed on the database for the tSQLt framework. Also in this dialogue, uncheck the “Add SQL Cop tests” option (shown below). SQL Cop is a great set of pre-defined tests that work within the tSQLt framework to check the general health of your SQL Server database. However, we won’t be using them in this particular simple example: Once you’ve clicked on the OK button, the changes described in the dialogue will be made to your database. Some of these are shown in the left-hand-side below: We’ve now installed the framework. However, we haven’t actually created any tests, so this will be the next step. But, before we proceed, we’ve made an update to our database so should, again check this in to source control, adding comments as required:   Also worth a quick check that your build still runs with the new additions!: (And a quick check of the RedGateAppCI database shows that the changes have been made).   Creating and Testing a Unit Test There are, of course, a lot of very interesting unit tests that you could and should set up for a database. The great thing about the tSQLt framework is that you can write these in SQL. The example I’m going to use here is pretty Mickey Mouse – our database table is going to include some email addresses as reference data and I want to check whether these are all in a correct email format. Nothing clever but it illustrates the process and hopefully shows the method by which more interesting tests could be set up. Adding Reference Data to our Database To start, I want to add some reference data to my database, and have this source controlled (as well as the schema). First of all I need to add some data in to my solitary table – this can be done a number of ways, but I’ll do this in SSMS for simplicity: I then add some reference data to my table: Currently this reference data just exists in the database. For proper integration testing, this needs to form part of the source-controlled version of the database – and so needs to be added to the Git repository. This can be done via SQL Source Control, though first a Primary Key needs to be added to the table. Right click the table, select Design, then right-click on the first “id” row. Then click on “Set Primary Key”: NB: once this change is made, click Save to save the change to the table. Then, to source control this reference data, right click on the table (dbo.Email) and selecting the following option:   In the next screen, link the data in the Email table, by selecting it from the list and clicking “save and close”: We should at this point re-commit the changes (both the addition of the Primary Key, and the data) to the Git repo. NB: From here on, I won’t show screenshots for the GitHub side of things – it’s the same each time: whenever a change is made in SQL Source Control and committed to your local folder, you then need to sync this in the GitHub Windows client (as this is where the build server, Bamboo is taking it from). An interesting point to note here, when these changes are committed in SQL Source Control (right-click database and select “Commit Changes to Source Control..”): The display gives a warning about possibly needing a migration script for the “Add Primary Key” step of the changes. This isn’t actually necessary in this case, but this mechanism would allow you to create override scripts to replace the default change scripts created by the SQL Compare engine (which runs underneath SQL Source Control). Ignoring this message (!), we add a comment and commit the changes to Git. I then sync these, run a build (or the build gets run automatically), and check that the data is being deployed over to the target RedGateAppCI database:   Creating and Running the Test As I mention, the test I’m going to use here is a very simple one - are the email addresses in my reference table valid? This isn’t of course, a full test of email validation (I expect the email addresses I’ve chosen here aren’t really the those of the Fab Four) – but just a very basic check of format used. I’ve taken the relevant SQL from this Stack Overflow article. In SSMS select “SQL Test” from the Tools menu, then click on + New Test: In the next screen, give your new test a name, and also enter a name in the Test Class box (test classes are schemas that help you keep things organised). Also check that the database in which the test is going to be created is correct – RedGateApp in this example: Click “Create Test”. After closing a couple of subsequent dialogues, you’ll see a dummy script for the test, that needs filling in:   We now need to define the SQL for our test. As mentioned before, tSQLt allows you to write your unit tests in T-SQL, and the code I’m going to use here is as below. This needs to be copied and pasted in to the query window, to replace the default given by tSQLt: –  Basic email check test ALTER PROCEDURE [MyChecks].[test Check Email Addresses] AS BEGIN SET NOCOUNT ON         Declare @Output VarChar(max)     Set @Output = ”       SELECT  @Output = @Output + Email +Char(13) + Char(10) FROM dbo.Email WHERE email NOT LIKE ‘%_@__%.__%’       If @Output > ”         Begin             Set @Output = Char(13) + Char(10)                           + @Output             EXEC tSQLt.Fail@Output         End   END;   Once this script is entered, hit execute to add the Stored Procedure to the database. Before committing the test to source control,  it’s worth just checking that it works! For a positive test, click on “SQL Test” from the Tools menu, then click Run Tests. You should see output like the following: - a green tick to indicate success! But of course, what we also need to do is test that this is actually doing something by showing a failed test. Edit one of the email addresses in your table to an incorrect format: Now, re-run the same SQL Test as before and you’ll see the following: Great – we now know that our test is really doing something! You’ll also see a useful error message at the bottom of SSMS: (leave the email address as invalid for now, for the next steps). The next stage is to check this new test in to source control again, by right-clicking on the database and checking in the changes with a commit message (and not forgetting to sync in the GitHub client):   Checking that the Tests are Running as Integration Tests After the changes above are made, and after a build has run on Bamboo (manual or automatic), looking at the Stored Procedures for the RedGateAppCI, the SPROC for the new test has been moved over to the database. However this is not exactly what we were after. We didn’t want to just copy objects from one database to another, but actually run the tests as part of the build/integration test process. I.e. we’re continuously checking any changes we make (in this case, to the reference data emails), to ensure we’re not breaking a test that we’ve set up. The behaviour we want to see is that, if we check in static data that is incorrect (as we did in step 9 above) and we have the tSQLt test set up, then our build in Bamboo should fail. However, re-running the build shows the following: - sadly, a successful build! To make sure the tSQLt tests are run as part of the integration test, we need to amend a switch in the Red Gate CI config file. First, navigate to file sqlCI.targets in your working folder: Edit this document, make the following change, save the document, then commit and sync this change in the GitHub client: <!-- tSQLt tests --> <!-- Optional --> <!-- To run tSQLt tests in source control for the database, enter true. --> <enableTsqlt>true</enableTsqlt> Now, if we re-run the build in Bamboo (NB: I’ve moved to a new server here, hence different address and build number): - superb, a broken build!! The error message isn’t great here, so to get more detailed info, click on the full build log link on this page (below the fold). The interesting part of the log shown is towards the bottom. Pulling out this part:   21-Jun-2013 11:35:19 Build FAILED. 21-Jun-2013 11:35:19 21-Jun-2013 11:35:19 "C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj" (default target) (1) -> 21-Jun-2013 11:35:19 (sqlCI target) -> 21-Jun-2013 11:35:19 EXEC : sqlCI error occurred: RedGate.Deploy.SqlServerDbPackage.Shared.Exceptions.InvalidSqlException: Test Case Summary: 1 test case(s) executed, 0 succeeded, 1 failed, 0 errored. [C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj] 21-Jun-2013 11:35:19 EXEC : sqlCI error occurred: [MyChecks].[test Check Email Addresses] failed: [C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj] 21-Jun-2013 11:35:19 EXEC : sqlCI error occurred: ringo.starr@beatles [C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj] 21-Jun-2013 11:35:19 EXEC : sqlCI error occurred: [C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj] 21-Jun-2013 11:35:19 EXEC : sqlCI error occurred: +----------------------+ [C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj] 21-Jun-2013 11:35:19 EXEC : sqlCI error occurred: |Test Execution Summary| [C:\Users\Administrator\bamboo-home\xml-data\build-dir\RGA-RGP-JOB1\sqlCI.proj]   As a final check, we should make sure that, if we now fix this error, the build succeeds. So in SSMS, I’m going to correct the invalid email address, then check this change in to SQL Source Control (with a comment), commit to GitHub, and re-run the build:   This should have fixed the build: It worked! Summary This has been a very quick run through the implementation of CI for databases, including tSQLt tests to test whether your database updates are working. The next post in this series will focus on automated deployment – we’ve tested our database changes, how can we now deploy these to target sites?  

    Read the article

  • Regression testing with Selenium GRID

    - by Ben Adderson
    A lot of software teams out there are tasked with supporting and maintaining systems that have grown organically over time, and the web team here at Red Gate is no exception. We're about to embark on our first significant refactoring endeavour for some time, and as such its clearly paramount that the code be tested thoroughly for regressions. Unfortunately we currently find ourselves with a codebase that isn't very testable - the three layers (database, business logic and UI) are currently tightly coupled. This leaves us with the unfortunate problem that, in order to confidently refactor the code, we need unit tests. But in order to write unit tests, we need to refactor the code :S To try and ease the initial pain of decoupling these layers, I've been looking into the idea of using UI automation to provide a sort of system-level regression test suite. The idea being that these tests can help us identify regressions whilst we work towards a more testable codebase, at which point the more traditional combination of unit and integration tests can take over. Ending up with a strong battery of UI tests is also a nice bonus :) Following on from my previous posts (here, here and here) I knew I wanted to use Selenium. I also figured that this would be a good excuse to put my xUnit [Browser] attribute to good use. Pretty quickly, I had a raft of tests that looked like the following (this particular example uses Reflector Pro). In a nut shell the test traverses our shopping cart and, for a particular combination of number of users and months of support, checks that the price calculations all come up with the correct values. [BrowserTheory] [Browser(Browsers.Firefox3_6, "http://www.red-gate.com")] public void Purchase1UserLicenceNoSupport(SeleniumProvider seleniumProvider) {     //Arrange     _browser = seleniumProvider.GetBrowser();     _browser.Open("http://www.red-gate.com/dynamic/shoppingCart/ProductOption.aspx?Product=ReflectorPro");                  //Act     _browser = ShoppingCartHelpers.TraverseShoppingCart(_browser, 1, 0, ".NET Reflector Pro");     //Assert     var priceResult = PriceHelpers.GetNewPurchasePrice(db, "ReflectorPro", 1, 0, Currencies.Euros);         Assert.Equal(priceResult.Price, _browser.GetText("ctl00_content_InvoiceShoppingItemRepeater_ctl01_Price"));     Assert.Equal(priceResult.Tax, _browser.GetText("ctl00_content_InvoiceShoppingItemRepeater_ctl02_Tax"));     Assert.Equal(priceResult.Total, _browser.GetText("ctl00_content_InvoiceShoppingItemRepeater_ctl02_Total")); } These tests are pretty concise, with much of the common code in the TraverseShoppingCart() and GetNewPurchasePrice() methods. The (inevitable) problem arose when it came to execute these tests en masse. Selenium is a very slick tool, but it can't mask the fact that UI automation is very slow. To give you an idea, the set of cases that covers all of our products, for all combinations of users and support, came to 372 tests (for now only considering purchases in dollars). In the world of automated integration tests, that's a very manageable number. For unit tests, it's a trifle. However for UI automation, those 372 tests were taking just over two hours to run. Two hours may not sound like a lot, but those cases only cover one of the three currencies we deal with, and only one of the many different ways our systems can be asked to calculate a price. It was already pretty clear at this point that in order for this approach to be viable, I was going to have to find a way to speed things up. Up to this point I had been using Selenium Remote Control to automate Firefox, as this was the approach I had used previously and it had worked well. Fortunately,  the guys at SeleniumHQ also maintain a tool for executing multiple Selenium RC tests in parallel: Selenium Grid. Selenium Grid uses a central 'hub' to handle allocation of Selenium tests to individual RCs. The Remote Controls simply register themselves with the hub when they start, and then wait to be assigned work. The (for me) really clever part is that, as far as the client driver library is concerned, the grid hub looks exactly the same as a vanilla remote control. To create a new browser session against Selenium RC, the following C# code suffices: new DefaultSelenium("localhost", 4444, "*firefox", "http://www.red-gate.com"); This assumes that the RC is running on the local machine, and is listening on port 4444 (the default). Assuming the hub is running on your local machine, then to create a browser session in Selenium Grid, via the hub rather than directly against the control, the code is exactly the same! Behind the scenes, the hub will take this request and hand it off to one of the registered RCs that provides the "*firefox" execution environment. It will then pass all communications back and forth between the test runner and the remote control transparently. This makes running existing RC tests on a Selenium Grid a piece of cake, as the developers intended. For a more detailed description of exactly how Selenium Grid works, see this page. Once I had a test environment capable of running multiple tests in parallel, I needed a test runner capable of doing the same. Unfortunately, this does not currently exist for xUnit (boo!). MbUnit on the other hand, has the concept of concurrent execution baked right into the framework. So after swapping out my assembly references, and fixing up the resulting mismatches in assertions, my example test now looks like this: [Test] public void Purchase1UserLicenceNoSupport() {    //Arrange    ISelenium browser = BrowserHelpers.GetBrowser();    var db = DbHelpers.GetWebsiteDBDataContext();    browser.Start();    browser.Open("http://www.red-gate.com/dynamic/shoppingCart/ProductOption.aspx?Product=ReflectorPro");                 //Act     browser = ShoppingCartHelpers.TraverseShoppingCart(browser, 1, 0, ".NET Reflector Pro");    var priceResult = PriceHelpers.GetNewPurchasePrice(db, "ReflectorPro", 1, 0, Currencies.Euros);    //Assert     Assert.AreEqual(priceResult.Price, browser.GetText("ctl00_content_InvoiceShoppingItemRepeater_ctl01_Price"));     Assert.AreEqual(priceResult.Tax, browser.GetText("ctl00_content_InvoiceShoppingItemRepeater_ctl02_Tax"));     Assert.AreEqual(priceResult.Total, browser.GetText("ctl00_content_InvoiceShoppingItemRepeater_ctl02_Total")); } This is pretty much the same as the xUnit version. The exceptions are that the attributes have changed,  the //Arrange phase now has to handle setting up the ISelenium object, as the attribute that previously did this has gone away, and the test now sets up its own database connection. Previously I was using a shared database connection, but this approach becomes more complicated when tests are being executed concurrently. To avoid complexity each test has its own connection, which it is responsible for closing. For the sake of readability, I snipped out the code that closes the browser session and the db connection at the end of the test. With all that done, there was only one more step required before the tests would execute concurrently. It is necessary to tell the test runner which tests are eligible to run in parallel, via the [Parallelizable] attribute. This can be done at the test, fixture or assembly level. Since I wanted to run all tests concurrently, I marked mine at the assembly level in the AssemblyInfo.cs using the following: [assembly: DegreeOfParallelism(3)] [assembly: Parallelizable(TestScope.All)] The second attribute marks all tests in the assembly as [Parallelizable], whilst the first tells the test runner how many concurrent threads to use when executing the tests. I set mine to three since I was using 3 RCs in separate VMs. With everything now in place, I fired up the Icarus* test runner that comes with MbUnit. Executing my 372 tests three at a time instead of one at a time reduced the running time from 2 hours 10 minutes, to 55 minutes, that's an improvement of about 58%! I'd like to have seen an improvement of 66%, but I can understand that either inefficiencies in the hub code, my test environment or the test runner code (or some combination of all three most likely) contributes to a slightly diminished improvement. That said, I'd love to hear about any experience you have in upping this efficiency. Ultimately though, it was a saving that was most definitely worth having. It makes regression testing via UI automation a far more plausible prospect. The other obvious point to make is that this approach scales far better than executing tests serially. So if ever we need to improve performance, we just register additional RC's with the hub, and up the DegreeOfParallelism. *This was just my personal preference for a GUI runner. The MbUnit/Gallio installer also provides a command line runner, a TestDriven.net runner, and a Resharper 4.5 runner. For now at least, Resharper 5 isn't supported.

    Read the article

  • Streaming Netflix Media with My Wii

    - by Ben Griswold
    Late last year, I wrote about Streaming Media with my Sony Blu-ray Disc Player. I am still digging the Blu-ray player setup but guess what showed up in the mail yesterday?   That’s right!  A free Netflix disc which now let’s me instantly watch TV episodes and movies via my Wii console.  I popped the disc into the console and in less than 2 minutes the brain-numbingly simple activation was complete.  (Full-disclosure: I already had my Wi-Fi connection configured, but I’m confident that the Netflix installation disc would have helpfully walked me through this additional step if need be.) As it turns out, the Wii Netflix UI offers far more options than what one gets with the Blu-ray setup.  Not only can I view my Instant Queue, but there’s a list of recently watched movies, a list of recommended titles by category, the star rating system, movies information and nearly everything you find on the web.  I reread Steve Krug’s Don’t Make Me Think: A Common Sense Approach to Web Usability on a flight back from Orlando on Wednesday, so my current view of the world may be a little skewed but, the brilliance of Netflix Wii’s user interface is undeniable. It’s not like the Blu-ray navigation is complicated but the Wii navigation feels familiar and intuitive. How intuitive?  Well, you won’t find a single bit of help text on any of the Wii screens – just a simple and obvious point-and-click navigation system.  And the UI is really pretty (which is still very important if you ask me) and so easy it became fun. Did I mention the media streaming works!  Yep, we watched 2 half-hour kid videos yesterday without any streaming issues at all.  If you have a Netflix account and a Wii, order your disc and give it a go. It’s good stuff.

    Read the article

  • New .NET Library for Accessing the Survey Monkey API

    - by Ben Emmett
    I’ve used Survey Monkey’s API for a while, and though it’s pretty powerful, there’s a lot of boilerplate each time it’s used in a new project, and the json it returns needs a bunch of processing to be able to use the raw information. So I’ve finally got around to releasing a .NET library you can use to consume the API more easily. The main advantages are: Only ever deal with strongly-typed .NET objects, making everything much more robust and a lot faster to get going Automatically handles things like rate-limiting and paging through results Uses combinations of endpoints to get all relevant data for you, and processes raw response data to map responses to questions To start, either install it using NuGet with PM> Install-Package SurveyMonkeyApi (easier option), or grab the source from https://github.com/bcemmett/SurveyMonkeyApi if you prefer to build it yourself. You’ll also need to have signed up for a developer account with Survey Monkey, and have both your API key and an OAuth token. A simple usage would be something like: string apiKey = "KEY"; string token = "TOKEN"; var sm = new SurveyMonkeyApi(apiKey, token); List<Survey> surveys = sm.GetSurveyList(); The surveys object is now a list of surveys with all the information available from the /surveys/get_survey_list API endpoint, including the title, id, date it was created and last modified, language, number of questions / responses, and relevant urls. If there are more than 1000 surveys in your account, the library pages through the results for you, making multiple requests to get a complete list of surveys. All the filtering available in the API can be controlled using .NET objects. For example you might only want surveys created in the last year and containing “pineapple” in the title: var settings = new GetSurveyListSettings { Title = "pineapple", StartDate = DateTime.Now.AddYears(-1) }; List<Survey> surveys = sm.GetSurveyList(settings); By default, whenever optional fields can be requested with a response, they will all be fetched for you. You can change this behaviour if for some reason you explicitly don’t want the information, using var settings = new GetSurveyListSettings { OptionalData = new GetSurveyListSettingsOptionalData { DateCreated = false, AnalysisUrl = false } }; Survey Monkey’s 7 read-only endpoints are supported, and the other 4 which make modifications to data might be supported in the future. The endpoints are: Endpoint Method Object returned /surveys/get_survey_list GetSurveyList() List<Survey> /surveys/get_survey_details GetSurveyDetails() Survey /surveys/get_collector_list GetCollectorList() List<Collector> /surveys/get_respondent_list GetRespondentList() List<Respondent> /surveys/get_responses GetResponses() List<Response> /surveys/get_response_counts GetResponseCounts() Collector /user/get_user_details GetUserDetails() UserDetails /batch/create_flow Not supported Not supported /batch/send_flow Not supported Not supported /templates/get_template_list Not supported Not supported /collectors/create_collector Not supported Not supported The hierarchy of objects the library can return is Survey List<Page> List<Question> QuestionType List<Answer> List<Item> List<Collector> List<Response> Respondent List<ResponseQuestion> List<ResponseAnswer> Each of these classes has properties which map directly to the names of properties returned by the API itself (though using PascalCasing which is more natural for .NET, rather than the snake_casing used by SurveyMonkey). For most users, Survey Monkey imposes a rate limit of 2 requests per second, so by default the library leaves at least 500ms between requests. You can request higher limits from them, so if you want to change the delay between requests just use a different constructor: var sm = new SurveyMonkeyApi(apiKey, token, 200); //200ms delay = 5 reqs per sec There’s a separate cap of 1000 requests per day for each API key, which the library doesn’t currently enforce, so if you think you’ll be in danger of exceeding that you’ll need to handle it yourself for now.  To help, you can see how many requests the current instance of the SurveyMonkeyApi object has made by reading its RequestsMade property. If the library encounters any errors, including communicating with the API, it will throw a SurveyMonkeyException, so be sure to handle that sensibly any time you use it to make calls. Finally, if you have a survey (or list of surveys) obtained using GetSurveyList(), the library can automatically fill in all available information using sm.FillMissingSurveyInformation(surveys); For each survey in the list, it uses the other endpoints to fill in the missing information about the survey’s question structure, respondents, and responses. This results in at least 5 API calls being made per survey, so be careful before passing it a large list. It also joins up the raw response information to the survey’s question structure, so that for each question in a respondent’s set of replies, you can access a ProcessedAnswer object. For example, a response to a dropdown question (from the /surveys/get_responses endpoint) might be represented in json as { "answers": [ { "row": "9384627365", } ], "question_id": "615487516" } Separately, the question’s structure (from the /surveys/get_survey_details endpoint) might have several possible answers, one of which might look like { "text": "Fourth item in dropdown list", "visible": true, "position": 4, "type": "row", "answer_id": "9384627365" } The library understands how this mapping works, and uses that to give you the following ProcessedAnswer object, which first describes the family and type of question, and secondly gives you the respondent’s answers as they relate to the question. Survey Monkey has many different question types, with 11 distinct data structures, each of which are supported by the library. If you have suggestions or spot any bugs, let me know in the comments, or even better submit a pull request .

    Read the article

  • My Mother Bought a Droid

    - by Ben Griswold
    I converted to iPhone two years ago when I left my former employer and my Blackberry behind. The truth is I half-heartedly purchased my iPhone. It was a great looking device, but as far as I was concerned, it was a mere toy compared to my Blackberry.  I remember hiding the toy in my briefcase when attending business meetings because I didn’t consider it to be professional enough.  I’ve since owned all three generations of the iPhone and, well, iPhone seems to have caught on. I still miss the click of the Blackberry keyboard and the blinking red light letting me know that someone/something requires my attention, but I’m officially an iPhone fanboy now.  My mom called last weekend and asked if she should buy an iPhone. I talked her ear off about everything I love about iPhone. I went on for about twenty minutes. I couldn’t help myself. I mentioned everything from my podcast subscriptions to the application which manages my workouts.  I went as far as to say that someday all smart phones will be referred to as iPhones just like all tissues are referred to as Kleenex and all sodas are referred to as Cokes.  I was really on a roll and then I stopped. I had to…the call dropped.  There I was, strategically standing in the far corner of my backyard where I get the most reliable AT&T reception and the call drops in middle of my iPhone pitch.  Folks, I don’t care how good a salesperson you are, it’s tough to recover from a situation like this.   I dialed my mom back and jokingly asked if she was planning to make calls with her new phone. I explained that AT&T is bound to provide better service eventually but I’m not sure she should wait. After all, I have troubles with the network in San Diego and I can only image how bad it would be for her in Western Massachusetts. Mom called back a few days later exclaiming, “I bought a Droid! I love this phone! I haven’t done anything with it but make phone calls, but I love it.”  I had to laugh.  My mom made the right call (pun intended.)  The iPhone is an amazing device, but owners are constantly reminded that its core function (it’s a phone, remember?) is subpar.  If you love gadgets, you’re probably enthralled by iPhone’s many bells and whistles and, relatively speaking, the terrible phone service might not amount to much.  (Maybe it amounts to a rant on your blog.) The overall iPhone offering is so attractive that consumers are willing to wait for AT&T to straighten up their act or wait until Apple grants a choice of carriers.  But I don’t see either of these remedies coming soon. In the interim, I’m willing to take my iPhone for what it is and just continue to enjoy my favorite features while pretending that poor coverage isn’t a big deal. With any luck, more and more reasonable folks will recognize that Android Phones are legitimate players in the smart phone space, they will buy loads of them and there will become plenty of functional phones to borrow when my “phone” is showing zero bars.  Heck, I’m already covered when I visit my mom.

    Read the article

  • Test-Drive ASP.NET MVC Review

    - by Ben Griswold
    A few years back I started dallying with test-driven development, but I never fully committed to the practice. This wasn’t because I didn’t believe in the value of TDD; it was more a matter of not completely understanding how to incorporate “test first” into my everyday development. Back in my web forms days, I could point fingers at the framework for my ignorance and laziness. After all, web forms weren’t exactly designed for testability so who could blame me for not embracing TDD in those conditions, right? But when I switched to ASP.NET MVC and quickly found myself fresh out of excuses and it became instantly clear that it was time to get my head around red-green-refactor once and for all or I would regretfully miss out on one of the biggest selling points the new framework had to offer. I have previously written about how I learned ASP.NET MVC. It was primarily hands on learning but I did read a couple of ASP.NET MVC books along the way. The books I read dedicated a chapter or two to TDD and they certainly addressed the benefits of TDD and how MVC was designed with testability in mind, but TDD was merely an afterthought compared to, well, teaching one how to code the model, view and controller. This approach made some sense, and I learned a bunch about MVC from those books, but when it came to TDD the books were just a teaser and an opportunity missed.  But then I got lucky – Jonathan McCracken contacted me and asked if I’d review his book, Test-Drive ASP.NET MVC, and it was just what I needed to get over the TDD hump. As the title suggests, Test-Drive ASP.NET MVC takes a different approach to learning MVC as it focuses on testing right from the very start. McCracken wastes no time and swiftly familiarizes us with the framework by building out a trivial Quote-O-Matic application and then dedicates the better part of his book to testing first – first by explaining TDD and then coding a full-featured Getting Organized application inspired by David Allen’s popular book, Getting Things Done. If you are a learn-by-example kind of coder (like me), you will instantly appreciate and enjoy McCracken’s style – its fast-moving, pragmatic and focused on only the most relevant information required to get you going with ASP.NET MVC and TDD. The book continues with the test-first theme but McCracken moves away from the sample application and incorporates other practical skills like persisting models with NHibernate, leveraging Inversion of Control with the IControllerFactory and building a RESTful web service. What I most appreciated about this section was McCracken’s use of and praise for open source libraries like Rhino Mocks, SQLite and StructureMap (to name just a few) and productivity tools like ReSharper, Web Platform Installer and ASP.NET SQL Server Setup Wizard.  McCracken’s emphasis on real world, pragmatic development was clearly demonstrated in every tool choice, straight-forward code block and developer tip. Whether one is already familiar with the tools/tips or not, McCracken’s thought process is easily understood and appreciated. The final section of the book walks the reader through security and deployment – everything from error handling and logging with ELMAH, to ASP.NET Health Monitoring, to using MSBuild with automated builds, to the deployment  of ASP.NET MVC to various web environments. These chapters, like those prior, offer enough information and explanation to simply help you get the job done.  Do I believe Test-Drive ASP.NET MVC will turn you into an expert MVC developer overnight?  Well, no.  I don’t think any book can make that claim.  If that were possible, I think book list prices would skyrocket!  That said, Test-Drive ASP.NET MVC provides a solid foundation and a unique (and dare I say necessary) approach to learning ASP.NET MVC.  Along the way McCracken shares loads of very practical software development tips and references numerous tools and libraries. The bottom line is it’s a great ASP.NET MVC primer – if you’re new to ASP.NET MVC it’s just what you need to get started.  Do I believe Test-Drive ASP.NET MVC will give you everything you need to start employing TDD in your everyday development?  Well, I used to think that learning TDD required a lot of practice and, if you’re lucky enough, the guidance of a mentor or coach.  I used to think that one couldn’t learn TDD from a book alone. Well, I’m still no pro, but I’m testing first now and Jonathan McCracken and his book, Test-Drive ASP.NET MVC, played a big part in making this happen.  If you are an MVC developer and a TDD newb, Test-Drive ASP.NET MVC is just the book for you.

    Read the article

  • What exactly is an SEO badge and how does it help my SEO?

    - by ben geit
    Hi everyone! I'm using attracta.com as an SEO tool to try and improve my SEO. They say that I should use an SEO badge to improve my ranking but I don't really understand what it is or what it does (a quick Google search didn't provide too much information). I was just wondering if any of you guys here at stackoverflow had any information they would be able to impart to me. Thanks for your help!

    Read the article

  • Do you have any “Family Feud” style questions and answers for a game for high school students?

    - by Ben Jakuben
    I am gathering questions and responses in math, science, and technology for a "Family Feud" style game for high school students. I am having trouble finding and thinking of questions, especially in the technology realm. Technology (programming or general tech) questions are preferred. If you have never seen the game show, "Family Feud" involves two teams trying to guess the most popular responses to questions asked to a group of 100 respondents. The team must guess all the popular responses to get the points for the question. For example, if the question is, "What are the major tags in HTML 4.0?", the responses might be: P (64 votes) DIV (16 votes) TABLE (8 votes) BLINK (4 votes)

    Read the article

  • How to Organize a Programming Language Club

    - by Ben Griswold
    I previously noted that we started a language club at work.  You know, I searched around but I couldn’t find a copy of the How to Organize a Programming Language Club Handbook. Maybe it’s sold out?  Yes, Stack Overflow has quite a bit of information on how to learn and teach new languages and there’s also a good number of online tutorials which provide language introductions but I was interested in group learning.  After   two months of meetings, I present to you the Unofficial How to Organize a Programming Language Club Handbook.  1. Gauge interest. Start by surveying prospects. “Excuse me, smart-developer-whom-I-work-with-and-I-think-might-be-interested-in-learning-a-new-coding-language-with-me. Are you interested in learning a new language with me?” If you’re lucky, you work with a bunch of really smart folks who aren’t shy about teaching/learning in a group setting and you’ll have a collective interest in no time.  Simply suggesting the idea is the only effort required.  If you don’t work in this type of environment, maybe you should consider a new place of employment.  2. Make it official. Send out a “Welcome to the Club” email: There’s been talk of folks itching to learn new languages – Python, Scala, F# and Haskell to name a few.  Rather than taking on new languages alone, let’s learn in the open.  That’s right.  Let’s start a languages club.  We’ll have everything a real club needs – secret handshake, goofy motto and a high-and-mighty sense that we’re better than everybody else. T-shirts?  Hell YES!  Anyway, I’ve thrown this idea around the office and no one has laughed at me yet so please consider this your very official invitation to be in THE club. [Insert your ideas about how the club might be run, solicit feedback and suggestions, ask what other folks would like to get out the club, comment about club hazing practices and talk up the T-shirts even more. Finally, call out the languages you are interested in learning and ask the group for their list.] 3.  Send out invitations to the first meeting.  Don’t skimp!  Hallmark greeting cards for everyone.  Personalized.  Hearts over the I’s and everything.  Oh, and be sure to include the list of suggested languages with vote count.  Here the list of languages we are interested in: Python 5 Ruby 4 Objective-C 3 F# 2 Haskell 2 Scala 2 Ada 1 Boo 1 C# 1 Clojure 1 Erlang 1 Go 1 Pi 1 Prolog 1 Qt 1 4.  At the first meeting, there must be cake.  Lots of cake. And you should tackle some very important questions: Which language should we start with?  You can immediately go with the top vote getter or you could do as we did and designate each person to provide a high-level review of each of the proposed languages over the next two weeks.  After all presentations are completed, vote on the language. Our high-level review consisted of answers to a series of questions. Decide how often and where the group will meet.  We, for example, meet for a brown bag lunch every Wednesday.  Decide how you’re going to learn.  We determined that the best way to learn is to just dive in and write code.  After choosing our first language (Python), we talked about building an application, or performing coding katas, but we ultimately choose to complete a series of Project Euler problems.  We kept it simple – each member works out the same two problems each week in preparation of a code review the following Wednesday. 5.  Code, Review, Learn.  Prior to the weekly meeting, everyone uploads their solutions to our internal wiki.  Each Project Euler problem has a dedicated page.  In the meeting, we use a really fancy HD projector to show off each member’s solution.  It is very important to use an HD projector.  Again, don’t skimp!  Each code author speaks to their solution, everyone else comments, applauds, points fingers and laughs, etc.  As much as I’ve learned from solving the problems on my own, I’ve learned at least twice as much at the group code review.  6.  Rinse. Lather. Repeat.  We’ve hosted the language club for 7 weeks now.  The first meeting just set the stage.  The next two meetings provided a review of the languages followed by a first language selection.  The remaining meetings focused on Python and Project Euler problems.  Today we took a vote as to whether or not we’re ready to switch to another language and/or another problem set.  Pretty much everyone wants to stay the course for a few more weeks at least.  Until then, we’ll continue to code the next two solutions, review and learn. Again, we’ve been having a good time with the programming language club.  I’m glad it got off the ground.  What do you think?  Would you be interested in a language club?  Any suggestions on what we might do better?

    Read the article

  • Pro ASP.NET MVC Framework Review

    - by Ben Griswold
    Early in my career, when I wanted to learn a new technology, I’d sit in the bookstore aisle and I’d work my way through each of the available books on the given subject.  Put in enough time in a bookstore and you can learn just about anything. I used to really enjoy my time in the bookstore – but times have certainly changed.  Whereas books used to be the only place I could find solutions to my problems, now they may be the very last place I look.  I have been working with the ASP.NET MVC Framework for more than a year.  I have a few projects and a couple of major deployments under my belt and I was able to get up to speed with the framework without reading a single book*.  With so many resources at our fingertips (podcasts, screencasts, blogs, stackoverflow, open source projects, www.asp.net, you name it) why bother with a book? Well, I flipped through Steven Sanderson’s Pro ASP.NET MVC Framework a few months ago. And since it is prominently displayed in my co-worker’s office, I tend to pick it up as a reference from time to time.  Last week, I’m not sure why, I decided to read it cover to cover.  Man, did I eat this book up.  Granted, a lot of what I read was review, but it was only review because I had already learned lessons by piecing the puzzle together for myself via various sources. If I were starting with ASP.NET MVC (or ASP.NET Web Deployment in general) today, the first thing I would do is buy Steven Sanderson’s Pro ASP.NET MVC Framework and read it cover to cover. Steven Sanderson did such a great job with this book! As much as I appreciated the in-depth model, view, and controller talk, I was completely impressed with all the extra bits which were included.  There a was nice overview of BDD, view engine comparisons, a chapter dedicated to security and vulnerabilities, IoC, TDD and Mocking (of course), IIS deployment options and a nice overview of what the .NET platform and C# offers.  Heck, Sanderson even include bits about webforms! The book is fantastic and I highly recommend it – even if you think you’ve already got your head around ASP.NET MVC.  By the way, procrastinators may be in luck.  ASP.NET MVC V2 Framework can be pre-ordered.  You might want to jump right into the second edition and find out what Sanderson has to say about MVC 2. * Actually, I did read through the free bits of Professional ASP.NET MVC 1.0.  But it was just a chapter – albeit a really long chapter.

    Read the article

  • Project Euler 53: Ruby

    - by Ben Griswold
    In my attempt to learn Ruby out in the open, here’s my solution for Project Euler Problem 53.  I first attempted to solve this problem using the Ruby combinations libraries. That didn’t work out so well. With a second look at the problem, the provided formula ended up being just the thing to solve the problem effectively. As always, any feedback is welcome. # Euler 53 # http://projecteuler.net/index.php?section=problems&id=53 # There are exactly ten ways of selecting three from five, # 12345: 123, 124, 125, 134, 135, 145, 234, 235, 245, # and 345 # In combinatorics, we use the notation, 5C3 = 10. # In general, # # nCr = n! / r!(n-r)!,where r <= n, # n! = n(n1)...321, and 0! = 1. # # It is not until n = 23, that a value exceeds # one-million: 23C10 = 1144066. # In general: nCr # How many, not necessarily distinct, values of nCr, # for 1 <= n <= 100, are greater than one-million timer_start = Time.now # There's no factorial method in Ruby, I guess. class Integer # http://rosettacode.org/wiki/Factorial#Ruby def factorial (1..self).reduce(1, :*) end end def combinations(n, r) n.factorial / (r.factorial * (n-r).factorial) end answer = 0 100.downto(3) do |c| (2).upto(c-1) { |r| answer += 1 if combinations(c, r) > 1_000_000 } end puts answer puts "Elapsed Time: #{(Time.now - timer_start)*1000} milliseconds"

    Read the article

  • Project Euler 51: Ruby

    - by Ben Griswold
    In my attempt to learn Ruby out in the open, here’s my solution for Project Euler Problem 51.  I know I started back up with Python this week, but I have three more Ruby solutions in the hopper and I wanted to share. For the record, Project Euler 51 was the second hardest Euler problem for me thus far. Yeah. As always, any feedback is welcome. # Euler 51 # http://projecteuler.net/index.php?section=problems&id=51 # By replacing the 1st digit of *3, it turns out that six # of the nine possible values: 13, 23, 43, 53, 73, and 83, # are all prime. # # By replacing the 3rd and 4th digits of 56**3 with the # same digit, this 5-digit number is the first example # having seven primes among the ten generated numbers, # yielding the family: 56003, 56113, 56333, 56443, # 56663, 56773, and 56993. Consequently 56003, being the # first member of this family, is the smallest prime with # this property. # # Find the smallest prime which, by replacing part of the # number (not necessarily adjacent digits) with the same # digit, is part of an eight prime value family. timer_start = Time.now require 'mathn' def eight_prime_family(prime) 0.upto(9) do |repeating_number| # Assume mask of 3 or more repeating numbers if prime.count(repeating_number.to_s) >= 3 ctr = 1 (repeating_number + 1).upto(9) do |replacement_number| family_candidate = prime.gsub(repeating_number.to_s, replacement_number.to_s) ctr += 1 if (family_candidate.to_i).prime? end return true if ctr >= 8 end end false end # Wanted to loop through primes using Prime.each # but it took too long to get to the starting value. n = 9999 while n += 2 next if !n.prime? break if eight_prime_family(n.to_s) end puts n puts "Elapsed Time: #{(Time.now - timer_start)*1000} milliseconds"

    Read the article

  • Introduction to Lean Software Development and Kanban Systems – Defer Commitment and Decide As Late A

    - by Ben Griswold
    In this post, we’ll continue the series by concentrating on Principle #4: Defer Commitment and Decide As Late As Possible.   In the next part of the series, we’ll dive into Principle #5: Deliver As Fast As Possible. And I am going to be a little obnoxious about listing my Lean and Kanban references with every series post.  The references are great and they deserve this sort of attention.  

    Read the article

  • Language Club – Battle of the Dynamic Languages

    - by Ben Griswold
    After dedicating the last eight weeks to learning Ruby, it’s time to move onto another language.  I really dig Ruby.  I really enjoy its dynamism and expressiveness and always-openness and it’s been the highlight of our coding club for me so far. But that’s just my take on the language.  I know a lot of coders who’s stomachs turn with the mere thought of Ruby.  They say it’s Ruby’s openness which has them feeling uneasy.  I’d say “write a bunch of tests and get over it,” but I figure there must be more to it than always open classes and possible method collisions. Yes, there’s something else to it alright. The folks who didn’t fall head over heals for Ruby are already in love with Python.  You might remember that Python was the first language we tackled in our coding club.  My time with Python was okay but it didn’t feel as natural to me as Ruby.  But let’s say we started with Ruby and then moved onto Python.  Would I see Python in a different light right now.  Might I even prefer Python over Ruby?  I suppose it’s possible but it’s pretty tough to test that theory – unless we visit Python for a second time. That’s right. The language club is going to focus on Python again and in my attempt to learn Python – yet again – in the open, I’ll be posting my solutions here just as I did for Ruby.  We don’t always have second chances so I going about this relearning with two primary goals in mind:  First, I’m going to use IronPython and the IronPython tools which provide a Python code editor, a file-based project system, and an interactive Python interpreter, all inside Visual Studio 2010.  As a note, the IronPython tools are now part of the main IronPython installer which is Version 2.7 Alpha 1 (not the latest stable version, 2.6.1) and I’d be crazy not to use them.  Second, I’d like to make sure I’m still learning Python without a complete MS skew so I’m going to run my code through Eclipse using the PyDev plugin as well.  Heck, I might use IDLE too. I already have this setup on my machine so it’s no big deal. Okay, that’s it for now.  I worked on the first ten Euler problems last night and the solutions will be posted shortly. Wish me luck.

    Read the article

  • ASP.NET Meta Keywords and Description

    - by Ben Griswold
    Some of the ASP.NET 4 improvements around SEO are neat.  The ASP.NET 4 Page.MetaKeywords and Page.MetaDescription properties, for example, are a welcomed change.  There’s nothing earth-shattering going on here – you can now set these meta tags via your Master page’s code behind rather than relying on updates to your markup alone.  It isn’t difficult to manage meta keywords and descriptions without these ASP.NET 4 properties but I still appreciate the attention SEO is getting.  It’s nice to get gentle reminder via new coding features that some of the more subtle aspects of one’s application deserve thought and attention too.  For the record, this is how I currently manage my meta: <meta name="keywords"     content="<%= Html.Encode(ConfigurationManager.AppSettings["Meta.Keywords"]) %>" /> <meta name="description"     content="<%= Html.Encode(ConfigurationManager.AppSettings["Meta.Description"]) %>" /> All Master pages assume the same keywords and description values as defined by the application settings.  Nothing fancy. Nothing dynamic. But it’s manageable.  It works, but I’m looking forward to the new way in ASP.NET 4.

    Read the article

  • Project Euler 10: (Iron)Python

    - by Ben Griswold
    In my attempt to learn (Iron)Python out in the open, here’s my solution for Project Euler Problem 10.  As always, any feedback is welcome. # Euler 10 # http://projecteuler.net/index.php?section=problems&id=10 # The sum of the primes below 10 is 2 + 3 + 5 + 7 = 17. # Find the sum of all the primes below two million. import time start = time.time() def primes_to_max(max): primes, number = [2], 3 while number < max: isPrime = True for prime in primes: if number % prime == 0: isPrime = False break if (prime * prime > number): break if isPrime: primes.append(number) number += 2 return primes primes = primes_to_max(2000000) print sum(primes) print "Elapsed Time:", (time.time() - start) * 1000, "millisecs" a=raw_input('Press return to continue')

    Read the article

  • Streaming Media with Sony Blu-ray Disc Player

    - by Ben Griswold
    The best gift under the tree this year? A Sony Blu-ray Disc player: The BDP-N460 allows you to instantly stream thousands of movies, videos and music from the largest selection of leading content providers including Netflix, Amazon Video On Demand, YouTube™, Slacker® Radio and many, many more. Plus, enjoy the ultimate in high-definition entertainment and watch Blu-ray Disc movies in Full HD 1080p quality with HD audio. The BDP-N460 includes built-in software that makes it easy to connect this player to your existing wireless network.  So I did… I paired the disc player with the recommended Linksys Wireless Ethernet Bridge (WET-610N) and I was streaming the last season of Lost episodes in no time. Really cool. Highly recommended.

    Read the article

  • Set up printing with a networked LPR printer?

    - by Ben
    I'm trying to set up printing to a networked LPR printer, but I can't seem to make it work. The connexion information that IT gives is: Print Spooler: printing.domain.edu Printer Name: PrinterName Print port: 515 Bidirectional: Disabled It's an HP LaserJet p4015x, and I'm using the "HP LaserJet p4015x, hpcups 3.10.6 (color, 2-sided printing)" driver, according to CUPS. I have tried installing via the CUPS admin interface as lpd://printing.domain.edu:515/PrinterName, lpd://printing.domain.edu/PrinterName, and http://printing.domain.edu:515/PrinterName, and none of these has worked (i.e. test pages show up as "stopped").

    Read the article

  • Introduction to Lean Software Development and Kanban Systems – Build Integrity and Quality In

    - by Ben Griswold
    In this post, we’ll continue the series by concentrating on Principle #3: Build Integrity and Quality In.   In the next part of the series, we’ll dive into Principle #4: Defer Commitment and Decide As Late As Possible. And I am going to be a little obnoxious about listing my Lean and Kanban references with every series post.  The references are great and they deserve this sort of attention.  

    Read the article

  • Custom Profile Provider with Web Deployment Project

    - by Ben Griswold
    I wrote about implementing a custom profile provider inside of your ASP.NET MVC application yesterday. If you haven’t read the article, don’t sweat it.  Most of the stuff I write is rubbish anyway. Since you have joined me today, though, I might as well offer up a little tip: you can run into trouble, like I did, if you enable your custom profile provider inside of an application which is deployed using a Web Deployment Project.  Everything will run great on your local machine and you’ll probably take an early lunch because you got the code running in no time flat and the build server is happy and all tests pass and, gosh, maybe you’ll just cut out early because it is Friday after all.  But then the first user hits the integration machine and, that’s right, yellow screen of death. Lucky you, just as you’re walking out the door, the user kindly sends the exception message and stack trace: Value cannot be null. Parameter name: type Description: An unhandled exception occurred during the execution of the current web request. Please review the stack trace for more information about the error and where it originated in the code. Stack Trace: [ArgumentNullException: Value cannot be null. Parameter name: type] System.Activator.CreateInstance(Type type, Boolean nonPublic) +2796915 System.Web.Profile.ProfileBase.CreateMyInstance(String username, Boolean isAuthenticated) +76 System.Web.Profile.ProfileBase.Create(String username, Boolean isAuthenticated) +312 User error?  Not this time. Damn! One hour later… you notice the harmless “Treat as library component (remove the App_Code.compiled file)” setting on the Output Assemblies Tab of your Web Deployment Project. You have no idea why, but you uncheck it.  You test and everything works great both locally and on the integration machine.  Application users think you’re the best and you’re still going to catch the last half hour of happy hour.  Happy Friday.

    Read the article

  • Why’s (Poignant) Guide to Ruby

    - by Ben Griswold
    You’re familiar with O’Reilly’s brilliant Head First Series, right?  Great.  Then you know how every book begins with an explanation of the Head First teaching style and you know the teaching format which Kathy Sierra and Bert Bates developed is based on research in cognitive science, neurobiology and educational psychology and it’s all about making learning visual and conversational and attractive and emotional and it’s highly effective.  Anyway, it’s a great series and you should read every last one of the books. Moving on… I’ve been wanting to learn more about Ruby and Why’s (Poignant) Guide to Ruby has been on my reading list for a while and there was talk about cartoon foxes and other silliness and I figured Why’s (Poignant) Guide to Ruby probably takes the same unorthodox teaching style as the Head First books – and that’s great – so I read the book, in piecemeal, over the last couple of weeks and, well, I figured wrong. Now having read the book, here’s my take on Why’s (Poignant) Guide – it’s very creative and clever and it does a darn good job of introducing one to Ruby.  If you’re interested in Ruby or simply interested, the online book is worth your time.  If you’re thinking (like me) that cartoon foxes will be doing the teaching, that’s simple not the case.  However, the cartoons and the random stories in the sidebar may serve a purpose. Unlike the Head First books where images and captions are used to further explain the teachings, the cartoons and stories in Why’s Guide serve as intermission and offer your brain a brief moment of rest before the next Ruby concept is explained.  It’s not a bad strategy, but definitely not as effective as the Head First techniques.  

    Read the article

  • F# in 90 Seconds

    - by Ben Griswold
    I mentioned in a previous post that we’ve started a languages club at the office.  In an effort to decide which language we will first concentrate on, I volunteered to give the rundown on F#.  Rather than providing a summary here, I’ve provided my slide deck for your viewing enjoyment.  There’s nothing special here outside of a some pretty cool characters from The 56 Geeks Project by Scott Johnson and collection of information from my prior functional programming presentations.   Download F# in 90 Seconds

    Read the article

< Previous Page | 4 5 6 7 8 9 10 11 12 13 14 15  | Next Page >