Search Results

Search found 18546 results on 742 pages for 'evaluation order'.

Page 395/742 | < Previous Page | 391 392 393 394 395 396 397 398 399 400 401 402  | Next Page >

  • What is the best objective way to measure language popularity trends? (What's better than TIOBE?)

    - by Eric Wilson
    The best way to get data on computer language popularity that I know is the TIOBE index. But everyone knows that TIOBE is hopelessly flawed. (If someone provides a link to support this, I'll add it here.) So is there any data on programming language popularity that is generally considered meaningful? The only other option I know is to look at the trends at indeed.com, which is inherently flawed, being based on job postings. It isn't like I would make a future language decision solely based on an index, but it might provide a useful balance to the skewed perspective one obtains by talking to ones friends and colleagues. To illustrate that bias, I'll point out that based on the experience of those I personally know, the only languages used professionally today (in order of popularity) are Java, C#, Groovy, JavaScript, Ruby, Objective C, and Perl. (Though it is evident that C, C++ and PHP were used in the past.) So my question is, everyone bashes TIOBE, but is there anything else? If so, can anyone explain how we know the alternative has better methodology? Thanks.

    Read the article

  • Url for search withing Exchange Server OWA

    - by Martin Vobr
    I can easily create url for searching the Internet for specific term: http://www.bing.com/search?q=example Is is possible to create a similar url for searching my Exchange server mailbox using Outlook Web App? Something like: http://my.exchange.server.tld/owa/search.aspx?q=example EDIT Some of comments asked for target, so here is the clarification: There is a web-based backoffice system which includes customer's email addresses. I want to provide an easy way to show emails related to this customer in a support mailbox. Adding a link to an OWA search result page seemed a way to accomplish quickly. I can lookup emails either via EWS or IMAP. I wanted to reuse OWA for displaying them instead of reinventing the wheel. If creating a search link is not possible what would be best alternative approach? I'm thinking about getting message list via EWS/IMAP, showing them and (at least now) redirecting to OWA in order to display the message content.

    Read the article

  • Packaging MATLAB (or, more generally, a large binary, proprietary piece of software)

    - by nfirvine
    I'm trying to package MATLAB for internal distribution, but this could apply to any piece of software with the same architecture. In fact, I'm packaging multiple releases of MATLAB to be installed concurrently. Key things Very large installation size (~4 GB) Composed of a core, and several plugins (toolboxes) Initially, I created a single "source" package (matlab2011b) that builds several .debs (mainly matlab2011b-core and matlab2011b-toolbox-* for each toolbox). The control file is just the standard all: dh $@ There is no Makefile; only copying files. I use a number of debian/*.install files to specify files to copy from a copy of an installation to /usr/lib/. The problem is, every time I build the thing (say, to make a correction to the core package), it recopies every file listed in the *.install file to e.g debian/$packagename/usr/ (the build phase), and then has to bundle that into a .deb file. It takes a long time, on the order of hours, and is doing a lot of extra work. So my questions are: Can you make dh_install do a hardlink copy (like cp -l) to save time? (AFAICT from the man page, no.) Maybe I should just get it to do this in the Makefile? (That's gonna b e big Makefile.) Can you make debuild only rebuild .debs that need rebuilding? Or specify which .debs to rebuild? Is my approach completely stupid? Should I break each of the toolboxes into its own source package too? (I'll have to do some silly templating or something, because there's hundreds of them. :/)

    Read the article

  • Advanced TSQL training

    - by Dave Ballantyne
    Over the past few years, Ive had it on my to do list to write and deliver and full-scale SQLServer training course and not just an hour long bite size session at user groups and conferences.  To me, SQLServer development is not just knowing and remembering the syntax of commands.  Sometimes I semi-jest that i have “Written a merge statement without looking up the syntax”, but I know from my interactions on and off line that I am far from alone in this.  In any case we have an awesome tool in the internet which is great at looking things up. When developing SQL Server based solutions,  of more importance is knowing the internals of the engine.  SQL Server is a complex piece of software and we need to be able to understand to a fairly low level ( you can always dive deeper ) the choices that it makes and why it makes them in order to deliver performant, reliable, predictable and scalable systems to our customers and end-users. This is the view i shall be taking over two days in March (19th and 20th) in London and ,TBH, one I dont see taken often enough. Early bird discounts are available until 31Dec. Full details of the course and a high level view of the bullet points we shall be covering are available at the Technitrain site ( http://tinyurl.com/TSQLTraining )

    Read the article

  • How can I render a semi transparent model with OpenGL correctly?

    - by phobitor
    I'm using OpenGL ES 2 and I want to render a simple model with some level of transparency. I'm just starting out with shaders, and I wrote a simple diffuse shader for the model without any issues but I don't know how to add transparency to it. I tried to set my fragment shader's output (gl_FragColor) to a non opaque alpha value but the results weren't too great. It sort of works, but it looks like certain model triangles are only rendered based on the camera position... It's really hard to describe what's wrong so please watch this short video I recorded: http://www.youtube.com/watch?v=s0JqA0rZabE I thought this was a depth testing issue so I tried playing around with enabling/disabling depth testing and back face culling. Enabling back face culling changes the output slightly but the problem in the video is still there. Enabling/disabling depth testing doesn't seem to do anything. Could anyone explain what I'm seeing and how I can add some simple transparency to my model with the shader? I'm not looking for advanced order independent transparency implementations. edit: Vertex Shader: // color varying for fragment shader varying mediump vec3 LightIntensity; varying highp vec3 VertexInModelSpace; void main() { // vec4 LightPosition = vec4(0.0, 0.0, 0.0, 1.0); vec3 LightColor = vec3(1.0, 1.0, 1.0); vec3 DiffuseColor = vec3(1.0, 0.25, 0.0); // find the vector from the given vertex to the light source vec4 vertexInWorldSpace = gl_ModelViewMatrix * vec4(gl_Vertex); vec3 normalInWorldSpace = normalize(gl_NormalMatrix * gl_Normal); vec3 lightDirn = normalize(vec3(LightPosition-vertexInWorldSpace)); // save vertexInWorldSpace VertexInModelSpace = vec3(gl_Vertex); // calculate light intensity LightIntensity = LightColor * DiffuseColor * max(dot(lightDirn,normalInWorldSpace),0.0); // calculate projected vertex position gl_Position = gl_ModelViewProjectionMatrix * gl_Vertex; } Fragment Shader: // varying to define color varying vec3 LightIntensity; varying vec3 VertexInModelSpace; void main() { gl_FragColor = vec4(LightIntensity,0.5); }

    Read the article

  • Help me with this logic (newbie) [migrated]

    - by Surendra
    I need to generate a half pyramid number series with the entered starting number and the number of lines in a html page using Javascript and show the result in html page . I have done the Java scripting and stuff . What I don't get is the logic to it. Take a look at this you may get an idea what I'm talking about: Here is my function in Javascript that will be triggered on a button click function doFunction(){ var enteredNumber=document.getElementById("start"); var lines=document.getElementById("lines"); var result; for(i=0;i<=lines.value;i++) { for(j=enteredNumber.value;j<=i;j++) { document.write(j + "&nbsp;" + "&nbsp;"); } document.write("<br />"); } } Help me with the logic to print following order: 1 1 2 1 2 3 1 2 3 4 1 2 3 4 5 There is a condition. I will specify $start and $lines. If $start = 5 and $lines = 3 then output should be like: 5 5 6 5 6 7 I have had used the for loop , but that doesn't work if I give my own start number that is higher than the number of lines. I actually need it done with Javascript, I have had done the necessary but I'm confused with the logic to generate such series (with the user given values) I had actually used two for loops to generate the regular number series like below 1 1 2 1 2 3 and so on.

    Read the article

  • Why is Windows Update telling me to install Internet Explorer 10 even though it is already installed?

    - by gparyani
    I have a Windows 7 laptop. For some reason, Windows Update keeps telling me to install Internet Explorer 10 for Windows 7 even though it is already installed (downloaded from Microsoft website). Whenever I try and install it from Windows Update in order to get rid of it, I receive error 9C48. What is the problem and how do I solve it? (I've noticed several entries for similar problems online, but those are all for Internet Explorer 9). As I said in the comments, my system keeps trying to install the same update when it is shutting down, only failing with the same error. It is hogging my shut down button and wasting time during shutdown. Update: I just went and installed Internet Explorer 11 from the Windows website. However, Windows Update continues to offer me Internet Explorer 10. How can I tell it that the update is in fact installed so that it instead starts offering me updates?

    Read the article

  • Will Windows 7 setup overwrite Windows 8 and Ubuntu boot records?

    - by Jens
    I have Windows 8 and Ubuntu installed, both an the same physical hd (2 partitions). Windows 8 was installed after a format, Ubuntu afterwards. I used EasyBCD to setup the boot records and the boot menu order. I want to install Windows 7 as a third operating system and just add its boot record with EasyBCD. Will there be any unwanted behavior when I install it on a third partition? I already have 100 GB of unallocated space on the disk, though have not yet made the partition. I red on this site that I should install the oldest OS first, but than I would have to do a format and start again from scratch (which I clearly do not want to do). edit: On this site, the opposite is claimed, saying it's also possible to install Windows 7 after Windows 8. However, it will restore a Windows 7 boot menu instead one of Windows 8.

    Read the article

  • Why does compressing and decompressing my SSD hard drive free up space?

    - by Paperflyer
    I bought an SSD (SandForce 2), created a tiny 25GB partition on it for Windows and installed Windows 7 64-bit. In order to free disk space, I enabled compression on the drive using the Properties entry in the context menu for the drive in Explorer. Prior to compressing I had around 5GB of free space. After compression I had 4GB, so compression was not working for me. I figured this might have happened because of the built-in data compression of the SSD. I decompressed the files again - after decompression, it left me with 7GB of free space! Better yet, after restarting, I had 10GB. What is happening here?

    Read the article

  • OSSEC HIDS Notification "Unknown problem somewhere in the system." (seems like hdd issue)

    - by John
    from what i understand somethings is wrong with hdd i am trying to find some commands in order to run some tests to check if hard disk is OK I will post a full list of logs after REBOOT of system: "Unknown problem somewhere in the system." kernel: ata2.00: failed command: READ FPDMA QUEUED kernel: res 51/40:c8:38:5c:16/00:00:00:00:00/40 Emask 0x409 (media error) <F> kernel: ata2.00: error: { UNC } kernel: ata2.00: failed command: READ FPDMA QUEUED kernel: res 51/40:78:88:5c:16/00:00:00:00:00/40 Emask 0x409 (media error) <F> kernel: sd 1:0:0:0: [sda] Sense Key : Medium Error [current] [descriptor] kernel: sd 1:0:0:0: [sda] Add. Sense: Unrecovered read error - auto reallocate failed kernel: md/raid1:md1: read error corrected (8 sectors at 1461400 on sda1) kernel: sd 1:0:0:0: [sda] Add. Sense: Unrecovered read error - auto reallocate failed kernel: sd 1:0:0:0: [sda] Add. Sense: Unrecovered read error - auto reallocate failed kernel: md/raid1:md1: read error corrected (8 sectors at 1461672 on sda1) Also some of this logs are duplicate or even more. Thanks.

    Read the article

  • Need help: jBPM installation issue

    - by Kouky
    This is Ms Kahina, I’m getting started with jBPM & Java EE, I tried to install the jBPM 5.4 version by following all the steps stated in the iBPM user guide as below: 1- I’ve installed Java (JDK 1.7) and ant ( Apache-ant-1.9.4). 2- I’ve set both JAVA_HOME and ANT_HOME environement variables. 3- I’ve downloaded the jBPM 5.4 full installer and run the installation script “ant install.demo” and after that the start script “ant start.demo”. But unfortunately jBPM 5.4 is not working. When running the installation Script I’m getting successful message but with the following warnings : [copy] Warning: Could not find file C:\jbpm-installer\db\task-persistence.xml to copy. [copy] Warning: Could not find file C:\jbpm-installer\db\Taskorm.xml to copy. The start demo was successful. Actually when I try to open any tool provided with the jBPM such us the jBPM-console for example I’m getting the following messages: Address not found or sometimes the http status 404 occurred even if the Welcome page of jBoss AS was opened at http://localhost:8080/. Please need your assistance to sort out this issue in order to move forward in my project as I'm blocked in the installation stage since more than a week now, I don’t know if this is related to the jBoss AS7 or to any other thing that I didn’t find out. Looking forward to hearing from you. Thanks Kahina

    Read the article

  • A space-efficient filesystem for grow-as-needed virtual disks ?

    - by Steve Schnepp
    A common practice is to use non-preallocated virtual disks. Since they only grow as needed, it makes them perfect for fast backup, overallocation and creation speed. Since file systems are usually based on physical disks they have the tendency to use the whole area available1 in order to increase the speed2 or reliability3. I'm searching a filesystem that does the exact opposite : try to touch the minimum blocks need by an aggressive block reuse. I would happily trade some performance for space usage. There is already a similar question, but it is rather general. I have very specific goal : space-efficiency. 1. Like page caching uses all the free physical memory 2. Canonical example : online defragmentation 3. Canonical example : snapshotting

    Read the article

  • LDAP authentication: Windows Server2k3 vs. 2k8

    - by wolfgangsz
    We have around 70% linux users, all of which are configured to authenticate against Active Directory through LDAP. In order for this to work, we used the "Windows Services for Unix" under Windows Server 2003, and it all works fine. We are now at a point where the server running this contraption is getting a bit tired and will be replaced with a newer machine, running Windows Server 2008 (where the relevant services such as user name mapping and password changes, etc., are integrated with the OS). And here's the rub: If a new user is configured through the Win2k3 server, then it all works fine. If the same thing is done through the Win2k8 server, then : The ADS plugin on the 2k3 server does not recognize it and behaves as if the UNIX attributes were never set. The user cannot authenticate against ADS using LDAP. Has anybody encountered this problem? If so, how did you overcome this? If you need any additional information to provide further help, just ask and I shall provide it.

    Read the article

  • Permissions issue: how can Apache access files in my Home directory?

    - by richzilla
    I know file permissions have been covered on here before, but im struggling to get my head around the concept for my scenario. I created the files on an old ubuntu installation. Ive copied the files into my new ubuntu installation and put them in my webroot. When i attempt to run the files (theyre PHP files) i get an error relating to permissions in an attempt to fix this, i assumed that they must still be owned by the previous owner, so i ran chown -R on the directory, with my username as an argument, in order to take ownership of all of the files in the directory. It should be noted that the usernames between new and old ubuntu installations were the same. When i attempt to run the files again, same problem: 500 error due to permissions problems. Can anyone tell me what other steps i should take? The webroot for my apache installation is inside my home folder. If i create new files in my webroot, they also work as expected, its only the old files that are causing the problem.

    Read the article

  • GRUB2 stuck at rescue console, showing "unknown filesystem" for all partitions

    - by AndiDog
    I installed Ubuntu 12.04 on my external USB drive, where I have a 700GB NTFS partition followed by the new 6GB ext4 partition and a swap partition (all primary). The GRUB MBR is also installed to the external hard disk. Since my BIOS puts the external drive as first disk when booting, I removed my internal hard disk before installation in order to avoid ordering problems. Now when I boot from the external drive, GRUB is stuck at the rescue console with the error "unknown filesystem". grub rescue> ls (hd0) (hd0,msdos3) (hd0,msdos2) (hd0,msdos1) ls (hd0,<any of them>)/ gives me "unknown filesystem", thus also "insmod normal" GRUB doesn't seem to be able to read my Linux partition as you can see above?! How can I solve this? Additional info: bootinfoscript says (this is with the internal drive in again, but that does not make a difference): Grub2 (v1.99) is installed in the MBR of /dev/sdb and looks at sector 1 of the same hard drive for core.img. core.img is at this location and looks for (,msdos2)/boot/grub on this drive. sdb1: __________________________________________________________________________ File system: ntfs Boot sector type: Windows Vista/7: NTFS Boot sector info: No errors found in the Boot Parameter Block. Operating System: Boot files: sdb2: __________________________________________________________________________ File system: ext4 Boot sector type: - Boot sector info: Operating System: Ubuntu 12.04 LTS Boot files: /boot/grub/grub.cfg /etc/fstab /boot/grub/core.img sdb3: __________________________________________________________________________ File system: swap Boot sector type: - Boot sector info:

    Read the article

  • Small 64-bit linux server distro with low memory footprint?

    - by djangofan
    I am looking for a linux server distro with a low memory footprint. I usually use Ubuntu but I need something with a smaller footprint in order to run a large Java JVM service inside of it and also run X-windows. Any ideas? The Java service needs to handle a 3GB memory heap and so I require a 64-bit OS and JRE. http://en.wikipedia.org/wiki/Comparison%5Fof%5FLinux%5Fdistributions I am thinking that ArchLinux is the only one that I can find right now. It uses 250MB out of the box (without X-win). Any better suggestions?

    Read the article

  • Google Chrome Automatically Save Passwords

    - by benhc911
    I have seen, and at one point implemented, methods of modifying firefox's password manager popup such that it would automatically save usernames and passwords without prompt. I was wondering if there is a similar feature for Chrome. I just recently switched to chrome and it is one of a few features/modifications that didnt survive the transition, and while I am moderately capable in programming in certain languages (C+, visual etc) I don't really feel confident enough to muck around unguided. For anyone that is particularly interested, the other features I miss are things such as the slash based 'quick search' in firefox... and I am currently annoyed by a different right click menu order for opening in new tabs and new windows (the reverse of firefox). Any help would be greatly appreciated!

    Read the article

  • Ubuntu 12.04 menu bar, nautilus, terminal, and gtk themes not working after installation of Gimp 2.8

    - by Chris
    I installed gimp2.8 from this ppa: ppa:otto-kesselgulasch/gimp after that, my system began having problems. This is my thought process in trying to fix what's happened and the order in which it happened: I noticed the menu bar at the top changed from an opaque black to perfectly clear and the titles of applications and the hidden buttons reacted slowly. No big deal, restarted to see if it fixed it. It didn't, in fact, when the logon screen came up, the password field was grey and boxy like a default windows 98 theme (that's the best I can describe it) as were all the option buttons for gtk programs. I open terminal to try and reinstall gtk, but the terminal is just a black screen with no ability to input commands. I go to a tty and I reinstalled gtk3 and gtk2 (I have both on my system. I don't think they're in conflict, they hadn't been before hand). I restarted. Nothing doing. Log in, nautilus isn't placing icons on my desktop. I click the launcher. It flashes, but no window opens. Try to open by Alt+f2, nothing. I purge ubuntu-desktop, restart, reinstall ubuntu-desktop. Nothing. I have no clue what to do at this point so I'm asking for any help diagnosing the problem and fixing it.

    Read the article

  • How to specify multiple HostName/Port combinations in .ssh/config

    - by leoluk
    I have multiple notebooks and workstations which pull and push from multiple Mercurial repositories on a central server. I usually use .ssh/config to set an alias: Host repo-server HostName server.somedomain User user143 IdentityOnly yes IdentityFile ~/hgkey Port 156 ... and some more options, you get the idea. I can then simply do a hg push ssh://repo-server//hgroot/someproject on every local repository, and I can change the server address and port in one place. For workstations, this works fine, but the notebooks can access the server either from inside the network or from outside, using a different address and a different port. Is there any way I can specify multiple HostName/Port combinations so that SSH automatically tries them in order? This way, the users could push and pull without having to care about the correct address. (of course, using a VPN would be the most correct solution)

    Read the article

  • Apple Remote Desktop and Screen Sharing

    - by jfm429
    We have a Mac OS 10.8.2 server that we want to be able to administer with Apple Remote Desktop. At the same time, we want normal users to be able to access their account screens (through background login) without being able to view the current screen. However, in order to enable this (by enabling the "normal" Screen Sharing option in System Preferences) Remote Desktop needs to be disabled. The question is - how can we run both Remote Desktop for administrators and VNC screen sharing for normal users while restricting normal users to logging in on a background window instead of viewing the front screen?

    Read the article

  • Music player for ubuntu similar to iTunes

    - by Allwar
    Hi, I really like Itunes as a music player but i can't get it to work in ubuntu so I need another one! functions i want: shuffle, low cpu usage, a friendly interface and free of bugs. (friendly interface = esay to use) why not use rythmbox, well it has bugs, the shuffle button doesn't work because it shuffles the next song after each song. So when i use it it's just a big loop over three different songs! i want the itunes look, like it shuffles once and lets you see in what order the songs will be played. //Alvar

    Read the article

  • Hello!

    - by barryoreilly
    After many months of deliberating I have finally gotten around to starting this blog! The reason for doing this is the large number of half finished articles lying around on my hard disk, unpublished and unloved. These articles have been of huge benefit to me, and have been written in an attempt to consolidate my own thinking, in order to help me structure my thoughts and ideas as I have tried to digest new ideas and understand abstract theories. It is my hope that by tidying up these articles and publishing them here that I can continue this learning process by getting feedback on the ideas from within the developer community. i have worked with .NET for 8 years now, and have worked with ASP.NET, SQL Server, Windows programming as well as general network administration. Since 2004 my focus has been on integration, web services, and more often than not Biztalk Server. The last two years have seen me focus on SOA and WCF, and the Managed Services Engine, so this is probably where the main focus of the blog will to start with, but there are so many fun things to play with these days that i have no idea where it will end up.....   Barry

    Read the article

  • what to use instead of laptop-mode?

    - by playcat
    hello, i have ubuntu 10.10 64bit on hp 6735s (turion processor). it overheats, and i'm forced to use turion power control in order to keep core temperature to a reasonable level. one more measure that i use is putting my processors to conservative mode. that way, i'm perfectly happy with its performance, and heat is where it should be. however, after my latest upgrade, something happened - cores are back to ondemand by default, and i'm not sure if turionpowercontrol is working any more (ps axu | grep urion shows no process). in addition, i read somewhere that laptop-mode uses hdd spindown for preserving data/energy, and that hdds have only a limited amount of those spindowns, so laptop-mode usage can actually shorten the life of my hdd. i'm wondering if there is a good way to set my cores to automatically go to conservative mode? also, what's the good way to see what is the voltage my cores use? on windows i use cpuz tools. thx and sorry for the long explanation.

    Read the article

  • Preferred way for dealing with customer-defined data in enterprise application

    - by Axarydax
    Let's say that we have a small enterprise web (intranet) application for managing data for car dealers. It has screens for managing customers, inventory, orders, warranties and workshops. This application is installed at 10 customer sites for different car dealers. First version of this application was created without any way to provide for customer-specific data. For example, if dealer A wanted to be able to attach a photo to a customer, dealer B wanted to add e-mail contact to each workshop, and dealer C wanted to attach multiple PDF reports to a warranty, each and every feature like this was added to the application, so all of the customers received everything on new update. However, this will inevitably lead to conflicts as the number of customers grow as their usage patterns are unique, and if, for instance, a specific dealer requested to have an ability to attach (for some reason) a color of inventory item (and be able to search by this color) as a required item, others really wouldn't need this feature, and definitely will not want it to be a required item. Or, one dealer would like to manage e-mail contacts for their employees on a separate screen of the application. I imagine that a solution for this is to use a kind of plugin system, where we would have a core of the application that provides for standard features like customers, inventory, etc, and all of the customer's installed plugins. There would be different kinds of plugins - standalone screens like e-mail contacts for employees, with their own logic, and customer plugin which would extend or decorate inventory items (like photo or color). Inventory (customer,order,...) plugins would require to have installation procedure, hooks for plugging into the item editor, item displayer, item filtering for searching, backup hook and such. Is this the right way to solve this problem?

    Read the article

  • MVVM Light V4 preview (BL0014) release notes

    - by Laurent Bugnion
    I just pushed to Codeplex an update to the MVVM Light source code. This is an early preview containing some of the features that I want to release later under the version 4. If you find these features useful for your project, please download the source code and build the assemblies. I will appreciate greatly any issue report. This version is labeled “V4.0.0.0/BL0014”. The “BL” string is an old habit that we used in my days at Siemens Building Technologies, called a “base level”. Somehow I like this way of incrementing the “base level” independently of any other consideration (such as alpha, beta, CTP, RTM etc) and continue to use it to tag my software versions. In Microsoft parlance, you could say that this is an early CTP of MVVM Light V4. Caveat The code is unit tested, but as we all know this does not mean that there are no bugs This code has not yet been used in production. Again, your help in testing this is greatly appreciated, so please report all bugs to me! What’s new? The following features have been implemented: Misc Various “maintenance work”. All WPF assemblies (that is .NET35 and .NET4) now allow partially trusted callers. It means that you can use them in am XBAP in partial trust mode. Testing Various test updates Added Windows Phone 7 unit tests Note: For Windows Phone 7, due to an issue in the unit test framework, not all tests can be executed. I had to isolate those tests for the moment. The error was reported to Microsoft. ViewModelBase The constructor is now public to allow serialization (especially useful on the phone to tombstone the state). ViewModelBase.MessengerInstance now returns Messenger.Default unless it is set explicitly. Previously, MessengerInstance was returning null, which was complicating the code. Two new ways to raise the PropertyChanged event have been added. See below for details. Messenger Updated the IMessenger interface with all public members from the Messenger class. Previously some members were missing. A new Unregister method is now available, allowing to unregister a recipient for a given token. RelayCommand RaiseCanExecuteChanged now acts the same in Windows Presentation Foundation than in Silverlight. In previous versions, I was relying on the CommandManager to raise the CanExecuteChanged event in WPF. However, it was found to be too unreliable, and a more direct way of raising the event was found preferable. See below for details. Raising the PropertyChanged event A very much requested update is now included: the ability to raise the PropertyChanged event in a viewmodel without using “magic strings”. Personally, I don’t see strings as a major issue, thanks to two features of the MVVM Light Toolkit: In the DEBUG configuration, every time that the RaisePropertyChanged method is called, the name of the property is checked against all existing properties of the viewmodel. Should the property name be misspelled (because of a typo or refactoring), an exception is thrown, notifying the developer that something is wrong. To avoid impacting the performance, this check is only made in DEBUG configuration, but that should be enough to warn the developers in case they miss a rename. The property name is defined as a public constant in the “mvvminpc” code snippet. This allows checking the property name from another class (for example if the PropertyChanged event is handled in the view). It also allows changing the property name in one place only. However, these two safeguards didn’t satisfy some of the users, who requested another way to raise the PropertyChanged event. In V4, you can now do the following: Using lambdas private int _myProperty; public int MyProperty { get { return _myProperty; } set { if (_myProperty == value) { return; } _myProperty = value; RaisePropertyChanged(() => MyProperty); } } This raises the property changed event using a lambda expression instead of the property name. Light reflection is used to get the name. This supports Intellisense and can easily be refactored. You can also broadcast a PropertyChangedMessage using the Messenger.Default instance with: private int _myProperty; public int MyProperty { get { return _myProperty; } set { if (_myProperty == value) { return; } var oldValue = _myProperty; _myProperty = value; RaisePropertyChanged(() => MyProperty, oldValue, value, true); } } Using no arguments When the RaisePropertyChanged method is called within a setter, you can also omit the property name altogether. This will fail if executed outside of the setter however. Also, to avoid confusion, there is no way to broadcast the PropertyChangedMessage using this syntax. private int _myProperty; public int MyProperty { get { return _myProperty; } set { if (_myProperty == value) { return; } _myProperty = value; RaisePropertyChanged(); } } The old way Of course the “old” way is still supported, without broadcast: public const string MyPropertyName = "MyProperty"; private int _myProperty; public int MyProperty { get { return _myProperty; } set { if (_myProperty == value) { return; } _myProperty = value; RaisePropertyChanged(MyPropertyName); } } And with broadcast: public const string MyPropertyName = "MyProperty"; private int _myProperty; public int MyProperty { get { return _myProperty; } set { if (_myProperty == value) { return; } var oldValue = _myProperty; _myProperty = value; RaisePropertyChanged(MyPropertyName, oldValue, value, true); } } Performance considerations It is notorious that using reflection takes more time than using a string constant to get the property name. However, after measuring for all platforms, I found the differences to be very small. I will measure more and submit the results to the community for evaluation, because some of the results are actually surprising (for example, using the Messenger to broadcast a PropertyChangedMessage does not significantly increase the time taken to raise the PropertyChanged event and update the bindings). For now, I submit this code to you, and would be delighted to hear about your own results. Raising the CanExecuteChanged event manually In WPF, until now, the CanExecuteChanged event for a RelayCommand was raised automatically. Or rather, it was attempted to be raised, using a feature that is only available in WPF called the CommandManager. This class monitors the UI and when something occurs, it queries the state of the CanExecute delegate for all the commands. However, this proved unreliable for the purpose of MVVM: Since very often the value of the CanExecute delegate changes according to non-UI events (for example something changing in the viewmodel or in the model), raising the CanExecuteChanged event manually is necessary. In Silverlight, the CommandManager does not exist, so we had to raise the event manually from the start. This proved more reliable, and I now changed the WPF implementation of the RaiseCanExecuteChanged method to be the exact same in WPF than in Silverlight. For instance, if a command must be enabled when a string property is set to a value other than null or empty string, you can do: public MainViewModel() { MyTestCommand = new RelayCommand( () => DoSomething(), () => !string.IsNullOrEmpty(MyProperty)); } public const string MyPropertyName = "MyProperty"; private string _myProperty = string.Empty; public string MyProperty { get { return _myProperty; } set { if (_myProperty == value) { return; } _myProperty = value; RaisePropertyChanged(MyPropertyName); MyTestCommand.RaiseCanExecuteChanged(); } } Logo update I made a minor change to the logo: Some people found the lack of the word “light” (as in MVVM Light Toolkit) confusing. I thought it was cool, because the feather suggests the idea of lightness, however I can see the point. So I added the word “light” to the logo. Things should be quite clear now. What’s next? This is only the first of a series of releases that will bring MVVM Light to V4. In the next weeks, I will continue to add some very requested features and correct some issues in the code. I will probably continue this fashion of releasing the changes to the public as source code through Codeplex. I would be very interested to hear what you think of that, and to get feedback about the changes. Cheers, Laurent   Laurent Bugnion (GalaSoft) Subscribe | Twitter | Facebook | Flickr | LinkedIn

    Read the article

< Previous Page | 391 392 393 394 395 396 397 398 399 400 401 402  | Next Page >