Search Results

Search found 13727 results on 550 pages for 'target platform'.

Page 444/550 | < Previous Page | 440 441 442 443 444 445 446 447 448 449 450 451  | Next Page >

  • How to deal with calculated values with Dependency Properties on a custom WPF control

    - by jpierson
    To summarize what I'm doing, I have a custom control that looks like a checked listbox and that has two dependency properties one that provides a list of available options and the other that represents a enum flag value that combines the selection options. So as I mentioned my custom control exposes two different DependencyProperties, one of which is a list of options called Options and the other property called SelectedOptions is of a specific Enum type that uses the [Flags] attribute to allow combinations of values to be set. My UserControl then contains an ItemsControl similar to a ListBox that is used to display the options along with a checkbox. When the check box is checked or unchecked the SelectedOptions property should be updated accordingly by using the corresponding bitwise operation. The problem I'm experiencing is that I have no way other than resorting to maintaining private fields and handling property change events to update my properties which just feels unatural in WPF. I have tried using ValueConverters but have run into the problem that I can't really using binding with the value converter binding so I would need to resort to hard coding my enum values as the ValueConverter parameter which is not acceptable. If anybody has seen a good example of how to do this sanely I would greatly appreciate any input. Side Note: This has been a problem I've had in the past too while trying to wrap my head around how dependency properties don't allow calculated or deferred values. Another example is when one may want to expose a property on a child control as a property on the parent. Most suggest in this case to use binding but that only works if the child controls property is a Dependency Property since placing the binding so that the target is the parent property it would be overwritten when the user of the parent control wants to set their own binding for that property.

    Read the article

  • How to forward a 'saved' request stream to another Action within the same controller?

    - by Moe Howard
    We have a need to chunk-up large http requests sent by our mobile devices. These smaller chunk streams are merged to a file on the server. Once all chunks are received we need a way to submit the saved merged request to an another method(Action) within the same controller that will process this large http request. How can this be done? The code we tried below results in the service hanging. Is there a way to do this without a round-trip? //Open merged chunked file FileStream fileStream = new FileStream(fileName, FileMode.Open, FileAccess.Read); //Read steam support variables int bytesRead = 0; byte[] buffer = new byte[1024]; //Build New Web Request. The target Action is called "Upload", this method we are in is called "UploadChunk" HttpWebRequest webRequest; webRequest = (HttpWebRequest)WebRequest.Create(Request.Url.ToString().Replace("Chunk", string.Empty)); webRequest.Method = "POST"; webRequest.ContentType = "text/xml"; webRequest.KeepAlive = true; webRequest.Timeout = 600000; webRequest.ReadWriteTimeout = 600000; webRequest.Credentials = System.Net.CredentialCache.DefaultCredentials; Stream webStream = webRequest.GetRequestStream(); //Hangs here, no errors, just hangs I have looked into using RedirectToAction and RedirecctToRoute but these methods don't fit well with what we are looking to do as we cannot edit the Request.InputStream (as it is read-only) to carry out large request stream. Thanks, Moe

    Read the article

  • Overlay an HTML page with an HTML form

    - by jah
    Hi folks, this is a question about the best way (or least effort of the best ways) to overlay an html page with a form. Best in this context meaning best user experience whilst meeting the functional requirements. Let's say I have a page with a short form on it; the user has to enter some financial details. To assist the user to enter an accurate value for one of the fields there's another, much longer form. The longer form needs to be displayed only if the user requests the help. For users without javascript, clicking a link will submit the short form (persisting already filled fields in a session) and the server will respond with the long form. They'll submit the long form and the server will combine the submitted data with the persisted data and serve the short form again - with the fields populated. For users with javascript I want to overlay the short form page (in a lightbox stylee) with the long form, allow them to populate the long form and then go back to the short form with less round-trips to the server. Do I: Overlay the short form page with an iframe whose target is the long form? Request the long form over ajax and stuff it into a div? Generate the long form entirely on the client-side? Some other wizadry I haven't thought of? A short explanation of the best mechanism will do me very nicely indeed. Thank you very much!

    Read the article

  • Reference for proper handling of PID file on Unix

    - by bignose
    Where can I find a well-respected reference that details the proper handling of PID files on Unix? On Unix operating systems, it is common practice to “lock” a program (often a daemon) by use of a special lock file: the PID file. This is a file in a predictable location, often ‘/var/run/foo.pid’. The program is supposed to check when it starts up whether the PID file exists and, if the file does exist, exit with an error. So it's a kind of advisory, collaborative locking mechanism. The file contains a single line of text, being the numeric process ID (hence the name “PID file”) of the process that currently holds the lock; this allows an easy way to automate sending a signal to the process that holds the lock. What I can't find is a good reference on expected or “best practice” behaviour for handling PID files. There are various nuances: how to actually lock the file (don't bother? use the kernel? what about platform incompatibilities?), handling stale locks (silently delete them? when to check?), when exactly to acquire and release the lock, and so forth. Where can I find a respected, most-authoritative reference (ideally on the level of W. Richard Stevens) for this small topic?

    Read the article

  • Why isn’t my autoreleased object getting released?

    - by zoul
    Hello. I am debugging a weird memory management error and I can’t figure it out. I noticed that some of my objects are staying in memory longer than expected. I checked all my memory management and finally got to the very improbable conclusion that some of my autorelease operations don’t result in a release. Under what circumstances is that possible? I created a small testing Canary class that logs a message in dealloc and have the following testing code in place: NSLog(@"On the main thread: %i.", [NSThread isMainThread]); [[[Canary alloc] init] autorelease]; According to the code we’re really on the main thread, but the dealloc in Canary does not get called until much later. The delay is not deterministic and can easily take seconds or more. How is that possible? The application runs on a Mac, the garbage collection is turned off (Objective-C Garbage Collection is set to Unsupported on the target.) I am mostly used to iOS, is memory management on OS X different in some important way?

    Read the article

  • ASP.NET MVC - Entending the Authorize Attribute

    - by Mad Halfling
    Hi folks, currently I use [Authorize(Roles = ".....")] to secure my controller actions on my ASP.NET MVC 1 app, and this works fine. However, certain search views need to have buttons that route to these actions that need to be enabled/disabled based on the record selected on the search list, and also the security privs of the user logged in. Therefore I think I need to have a class accessing a DB table which cross-references these target controller/actions with application roles to determine the state of these buttons. This will, obviously, make things messy as privs will need to be maintained in 2 places - in that class/DB table and also on the controller actions (plus, if I want to change the access to the action I will have to change the code and compile rather than just change a DB table entry). Ideally I would like to extend the [Authorize] functionality so that instead of having to specify the roles in the [Authorize] code, it will query the security class based on the user, controller and action and that will then return a boolean allowing or denying access. Are there any good articles on this - I can't imagine it's an unusual thing to want to do, but I seem to be struggling to find anything on how to do it (could be Monday-morning brain). I've started some code doing this, looking at article http://schotime.net/blog/index.php/2009/02/17/custom-authorization-with-aspnet-mvc/ , and it seems to be starting off ok but I can't find the "correct" way to get the calling controller and action values from the httpContext - I could possibly fudge a bit of code to extract them from the request url, but that doesn't seem right to me and I'd rather do it properly. Cheers MH

    Read the article

  • WinForms Load Event / Static Initialization Strangeness

    - by Eric J.
    Background I'm troubleshooting an WinForms 2.0 program that's already been burned to CD for distribution to an internet-challenged target audience. Some users are experiencing a fatal error that I can reproduce locally. Reproducing the Error I get the fatal error when I log into my Vista box using a standard user that I just created, even if I run the program as administrator. I do not get the fatal error when I log in as local administrator. I'm not sure that being administrator is necessarily the trigger (since runas did not help). I have reproduced this half a dozen times under each account with consistent results. The faulty code Base.cs (base class for several user controls, only one of which is shown on first screen) private void BaseWindow_Load(object sender, EventArgs e) { // This message shown once in both cases MessageBox.Show("BaseWindow_Load for " + this.GetType().FullName); SkinManager.ApplySkin(this); } SkinManager.cs private static Skin skin = null; public static void ApplySkin(UserControl applyTo) { if (skin == null) { skin = new Skin(SkinsDirectory, "Default"); } } Skin.cs internal Skin(string skinPath, string skinName) { config = SkinConfig.Load(path); } SkinConfig.cs public static SkinConfig Load(string path) { // This message shown only once running as Admin but twice running as standard user System.Windows.Forms.MessageBox.Show("@1"); // !!! LOCK path HERE !!! } A user control loads on the first form, which triggers a call to SkinManager.ApplySkin, which checks if skin is null and, if so assigns it (without thread synchronization or recursion protection), which ultimately causes a file to be opened. When logged in as local admin, that sequence completes just fine. When logged in as my test standard user, ApplySkin is always called a second time while skin is still null, causing a second attempt to load, causing the file to be locked on the second attempt. The error handling is draconian at this point and the program terminates. The Question While this code can be easily fixed, I would like to understand why the error is happening only in some cases.

    Read the article

  • Javascript add a PHP file.

    - by Jordan Pagaduan
    var editing = false; if (document.getElementById && document.createElement) { var butt = document.createElement('BUTTON'); var buttext = document.createTextNode('Ready!'); butt.appendChild(buttext); butt.onclick = saveEdit; } function catchIt(e) { if (editing) return; if (!document.getElementById || !document.createElement) return; if (!e) var obj = window.event.srcElement; else var obj = e.target; while (obj.nodeType != 1) { obj = obj.parentNode; } if (obj.tagName == 'TEXTAREA' || obj.tagName == 'A') return; while (obj.nodeName != 'P' && obj.nodeName != 'HTML') { obj = obj.parentNode; } if (obj.nodeName == 'HTML') return; var x = obj.innerHTML; var y = document.createElement('TEXTAREA'); var z = obj.parentNode; z.insertBefore(y,obj); z.insertBefore(butt,obj); z.removeChild(obj); y.value = x; y.focus(); editing = true; } function saveEdit() { var area = document.getElementsByTagName('TEXTAREA')[0]; var y = document.createElement('P'); var z = area.parentNode; y.innerHTML = area.value; z.insertBefore(y,area); z.removeChild(area); z.removeChild(document.getElementsByTagName('button')[0]); editing = false; } document.onclick = catchIt; This code is a quick edit and I want to add a PHP script that will UPDATE my database base on the changes on the text.

    Read the article

  • Rails deployment strategies with Bundler and JRuby

    - by brad
    I have a jruby rails app and I've just started using bundler for gem dependency management. I'm interested in hearing peoples' opinions on deployment strategies. The docs say that bundle package will package your gems locally so you don't have to fetch them on the server (and I believe warbler does this by default), but I personally think (for us) this is not the way to go as our deployed code (in our case a WAR file) becomes much larger. My preference would be to mimic our MVN setup which fetches all dependencies directly on the server AFTER the code has been copied there. Here's what I'm thinking, all comments are appreciated: Step1: Build war file, copy to server Step2: Unpack war on server, fetch java dependencies with mvn Step3: use Bundler to fetch Gem deps (Where should these be placed??) * Step 3 is the step I'm a bit unclear on. Do I run bundle install with a particular target in mind?? Step4: Restart Tomcat Again my reasoning here is that I'd like to keep the dependencies separate from the code at deploy time. I'd also like to place all gem dependencies in the app itself so they are contained, rather than installing them in the app user's home directory (as, again, I believe is the default for Bundler)

    Read the article

  • How do I log from inside my web application in Tomcat 6.

    - by Carlos
    How do I log from within my web application deployed on Tomcat 6? Where should I expect the logging output to go (internal tomcat log files, or will another logfile be generated)? I see a ton of documentation but am having a hard time finding a direct answer to the above questions. Where should I expect the logging to show up (currently it is log4j is not generating a log file and it is not showing up in my console). I am trying to follow http://www.laliluna.de/articles/log4j-tutorial.html . ### direct log messages to stdout ### log4j.appender.stdout=org.apache.log4j.ConsoleAppender log4j.appender.stdout.Target=System.out log4j.appender.stdout.layout=org.apache.log4j.PatternLayout log4j.appender.stdout.layout.ConversionPattern=%d{ABSOLUTE} %5p %c{1}:%L - %m%n ### file appender log4j.appender.file=org.apache.log4j.RollingFileAppender log4j.appender.file.maxFileSize=100KB log4j.appender.file.maxBackupIndex=5 log4j.appender.file.File=test.log log4j.appender.file.threshold=info log4j.appender.file.layout=org.apache.log4j.PatternLayout log4j.appender.file.layout.ConversionPattern=%d{ABSOLUTE} %5p %c{1}:%L - %m%n log4j.rootLogger=debug, stdout In my application I define a log object: private static org.apache.log4j.Logger log = Logger.getLogger(MyClass.class); log.error("LOGGING!"); Thanks for the help.

    Read the article

  • Cocos2D: Problem Rotating a CCMenu

    - by srikanth rongali
    If I try to do actions over menuItems but the actions are not running as expected. I think code below should make the menuItem rotate by 90 degrees but when I run it, the menuItem translates from its coordinates to another coordinate then returns to its original coordinate. The complete translation takes 3 seconds. What I need is for the menuItem to rotate by 90 degrees in place within a 3 seconds duration. Please explain where I have done wrong? CCMenuItemImage *targetE;//Globally declared CCMenu *menu;//Globally declared -(id)init { if( (self = [super init]) ) { isTouchEnabled = YES; CGSize windowSize = [[CCDirector sharedDirector] winSize]; targetE = [CCMenuItemImage itemFromNormalImage:@"grossinis_sister1.png" selectedImage:@"grossinis_sister1.png" target:self selector:@selector(touch:)]; menu = [CCMenu menuWithItems:targetE,nil]; id action4 = [CCRotateBy actionWithDuration:3.0 angle:90]; [menu runAction: [CCSequence actions: action4, nil]]; menu.position = ccp(windowSize.width/2 + 200, windowSize.height/2); [self addChild: menu z:10]; } return self; } @end Thank You.

    Read the article

  • ejb lookup failing with NamingException

    - by Drake
    I've added the following in my web.xml: <ejb-ref> <ejb-ref-name>ejb/userManagerBean</ejb-ref-name> <ejb-ref-type>Session</ejb-ref-type> <home>gha.ywk.name.entry.ejb.usermanager.UserManagerHome</home> <remote>what should go here??</remote> </ejb-ref> The following java code is giving me NamingException: public UserManager getUserManager () throws HUDException { String ROLE_JNDI_NAME = "ejb/userManagerBean"; try { Properties props = System.getProperties(); Context ctx = new InitialContext(props); UserManagerHome userHome = (UserManagerHome) ctx.lookup(ROLE_JNDI_NAME); UserManager userManager = userHome.create(); WASSSecurity user = userManager.getUserProfile("user101", null); return userManager; } catch (NamingException e) { log.error("Error Occured while getting EJB UserManager" + e); return null; } catch (RemoteException ex) { log.error("Error Occured while getting EJB UserManager" + ex); return null; } catch (CreateException ex) { log.error("Error Occured while getting EJB UserManager" + ex); return null; } } The code is used inside the container. By that I mean that the .WAR is deployed on the server (Sun Application Server). StackTrace (after jsight's suggestion): >Exception occurred in target VM: com.sun.enterprise.naming.java.javaURLContext.<init>(Ljava/util/Hashtable;Lcom/sun/enterprise/naming/NamingManagerImpl;)V java.lang.NoSuchMethodError: com.sun.enterprise.naming.java.javaURLContext.<init>(Ljava/util/Hashtable;Lcom/sun/enterprise/naming/NamingManagerImpl;)V at com.sun.enterprise.naming.java.javaURLContextFactory.getObjectInstance(javaURLContextFactory.java:32) at javax.naming.spi.NamingManager.getURLObject(NamingManager.java:584) at javax.naming.spi.NamingManager.getURLContext(NamingManager.java:533) at javax.naming.InitialContext.getURLOrDefaultInitCtx(InitialContext.java:279) at javax.naming.InitialContext.lookup(InitialContext.java:351) at gov.hud.pih.eiv.web.EjbClient.EjbClient.getUserManager(EjbClient.java:34)

    Read the article

  • Java: How do you really force a GC using JVMTI's ForceGargabeCollection?

    - by WizardOfOdds
    I'm not looking for the usual "you can only hint the GC in Java using System.gc()" answers, this is not at all what this question is about. My questions is not subjective and is based on a reality: GC can be forced in Java for a fact. A lot of programs that we use daily do it: IntelliJ IDEA, NetBeans, VisualVM. They all can force GC to happen. How is it done? I take it they're all using JVMTI and more specifically the ForceGarbabeCollection (notice the "Force") but how can I try it for myself? http://java.sun.com/javase/6/docs/platform/jvmti/jvmti.html#ForceGarbageCollection Also note that this question is not about "why" I'd want to do this: the "why" may be "curiosity" or "we're writing a program similar to VisualVM", etc. The question is really "how do you force a GC using JVMTI's ForceGarbageCollection"? Does the JVM needs to be launched with any special parameters? Is any JNI required? If so, what code exactly? Does it only work on Sun VMs? Any complete and compilable example would be most welcome.

    Read the article

  • Bad linking in Qt unit test -- missing the link to the moc file?

    - by dwj
    I'm trying to unit test a class that inherits QObject; the class itself is located up one level in my directory structure. When I build the unit test I get the standard unresolved errors if a class' MOC file cannot be found: test.obj : error LNK2001: unresolved external symbol "public: virtual void * __thiscall UnitToTest::qt_metacast(char const *)" (?qt_metacast@UnitToTest@@UAEPAXPBD@Z) + 2 missing functions The MOC file is created but appears to not be linking. I've been poking around SO, the web, and Qt's docs for quite a while and have hit a wall. How do I get the unit test to include the MOC file in the link? ==== My project file is dead simple: TEMPLATE = app TARGET = test DESTDIR = . CONFIG += qtestlib INCLUDEPATH += . .. DEPENDPATH += . HEADERS += test.h SOURCES += test.cpp ../UnitToTest.cpp stubs.cpp DEFINES += UNIT_TEST My directory structure and files: C:. | UnitToTest.cpp | UnitToTest.h | \---test | test.cpp (Makefiles removed for clarity) | test.h | test.pro | stubs.cpp | +---debug | UnitToTest.obj | test.obj | test.pdb | moc_test.cpp | moc_test.obj | stubs.obj Edit: Additional information The generated Makefile.Debug shows the moc file missing: SOURCES = test.cpp \ ..\test.cpp \ stubs.cpp debug\moc_test.cpp OBJECTS = debug\test.obj \ debug\UnitToTest.obj \ debug\stubs.obj \ debug\moc_test.obj

    Read the article

  • Creating customized .dmg files upon download

    - by Marten
    I want to distribute a cross-platform application for which the executable file is slightly different, depending on the user who downloaded it. This is done by having a placeholder string somewhere in the executable that is replaced with something user-specific upon download. The webserver that has to do these string replacements is a Linux machine. For Windows, the executable is not compressed in the installer .exe, so the string replacement is easy. For uncompressed Mac OS X .dmg files, this is also easy. However, .dmg files that are compressed with either gzip or bzip2 are not so easy. For example, in the latter case, the compressed .dmg is not one big bzip2-compressed disk image, but instead consists of a few different bzip2-compressed parts (with different block sizes) and a plist suffix. Also, decompressing and recompressing the different parts with bzip2 does not result in the original data, so I'm guessing Apple uses some different parameters to bzip2 than the command-line tool. Is there a way to generate a compressed .dmg from an uncompressed one on Linux (which does not have hdiutil)? Or maybe another suggestion for creating customized applications without pregenerating them? It should work without any input by the user.

    Read the article

  • Error displaying a WinForm in Design mode with a custom control on it.

    - by George
    I have a UserControl that is part of a Class library. I reference this project from my solution. This adds a control from the referenced project to my toolbox. I add tghe control to a form. Everything looks good, I compile all and run. Perfect... But when I close the .frm with the control on it and re-open it, I get this error. The code continues to run. It may have something to do with namespaces. The original namespace was simply "Design" and this was ambiguous and conflicting so i decided to rename it. I think that's when my problems began. To prevent possible data loss before loading the designer, the following errors must be resolved: 2 Errors Ignore and Continue Why am I seeing this page? Could not find type 'Besi.Winforms.HtmlEditor.Editor'. Please make sure that the assembly that contains this type is referenced. If this type is a part of your development project, make sure that the project has been successfully built using settings for your current platform or Any CPU. Instances of this error (1) 1. There is no stack trace or error line information available for this error. Help with this error Could not find an associated help topic for this error. Check Windows Forms Design-Time error list Forum posts about this error Search the MSDN Forums for posts related to this error The variable 'Editor1' is either undeclared or was never assigned. Go to code Instances of this error (1) 1. BesiAdmin frmOrder.Designer.vb Line:775 Column:1 Show Call Stack at System.ComponentModel.Design.Serialization.CodeDomSerializerBase.Error(IDesignerSerializationManager manager, String exceptionText, String helpLink) at System.ComponentModel.Design.Serialization.CodeDomSerializerBase.DeserializeExpression(IDesignerSerializationManager manager, String name, CodeExpression expression) at System.ComponentModel.Design.Serialization.CodeDomSerializerBase.DeserializeExpression(IDesignerSerializationManager manager, String name, CodeExpression expression) at System.ComponentModel.Design.Serialization.CodeDomSerializerBase.DeserializeStatement(IDesignerSerializationManager manager, CodeStatement statement) Help with this error MSDN Help Forum posts about this error Search the MSDN Forums for posts related to this error

    Read the article

  • Flash video slooow in AIR 2 HTMLLoader component

    - by shane
    I am working on a full screen kiosk application in Flex 4/Air 2 using Flash Builder 4. We have a company training website which staff can access via the kiosk, and the main content is interactive flash training videos. Our target machines are by no means 'beefy', they are Atom n270s @ 1.6Ghz with 1Gb RAM. As it stands the videos are all but unusable when used from within the Air application, the application becomes completely unresponsive (100% cpu usage, click events take approx 5-10 seconds to register). So far I have tried: increasing the default frame rate from 24fps to 60. No improvement. nativeWindow.stage.frameRate = 60; running the videos in a stripped down version of my app, just a full screen HTMLLoader component pointed at the training website. No better than before. disabled hyper threading. The Atom CPU is split into two virtual cores, and the AIR app was only able to use one thread so maxed out at 50% CPU usage. Since the kiosk will only run the AIR app I am happy to loose hyper threading to increase the performance of the Air app. Marginal Improvement. The same website with the same videos is responsive if viewed in ie7 on the same machine, although Internet Explorer takes advantage of the CPU’s hyper threading. The flash videos are built with Adobe Captivate and from what I understand employee JavaScript to relay results back to the server. I will add more information about the video content asap as the training guru is back in the office later this week.

    Read the article

  • Create a Desktop Shortcut to an existing FOLDER using WiX

    - by Sach
    I have the need to create a Desktop Shortcut to an existing FOLDER (NOT to a file) using Wix. To elaborate more, my installer program has a CustomAction program written using C# associated with it. This CustomAction program creates a folder named "BSS" of which the path is selected by user. C:\ProgramData\MT\BSS Now I need to place a Desktop Shortcut to this folder using WiX. However, I encounter a problem since this folder does not have a folder structure within WiX. The closest code I could find was the following. <Directory Id="DesktopFolder" Name="Desktop"/> <Directory Id="CommonAppDataFolder" Name="ProgramDataFolder"/> <Component Id="ComponentBSStrageShortcut" Guid="{8436995c-2e76-4030-b92d-c6b4bc243c43}"> <Shortcut Id="ShortcutBSStrageShortcut" Directory="DesktopFolder" WorkingDirectory="APPLICATIONFOLDER" Target="[CommonAppDataFolder]/MTK/BSStrage" Name="BSStrage" Show="normal"/> <RegistryValue Action="write" Key="SOFTWARE/MTK/BackStreet" Root="HKCU" Type="string" KeyPath="yes" Value="ApplicationFolderName"/> </Component> When I build the installer this way, it actually creates a shortcut on Desktop. However, WiX seems to think that BSStrage is a file/application so it places a shortcut to an imaginary application called BSStrage in the location C:\ProgramData\MT. But double clicking on it dosen't help as there is no program that can be used to open it. Obviously I'm doing it wrong here. Can someone please help me with this, so as how to overcome this problem. Note that I'm extremely new to Wix (it's been only two days) and has never worked with it before. Any code sample would be of great help.

    Read the article

  • Adding BarButtons to a UINavigationBar after presenting a modal view controller.

    - by yujean
    I'm using the template for a "Utility Application". In the "FlipSideViewController", I added an IBOutlet for a UINavigationController, navController. In the code, I added the navController just fine. The rootViewController loads perfectly: navController.viewControllers = [[NSArray arrayWithObject:rootViewController] retain]; [self.view addSubview:navController.view]; I changed the color of the navController just fine: navController.navigationBar.tintColor = [UIColor colorWithRed:0.6 green:0.75 blue:0.6 alpha:1.0]; navController.navigationBar.translucent = NO; I make a button (note: "done" refers to a IBAction that dismisses the modalviewcontroller): UIBarButtonItem *backButton = [[UIBarButtonItem alloc] initWithTitle:@"TEST" style:UIBarButtonItemStyleDone target:self action:@selector(done:)]; I make a navItem using that button: UINavigationItem *backNavItem = [[UINavigationItem alloc] initWithTitle:@"TESTTEST"]; [backNavItem setRightBarButtonItem:backButton animated:YES]; I try to add that button: [navController.navigationBar pushNavigationItem:backNavItem animated:YES]; This above code fails miserably. I apparently can't add buttons to the navController because: * Terminating app due to uncaught exception 'NSInternalInconsistencyException', reason: 'Cannot call pushNavigationItem:animated: directly on a UINavigationBar managed by a controller.' Do I have to make a separate UINavigationBar that's somehow connected to my navController? I tried going that route but with no avail.

    Read the article

  • How to determine if "html" or "body" scrolls the window.

    - by David Murdoch
    The code below is used to find the element that can be scrolled (body or html) via javascript. var scrollElement = (function (tags) { var el, $el, init; // iterate through the tags... while (el = tags.pop()) { $el = $(el); // if the scrollTop value is already > 0 then this element will work if ( $el.scrollTop() > 0){ return $el; } // if scrollTop is 0 try to scroll. else if($el.scrollTop( 1 ).scrollTop() > 0) { // if that worked reset the scroll top and return the element return $el.scrollTop(0); } } return $(); } (["html", "body"])); // do stuff with scrollElement...like: // scrollElement.animate({"scrollTop":target.offset().top},1000); This code works perfectly when the height of the document is greater than the height of the window. However, when the height of the document is the same or less than the window the method above will not work because scrollTop() will always be equal to 0. This becomes a problem if the DOM is updated and the height of the document grows beyond the height of the window after the code runs. Also, I generally don't wait until document.ready to set up my javascript handlers (this generally works). I could append a tall div to the body temporarily to force the method above to work BUT that would required the document to be ready in IE (you can't add a node to the body element before the tag is closed). For more reading on the document.ready "anti-pattern" topic read this. So, I'd love to find a solution that finds the scrollable element even when the document is short. Any ideas?

    Read the article

  • Why does perl crash with "*** glibc detected *** perl: munmap_chunk(): invalid pointer"?

    - by sid_com
    #!/usr/bin/env perl use warnings; use strict; use 5.012; use XML::LibXML::Reader; my $reader = XML::LibXML::Reader->new( location => 'http://www.heise.de/' ) or die $!; while ( $reader->read ) { say $reader->name; } At the end of the output from this script I get this error-messages: * glibc detected * perl: munmap_chunk(): invalid pointer: 0x0000000000b362e0 * ======= Backtrace: ========= /lib64/libc.so.6[0x7fb84952fc76] ... ======= Memory map: ======== 00400000-0053d000 r-xp 00000000 08:01 182002 /usr/local/bin/perl ... Is this due a bug? perl -V: Summary of my perl5 (revision 5 version 12 subversion 0) configuration: Platform: osname=linux, osvers=2.6.31.12-0.2-desktop, archname=x86_64-linux uname='linux linux1 2.6.31.12-0.2-desktop #1 smp preempt 2010-03-16 21:25:39 +0100 x86_64 x86_64 x86_64 gnulinux ' config_args='-Dnoextensions=ODBM_File' hint=recommended, useposix=true, d_sigaction=define useithreads=undef, usemultiplicity=undef useperlio=define, d_sfio=undef, uselargefiles=define, usesocks=undef use64bitint=define, use64bitall=define, uselongdouble=undef usemymalloc=n, bincompat5005=undef Compiler: cc='cc', ccflags ='-fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64', optimize='-O2', cppflags='-fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include' ccversion='', gccversion='4.4.1 [gcc-4_4-branch revision 150839]', gccosandvers='' intsize=4, longsize=8, ptrsize=8, doublesize=8, byteorder=12345678 d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=16 ivtype='long', ivsize=8, nvtype='double', nvsize=8, Off_t='off_t', lseeksize=8 alignbytes=8, prototype=define Linker and Libraries: ld='cc', ldflags =' -fstack-protector -L/usr/local/lib' libpth=/usr/local/lib /lib /usr/lib /lib64 /usr/lib64 /usr/local/lib64 libs=-lnsl -ldl -lm -lcrypt -lutil -lc perllibs=-lnsl -ldl -lm -lcrypt -lutil -lc libc=/lib/libc-2.10.1.so, so=so, useshrplib=false, libperl=libperl.a gnulibc_version='2.10.1' Dynamic Linking: dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E' cccdlflags='-fPIC', lddlflags='-shared -O2 -L/usr/local/lib -fstack-protector' Characteristics of this binary (from libperl): Compile-time options: PERL_DONT_CREATE_GVSV PERL_MALLOC_WRAP USE_64_BIT_ALL USE_64_BIT_INT USE_LARGE_FILES USE_PERLIO USE_PERL_ATOF Built under linux Compiled at Apr 15 2010 13:25:46 @INC: /usr/local/lib/perl5/site_perl/5.12.0/x86_64-linux /usr/local/lib/perl5/site_perl/5.12.0 /usr/local/lib/perl5/5.12.0/x86_64-linux /usr/local/lib/perl5/5.12.0 .

    Read the article

  • How to make Nlog archive a file with the date when the logging took place.

    - by Carl Bergquist
    Hi, We are using Nlog as our logging framework and I cannot find a way to archive files the way I want. I would like to have the date of when the logging took place in the logging file name. Ex All logging that happend from 2009-10-01 00:00 -> 2009-10-01:23:59 should be placed in Log.2009-10-01.log The current NLog.config that I use looks like this. <?xml version="1.0" encoding="utf-8" ?> <nlog xmlns="http://www.nlog-project.org/schemas/NLog.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" > <extensions> <add assembly="My.Awesome.LoggingExentions"/> </extensions> <targets> <target name="file1" xsi:type="File" fileName="${basedir}/Logs/Log.log" layout="${longdate} ${level:uppercase=true:padding=5} ${session} ${storeid} ${msisdn} - ${logger:shortName=true} - ${message} ${exception:format=tostring}" archiveEvery="Day" archiveFileName="${basedir}/Logs/Log${shortdate}-{#}.log" archiveNumbering="Sequence" maxArchiveFiles="99999" keepFileOpen="true" /> </targets> <rules> <logger name="*" minlevel="Trace" writeTo="file1" /> </rules> </nlog> This however sets the date in the logfile to the date when the new logfile is created. Which cause frustration when you want to read logs later. It also seems like I have to have atleast on # in the archiveFileName, which I rather not. So if you got a solution for that also I would be twice as thankful =)

    Read the article

  • Can't build pyxpcom on OS X 10.6

    - by Gj
    I've been following these instructions at https://developer.mozilla.org/en/Building_PyXPCOM but getting this: $ make make export make[2]: Nothing to be done for `export'. make[4]: Nothing to be done for `export'. make[4]: Nothing to be done for `export'. /opt/local/bin/python2.5 ../../../src/config/nsinstall.py -L /usr/local/pyxpcom/build/xpcom/src -m 644 ../../../src/xpcom/src/PyXPCOM.h ../../dist/include make[3]: Nothing to be done for `export'. /opt/local/bin/python2.5 ../../../../src/config/nsinstall.py -D ../../../dist/idl /opt/local/bin/python2.5 ../../../../src/config/nsinstall.py -D ../../../dist/idl make[4]: *** No rule to make target `_xpidlgen/py_test_component.h', needed by `export'. Stop. make[3]: *** [export] Error 2 make[2]: *** [export] Error 2 make[1]: *** [export] Error 2 make: *** [default] Error 2 Any ideas? An interesting anomaly is that despite me setting the PYTHON env variable to Python 2.6, the configure and make both seem to go after the 2.5... Thanks for any advice! PS here's the configure output: $ ../src/configure --with-libxul-sdk=/Users/me/xulrunner-sdk/ loading cache ./config.cache checking host system type... i386-apple-darwin10.3.0 checking target system type... i386-apple-darwin10.3.0 checking build system type... i386-apple-darwin10.3.0 checking for mawk... (cached) gawk checking for perl5... (cached) /opt/local/bin/perl5 checking for gcc... (cached) gcc checking whether the C compiler (gcc ) works... yes checking whether the C compiler (gcc ) is a cross-compiler... no checking whether we are using GNU C... (cached) yes checking whether gcc accepts -g... (cached) yes checking for c++... (cached) c++ checking whether the C++ compiler (c++ ) works... yes checking whether the C++ compiler (c++ ) is a cross-compiler... no checking whether we are using GNU C++... (cached) yes checking whether c++ accepts -g... (cached) yes checking for ranlib... (cached) ranlib checking for as... (cached) /usr/bin/as checking for ar... (cached) ar checking for ld... (cached) ld checking for strip... (cached) strip checking for windres... no checking whether gcc and cc understand -c and -o together... (cached) yes checking how to run the C preprocessor... (cached) gcc -E checking how to run the C++ preprocessor... (cached) c++ -E checking for a BSD compatible install... (cached) /usr/bin/install -c checking whether ln -s works... (cached) yes checking for minimum required perl version >= 5.006... 5.008009 checking for full perl installation... yes checking for /opt/local/bin/python... (cached) /opt/local/bin/python2.5 checking for doxygen... (cached) : checking for whoami... (cached) /usr/bin/whoami checking for autoconf... (cached) /opt/local/bin/autoconf checking for unzip... (cached) /usr/bin/unzip checking for zip... (cached) /usr/bin/zip checking for makedepend... (cached) /opt/local/bin/makedepend checking for xargs... (cached) /usr/bin/xargs checking for pbbuild... (cached) /usr/bin/xcodebuild checking for sdp... (cached) /usr/bin/sdp checking for gmake... (cached) /opt/local/bin/gmake checking for X... (cached) no checking whether the compiler supports -Wno-invalid-offsetof... yes checking whether ld has archive extraction flags... (cached) no checking that static assertion macros used in autoconf tests work... (cached) yes checking for 64-bit OS... yes checking for minimum required Python version >= 2.4... yes checking for -dead_strip option to ld... yes checking for ANSI C header files... (cached) yes checking for working const... (cached) yes checking for mode_t... (cached) yes checking for off_t... (cached) yes checking for pid_t... (cached) yes checking for size_t... (cached) yes checking for st_blksize in struct stat... (cached) yes checking for siginfo_t... (cached) yes checking for int16_t... (cached) yes checking for int32_t... (cached) yes checking for int64_t... (cached) yes checking for int64... (cached) no checking for uint... (cached) yes checking for uint_t... (cached) no checking for uint16_t... (cached) no checking for uname.domainname... (cached) no checking for uname.__domainname... (cached) no checking for usable char16_t (2 bytes, unsigned)... (cached) no checking for usable wchar_t (2 bytes, unsigned)... (cached) no checking for compiler -fshort-wchar option... (cached) yes checking for visibility(hidden) attribute... (cached) yes checking for visibility(default) attribute... (cached) yes checking for visibility pragma support... (cached) yes checking For gcc visibility bug with class-level attributes (GCC bug 26905)... (cached) yes checking For x86_64 gcc visibility bug with builtins (GCC bug 20297)... (cached) no checking for dirent.h that defines DIR... (cached) yes checking for opendir in -ldir... (cached) no checking for sys/byteorder.h... (cached) no checking for compat.h... (cached) no checking for getopt.h... (cached) yes checking for sys/bitypes.h... (cached) no checking for memory.h... (cached) yes checking for unistd.h... (cached) yes checking for gnu/libc-version.h... (cached) no checking for nl_types.h... (cached) yes checking for malloc.h... (cached) no checking for X11/XKBlib.h... (cached) yes checking for io.h... (cached) no checking for sys/statvfs.h... (cached) yes checking for sys/statfs.h... (cached) no checking for sys/vfs.h... (cached) no checking for sys/mount.h... (cached) yes checking for sys/quota.h... (cached) yes checking for mmintrin.h... (cached) yes checking for new... (cached) yes checking for sys/cdefs.h... (cached) yes checking for gethostbyname_r in -lc_r... (cached) no checking for dladdr... (cached) yes checking for socket in -lsocket... (cached) no checking whether mmap() sees write()s... yes checking whether gcc needs -traditional... (cached) no checking for 8-bit clean memcmp... (cached) yes checking for random... (cached) yes checking for strerror... (cached) yes checking for lchown... (cached) yes checking for fchmod... (cached) yes checking for snprintf... (cached) yes checking for statvfs... (cached) yes checking for memmove... (cached) yes checking for rint... (cached) yes checking for stat64... (cached) yes checking for lstat64... (cached) yes checking for truncate64... (cached) no checking for statvfs64... (cached) no checking for setbuf... (cached) yes checking for isatty... (cached) yes checking for flockfile... (cached) yes checking for getpagesize... (cached) yes checking for localtime_r... (cached) yes checking for strtok_r... (cached) yes checking for wcrtomb... (cached) yes checking for mbrtowc... (cached) yes checking for res_ninit()... (cached) no checking for gnu_get_libc_version()... (cached) no ../src/configure: line 9881: AM_LANGINFO_CODESET: command not found checking for an implementation of va_copy()... (cached) yes checking for an implementation of __va_copy()... (cached) yes checking whether va_lists can be copied by value... (cached) no checking for C++ exceptions flag... (cached) -fno-exceptions checking for gcc 3.0 ABI... (cached) yes checking for C++ "explicit" keyword... (cached) yes checking for C++ "typename" keyword... (cached) yes checking for modern C++ template specialization syntax support... (cached) yes checking whether partial template specialization works... (cached) yes checking whether operators must be re-defined for templates derived from templates... (cached) no checking whether we need to cast a derived template to pass as its base class... (cached) no checking whether the compiler can resolve const ambiguities for templates... (cached) yes checking whether the C++ "using" keyword can change access... (cached) yes checking whether the C++ "using" keyword resolves ambiguity... (cached) yes checking for "std::" namespace... (cached) yes checking whether standard template operator!=() is ambiguous... (cached) unambiguous checking for C++ reinterpret_cast... (cached) yes checking for C++ dynamic_cast to void*... (cached) yes checking whether C++ requires implementation of unused virtual methods... (cached) yes checking for trouble comparing to zero near std::operator!=()... (cached) no checking for LC_MESSAGES... (cached) yes checking for tar archiver... checking for gnutar... (cached) gnutar gnutar checking for wget... checking for wget... (cached) wget wget checking for valid optimization flags... yes checking for gcc -pipe support... yes checking whether compiler supports -Wno-long-long... yes checking whether C compiler supports -fprofile-generate... yes checking for correct temporary object destruction order... yes checking for correct overload resolution with const and templates... no Building Python extensions using python-2.5 from /opt/local/Library/Frameworks/Python.framework/Versions/2.5 creating ./config.status creating config/autoconf.mk creating Makefile creating xpcom/Makefile creating xpcom/src/Makefile creating xpcom/src/loader/Makefile creating xpcom/src/module/Makefile creating xpcom/components/Makefile creating xpcom/test/Makefile creating xpcom/test/test_component/Makefile creating dom/Makefile creating dom/src/Makefile creating dom/test/Makefile creating dom/test/pyxultest/Makefile creating dom/nsdom/Makefile creating dom/nsdom/test/Makefile

    Read the article

  • Display ñ on a C# .NET application

    - by mmr
    I have a localization issue. One of my industrious coworkers has replaced all the strings throughout our application with constants that are contained in a dictionary. That dictionary gets various strings placed in it once the user selects a language (English by default, but target languages are German, Spanish, French, Portuguese, Mandarin, and Thai). For our test of this functionality, we wanted to change a button to include text which has a ñ character, which appears both in Spanish and in the Arial Unicode MS font (which we're using throughout the application). Problem is, the ñ is appearing as a square block, as if the program did not know how to display it. When I debug into that particular string being read from disk, the debugger reports that character as a square block as well. So where is the failure? I think it could be in a few places: 1) Notepad may not be unicode aware, so the ñ displayed there is not the same as what vs2008 expects, and so the program interprets the character as a square (EDIT: notepad shows the same characters as vs; ie, they both show the ñ. In the same place.). 2) vs2008 can't handle ñ. I find that very, very hard to believe. 3) The text is read in properly, but the default font for vs2008 can't display it, which is why the debugger shows a square. 4) The text is not read in properly, and I should use something other than a regular StreamReader to get strings. 5) The text is read in properly, but the default String class in C# doesn't handle ñ well. I find that very, very hard to believe. 6) The version of Arial Unicode MS I have doesn't have ñ, despite it being listed as one of the 50k characters by http://www.fileinfo.info. Anything else I could have left out? Thanks for any help!

    Read the article

  • Sharepoint: Integrity of lookup fields after a list import

    - by driAn
    Hi there I got a question about the behavior of lookup fields when importing data. I wonder how the lookup fields behave when the list they point to is being replaced/imported. To explain the issue, I will provide a quick example below: As example, assume we have these two sharepoint lists: Product Types ------------- + Type Name + Code Nr + etc Products -------- + Product Name + Product Type (Lookup field to list "Product Types") + etc In my scenario, the Products List contains production data on the production Sharepoint platform. It is filled with data by the business users. However the Product Types list contains rather static data and is maintained by the developer. Now after a development cycle, the developer wants to deploy his new webparts and his new data (product types list). The developer performs the following procedure: On the dev machine: Export "product type" list using stsadm On the production machine: Delete all items in the "product type" list On the production machine: Import the "product type" list using stsadm This means we basically replace the "product type" list on the production server while keeping the "product" list as it is. Now the question: Is this safe? Will the lookup references break under certain circumstances? Any downside of this import/export procedure? What happens if someone accesses a "product" during the import? Will the (now invalid) reference clear its own content (become a null value). What happens if the schema of the "product type" list changes (new column)? Will this cause any troubles? Thanks for all feedback and suggestions!

    Read the article

< Previous Page | 440 441 442 443 444 445 446 447 448 449 450 451  | Next Page >