Search Results

Search found 24391 results on 976 pages for 'static methods'.

Page 119/976 | < Previous Page | 115 116 117 118 119 120 121 122 123 124 125 126  | Next Page >

  • Why has the accessor methods from the JavaBean specification become the standard for Java development?

    - by Dakotah North
    The JavaBeans Specification describes a JavaBean as A Java Bean is a reusable software component that can be manipulated visually in a builder tool Since the majority of the lines of code that are written seem to have nothing to do with being manipulated visually in a builder tool, why has the JavaBean specification been the "way" to write object oriented code? I would like to forgo the traditional getter/setter in favor of Fluent Interfaces all throughout the code, not just in builders but fear doing so since this is traditionally not the way way object oriented code is written in Java.

    Read the article

  • In C and C++, what methods can prevent accidental use of the assignment(=) where equivalence(==) is needed?

    - by DeveloperDon
    In C and C++, it is very easy to write the following code with a serious error. char responseChar = getchar(); int confirmExit = 'y' == tolower(responseChar); if (confirmExit = 1) { exit(0); } The error is that the if statement should have been: if (confirmExit == 1) As coded, it will exit every time, because the assignment of the confirmExit variable occurs, then confirmExit is used as the result of the expression. Are there good ways to prevent this kind of error?

    Read the article

  • What methods of requirements elicitation are suitable when I do not have direct access to the user base?

    - by metadice
    I am working on an application to create invoices. There are some features that are required based on the type of the application and are common to all invoice applications. However, we still need to determine what unique needs the user base might have. We do not have direct access to the users to obtain requirements or user stories. What techniques are most suitable for eliciting high-quality requirements from users when direct or frequent access is not possible?

    Read the article

  • Can Ruby on Rails handle AJAX Response by Static Javascript code? by using RJS?

    - by Jian Lin
    So it looks like on RoR, when Ajax (using form_remote_tag) returns a success code, Javascript is also returned to handle the visual effects. (this is the RJS mechanism) using Fiddler, I do see the following response: try { Element.update("vote_score", "Score 58"); $("vote_score").visualEffect("highlight"); } catch (e) { alert('RJS error:\n\n' + e.toString()); alert('Element.update(\"vote_score\", \"Score 58\");\n$(\"vote_score\").visualEffect(\"highlight\");'); throw e } Will there be situation where the code is quite big (like 1 or 2k) that RJS can use some sort of static Javascript already loaded to handle the effect? Is there any other ways to use static Javascript already loaded with the page? Must it be just raw Javascript and use the Prototype's Ajax success response to call the static Javascript code, or can it be some Rails mechanism related to RJS?

    Read the article

  • Pass a Delphi class to a C++ function/method that expects a class with __thiscall methods.

    - by Alan G.
    I have some MSVC++ compiled DLL's for which I have created COM-like (lite) interfaces (abstract Delphi classes). Some of those classes have methods that need pointers to objects. These C++ methods are declared with the __thiscall calling convention (which I cannot change), which is just like __stdcall, except a this pointer is passed on the ECX register. I create the class instance in Delphi, then pass it on to the C++ method. I can set breakpoints in Delphi and see it hitting the exposed __stdcall methods in my Delphi class, but soon I get a STATUS_STACK_BUFFER_OVERRUN and the app has to exit. Is it possible to emulate/deal with __thiscall on the Delphi side of things? If I pass an object instantiated by the C++ system then all is good, and that object's methods are called (as would be expected), but this is useless - I need to pass Delphi objects. Edit 2010-04-19 18:12 This is what happens in more detail: The first method called (setLabel) exits with no error (though its a stub method). The second method called (init), enters then dies when it attempts to read the vol parameter. C++ Side #define SHAPES_EXPORT __declspec(dllexport) // just to show the value class SHAPES_EXPORT CBox { public: virtual ~CBox() {} virtual void init(double volume) = 0; virtual void grow(double amount) = 0; virtual void shrink(double amount) = 0; virtual void setID(int ID = 0) = 0; virtual void setLabel(const char* text) = 0; }; Delphi Side IBox = class public procedure destroyBox; virtual; stdcall; abstract; procedure init(vol: Double); virtual; stdcall; abstract; procedure grow(amount: Double); virtual; stdcall; abstract; procedure shrink(amount: Double); virtual; stdcall; abstract; procedure setID(val: Integer); virtual; stdcall; abstract; procedure setLabel(text: PChar); virtual; stdcall; abstract; end; TMyBox = class(IBox) protected FVolume: Double; FID: Integer; FLabel: String; // public constructor Create; destructor Destroy; override; // BEGIN Virtual Method implementation procedure destroyBox; override; stdcall; // empty - Dont need/want C++ to manage my Delphi objects, just call their methods procedure init(vol: Double); override; stdcall; // FVolume := vol; procedure grow(amount: Double); override; stdcall; // Inc(FVolume, amount); procedure shrink(amount: Double); override; stdcall; // Dec(FVolume, amount); procedure setID(val: Integer); override; stdcall; // FID := val; procedure setLabel(text: PChar); override; stdcall; // Stub method; empty. // END Virtual Method implementation property Volume: Double read FVolume; property ID: Integer read FID; property Label: String read FLabel; end; I would have half expected using stdcall alone to work, but something is messing up, not sure what, perhaps something to do with the ECX register being used? Help would be greatly appreciated. Edit 2010-04-19 17:42 Could it be that the ECX register needs to be preserved on entry and restored once the function exits? Is the this pointer required by C++? I'm probably just reaching at the moment based on some intense Google searches. I found something related, but it seems to be dealing with the reverse of this issue.

    Read the article

  • What does a Java static method look like in Ruby?

    - by Hosh
    In Java, a 'static method' would look like this: class MyUtils { . . . public static double mean(int[] p) { int sum = 0; // sum of all the elements for (int i=0; i<p.length; i++) { sum += p[i]; } return ((double)sum) / p.length; } . . . } // Called from outside the MyUtils class. double meanAttendance = MyUtils.mean(attendance); What's the equivalent 'Ruby way' of writing a 'static method'?

    Read the article

  • C# : What if a static method is called from multiple threads?

    - by Holli
    In my Application I have a static method that is called from multiple threads at the same time. Is there any danger of my data being mixed up? In my first attempt the method was not static and I was creating multiple instance of the class. In that case my data got mixed up somehow. I am not sure how this happens because it only happens sometimes. I am still debugging. But now the method is static on I have no problems so far. Maybe it's just luck. I don't know for sure.

    Read the article

  • How do I get class of an internal static class in another assembly?

    - by Echiban
    I have a class C in Assembly A like this: internal class C { internal static string About_Name { get { return "text"; } ... } I have about 20 such static properties. Is there a way, in an outside assembly, without using friend assembly attribute (.Net reflection only), get class C so I can invoke any of the static string properties like this: Class C = <some .Net reflection code>; string expected = C.About_Name; If that is not possible, a .Net reflection code to get the string property value directly will suffice, but not ideal. Thanks in advance!

    Read the article

  • C++: static function member shared between threads, can block all?

    - by mhambra
    Hi all, I have a class, which has static function defined to work with C-style extern C { static void callback(foo bar) { } }. // static is defined in header. Three objects (each in separate pthread) are instantiated from this class, each of them has own loop (in class constructor), which can receive the callback. The pointer to function is passed as: x = init_function(h, queue_id, &callback, NULL); while(1) { loop_function(x); } So each thread has the same pointer to &callback. Callback function can block for minutes. Each thread object, excluding the one which got the blocking callback, can call callback again. If the callback function exists only once, then any thread attempting to callback will also block. This would give me an undesired bug, circa is interesting to ask: can anything in C++ become acting this way? Maybe, due to extern { } or some pointer usage?

    Read the article

  • Complex event system for DungeonKeeper like game

    - by paul424
    I am working on opensource GPL3 game. http://opendungeons.sourceforge.net/ , new coders would be welcome. Now there's design question regarding Event System: We want to improve the game logic, that is program a new event system. I will just repost what's settled up already on http://forum.freegamedev.net/viewtopic.php?f=45&t=3033. From the discussion came the idea of the Publisher / Subscriber pattern + "domains": My current idea is to use the subscirbers / publishers model. Its similar to Observable pattern, but instead one subscribes to Events types, not Object's Events. For each Event would like to have both static and dynamic type. Static that is its's type would be resolved by belonging to the proper inherited class from Event. That is from Event we would have EventTile, EventCreature, EvenMapLoader, EventGameMap etc. From that there are of course subtypes like EventCreature would be EventKobold, EventKnight, EventTentacle etc. The listeners would collect the event from publishers, and send them subcribers , each of them would be a global singleton. The Listeners type hierachy would exactly mirror the type hierarchy of Events. In each constructor of Event type, the created instance would notify the proper listeners. That is when calling EventKnight the proper ctor would notify the Listeners : EventListener, CreatureLisener and KnightListener. The default action for an listner would be to notify all subscribers, but there would be some exceptions , like EventAttack would notify AttackListener which would dispatch event by the dynamic part ( that is the Creature pointer or hash). Any comments ? #include <vector> class Subscriber; class SubscriberAttack; class Event{ private: int foo; int bar; protected: // static std::vector<Publisher*> publishersList; static std::vector<Subscriber*> subscribersList; static std::vector<Event*> eventQueue; public: Event(){ eventQueue.push_back(this); } static int subscribe(Subscriber* ss); static int unsubscribe(Subscriber* ss); //static int reg_publisher(Publisher* pp); //static int unreg_publisher(Publisher* pp); }; // class Publisher{ // }; class Subscriber{ public: int (*newEvent) (Event* ee); Subscriber( ){ Event::subscribe(this); } Subscriber( int (*fp) (Event* ee) ):newEvent(fp){ Subscriber(); } ~Subscriber(){ Event::unsubscribe(this); } }; class EventAttack: Event{ private: int foo; int bar; protected: // static std::vector<Publisher*> publishersList; static std::vector<SubscriberAttack*> subscribersList; static std::vector<EventAttack*> eventQueue; public: EventAttack(){ eventQueue.push_back(this); } static int subscribe(SubscriberAttack* ss); static int unsubscribe(SubscriberAttack* ss); //static int reg_publisher(Publisher* pp); //static int unreg_publisher(Publisher* pp); }; class AttackSubscriber :Subscriber{ public: int (*newEvent) (EventAttack* ee); AttackSubscriber( ){ EventAttack::subscribe(this); } AttackSubscriber( int (*fp) (EventAttack* ee) ):newEventAttack(fp){ AttackSubscriber(); } ~AttackSubscriber(){ EventAttack::unsubscribe(this); } }; From that point, others wanted the Subject-Observer pattern, that is one would subscribe to all event types produced by particular object. That way it came out to add the domain system : Huh, to meet the ability to listen to particular game's object events, I though of introducing entity domains . Domains are trees, which nodes are labeled by unique names for each level. ( like the www addresses ). Each Entity wanting to participate in our event system ( that is be able to publish / produce events ) should at least now its domain name. That would end up in Player1/Room/Treasury/#24 or Player1/Creature/Kobold/#3 producing events. The subscriber picks some part of a tree. For example by specifiing subtree with the root in one of the nodes like Player1/Room/* ,would subscribe us to all Players1's room's event, and Player1/Creature/Kobold/#3 would subscribe to Players' third kobold's event. Does such event system make sense to you ? I have many implementation details to ask as well, but first let's start some general discussion. Note1: Notice that in the case of a fight between two creatues fight , the creature being attacked would have to throw an event, becuase it is HE/SHE/IT who have its domain address. So that would be BeingAttackedEvent() etc. I will edit that post if some other reflections on this would come out. Note2: the existing class hierarchy might be used to get the domains addresses being build in constructor . In a ctor you would just add + ."className" to domain address. If you are in a class'es hierarchy leaf constructor one might use nextID , hash or any other charactteristic, just to make the addresses distinguishable . Note3:subscribing to all entity's Events would require knowledge of all possible events produced by this entity . This could be done in one function call, but information on E produced would have to be handled for every Entity. SmartNote4 : Finding proper subscribers in a tree would be easy. One would start in particular Leaf for example Player1/Creature/Kobold/#3 and go up one parent a time , notifiying each Subscriber in a Node ie. : Player1/Creature/Kobold/* , Player1/Creature/* , Player1/* etc, , up to a root that is /* .<<<< Note5: The Event system was needed to have some way of incorporating Angelscript code into application. So the Event dispatcher was to be a gate to A-script functions. But it came out to this one.

    Read the article

  • Is Cherokee (probably) the best static content server for beginner sysadmins?

    - by Bad Learner
    I have read the pros and cons of most of the popular web servers and have come to a conclusion that Apache would (probably) be the best web server for serving dynamic content - - no wonder YouTube, Flickr and Facbook, among many others, use it. I do not know if that C10K problem applies to Apache even when serving dynamic content only, but I think any web server used to serve dynamic content needs some good tweaking for optimized performance, and the fact that nothing beats Apache when it comes to documentation, resources and support on the web, I think should will go with Apache for dynamic content. That apart, the confusion begins when it comes to choosing web servers for static content (including streaming videos). I see that Nginx, Cherokee and Lighttpd are among the best (I am not considering non-open source or non-linux stuff here). So, which too choose? I know one cannot go wrong with any of the three (Nginx, Cherokee, Lighttpd). Lighttpd's development has evidently gotten slower than it was a good time ago. The documentation is pretty good for all the three, and hopefully, so are the resources (knowledge of these among the users of Stackoverflow/Serverfault sites, the web etc). Precisely, and noting point [2] and [3], if I am not wrong, I should either go with Nginx or Cherokee. I would love to see someone clarify these... is Cherokee just as fast (mb/s), performant (connections/s), and reliable (think downtime/restarting server) as Nginx for serving static content and load balancing, for small, medium to large (and really large) websites and applications? (Think, the size of YouTube, Apache or Facebook.) if the answer for the Q above is a big "hell, yes!" then, I should probably prefer Cherokee, right? Because, since I am a beginner, it would a lot easier to setup Cherokee as it has a graphical admin user interface + really good documentation. Yes? I could be wrong, I could be right. I put down what I know so that you can offer most relevant advise. Pardon if anything I've said is offensive.

    Read the article

  • How to create a static IP on Windows Server 2008 R2 so I can access the server remotely

    - by Aesir
    I have just purchased a HP Proliant N40L which I am intending to use as a NAS, learning tool and just in general something to mess around with. As a student via the Microsoft dreamspark program I can get a free copy of Windows Server 2008 R2 which I am using as the OS. So that I can remote to the box from outside of my local network and so that I can stream media from it to my PS3, I have read that I need to create a static IP for the server and use port forwarding to forward to this IP so I can remote in. Is this correct? I am not really sure how to do this and if I need to make these changes on my router configuration, on the OS or both. I am a novice when it comes to networking however most resources for Windows server 2008 R2 seem to assume a fair amount of experience already. I realise that using this particular OS may seem like overkill for what I currently wish to do with it (stream content to other devices and backup) but as I can get a copy for free it seems sensible. Edit: From reading answers posted I feel I should give more information. I have now tried to add a static IP address using my router configuration settings. I have used the getmac command to get the mac address of the server. My ISP is Virgin Media and I have gone to the LAN IP section and I have added an IP address to the DHCP Reservation Lease Info. I can now use remote desktop connection internally to remote to the server (so I am assuming assigning this IP has worked). How do I configure this on the OS as well? I am also unsure on how I would remote to this machine outside of my local network?

    Read the article

  • How to set static ip address on vmware for NAT guest vms from an ubuntu Host dhcp server?

    - by javadba
    I need to configure various linux flavor NAT'ed guest vm's to have static ip addresses provided by the Ubuntu host. The vmware documentation punts on this topic, deferring to "see the man pages for your linux distribution". But the generic pages for "my linux distro" do not know about the special stuff for vmware e.g. vmnet8. Pointers from someone who just knows how to do this would be much appreciated. Here is the /etc/vmware/vmnet8/dhcpd/dhcpd.conf: allow unknown-clients; default-lease-time 1800; # default is 30 minutes max-lease-time 7200; # default is 2 hours subnet 192.168.238.0 netmask 255.255.255.0 { range 192.168.238.128 192.168.238.254; option broadcast-address 192.168.238.255; option domain-name-servers 192.168.238.2; option domain-name localdomain; default-lease-time 1800; # default is 30 minutes max-lease-time 7200; # default is 2 hours option netbios-name-servers 192.168.238.2; option routers 192.168.238.2; } host vmnet8 { hardware ethernet 00:50:56:C0:00:08; fixed-address 192.168.238.1; option domain-name-servers 0.0.0.0; option domain-name ""; option routers 0.0.0.0; } Fromt the dhcpd.conf documentation, we are supposed to add an entry for static hosts similar to the following: host mystatichostonee { hardware ethernet 00:20:6B:C7:9B:E4; fixed-address 192.168.238.101; } host mystatichosttwo { hardware ethernet 00:23:7a:C7:9c:F2; fixed-address 192.168.238.102; } But notice that the vmnet8 entry in the vmware-generated dhcpd.conf already is set to fixed-address. I don't know how to add the specifics for my hosts to that vmnet8 entry: do they become nested?

    Read the article

  • Why does lighttpd keep static files in cache, even when modified on disk ?

    - by Pixelastic
    I am using lighttpd to serve static files. I have a bunch of images in a dir that I regularly update. This will change the file content (and filesize) as well as the modification date, but not their filename. When I access the files through http, the updates are not taken into account and lighty serves the old file. I can manually rename the file to something different, then lighttpd will return a 404 error, and if I rename my file back, I will get the correct updated version. Seems like lightty is using some kind of cache mechanism of its own (which is fine) to return static files. Unfortunatly, it seems that this mechanism doesn't update itself when files are modified. I checked through Wireshark, and my browser is really doing a request to the file, this is not a browser caching issue. It returns a 200 OK when requesting it from an empty cache, and a 304 Not Modified otherwise, as expected. But the file is returned with a wrong Last-Modified header that do not reflect the real last modification date. Maybe there is some config directive that I am not aware of ? I would like the files returned by lighty to reflect the changes made on disk directly, or at least being able to invalidate its cache.

    Read the article

  • What can an effective workflow for changing input methods be on Mac OS X?

    - by Martin Josefsson
    I switch between three languages input methods frequently, sometimes in the same typing session. The default shortcut on a Macintosh seems to be to set keyboard shortcuts for previous/next language (I hit opt-cmd-space to go to previous language) and if you're more than bi-lingual you have to cycle until you find the one for you. The ideal would be something like hitting fn-e for English, fn-j for Japanese and fn-g for German, but anything better than the current state would be a great improvement.

    Read the article

  • How do I prevent lighttpd from caching static files, even when modified on disk?

    - by Pixelastic
    I am using lighttpd to serve static files. I have a bunch of images in a dir that I regularly update. This will change the file content (and filesize) as well as the modification date, but not their filename. When I access the files through http, the updates are not taken into account and lighty serves the old file. I can manually rename the file to something different, then lighttpd will return a 404 error, and if I rename my file back, I will get the correct updated version. Seems like lightty is using some kind of cache mechanism of its own (which is fine) to return static files. Unfortunatly, it seems that this mechanism doesn't update itself when files are modified. I checked through Wireshark, and my browser is really doing a request to the file, this is not a browser caching issue. It returns a 200 OK when requesting it from an empty cache, and a 304 Not Modified otherwise, as expected. But the file is returned with a wrong Last-Modified header that do not reflect the real last modification date. Maybe there is some config directive that I am not aware of ? I would like the files returned by lighty to reflect the changes made on disk directly, or at least being able to invalidate its cache.

    Read the article

  • How does one properly redefine self while avoiding the "Multiple methods named ..." warning?

    - by Elise van Looij
    In Apple's The Objective-C Programming Language: Defining a Class the section named "Redifining self" recommends that that class methods allocate and return instances use 'self' only to allocate an instance and then refer only to that instance. Thus, I have a number of subclasses, that have class methods along the lines of: + (id)scrollViewWithFrame: (NSRect)rectFrame { id newInstance = [[[self alloc] initWithFrame:rectFrame] autorelease]; [newInstance setHasHorizontalScroller: YES]; [newInstance setHasVerticalScroller: YES]; [newInstance setBorderType: NSNoBorder]; [newInstance setAutoresizingMask: (NSViewWidthSizable | NSViewHeightSizable)]; return newInstance; } The above is, of course, a subclass of NSScrollView. Unfortunately, Xcode 3.x all these NSView subclasses now raise warnings: "Warning: Multiple methods named '-setAutoresizingMask' found". I believe it has something to do with GCC 4.2, the settings of which I have not changed. The warning is correct, of course, since NSView and its various subclasses all implement setAutoresizingMask, but it is also unnecessary. Since they're only warnings, I ignore them but there is a risk that in between the thirty or so unnecessary ones, a really useful warning lurks which I simply don't see. So, what to do? I do want to adhere to good coding practices and I want to build warning-free apps -- how can I do both?

    Read the article

  • When is Facebook Connect supposed to call its delegate methods?

    - by RickiG
    Hi The Facebook connect code is eluding me a bit. I have no problem doing a login, and a wall post, however, I simply can not figure out how the delegate methods for the FBDialog andFBStreamDialog is supposed to work. - (void)postToWall { FBStreamDialog *dialog = [[[FBStreamDialog alloc] init] autorelease]; dialog.delegate = self; dialog.userMessagePrompt = @"Enter your message:"; dialog.attachment = [NSString stringWithFormat:@"JSONpost code"]; [dialog show]; } I adhere to these protocols in my controller: <FBDialogDelegate, FBSessionDelegate, FBRequestDelegate> I then implement the two methods: - (void) dialogDidCancel:(FBDialog *)dialog { NSLog(@"Failed"); } - (void) dialogDidSucceed:(FBDialog *)dialog { NSLog(@"Success"); } After I tap "publish" and the postToWall methods is done executing the Facebook "pop up" in the UI is empty, except a small "X" in the top right corner and a "F" (facebook logo) in the top left corner. The UI will stay there until I tap the "X", this results in the dialogDidCancel delegate method being called. The post data is showing up on the Facebook page, everything seems to work. Why is thedialogDidSucceedmethod never called? I need this to release my facebook controller and restore the UI back to where the user was before "starting" FB Connect. Thank You:)

    Read the article

  • Zend Framework-where do calls to my methods go? Controller of Model?

    - by Joel
    Hi guys, I'm confused about exactly what I should have in my controller and what in my method. Specifically, I have this in the action method: public function upcomingshowsAction() { $gcal = $this->_validateCalendarConnection(); $uncleanedFeedArray = $this->_getCalendarFeed($gcal); $finishedFeedArray = $this->_cleanFeed($uncleanedFeedArray); $this->view->googleArray = $finishedFeedArray; } And then (incorrectly I know), I have my methods still in the bottom of my controller. So what I'm wondering, is for those methods in the upcomingshowsAction method, should all the actual methods just be in one model and then I'd have something like this: public function upcomingshowsAction() { $finishedFeedArray = new Application_Model_calendarModelPage(); $this->view->googleArray = $finishedFeedArray; } And then something like this in the model: class Application_Model_CalendarModelPage { $gcal = $this->_validateCalendarConnection(); $uncleanedFeedArray = $this->_getCalendarFeed($gcal); $finishedFeedArray = $this->_cleanFeed($uncleanedFeedArray); public functions { ... ... ... } } Am I on the right track here? Thanks!

    Read the article

  • How can I get my business objects layer to use the management layer in their methods?

    - by Tom Pickles
    I have a solution in VS2010 with several projects, each making up a layer within my application. I have business entities which are currently objects with no methods, and I have a management layer which references the business entities layer in it's project. I now think I have designed my application poorly and would like to move methods from helper classes (which are in another layer) into methods I'll create within the business entities themselves. For example I have a VirtualMachine object, which uses a helper class to call a Reboot() method on it which passes the request to the management layer. The static manager class talks to an API that reboots the VM. I want to move the Reboot() method into the VirtualMachine object, but I will need to reference the management layer: public void Reboot() { VMManager.Reboot(this.Name); } So if I add a reference to my management project in my entities project, I get the circular dependency error, which is how it should be. How can I sort this situation out? Do I need to an yet another layer between the entity layer and the management layer? Or, should I just forget it and leave it as it is. The application works ok now, but I am concerned my design isn't particularly OOP centric and I would like to correct this.

    Read the article

  • Class Loading Deadlocks

    - by tomas.nilsson
    Mattis follows up on his previous post with one more expose on Class Loading Deadlocks As I wrote in a previous post, the class loading mechanism in Java is very powerful. There are many advanced techniques you can use, and when used wrongly you can get into all sorts of trouble. But one of the sneakiest deadlocks you can run into when it comes to class loading doesn't require any home made class loaders or anything. All you need is classes depending on each other, and some bad luck. First of all, here are some basic facts about class loading: 1) If a thread needs to use a class that is not yet loaded, it will try to load that class 2) If another thread is already loading the class, the first thread will wait for the other thread to finish the loading 3) During the loading of a class, one thing that happens is that the <clinit method of a class is being run 4) The <clinit method initializes all static fields, and runs any static blocks in the class. Take the following class for example: class Foo { static Bar bar = new Bar(); static { System.out.println("Loading Foo"); } } The first time a thread needs to use the Foo class, the class will be initialized. The <clinit method will run, creating a new Bar object and printing "Loading Foo" But what happens if the Bar object has never been used before either? Well, then we will need to load that class as well, calling the Bar <clinit method as we go. Can you start to see the potential problem here? A hint is in fact #2 above. What if another thread is currently loading class Bar? The thread loading class Foo will have to wait for that thread to finish loading. But what happens if the <clinit method of class Bar tries to initialize a Foo object? That thread will have to wait for the first thread, and there we have the deadlock. Thread one is waiting for thread two to initialize class Bar, thread two is waiting for thread one to initialize class Foo. All that is needed for a class loading deadlock is static cross dependencies between two classes (and a multi threaded environment): class Foo { static Bar b = new Bar(); } class Bar { static Foo f = new Foo(); } If two threads cause these classes to be loaded at exactly the same time, we will have a deadlock. So, how do you avoid this? Well, one way is of course to not have these circular (static) dependencies. On the other hand, it can be very hard to detect these, and sometimes your design may depend on it. What you can do in that case is to make sure that the classes are first loaded single threadedly, for example during an initialization phase of your application. The following program shows this kind of deadlock. To help bad luck on the way, I added a one second sleep in the static block of the classes to trigger the unlucky timing. Notice that if you uncomment the "//Foo f = new Foo();" line in the main method, the class will be loaded single threadedly, and the program will terminate as it should. public class ClassLoadingDeadlock { // Start two threads. The first will instansiate a Foo object, // the second one will instansiate a Bar object. public static void main(String[] arg) { // Uncomment next line to stop the deadlock // Foo f = new Foo(); new Thread(new FooUser()).start(); new Thread(new BarUser()).start(); } } class FooUser implements Runnable { public void run() { System.out.println("FooUser causing class Foo to be loaded"); Foo f = new Foo(); System.out.println("FooUser done"); } } class BarUser implements Runnable { public void run() { System.out.println("BarUser causing class Bar to be loaded"); Bar b = new Bar(); System.out.println("BarUser done"); } } class Foo { static { // We are deadlock prone even without this sleep... // The sleep just makes us more deterministic try { Thread.sleep(1000); } catch(InterruptedException e) {} } static Bar b = new Bar(); } class Bar { static { try { Thread.sleep(1000); } catch(InterruptedException e) {} } static Foo f = new Foo(); }

    Read the article

< Previous Page | 115 116 117 118 119 120 121 122 123 124 125 126  | Next Page >