Search Results

Search found 17326 results on 694 pages for 'design pattern'.

Page 101/694 | < Previous Page | 97 98 99 100 101 102 103 104 105 106 107 108  | Next Page >

  • How accurate is "Business logic should be in a service, not in a model"?

    - by Jeroen Vannevel
    Situation Earlier this evening I gave an answer to a question on StackOverflow. The question: Editing of an existing object should be done in repository layer or in service? For example if I have a User that has debt. I want to change his debt. Should I do it in UserRepository or in service for example BuyingService by getting an object, editing it and saving it ? My answer: You should leave the responsibility of mutating an object to that same object and use the repository to retrieve this object. Example situation: class User { private int debt; // debt in cents private string name; // getters public void makePayment(int cents){ debt -= cents; } } class UserRepository { public User GetUserByName(string name){ // Get appropriate user from database } } A comment I received: Business logic should really be in a service. Not in a model. What does the internet say? So, this got me searching since I've never really (consciously) used a service layer. I started reading up on the Service Layer pattern and the Unit Of Work pattern but so far I can't say I'm convinced a service layer has to be used. Take for example this article by Martin Fowler on the anti-pattern of an Anemic Domain Model: There are objects, many named after the nouns in the domain space, and these objects are connected with the rich relationships and structure that true domain models have. The catch comes when you look at the behavior, and you realize that there is hardly any behavior on these objects, making them little more than bags of getters and setters. Indeed often these models come with design rules that say that you are not to put any domain logic in the the domain objects. Instead there are a set of service objects which capture all the domain logic. These services live on top of the domain model and use the domain model for data. (...) The logic that should be in a domain object is domain logic - validations, calculations, business rules - whatever you like to call it. To me, this seemed exactly what the situation was about: I advocated the manipulation of an object's data by introducing methods inside that class that do just that. However I realize that this should be a given either way, and it probably has more to do with how these methods are invoked (using a repository). I also had the feeling that in that article (see below), a Service Layer is more considered as a façade that delegates work to the underlying model, than an actual work-intensive layer. Application Layer [his name for Service Layer]: Defines the jobs the software is supposed to do and directs the expressive domain objects to work out problems. The tasks this layer is responsible for are meaningful to the business or necessary for interaction with the application layers of other systems. This layer is kept thin. It does not contain business rules or knowledge, but only coordinates tasks and delegates work to collaborations of domain objects in the next layer down. It does not have state reflecting the business situation, but it can have state that reflects the progress of a task for the user or the program. Which is reinforced here: Service interfaces. Services expose a service interface to which all inbound messages are sent. You can think of a service interface as a façade that exposes the business logic implemented in the application (typically, logic in the business layer) to potential consumers. And here: The service layer should be devoid of any application or business logic and should focus primarily on a few concerns. It should wrap Business Layer calls, translate your Domain in a common language that your clients can understand, and handle the communication medium between server and requesting client. This is a serious contrast to other resources that talk about the Service Layer: The service layer should consist of classes with methods that are units of work with actions that belong in the same transaction. Or the second answer to a question I've already linked: At some point, your application will want some business logic. Also, you might want to validate the input to make sure that there isn't something evil or nonperforming being requested. This logic belongs in your service layer. "Solution"? Following the guidelines in this answer, I came up with the following approach that uses a Service Layer: class UserController : Controller { private UserService _userService; public UserController(UserService userService){ _userService = userService; } public ActionResult MakeHimPay(string username, int amount) { _userService.MakeHimPay(username, amount); return RedirectToAction("ShowUserOverview"); } public ActionResult ShowUserOverview() { return View(); } } class UserService { private IUserRepository _userRepository; public UserService(IUserRepository userRepository) { _userRepository = userRepository; } public void MakeHimPay(username, amount) { _userRepository.GetUserByName(username).makePayment(amount); } } class UserRepository { public User GetUserByName(string name){ // Get appropriate user from database } } class User { private int debt; // debt in cents private string name; // getters public void makePayment(int cents){ debt -= cents; } } Conclusion All together not much has changed here: code from the controller has moved to the service layer (which is a good thing, so there is an upside to this approach). However this doesn't look like it had anything to do with my original answer. I realize design patterns are guidelines, not rules set in stone to be implemented whenever possible. Yet I have not found a definitive explanation of the service layer and how it should be regarded. Is it a means to simply extract logic from the controller and put it inside a service instead? Is it supposed to form a contract between the controller and the domain? Should there be a layer between the domain and the service layer? And, last but not least: following the original comment Business logic should really be in a service. Not in a model. Is this correct? How would I introduce my business logic in a service instead of the model?

    Read the article

  • Isn't MVC anti OOP?

    - by m3th0dman
    The main idea behind OOP is to unify data and behavior in a single entity - the object. In procedural programming there is data and separately algorithms modifying the data. In the Model-View-Controller pattern the data and the logic/algorithms are placed in distinct entities, the model and the controller respectively. In an equivalent OOP approach shouldn't the model and the controller be placed in the same logical entity?

    Read the article

  • What are DRY, KISS, SOLID, etc. classified as?

    - by Morgan Herlocker
    Is something like DRY a design pattern, a methodology, or something in between? They do not have specific implementations that could neccessarily be demonstrated(even if you can easily demonstrate a case NOT using something like KISS... see The Daily WTF for a plethora of examples), nor do they fully explain a development process like a methodology generally would. Where does that leave these types of "rule of thumb"'s?

    Read the article

  • Entity Framework 5, separating business logic from model - Repository?

    - by bnice7
    I am working on my first public-facing web application and I’m using MVC 4 for the presentation layer and EF 5 for the DAL. The database structure is locked, and there are moderate differences between how the user inputs data and how the database itself gets populated. I have done a ton of reading on the repository pattern (which I have never used) but most of my research is pushing me away from using it since it supposedly creates an unnecessary level of abstraction for the latest versions of EF since repositories and unit-of-work are already built-in. My initial approach is to simply create a separate set of classes for my business objects in the BLL that can act as an intermediary between my Controllers and the DAL. Here’s an example class: public class MyBuilding { public int Id { get; private set; } public string Name { get; set; } public string Notes { get; set; } private readonly Entities _context = new Entities(); // Is this thread safe? private static readonly int UserId = WebSecurity.GetCurrentUser().UserId; public IEnumerable<MyBuilding> GetList() { IEnumerable<MyBuilding> buildingList = from p in _context.BuildingInfo where p.Building.UserProfile.UserId == UserId select new MyBuilding {Id = p.BuildingId, Name = p.BuildingName, Notes = p.Building.Notes}; return buildingList; } public void Create() { var b = new Building {UserId = UserId, Notes = this.Notes}; _context.Building.Add(b); _context.SaveChanges(); // Set the building ID this.Id = b.BuildingId; // Seed 1-to-1 tables with reference the new building _context.BuildingInfo.Add(new BuildingInfo {Building = b}); _context.GeneralInfo.Add(new GeneralInfo {Building = b}); _context.LocationInfo.Add(new LocationInfo {Building = b}); _context.SaveChanges(); } public static MyBuilding Find(int id) { using (var context = new Entities()) // Is this OK to do in a static method? { var b = context.Building.FirstOrDefault(p => p.BuildingId == id && p.UserId == UserId); if (b == null) throw new Exception("Error: Building not found or user does not have access."); return new MyBuilding {Id = b.BuildingId, Name = b.BuildingInfo.BuildingName, Notes = b.Notes}; } } } My primary concern: Is the way I am instantiating my DbContext as a private property thread-safe, and is it safe to have a static method that instantiates a separate DbContext? Or am I approaching this all wrong? I am not opposed to learning up on the repository pattern if I am taking the total wrong approach here.

    Read the article

  • reference list for non-IT driven algorithmic patterns

    - by Quicker
    I am looking for a reference list for non-IT driven algorithmic patterns (which still can be helped with IT implementations of IT). An Example List would be: name; short desc; reference Travelling Salesman; find the shortest possible route on a multiple target path; http://en.wikipedia.org/wiki/Travelling_salesman_problem Ressource Disposition (aka Regulation); Distribute a limited/exceeding input on a given number output receivers based on distribution rules; http://database-programmer.blogspot.de/2010/12/critical-analysis-of-algorithm-sproc.html If there is no such list, but you instantly think of something specific, please 'put it on the desk'. Maybe I can compile something out of the input I get here (actually I am very frustrated as I did not find any such list via research by myself). Details on Scoping: I found it very hard to formulate what I want in a way everything is out that I do not need (which may be the issue why I did not find anything at google). There is a database centric definition for what I am looking for in the section 'Processes' of the second example link. That somehow fits, but the database focus sort of drifts away from the pattern thinking, which I have in mind. So here are my own thoughts around what's in and what's out: I am NOT looking for a foundational algo ref list, which is implemented as basis for any programming language. Eg. the php reference describes substr and strlen. That implements algos, but is not what I am looking for. the problem the algo does address would even exist, if there were no computers (or other IT components) the main focus of the algo is NOT to help other algo's chances are high, that there are implementions of the solution or any workaround without any IT support out there in the world however the algo could be benefitialy implemented/fully supported by a software application = means: the problem of the algo has to be addressed anyway, but running an algo implementation with software automates the process (that is why I posted it on stackoverflow and not somewhere else) typically such algo implementations have more than one input field value and more than one output field value - which implies it could not be implemented as simple function (which is fixed to produce not more than one output value) in a normalized data model often times such algo implementation outputs span accross multiple rows (sometimes multiple tables), whereby the number of output rows depends on the input paraters and rows in the table(s) at start time - which implies that any algo implementation/procedure must interact with a database (read and/or write) I am mainly looking for patterns, not for specific implementations. Example: The Travelling Salesman assumes any coordinates, however it does not say: You need a table targets with fields x and y. - however sometimes descriptions are focussed on examples with specific implementations very much - no worries, as long as the pattern gets clear

    Read the article

  • Overused or abused programming techniques

    - by Anto
    Are there any techniques in programming that you find to be overused (IE used way more excessively than what they should be) or abused, or used a bit for everything, while not being a really good solution to many of the problems which people attempt to solve with it. It could be regular expressions, some kind of design pattern or maybe an algorithm, or something completely different. Maybe you think people abuse multiple inheritance etc.

    Read the article

  • Need help on awk/sed/ perl pattern with regex / grep

    - by Jayakumar K
    Sample file output from grep file1:my $dbh = DBI->connect("dbi:mysql:$database_name", $DB_USER, $DB_PASSWD) file2:($dbc,$rc) = mysql_connect($mysql_host,$mysql_user,$mysql_password); The awk pattern should get values databasename, DB_USER And DB_PASSWD from line 1 and mysql_host,mysql_user and mysql_password from line 2 i.e all variables inside the function. Then it should search for the declaration of that variable in file before : (semicolon) ex: databasename in file1 may be $databasename = "dbweb" ; ex: mysql_user in file2 may be $mysql_user="root" ; Result: It should display variable declarations of all 6 variables along with filenames file2:$mysql_host = "db1"; file2:$mysql_user = "root"; file1:$DB_USER = 'user';

    Read the article

  • Windows freezes, showing a random color or pattern

    - by Manu
    I have a PC with windows 7, and I'm experiencing random freezes with increasing frequency. After some time (from 30 minutes to a few hours) the screen will shows a random color or pattern (vertical lines) and nothing works anymore, I need to reboot manually. I've checked the events log, but nothing is shown except for the unexpected reboot. I've tested the RAM with Memtest86+ 4.20, and the graphics card with FurMark 1.10.3, no errors have been found. I've updated the graphic drivers and openned the case to remove dust, but the issue is still there. Also, the problem doesn't seem to arise when I'm playing games fullscreen, but when I'm surfing the web, using itunes, or coding. My hardware is as follows : intel core i5 750 CPU, ATI Radeon HD 5670 graphics card, ASUSTeK P7P55D motherboard, two 2Gb KINGSTON DDR3 ram sticks, 2 SATA hard drives, a Netgear dongle for wifi.

    Read the article

  • Find command exclude files whose path match a certain pattern

    - by user40570
    I have a find command that looks for files that was modified recently and outputs the date find /path/on/server -mtime -1 -name '*.js' -exec ls -l {} \; I would like it to exclude any deeply nested folder that matches a certain pattern e.g. there are a number of folders that have a "statistics" directory and ".svn" directories. So i'd like to be able to say if the file that was modified yesterday is in a folder named statistics ignore it. Or perhaps not search for files in those folders at all.

    Read the article

  • How can we implement change notification propagation for WPF and SL in the MVVM pattern?

    - by Firoso
    Here's an example scenario targetting MVVM WPF/SL development: View data binds to view model Property "Target" "Target" exposes a field of an object called "data" that exists in the local application model, called "Original" when "Original" changes, it should raise notification to the view model and then propogate that change notification to the View. Here are the solutions I've come up with, but I don't like any of them all that much. I'm looking for other ideas, by the time we come up with something rock solid I'm certain Microsoft will have released .NET 5 with WPF/SL extensions for better tools for MVVM development. For now the question is, "What have you done to solve this problem and how has it worked out for you?" Option 1. Proposal: Attach a handler to data's PropertyChanged event that watches for string values of properties it cares about that might have changed, and raises the appropriate notification. Why I don't like it: Changes don't bubble naturally, objects must be explicitly watched, if data changes to a new source, events must be un-registered/registered. Why I kind of like it: I get explicit control over propogation of changes, and I don't have to use any types that belong at a higher level of the application such as dependancy properties. Option 2. Proposal: Attach a handler to data's PropertyChanged event that re-raises the event across all properties using the name property name. Why I don't like it: This is essentially the same as option 1, but less intelligent, and forces me to never change my property names, as they have to be the same as the property names on data Why I kind of like it: It's very easy to set up and I don't have to think about it... Then again if I try to think, and change names to things that make sense, I shoot myself in the foot, and then I have to think about it! Option 3. Proposal: Inherit my view model from dependancy object, and notify binding sources of changes directly. Why I don't like it: I'm not even 100% sure dependancy properties/objects can DO this, it was just a thought to look into. Also I don't personally feel that WPF/SL types like Dep Obj belong at the view model level. Why I kind of like it: IF it has the capability that I'm seeking then it's a good answer! minus that pesky layering issue. Option 4. Proposal: Use a consistant agent messaging system based off of Task Parallels DataFlow Library to propogate everything through linked pipelining. Why I don't like it: I've never tried this, and somehow I think it will be lacking, plus it requires me to think about my code completely differently all the way around. Why I kind of like it: It has the possiblity of allowing me to do some VERY fun manipulations with a push based data model and using ActionBlocks as validation AND setters to then privately change view model properties and explicitly control PropertyChanged notifications.

    Read the article

  • What is the basic pattern for using (N)Hibernate?

    - by Vilx-
    I'm creating a simple Windows Forms application with NHibernate and I'm a bit confused about how I'm supposed to use it. To quote the manual: ISession (NHibernate.ISession) A single-threaded, short-lived object representing a conversation between the application and the persistent store. Wraps an ADO.NET connection. Factory for ITransaction. Holds a mandatory (first-level) cache of persistent objects, used when navigating the object graph or looking up objects by identifier. Now, suppose I have the following scenario: I have a simple classifier which is a MSSQL table with two columns - ID (auto_increment) and Name (nvarchar). To edit this classifier I create a form which contains a single gridview and two buttons - OK and Cancel. The user can nearly directly edit the table in the gridview, and when he hits OK the changes he made are persisted to the DB (or if he hits cancel, nothing happens). Now, I have several questions about how to organize this: What should the lifetime of my ISession be? Should I create a single ISession for my whole application; an ISession for each of my forms (the application is single-threaded MDI); or an ISession for every DB operation/transaction? Does NHibernate offer some kind of built-in dirty tracking or must I do this myself? The manual mentions something like it here and there but does not go into details. How is this done? Is there not a huge overhead? Is it somehow tied with the cache(s) that NHibernate has? What are these caches for? Are they not specific to a single ISession? That is, if I use a seperate ISession for every transaction, won't it break the dirty tracking? How does the built-in dirty tracking detect deleted objects?

    Read the article

  • Which type of design pattern should be used to create an emulator?

    - by Facon
    I have programmed an emulator, but I have some doubts about how to organizate it properly, because, I see that it has some problems about classes connection (CPU <- Machine Board). For example: I/O ports, interruptions, communication between two or more CPU, etc. I need for the emulator to has the best performance and good understanding of the code. PD: Sorry for my bad English. EDITED: Asking for multiple patterns.

    Read the article

  • c++ Design pattern for CoW, inherited classes, and variable shared data?

    - by krunk
    I've designed a copy-on-write base class. The class holds the default set of data needed by all children in a shared data model/CoW model. The derived classes also have data that only pertains to them, but should be CoW between other instances of that derived class. I'm looking for a clean way to implement this. If I had a base class FooInterface with shared data FooDataPrivate and a derived object FooDerived. I could create a FooDerivedDataPrivate. The underlying data structure would not effect the exposed getters/setters API, so it's not about how a user interfaces with the objects. I'm just wondering if this is a typical MO for such cases or if there's a better/cleaner way? What peeks my interest, is I see the potential of inheritance between the the private data classes. E.g. FooDerivedDataPrivate : public FooDataPrivate, but I'm not seeing a way to take advantage of that polymorphism in my derived classes. class FooDataPrivate { public: Ref ref; // atomic reference counting object int a; int b; int c; }; class FooInterface { public: // constructors and such // .... // methods are implemented to be copy on write. void setA(int val); void setB(int val); void setC(int val); // copy constructors, destructors, etc. all CoW friendly private: FooDataPrivate *data; }; class FooDerived : public FooInterface { public: FooDerived() : FooInterface() {} private: // need more shared data for FooDerived // this is the ???, how is this best done cleanly? };

    Read the article

  • Which design pattern should be used to create an emulator?

    - by Facon
    I have programmed an emulator, but I have some doubts about how to organizate it properly, because, I see that it has some problems about classes connection (CPU <- Machine Board). For example: I/O ports, interruptions, communication between two or more CPU, etc. I need for the emulator to has the best performance and good understanding of the code. PD: Sorry for my bad English. EDITED - Asking for multiple patterns.

    Read the article

  • User Control as container

    - by Luca
    I'm designing a simple expander control. I've derived from UserControl, drawn inner controls, built, run; all ok. Since an inner Control is a Panel, I'd like to use it as container at design time. Indeed I've used the attributes: [Designer(typeof(ExpanderControlDesigner))] [Designer("System.Windows.Forms.Design.ParentControlDesigner, System.Design", typeof(IDesigner))] Great I say. But it isn't... The result is that I can use it as container at design time but: The added controls go back the inner controls already embedded in the user control Even if I push to top a control added at design time, at runtime it is back again on controls embedded to the user control I cannot restrict the container area at design time into a Panel area What am I missing? Here is the code for completeness... why this snippet of code is not working? [Designer(typeof(ExpanderControlDesigner))] [Designer("System.Windows.Forms.Design.ParentControlDesigner, System.Design", typeof(IDesigner))] public partial class ExpanderControl : UserControl { public ExpanderControl() { InitializeComponent(); .... [System.Security.Permissions.PermissionSet(System.Security.Permissions.SecurityAction.Demand, Name = "FullTrust")] internal class ExpanderControlDesigner : ControlDesigner { private ExpanderControl MyControl; public override void Initialize(IComponent component) { base.Initialize(component); MyControl = (ExpanderControl)component; // Hook up events ISelectionService s = (ISelectionService)GetService(typeof(ISelectionService)); IComponentChangeService c = (IComponentChangeService)GetService(typeof(IComponentChangeService)); s.SelectionChanged += new EventHandler(OnSelectionChanged); c.ComponentRemoving += new ComponentEventHandler(OnComponentRemoving); } private void OnSelectionChanged(object sender, System.EventArgs e) { } private void OnComponentRemoving(object sender, ComponentEventArgs e) { } protected override void Dispose(bool disposing) { ISelectionService s = (ISelectionService)GetService(typeof(ISelectionService)); IComponentChangeService c = (IComponentChangeService)GetService(typeof(IComponentChangeService)); // Unhook events s.SelectionChanged -= new EventHandler(OnSelectionChanged); c.ComponentRemoving -= new ComponentEventHandler(OnComponentRemoving); base.Dispose(disposing); } public override System.ComponentModel.Design.DesignerVerbCollection Verbs { get { DesignerVerbCollection v = new DesignerVerbCollection(); v.Add(new DesignerVerb("&asd", new EventHandler(null))); return v; } } } I've found many resources (Interaction, designed, limited area), but nothing was usefull for being operative...

    Read the article

  • What design pattern should be used to create an emulator?

    - by Facon
    I have programmed an emulator, but I have some doubts about how to organizate it properly, because, I see that it has some problems about classes connection (CPU <- Machine Board). For example: I/O ports, interruptions, communication between two or more CPU, etc. I need for the emulator to has the best performance and good understanding of the code. PD: Sorry for my bad English.

    Read the article

  • What types of objects should the ViewModel reference in the MVVM pattern?

    - by Blanthor
    I've seen quite a few examples of MVVM. I can see that the View should reference the ViewModel. I've seen recently an example of a ViewModel referencing a View, which seems wrong to me, as it would result in tighter coupling. Given that ViewModel is often described as an intermediary between the View and the Model, is there more to the ViewModel than a facade to domain objects? I hope I used the term "facade" correctly here.

    Read the article

  • How can I refactor this to work without breaking the pattern horribly?

    - by SnOrfus
    I've got a base class object that is used for filtering. It's a template method object that looks something like this. public class Filter { public void Process(User u, GeoRegion r, int countNeeded) { List<account> selected = this.Select(u, r, countNeeded); // 1 List<account> filtered = this.Filter(selected, u, r, countNeeded); // 2 if (filtered.Count > 0) { /* do businessy stuff */ } // 3 if (filtered.Count < countNeeded) this.SendToSuccessor(u, r, countNeeded - filtered) // 4 } } Select(...), Filter(...) are protected abstract methods and implemented by the derived classes. Select(...) finds objects in the based on x criteria, Filter(...) filters those selected further. If the remaining filtered collection has more than 1 object in it, we do some business stuff with it (unimportant to the problem here). SendToSuccessor(...) is called if there weren't enough objects found after filtering (it's a composite where the next class in succession will also be derived from Filter but have different filtering criteria) All has been ok, but now I'm building another set of filters, which I was going to subclass from this. The filters I'm building however would require different params and I don't want to just implement those methods and not use the params or just add to the param list the ones I need and have them not used in the existing filters. They still perform the same logical process though. I also don't want to complicated the consumer code for this (which looks like this) Filter f = new Filter1(); Filter f2 = new Filter2(); Filter f3 = new Filter3(); f.Sucessor = f2; f2.Sucessor = f3; /* and so on adding filters as successors to previous ones */ foreach (User u in users) { foreach (GeoRegion r in regions) { f.Process(u, r, ##); } } How should I go about it?

    Read the article

  • Can I use the decorator pattern to wrap a method body?

    - by mgroves
    I have a bunch of methods with varying signatures. These methods interact with a fragile data connection, so we often use a helper class to perform retries/reconnects, etc. Like so: MyHelper.PerformCall( () => { doStuffWithData(parameters...) }); And this works fine, but it can make the code a little cluttery. What I would prefer to do is decorate the methods that interact with the data connection like so: [InteractsWithData] protected string doStuffWithData(parameters...) { // do stuff... } And then essentially, whenever doStuffWithData is called, the body of that method would be passed in as an Action to MyHelper.PerformCall(). How do I do this?

    Read the article

  • Where does the query language sit within the MVC pattern?

    - by weesilmania
    I'd assume that since the query language sits within the controller (typically) that it belongs to that component, but if I play devil's advocate I'd argue that the query language is execute within the domain of the model, and is tightly coupled to that component so it might also be a part of it. Anyone know the answer? Is there a straight answer or is it technology specific?

    Read the article

  • How to implement the API/SPI Pattern in Java?

    - by Adam Tannon
    I am creating a framework that exposes an API for developers to use: public interface MyAPI { public void doSomeStuff(); public int getWidgets(boolean hasRun); } All the developers should have to do is code their projects against these API methods. I also want them to be able to place different "drivers"/"API bindings" on the runtime classpath (the same way JDBC or SLF4J work) and have the API method calls (doSomeStuff(), etc.) operate on different 3rd party resources (files, servers, whatever). Thus the same code and API calls will map to operations on different resources depending on what driver/binding the runtime classpath sees (i.e. myapi-ftp, myapi-ssh, myapi-teleportation). How do I write (and package) an SPI that allows for such runtime binding, and then maps MyAPI calls to the correct (concrete) implementation? In other words, if myapi-ftp allows you to getWidgets(boolean) from an FTP server, how would I could this up (to make use of both the API and SPI)? Bonus points for concrete, working Java code example! Thanks in advance!

    Read the article

  • Plagued by multithreaded bugs

    - by koncurrency
    On my new team that I manage, the majority of our code is platform, TCP socket, and http networking code. All C++. Most of it originated from other developers that have left the team. The current developers on the team are very smart, but mostly junior in terms of experience. Our biggest problem: multi-threaded concurrency bugs. Most of our class libraries are written to be asynchronous by use of some thread pool classes. Methods on the class libraries often enqueue long running taks onto the thread pool from one thread and then the callback methods of that class get invoked on a different thread. As a result, we have a lot of edge case bugs involving incorrect threading assumptions. This results in subtle bugs that go beyond just having critical sections and locks to guard against concurrency issues. What makes these problems even harder is that the attempts to fix are often incorrect. Some mistakes I've observed the team attempting (or within the legacy code itself) includes something like the following: Common mistake #1 - Fixing concurrency issue by just put a lock around the shared data, but forgetting about what happens when methods don't get called in an expected order. Here's a very simple example: void Foo::OnHttpRequestComplete(statuscode status) { m_pBar->DoSomethingImportant(status); } void Foo::Shutdown() { m_pBar->Cleanup(); delete m_pBar; m_pBar=nullptr; } So now we have a bug in which Shutdown could get called while OnHttpNetworkRequestComplete is occuring on. A tester finds the bug, captures the crash dump, and assigns the bug to a developer. He in turn fixes the bug like this. void Foo::OnHttpRequestComplete(statuscode status) { AutoLock lock(m_cs); m_pBar->DoSomethingImportant(status); } void Foo::Shutdown() { AutoLock lock(m_cs); m_pBar->Cleanup(); delete m_pBar; m_pBar=nullptr; } The above fix looks good until you realize there's an even more subtle edge case. What happens if Shutdown gets called before OnHttpRequestComplete gets called back? The real world examples my team has are even more complex, and the edge cases are even harder to spot during the code review process. Common Mistake #2 - fixing deadlock issues by blindly exiting the lock, wait for the other thread to finish, then re-enter the lock - but without handling the case that the object just got updated by the other thread! Common Mistake #3 - Even though the objects are reference counted, the shutdown sequence "releases" it's pointer. But forgets to wait for the thread that is still running to release it's instance. As such, components are shutdown cleanly, then spurious or late callbacks are invoked on an object in an state not expecting any more calls. There are other edge cases, but the bottom line is this: Multithreaded programming is just plain hard, even for smart people. As I catch these mistakes, I spend time discussing the errors with each developer on developing a more appropriate fix. But I suspect they are often confused on how to solve each issue because of the enormous amount of legacy code that the "right" fix will involve touching. We're going to be shipping soon, and I'm sure the patches we're applying will hold for the upcoming release. Afterwards, we're going to have some time to improve the code base and refactor where needed. We won't have time to just re-write everything. And the majority of the code isn't all that bad. But I'm looking to refactor code such that threading issues can be avoided altogether. One approach I am considering is this. For each significant platform feature, have a dedicated single thread where all events and network callbacks get marshalled onto. Similar to COM apartment threading in Windows with use of a message loop. Long blocking operations could still get dispatched to a work pool thread, but the completion callback is invoked on on the component's thread. Components could possibly even share the same thread. Then all the class libraries running inside the thread can be written under the assumption of a single threaded world. Before I go down that path, I am also very interested if there are other standard techniques or design patterns for dealing with multithreaded issues. And I have to emphasize - something beyond a book that describes the basics of mutexes and semaphores. What do you think? I am also interested in any other approaches to take towards a refactoring process. Including any of the following: Literature or papers on design patterns around threads. Something beyond an introduction to mutexes and semaphores. We don't need massive parallelism either, just ways to design an object model so as to handle asynchronous events from other threads correctly. Ways to diagram the threading of various components, so that it will be easy to study and evolve solutions for. (That is, a UML equivalent for discussing threads across objects and classes) Educating your development team on the issues with multithreaded code. What would you do?

    Read the article

  • How to Correct & Improve the Design of this Code?

    - by DaveDev
    HI Guys, I've been working on a little experiement to see if I could create a helper method to serialize any of my types to any type of HTML tag I specify. I'm getting a NullReferenceException when _writer = _viewContext.Writer; is called in protected virtual void Dispose(bool disposing) {/*...*/} I think I'm at a point where it almost works (I've gotten other implementations to work) and I was wondering if somebody could point out what I'm doing wrong? Also, I'd be interested in hearing suggestions on how I could improve the design? So basically, I have this code that will generate a Select box with a number of options: // the idea is I can use one method to create any complete tag of any type // and put whatever I want in the content area <% using (Html.GenerateTag<SelectTag>(Model, new { href = Url.Action("ActionName") })) { %> <%foreach (var fund in Model.Funds) {%> <% using (Html.GenerateTag<OptionTag>(fund)) { %> <%= fund.Name %> <% } %> <% } %> <% } %> This Html.GenerateTag helper is defined as: public static MMTag GenerateTag<T>(this HtmlHelper htmlHelper, object elementData, object attributes) where T : MMTag { return (T)Activator.CreateInstance(typeof(T), htmlHelper.ViewContext, elementData, attributes); } Depending on the type of T it'll create one of the types defined below, public class HtmlTypeBase : MMTag { public HtmlTypeBase() { } public HtmlTypeBase(ViewContext viewContext, params object[] elementData) { base._viewContext = viewContext; base.MergeDataToTag(viewContext, elementData); } } public class SelectTag : HtmlTypeBase { public SelectTag(ViewContext viewContext, params object[] elementData) { base._tag = new TagBuilder("select"); //base.MergeDataToTag(viewContext, elementData); } } public class OptionTag : HtmlTypeBase { public OptionTag(ViewContext viewContext, params object[] elementData) { base._tag = new TagBuilder("option"); //base.MergeDataToTag(viewContext, _elementData); } } public class AnchorTag : HtmlTypeBase { public AnchorTag(ViewContext viewContext, params object[] elementData) { base._tag = new TagBuilder("a"); //base.MergeDataToTag(viewContext, elementData); } } all of these types (anchor, select, option) inherit from HtmlTypeBase, which is intended to perform base.MergeDataToTag(viewContext, elementData);. This doesn't happen though. It works if I uncomment the MergeDataToTag methods in the derived classes, but I don't want to repeat that same code for every derived class I create. This is the definition for MMTag: public class MMTag : IDisposable { internal bool _disposed; internal ViewContext _viewContext; internal TextWriter _writer; internal TagBuilder _tag; internal object[] _elementData; public MMTag() {} public MMTag(ViewContext viewContext, params object[] elementData) { } public void Dispose() { Dispose(true /* disposing */); GC.SuppressFinalize(this); } protected virtual void Dispose(bool disposing) { if (!_disposed) { _disposed = true; _writer = _viewContext.Writer; _writer.Write(_tag.ToString(TagRenderMode.EndTag)); } } protected void MergeDataToTag(ViewContext viewContext, object[] elementData) { Type elementDataType = elementData[0].GetType(); foreach (PropertyInfo prop in elementDataType.GetProperties()) { if (prop.PropertyType.IsPrimitive || prop.PropertyType == typeof(Decimal) || prop.PropertyType == typeof(String)) { object propValue = prop.GetValue(elementData[0], null); string stringValue = propValue != null ? propValue.ToString() : String.Empty; _tag.Attributes.Add(prop.Name, stringValue); } } var dic = new Dictionary<string, object>(StringComparer.OrdinalIgnoreCase); var attributes = elementData[1]; if (attributes != null) { foreach (PropertyDescriptor descriptor in TypeDescriptor.GetProperties(attributes)) { object value = descriptor.GetValue(attributes); dic.Add(descriptor.Name, value); } } _tag.MergeAttributes<string, object>(dic); _viewContext = viewContext; _viewContext.Writer.Write(_tag.ToString(TagRenderMode.StartTag)); } } Thanks Dave

    Read the article

  • How to create multiple Repository object inside a Repository class using Unit Of Work?

    - by Santosh
    I am newbie to MVC3 application development, currently, we need following Application technologies as requirement MVC3 framework IOC framework – Autofac to manage object creation dynamically Moq – Unit testing Entity Framework Repository and Unit Of Work Pattern of Model class I have gone through many article to explore an basic idea about the above points but still I am little bit confused on the “Repository and Unit Of Work Pattern “. Basically what I understand Unit Of Work is a pattern which will be followed along with Repository Pattern in order to share the single DB Context among all Repository object, So here is my design : IUnitOfWork.cs public interface IUnitOfWork : IDisposable { IPermitRepository Permit_Repository{ get; } IRebateRepository Rebate_Repository { get; } IBuildingTypeRepository BuildingType_Repository { get; } IEEProjectRepository EEProject_Repository { get; } IRebateLookupRepository RebateLookup_Repository { get; } IEEProjectTypeRepository EEProjectType_Repository { get; } void Save(); } UnitOfWork.cs public class UnitOfWork : IUnitOfWork { #region Private Members private readonly CEEPMSEntities context = new CEEPMSEntities(); private IPermitRepository permit_Repository; private IRebateRepository rebate_Repository; private IBuildingTypeRepository buildingType_Repository; private IEEProjectRepository eeProject_Repository; private IRebateLookupRepository rebateLookup_Repository; private IEEProjectTypeRepository eeProjectType_Repository; #endregion #region IUnitOfWork Implemenation public IPermitRepository Permit_Repository { get { if (this.permit_Repository == null) { this.permit_Repository = new PermitRepository(context); } return permit_Repository; } } public IRebateRepository Rebate_Repository { get { if (this.rebate_Repository == null) { this.rebate_Repository = new RebateRepository(context); } return rebate_Repository; } } } PermitRepository .cs public class PermitRepository : IPermitRepository { #region Private Members private CEEPMSEntities objectContext = null; private IObjectSet<Permit> objectSet = null; #endregion #region Constructors public PermitRepository() { } public PermitRepository(CEEPMSEntities _objectContext) { this.objectContext = _objectContext; this.objectSet = objectContext.CreateObjectSet<Permit>(); } #endregion public IEnumerable<RebateViewModel> GetRebatesByPermitId(int _permitId) { // need to implment } } PermitController .cs public class PermitController : Controller { #region Private Members IUnitOfWork CEEPMSContext = null; #endregion #region Constructors public PermitController(IUnitOfWork _CEEPMSContext) { if (_CEEPMSContext == null) { throw new ArgumentNullException("Object can not be null"); } CEEPMSContext = _CEEPMSContext; } #endregion } So here I am wondering how to generate a new Repository for example “TestRepository.cs” using same pattern where I can create more then one Repository object like RebateRepository rebateRepo = new RebateRepository () AddressRepository addressRepo = new AddressRepository() because , what ever Repository object I want to create I need an object of UnitOfWork first as implmented in the PermitController class. So if I would follow the same in each individual Repository class that would again break the priciple of Unit Of Work and create multiple instance of object context. So any idea or suggestion will be highly appreciated. Thank you

    Read the article

< Previous Page | 97 98 99 100 101 102 103 104 105 106 107 108  | Next Page >