Search Results

Search found 26275 results on 1051 pages for 'multi view'.

Page 81/1051 | < Previous Page | 77 78 79 80 81 82 83 84 85 86 87 88  | Next Page >

  • PHP multi dimensional array manipulation

    - by atif089
    Hi, This is my array Array ( [0] => Array ( [sample_id] => 3 [time] => 2010-05-30 21:11:47 ) [1] => Array ( [sample_id] => 2 [time] => 2010-05-30 21:11:47 ) [2] => Array ( [sample_id] => 1 [time] => 2010-05-30 21:11:47 ) ) And I want to get all the sample_ids in one array. can someone please help ? Can this be done without for loops (because arrays are very large).

    Read the article

  • multi-core processing in R on windows XP - via doMC and foreach

    - by Jan
    Hi guys, I'm posting this question to ask for advice on how to optimize the use of multiple processors from R on a Windows XP machine. At the moment I'm creating 4 scripts (each script with e.g. for (i in 1:100) and (i in 101:200), etc) which I run in 4 different R sessions at the same time. This seems to use all the available cpu. I however would like to do this a bit more efficient. One solution could be to use the "doMC" and the "foreach" package but this is not possible in R on a Windows machine. e.g. library("foreach") library("strucchange") library("doMC") # would this be possible on a windows machine? registerDoMC(2) # for a computer with two cores (processors) ## Nile data with one breakpoint: the annual flows drop in 1898 ## because the first Ashwan dam was built data("Nile") plot(Nile) ## F statistics indicate one breakpoint fs.nile <- Fstats(Nile ~ 1) plot(fs.nile) breakpoints(fs.nile) # , hpc = "foreach" --> It would be great to test this. lines(breakpoints(fs.nile)) Any solutions or advice? Thanks, Jan

    Read the article

  • Stored Procedure for Multi-Table Insert Error: Cannot Insert the Value Null into Column

    - by SidC
    Good Evening All, I've created the following stored procedure: CREATE PROCEDURE AddQuote -- Add the parameters for the stored procedure here AS BEGIN -- SET NOCOUNT ON added to prevent extra result sets from -- interfering with SELECT statements. SET NOCOUNT ON; Declare @CompanyName nvarchar(50), @Addr nvarchar(50), @City nvarchar(50), @State nvarchar(2), @Zip nvarchar(5), @NeedDate datetime, @PartNumber float, @Qty int -- Insert statements for procedure here Insert into dbo.Customers (CompanyName, Address, City, State, ZipCode) Values (@CompanyName, @Addr, @City, @State, @Zip) Insert into dbo.Orders (NeedbyDate) Values(@NeedDate) Insert into dbo.OrderDetail (fkPartNumber,Qty) Values (@PartNumber,@Qty) END GO When I execute AddQuote, I receive an error stating: Msg 515, Level 16, State 2, Procedure AddQuote, Line 31 Cannot insert the value NULL into column 'ID', table 'Diel_inventory.dbo.OrderDetail'; column does not allow nulls. INSERT fails. The statement has been terminated. I understand that I've set Qty field to not allow nulls and want to continue doing so. However, are there other syntax changes I should make to ensure that this sproc works correctly? Thanks, Sid

    Read the article

  • Marionette multi user/roles application

    - by Fabrizio Fortino
    I have to build a pretty complex application using Backbone Marionette. The user interface has to handle multiple users with different roles. For example the 'admin' user will see the complete menu whereas the 'guest' user will access a subset of the same menu. Moreover some views will be accessible to all the users but the functions inside them (add, edit, delete) need to be profiled on the different roles. I am not sure about the right approach to use in order to solve this issue. I could have different templates for the different roles but in this case plenty of code will be duplicated inside them. Is there any best practice (or maybe some example) to sort my problem out using Marionette? Thanks in advance, Fabrizio

    Read the article

  • multi word query in mysql

    - by salmane
    Hi there , in order to make things easier for users i want to add multiple keyword search to my site. so that in the input the user would do something like : " keyword1 keyword 2" ( similar to google for example. would i need to write a code that would parse that string and do queries based on that or is there something built in mysql that could do it?

    Read the article

  • Multi-site WCF Service

    - by vault
    I have implemented a WCF service that will be used at Site A with 5 computers in the LAN accessing the service. What I would like to do is have an elevated user/administrator be able to use one of the 5 machines and query an identical service at Site B (also with 5 computers) that they will need to connect to by bypassing the LAN firewall.Accessing data from Site A to Site B (and vica-versa) need only be read-only Is there a standardised way to acheive this using WCF?

    Read the article

  • Clustered index - multi-part vs single-part index and effects of inserts/deletes

    - by Anssssss
    This question is about what happens with the reorganizing of data in a clustered index when an insert is done. I assume that it should be more expensive to do inserts on a table which has a clustered index than one that does not because reorganizing the data in a clustered index involves changing the physical layout of the data on the disk. I'm not sure how to phrase my question except through an example I came across at work. Assume there is a table (Junk) and there are two queries that are done on the table, the first query searches by Name and the second query searches by Name and Something. As I'm working on the database I discovered that the table has been created with two indexes, one to support each query, like so: --drop table Junk1 CREATE TABLE Junk1 ( Name char(5), Something char(5), WhoCares int ) CREATE CLUSTERED INDEX IX_Name ON Junk1 ( Name ) CREATE NONCLUSTERED INDEX IX_Name_Something ON Junk1 ( Name, Something ) Now when I looked at the two indexes, it seems that IX_Name is redundant since IX_Name_Something can be used by any query that desires to search by Name. So I would eliminate IX_Name and make IX_Name_Something the clustered index instead: --drop table Junk2 CREATE TABLE Junk2 ( Name char(5), Something char(5), WhoCares int ) CREATE CLUSTERED INDEX IX_Name_Something ON Junk2 ( Name, Something ) Someone suggested that the first indexing scheme should be kept since it would result in more efficient inserts/deletes (assume that there is no need to worry about updates for Name and Something). Would that make sense? I think the second indexing method would be better since it means one less index needs to be maintained. I would appreciate any insight into this specific example or directing me to more info on maintenance of clustered indexes.

    Read the article

  • ActiveRecord: Produce multi-line human-friendly json

    - by Mika
    Using ActiveRecord::Base.to_json I do: user = User.find_by_name 'Mika' {"created_at":"2011-07-10T11:30:49+03:00","id":5,"is_deleted":null,"name":"Mika"} Now, what I would like to have is: { "created_at":"2011-07-10T11:30:49+03:00", "id":5, "is_deleted":null, "name":"Mika" } Is there an option to do this? It would be great to have a global option, so that the behaviour be set depending on dev/live environment.

    Read the article

  • ViewModel Views relation/link/syncroniztion

    - by mehran
    Third try to describing problem: Try 1: Sunchronizing view model and view Try2: WPF ViewModel not active presenter Try3: I have some class for view models: public class Node : INotifyPropertyChanged { Guid NodeId { get; set; } public string Name { get; set; } } public class Connection: INotifyPropertyChanged { public Node StartNode { get; set; } public Node EndNode { get; set; } } public class SettingsPackModel { public List<Node> Nodes { get; private set; } public List<Connection> Connections { get; private set; } } I also have some templates to displays these models: <DataTemplate DataType="{x:Type vm:Node}">…</DataTemplate> <DataTemplate DataType="{x:Type vm:Connection}"> <my:ConnectionElment StartNodeElment="???" EndNodeElment="???"> </my:ConnectionElment> <DataTemplate> But the problem is that DataTemplate for Connection need reference ot two element of type UIElement , how can I pass these two, how can I fill ??? in above expression?

    Read the article

  • Multi-module maven build : different result from parent and from module

    - by Albaku
    I am migrating an application from ant build to maven 3 build. This app is composed by : A parent project specifying all the modules to build A project generating classes with jaxb and building a jar with them A project building an ejb project 3 projects building war modules 1 project building an ear Here is an extract from my parent pom : <groupId>com.test</groupId> <artifactId>P</artifactId> <packaging>pom</packaging> <version>04.01.00</version> <modules> <module>../PValidationJaxb</module> <-- jar <module>../PValidation</module> <-- ejb <module>../PImport</module> <-- war <module>../PTerminal</module> <-- war <module>../PWebService</module> <-- war <module>../PEAR</module> <-- ear </modules> I have several problems which I think have the same origin, probably a dependency management issue that I cannot figure out : The generated modules are different depending on if I build from the parent pom or a single module. Typically if I build PImport only, the generated war is similar to what I had with my ant build and if I build from the parent pom, my war took 20MB, a lot of dependencies from other modules had been added. Both wars are running well. My project PWebService has unit tests to be executed during the build. It is using mock-ejb which has cglib as dependency. Having a problem of ClassNotFound with this one, I had to exclude it and add a dependency to cglib-nodep (see last pom extract). If I then build only this module, it is working well. But if I build from the parent project, it fails because other dependencies in other modules also had an implicit dependency on cglib. I had to exclude it in every modules pom and add the dependency to cglib-nodep everywhere to make it run. Do I miss something important in my configuration ? The PValidation pom extract : It is creating a jar containing an ejb with interfaces generated by xdoclet, as well as a client jar. <parent> <groupId>com.test</groupId> <artifactId>P</artifactId> <version>04.01.00</version> </parent> <artifactId>P-validation</artifactId> <packaging>ejb</packaging> <dependencies> <dependency> <groupId>com.test</groupId> <artifactId>P-jaxb</artifactId> <version>${project.version}</version> </dependency> <dependency> <groupId>org.hibernate</groupId> <artifactId>hibernate</artifactId> <version>3.2.5.ga</version> <exclusions> <exclusion> <groupId>cglib</groupId> <artifactId>cglib</artifactId> </exclusion> </exclusions> </dependency> <dependency> <groupId>cglib</groupId> <artifactId>cglib-nodep</artifactId> <version>2.2.2</version> </dependency> ... [other libs] ... </dependencies> <build> ... <plugins> <plugin> <groupId>org.apache.maven.plugins</groupId> <artifactId>maven-ejb-plugin</artifactId> <configuration> <ejbVersion>2.0</ejbVersion> <generateClient>true</generateClient> </configuration> </plugin> <plugin> <groupId>org.codehaus.mojo</groupId> <artifactId>xdoclet-maven-plugin</artifactId> ... The PImport pom extract : It depends on both Jaxb generated jar and the ejb client jar. <parent> <groupId>com.test</groupId> <artifactId>P</artifactId> <version>04.01.00</version> </parent> <artifactId>P-import</artifactId> <packaging>war</packaging> <dependencies> <dependency> <groupId>com.test</groupId> <artifactId>P-jaxb</artifactId> <version>${project.version}</version> </dependency> <dependency> <groupId>com.test</groupId> <artifactId>P-validation</artifactId> <version>${project.version}</version> <type>ejb-client</type> </dependency> ... [other libs] ... </dependencies> The PWebService pom extract : <parent> <groupId>com.test</groupId> <artifactId>P</artifactId> <version>04.01.00</version> </parent> <artifactId>P-webservice</artifactId> <packaging>war</packaging> <properties> <jersey.version>1.14</jersey.version> </properties> <dependencies> <dependency> <groupId>com.sun.jersey</groupId> <artifactId>jersey-servlet</artifactId> <version>${jersey.version}</version> </dependency> <dependency> <groupId>com.rte.etso</groupId> <artifactId>etso-validation</artifactId> <version>${project.version}</version> <type>ejb-client</type> </dependency> ... [other libs] ... <dependency> <groupId>org.mockejb</groupId> <artifactId>mockejb</artifactId> <version>0.6-beta2</version> <scope>test</scope> <exclusions> <exclusion> <groupId>cglib</groupId> <artifactId>cglib-full</artifactId> </exclusion> </exclusions> </dependency> <dependency> <groupId>cglib</groupId> <artifactId>cglib-nodep</artifactId> <version>2.2.2</version> <scope>test</scope> </dependency> </dependencies> Many thanks

    Read the article

  • Multiple on-screen view controllers in iPhone apps

    - by Felixyz
    I'm creating a lot of custom views and controllers in a lot of my apps and so far I've mostly set them up programmatically, with adjustments and instantiations being controlled from plists. However, now I'm transitioning to using Interface Builder as much as possible (wish I had done that before, was always on my back-list). Apple is recommending against having many view controllers being simultaneously active in iPhone apps, with a couple of well-known exceptions. I've never fully understood why it should be bad to have different parts of the interface belong to different controllers, if they have no interdependent functionality at all. Does having multiple controllers risk messing up the responder chain, or is there some other reason that it's not recommended, except for the fact that it's usually not needed? What I want to be able to do is design reusable views and controls in IB, but since a nib is not associated with a view, but with a view controller, it seems I'd have to have different parts of the screen be connected to different controllers. I know it's possible to have other objects than view controllers being instantiated from nibs. Should I look into how to create my own alternative more light-weight controllers (that could be sub-controllers of a UIViewController) which could be instantiated from nibs?

    Read the article

  • Multi-platform development from one computer

    - by iama
    I am planning to build a new development computer for both Windows & Linux platforms. On Windows, my development would be primarily in .NET/C#/IIS/MSSQL Server. On Linux—preferably Ubuntu—my development would be in Ruby and Python. I am thinking of buying a laptop with Windows 7 pre-installed with 4GB RAM, Intel Core 2 Duo, and 320 GB HD; running 2 VMs for both Windows and Linux development with the host OS as my work station. Of course, I would be running DBs and web servers on the respective platforms. Is this a typical setup? My only concern is running two VMs side by side. Not sure if this configuration would be optimal. Alternative would be to do my Windows development on the host Windows 7 OS. What are your thoughts?

    Read the article

  • Bulding an multi-platform SWT application using Ant

    - by Mridang Agarwalla
    I'm writing an SWT application which can be used on Windows (32/64 bit) and Mac OSX (32/64 bit). Apart from the JRE I rely on the SWT library found here. I can find four versions of the SWT library depending upon my target platforms (as mentioned above). When building my application, how can I compile using the correct SWT Jar? If possible, I'd like to try and avoid hard-coding the Jar version, platform and architecture. The SWT Jars are named like this: swt-win32-x86_64.jar swt-win32-x86_32.jar swt-macosx-x86_32.jar swt-macosx-x86_64.jar (My project will be an open source project. I'd like people to be able to download the source and build it and therefore I've thought of including all the four versions of the SWT Jars in the source distribution. I hope this is the correct approach of publishing code relying on third-part libraries.) Thanks everyone.

    Read the article

  • How to read a file with variable multi-row data in Python

    - by dr.bunsen
    I have a file that is about 100Mb that looks like this: #meta data 1 skadjflaskdjfasljdfalskdjfl sdkfjhasdlkgjhsdlkjghlaskdj asdhfk #meta data 2 jflaksdjflaksjdflkjasdlfjas ldaksjflkdsajlkdfj #meta data 3 alsdkjflasdjkfglalaskdjf This file contains one row of meta data that corresponds to several, variable length data containing only alpha-numeric characters. What is the best way to read this data into a simple list like this: data = [[#meta data 1, skadjflaskdjfasljdfalskdjflsdkfjhasdlkgjhsdlkjghlaskdjasdhfk], [#meta data 2, jflaksdjflaksjdflkjasdlfjasldaksjflkdsajlkdfj], [#meta data 3, alsdkjflasdjkfglalaskdjf]] My initial idea was to use the read() method to read the whole file into memory and then use regular expressions to parse the data into the desired format. Is there a better more pythonic way? All metadata lines start with an octothorpe and all data lines are all alpha-numeric. Thanks!

    Read the article

  • Twisted - how to create multi protocol process and send the data between the protocols

    - by SpankMe
    Hey, Im trying to write a program that would be listening for data (simple text messages) on some port (say tcp 6666) and then pass them to one or more different protocols - irc, xmpp and so on. I've tried many approaches and digged the Internet, but I cant find easy and working solution for such task. The code I am currently fighting with is here: http://pastebin.com/ri7caXih I would like to know how to from object like: ircf = ircFactory('asdfasdf', '#asdf666') get access to self protocol methods, because this: self.protocol.dupa1(msg) returns error about self not being passed to active protocol object. Or maybe there is other, better, easier and more kosher way to create single reactor with multiple protocols and have actions triggeres when a message arrives on any of them, and then pass that message to other protocols for handling/processing/sending? Any help will be highly appreciated!

    Read the article

  • Give me a practical use-case of Multi-set

    - by Calm Storm
    I would like to know a few practical use-cases (if they are not related/tied to any programming language it will be better).I can associate Sets, Lists and Maps to practical use cases. For example if you wanted a glossary of a book where terms that you want are listed alphabetically and a location/page number is the value, you would use the collection TreeMap(OrderedMap which is a Map) Somehow, I can't associate MultiSets with any "practical" usecase. Does someone know of any uses? http://en.wikipedia.org/wiki/Multiset does not tell me enough :) PS: If you guys think this should be community-wiki'ed it is okay. The only reason I did not do it was "There is a clear objective way to answer this question".

    Read the article

  • Java multi Generic collection parameters complie error

    - by Geln Yang
    Hi, So strange!Please have a look the code first: public class A { } public class B extends A { } public class C extends A { } public class TestMain { public <T extends A> void test(T a, T b) { } public <T extends A> void test(List<T> a, List<T> b) { } public static void main(String[] args) { new TestMain().test(new B(), new C()); new TestMain().test(new ArrayList<B>(), new ArrayList<C>()); } } The statement "new TestMain().test(new ArrayList(), new ArrayList())" get a "Bound mismatch" compile error, while "new TestMain().test(new B(), new C())" is compiled ok. Bound mismatch: The generic method test(T, T) of type TestMain is not applicable for the arguments (ArrayList, ArrayList). The inferred type ArrayList is not a valid substitute for the bounded parameter It seems the type of the second generic List parameter is limited by the Type of the first.Is it a feature or a bug of the compile program? ps, jdk:1.6,IDE:Eclipse 3.5.1

    Read the article

  • JSF don't find component in view root with the form id

    - by kenzokujpn
    I have a t:inputFileUpload inside the form, in the html of the display page the id of this component is form:inputFile but when I tried to get the component from the view root using "form:inputFile" the return is null, but when the "form:" is removed the return is the component. The component don't set the value in my managed bean, someone have this problem? EDIT: <h:form id="form" enctype="multipart/form-data"> <t:inputFileUpload id="inputFile" size="40" value="#{managedBean.inputFile}"/> </h:form> In the managed bean: private UploadedFile inputFile; with the gets and sets provided by Eclipse. //This method scans the view root and returns the component with the id passed as parameter findComponentInRoot("form:inputFile"); This returns null, but when I use: //This method scans the view root and returns the component with the id passed as parameter findComponentInRoot("inputFile"); The return is the component I'm looking for, but when I use the View Source in Internet Explorer the id of this component is "form:inputFile". I don't know if this is related, but the component don't set the value in my managed bean and it's strange the fact that the id of the component is different from the HTML source. I'm using JSF 1.2 Mojarra. Someone else has this problem? Or know why this happens?

    Read the article

  • Sql server query using function and view is slower

    - by Lieven Cardoen
    I have a table with a xml column named Data: CREATE TABLE [dbo].[Users]( [UserId] [int] IDENTITY(1,1) NOT NULL, [FirstName] [nvarchar](max) NOT NULL, [LastName] [nvarchar](max) NOT NULL, [Email] [nvarchar](250) NOT NULL, [Password] [nvarchar](max) NULL, [UserName] [nvarchar](250) NOT NULL, [LanguageId] [int] NOT NULL, [Data] [xml] NULL, [IsDeleted] [bit] NOT NULL,... In the Data column there's this xml <data> <RRN>...</RRN> <DateOfBirth>...</DateOfBirth> <Gender>...</Gender> </data> Now, executing this query: SELECT UserId FROM Users WHERE data.value('(/data/RRN)[1]', 'nvarchar(max)') = @RRN after clearing the cache takes (if I execute it a couple of times after each other) 910, 739, 630, 635, ... ms. Now, a db specialist told me that adding a function, a view and changing the query would make it much more faster to search a user with a given RRN. But, instead, these are the results when I execute with the changes from the db specialist: 2584, 2342, 2322, 2383, ... This is the added function: CREATE FUNCTION dbo.fn_Users_RRN(@data xml) RETURNS varchar(100) WITH SCHEMABINDING AS BEGIN RETURN @data.value('(/data/RRN)[1]', 'varchar(max)'); END; The added view: CREATE VIEW vwi_Users WITH SCHEMABINDING AS SELECT UserId, dbo.fn_Users_RRN(Data) AS RRN from dbo.Users Indexes: CREATE UNIQUE CLUSTERED INDEX cx_vwi_Users ON vwi_Users(UserId) CREATE NONCLUSTERED INDEX cx_vwi_Users__RRN ON vwi_Users(RRN) And then the changed query: SELECT UserId FROM Users WHERE dbo.fn_Users_RRN(Data) = '59021626919-61861855-S_FA1E11' Why is the solution with a function and a view going slower?

    Read the article

  • Trouble in ActiveX multi-thread invoke javascript callback routine

    - by code0tt
    everyone. I'm get some trouble in ActiveX programming with ATL. I try to make a activex which can async-download files from http server to local folder and after download it will invoke javascript callback function. My solution: run a thread M to monitor download thread D, when D is finish the job, M is going to terminal themself and invoke IDispatch inferface to call javascript function. **************** THERE IS MY CODE: **************** /* javascript code */ funciton download() { var xfm = new ActiveXObject("XFileMngr.FileManager.1"); xfm.download( 'http://somedomain/somefile','localdev:\\folder\localfile',function(msg){alert(msg);}); } /* C++ code */ // main routine STDMETHODIMP CFileManager::download(BSTR url, BSTR local, VARIANT scriptCallback) { CString csURL(url); CString csLocal(local); CAsyncDownload download; download.Download(this, csURL, csLocal, scriptCallback); return S_OK; } // parts of CAsyncDownload.h typedef struct tagThreadData { CAsyncDownload* pThis; } THREAD_DATA, *LPTHREAD_DATA; class CAsyncDownload : public IBindStatusCallback { private: LPUNKNOWN pcaller; CString csRemoteFile; CString csLocalFile; CComPtr<IDispatch> spCallback; public: void onDone(HRESULT hr); HRESULT Download(LPUNKNOWN caller, CString& csRemote, CString& csLocal, VARIANT callback); static DWORD __stdcall ThreadProc(void* param); }; // parts of CAsyncDownload.cpp void CAsyncDownload::onDone(HRESULT hr) { if(spCallback) { TRACE(TEXT("invoke callback function\n")); CComVariant vParams[1]; vParams[0] = "callback is working!"; DISPPARAMS params = { vParams, NULL, 1, 0 }; HRESULT hr = spCallback->Invoke(0, IID_NULL, LOCALE_USER_DEFAULT, DISPATCH_METHOD, &params, NULL, NULL, NULL); if(FAILED(hr)) { CString csBuffer; csBuffer.Format(TEXT("invoke failed, result value: %d \n"),hr); TRACE(csBuffer); }else { TRACE(TEXT("invoke was successful\n")); } } } HRESULT CAsyncDownload::Download(LPUNKNOWN caller, CString& csRemote, CString& csLocal, VARIANT callback) { CoInitializeEx(NULL, COINIT_MULTITHREADED); csRemoteFile = csRemote; csLocalFile = csLocal; pcaller = caller; switch(callback.vt){ case VT_DISPATCH: case VT_VARIANT:{ spCallback = callback.pdispVal; } break; default:{ spCallback = NULL; } } LPTHREAD_DATA pData = new THREAD_DATA; pData->pThis = this; // create monitor thread M HANDLE hThread = CreateThread(NULL, 0, ThreadProc, (void*)(pData), 0, NULL); if(!hThread) { delete pData; return HRESULT_FROM_WIN32(GetLastError()); } WaitForSingleObject(hThread, INFINITE); CloseHandle(hThread); CoUninitialize(); return S_OK; } DWORD __stdcall CAsyncDownload::ThreadProc(void* param) { LPTHREAD_DATA pData = (LPTHREAD_DATA)param; // here, we will create http download thread D // when download job is finish, call onDone method; pData->pThis->onDone(S_OK); delete pData; return 0; } **************** CODE FINISH **************** OK, above is parts of my source code, if I call onDone method in sub-thread, I will get OLE ERROR(-2147418113 (8000FFFF) Catastrophic failure.). Did I miss something? please help me to figure it out.

    Read the article

  • What's the most efficient way to setup a multi-lingual website

    - by Jasper De Bruijn
    Hi, I'm developing a website that will be available in different languages. It is a LAMP (Linux, Apache, MySQL, PHP) setup, and it makes use of Smarty, mostly for the template engine. The way we currently translate is by a self-written smarty plugin, which will recognize certain tags in the HTML files, and will find the corresponding tag in an earlier defined language file. The HTML could look as follows: <p>Hi, welcome to $#gamedesc;!</p> And the language file could look like this: gamedesc:Poing 2009$; welcome:this is another tag$; Which would then output <p>Hi, welcome to Poing 2009!</p> This system is very basic, but it is pretty hard to control, if I f.e. would like to keep track of what has been translated so far, or give certain users the rights to translate only certain tags. I've been looking at some alternative ways to approach this, by either replacing the text-file with XML files which could store some extra meta-data, or by perhaps storing all the texts in the database, and retrieving it there. My question is, what would be the best way to make this system both maintainable and perform well with high user-traffic? Are there perhaps any (lightweight) plugins I could take a look at?

    Read the article

  • Controller class not available in Add View ASP.NET MVC

    - by Tassadaque
    Hi I have created the following controller i want to add view that should have the data access class UserMagnament.Controller.menuitems but when i add view by right clicking on view folder,it is not showing "UserMagnament.Controller.menuitems" in add data class using System; using System.Collections.Generic; using System.Linq; using System.Web; using System.Web.Mvc; using UserManagement.Models; namespace UserManagement.Controllers { public class menuitems { EvoLetDataContext db1 = new EvoLetDataContext(); public menuitems() { } public IQueryable<UMUserType> menuitems() { return db1.UMUserTypes; //this.Packages = _SysPackage; //this.Modules = _SysModule; } public List<SysPackage> Packages { get; private set; } public List<SysModule> Modules { get; private set; } } public class Default1Controller : Controller { // // GET: /Default1/ public ActionResult Index() { return View(new menuitems()); } } }

    Read the article

  • Looking for a fast, compact, streamable, multi-language, strongly typed serialization format

    - by sanity
    I'm currently using JSON (compressed via gzip) in my Java project, in which I need to store a large number of objects (hundreds of millions) on disk. I have one JSON object per line, and disallow linebreaks within the JSON object. This way I can stream the data off disk line-by-line without having to read the entire file at once. It turns out that parsing the JSON code (using http://www.json.org/java/) is a bigger overhead than either pulling the raw data off disk, or decompressing it (which I do on the fly). Ideally what I'd like is a strongly-typed serialization format, where I can specify "this object field is a list of strings" (for example), and because the system knows what to expect, it can deserialize it quickly. I can also specify the format just by giving someone else its "type". It would also need to be cross-platform. I use Java, but work with people using PHP, Python, and other languages. So, to recap, it should be: Strongly typed Streamable (ie. read a file bit by bit without having to load it all into RAM at once) Cross platform (including Java and PHP) Fast Free (as in speech) Any pointers?

    Read the article

< Previous Page | 77 78 79 80 81 82 83 84 85 86 87 88  | Next Page >