Search Results

Search found 24132 results on 966 pages for 'non clustered index'.

Page 214/966 | < Previous Page | 210 211 212 213 214 215 216 217 218 219 220 221  | Next Page >

  • Very different I/O performance in C++ on Windows

    - by Mr.Gate
    Hi all, I'm a new user and my english is not so good so I hope to be clear. We're facing a performance problem using large files (1GB or more) expecially (as it seems) when you try to grow them in size. Anyway... to verify our sensations we tryed the following (on Win 7 64Bit, 4core, 8GB Ram, 32 bit code compiled with VC2008) a) Open an unexisting file. Write it from the beginning up to 1Gb in 1Mb slots. Now you have a 1Gb file. Now randomize 10000 positions within that file, seek to that position and write 50 bytes in each position, no matter what you write. Close the file and look at the results. Time to create the file is quite fast (about 0.3"), time to write 10000 times is fast all the same (about 0.03"). Very good, this is the beginnig. Now try something else... b) Open an unexisting file, seek to 1Gb-1byte and write just 1 byte. Now you have another 1Gb file. Follow the next steps exactly same way of case 'a', close the file and look at the results. Time to create the file is the faster you can imagine (about 0.00009") but write time is something you can't believe.... about 90"!!!!! b.1) Open an unexisting file, don't write any byte. Act as before, ramdomizing, seeking and writing, close the file and look at the result. Time to write is long all the same: about 90"!!!!! Ok... this is quite amazing. But there's more! c) Open again the file you crated in case 'a', don't truncate it... randomize again 10000 positions and act as before. You're fast as before, about 0,03" to write 10000 times. This sounds Ok... try another step. d) Now open the file you created in case 'b', don't truncate it... randomize again 10000 positions and act as before. You're slow again and again, but the time is reduced to... 45"!! Maybe, trying again, the time will reduce. I actually wonder why... Any Idea? The following is part of the code I used to test what I told in previuos cases (you'll have to change someting in order to have a clean compilation, I just cut & paste from some source code, sorry). The sample can read and write, in random, ordered or reverse ordered mode, but write only in random order is the clearest test. We tryed using std::fstream but also using directly CreateFile(), WriteFile() and so on the results are the same (even if std::fstream is actually a little slower). Parameters for case 'a' = -f_tempdir_\casea.dat -n10000 -t -p -w Parameters for case 'b' = -f_tempdir_\caseb.dat -n10000 -t -v -w Parameters for case 'b.1' = -f_tempdir_\caseb.dat -n10000 -t -w Parameters for case 'c' = -f_tempdir_\casea.dat -n10000 -w Parameters for case 'd' = -f_tempdir_\caseb.dat -n10000 -w Run the test (and even others) and see... // iotest.cpp : Defines the entry point for the console application. // #include <windows.h> #include <iostream> #include <set> #include <vector> #include "stdafx.h" double RealTime_Microsecs() { LARGE_INTEGER fr = {0, 0}; LARGE_INTEGER ti = {0, 0}; double time = 0.0; QueryPerformanceCounter(&ti); QueryPerformanceFrequency(&fr); time = (double) ti.QuadPart / (double) fr.QuadPart; return time; } int main(int argc, char* argv[]) { std::string sFileName ; size_t stSize, stTimes, stBytes ; int retval = 0 ; char *p = NULL ; char *pPattern = NULL ; char *pReadBuf = NULL ; try { // Default stSize = 1<<30 ; // 1Gb stTimes = 1000 ; stBytes = 50 ; bool bTruncate = false ; bool bPre = false ; bool bPreFast = false ; bool bOrdered = false ; bool bReverse = false ; bool bWriteOnly = false ; // Comsumo i parametri for(int index=1; index < argc; ++index) { if ( '-' != argv[index][0] ) throw ; switch(argv[index][1]) { case 'f': sFileName = argv[index]+2 ; break ; case 's': stSize = xw::str::strtol(argv[index]+2) ; break ; case 'n': stTimes = xw::str::strtol(argv[index]+2) ; break ; case 'b':stBytes = xw::str::strtol(argv[index]+2) ; break ; case 't': bTruncate = true ; break ; case 'p' : bPre = true, bPreFast = false ; break ; case 'v' : bPreFast = true, bPre = false ; break ; case 'o' : bOrdered = true, bReverse = false ; break ; case 'r' : bReverse = true, bOrdered = false ; break ; case 'w' : bWriteOnly = true ; break ; default: throw ; break ; } } if ( sFileName.empty() ) { std::cout << "Usage: -f<File Name> -s<File Size> -n<Number of Reads and Writes> -b<Bytes per Read and Write> -t -p -v -o -r -w" << std::endl ; std::cout << "-t truncates the file, -p pre load the file, -v pre load 'veloce', -o writes in order mode, -r write in reverse order mode, -w Write Only" << std::endl ; std::cout << "Default: 1Gb, 1000 times, 50 bytes" << std::endl ; throw ; } if ( !stSize || !stTimes || !stBytes ) { std::cout << "Invalid Parameters" << std::endl ; return -1 ; } size_t stBestSize = 0x00100000 ; std::fstream fFile ; fFile.open(sFileName.c_str(), std::ios_base::binary|std::ios_base::out|std::ios_base::in|(bTruncate?std::ios_base::trunc:0)) ; p = new char[stBestSize] ; pPattern = new char[stBytes] ; pReadBuf = new char[stBytes] ; memset(p, 0, stBestSize) ; memset(pPattern, (int)(stBytes&0x000000ff), stBytes) ; double dTime = RealTime_Microsecs() ; size_t stCopySize, stSizeToCopy = stSize ; if ( bPre ) { do { stCopySize = std::min(stSizeToCopy, stBestSize) ; fFile.write(p, stCopySize) ; stSizeToCopy -= stCopySize ; } while (stSizeToCopy) ; std::cout << "Creating time is: " << xw::str::itoa(RealTime_Microsecs()-dTime, 5, 'f') << std::endl ; } else if ( bPreFast ) { fFile.seekp(stSize-1) ; fFile.write(p, 1) ; std::cout << "Creating Fast time is: " << xw::str::itoa(RealTime_Microsecs()-dTime, 5, 'f') << std::endl ; } size_t stPos ; ::srand((unsigned int)dTime) ; double dReadTime, dWriteTime ; stCopySize = stTimes ; std::vector<size_t> inVect ; std::vector<size_t> outVect ; std::set<size_t> outSet ; std::set<size_t> inSet ; // Prepare vector and set do { stPos = (size_t)(::rand()<<16) % stSize ; outVect.push_back(stPos) ; outSet.insert(stPos) ; stPos = (size_t)(::rand()<<16) % stSize ; inVect.push_back(stPos) ; inSet.insert(stPos) ; } while (--stCopySize) ; // Write & read using vectors if ( !bReverse && !bOrdered ) { std::vector<size_t>::iterator outI, inI ; outI = outVect.begin() ; inI = inVect.begin() ; stCopySize = stTimes ; dReadTime = 0.0 ; dWriteTime = 0.0 ; do { dTime = RealTime_Microsecs() ; fFile.seekp(*outI) ; fFile.write(pPattern, stBytes) ; dWriteTime += RealTime_Microsecs() - dTime ; ++outI ; if ( !bWriteOnly ) { dTime = RealTime_Microsecs() ; fFile.seekg(*inI) ; fFile.read(pReadBuf, stBytes) ; dReadTime += RealTime_Microsecs() - dTime ; ++inI ; } } while (--stCopySize) ; std::cout << "Write time is " << xw::str::itoa(dWriteTime, 5, 'f') << " (Ave: " << xw::str::itoa(dWriteTime/stTimes, 10, 'f') << ")" << std::endl ; if ( !bWriteOnly ) { std::cout << "Read time is " << xw::str::itoa(dReadTime, 5, 'f') << " (Ave: " << xw::str::itoa(dReadTime/stTimes, 10, 'f') << ")" << std::endl ; } } // End // Write in order if ( bOrdered ) { std::set<size_t>::iterator i = outSet.begin() ; dWriteTime = 0.0 ; stCopySize = 0 ; for(; i != outSet.end(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekp(stPos) ; fFile.write(pPattern, stBytes) ; dWriteTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Ordered Write time is " << xw::str::itoa(dWriteTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dWriteTime/stCopySize, 10, 'f') << ")" << std::endl ; if ( !bWriteOnly ) { i = inSet.begin() ; dReadTime = 0.0 ; stCopySize = 0 ; for(; i != inSet.end(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekg(stPos) ; fFile.read(pReadBuf, stBytes) ; dReadTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Ordered Read time is " << xw::str::itoa(dReadTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dReadTime/stCopySize, 10, 'f') << ")" << std::endl ; } }// End // Write in reverse order if ( bReverse ) { std::set<size_t>::reverse_iterator i = outSet.rbegin() ; dWriteTime = 0.0 ; stCopySize = 0 ; for(; i != outSet.rend(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekp(stPos) ; fFile.write(pPattern, stBytes) ; dWriteTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Reverse ordered Write time is " << xw::str::itoa(dWriteTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dWriteTime/stCopySize, 10, 'f') << ")" << std::endl ; if ( !bWriteOnly ) { i = inSet.rbegin() ; dReadTime = 0.0 ; stCopySize = 0 ; for(; i != inSet.rend(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekg(stPos) ; fFile.read(pReadBuf, stBytes) ; dReadTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Reverse ordered Read time is " << xw::str::itoa(dReadTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dReadTime/stCopySize, 10, 'f') << ")" << std::endl ; } }// End dTime = RealTime_Microsecs() ; fFile.close() ; std::cout << "Flush/Close Time is " << xw::str::itoa(RealTime_Microsecs()-dTime, 5, 'f') << std::endl ; std::cout << "Program Terminated" << std::endl ; } catch(...) { std::cout << "Something wrong or wrong parameters" << std::endl ; retval = -1 ; } if ( p ) delete []p ; if ( pPattern ) delete []pPattern ; if ( pReadBuf ) delete []pReadBuf ; return retval ; }

    Read the article

  • Very different IO performance in C/C++

    - by Roberto Tirabassi
    Hi all, I'm a new user and my english is not so good so I hope to be clear. We're facing a performance problem using large files (1GB or more) expecially (as it seems) when you try to grow them in size. Anyway... to verify our sensations we tryed the following (on Win 7 64Bit, 4core, 8GB Ram, 32 bit code compiled with VC2008) a) Open an unexisting file. Write it from the beginning up to 1Gb in 1Mb slots. Now you have a 1Gb file. Now randomize 10000 positions within that file, seek to that position and write 50 bytes in each position, no matter what you write. Close the file and look at the results. Time to create the file is quite fast (about 0.3"), time to write 10000 times is fast all the same (about 0.03"). Very good, this is the beginnig. Now try something else... b) Open an unexisting file, seek to 1Gb-1byte and write just 1 byte. Now you have another 1Gb file. Follow the next steps exactly same way of case 'a', close the file and look at the results. Time to create the file is the faster you can imagine (about 0.00009") but write time is something you can't believe.... about 90"!!!!! b.1) Open an unexisting file, don't write any byte. Act as before, ramdomizing, seeking and writing, close the file and look at the result. Time to write is long all the same: about 90"!!!!! Ok... this is quite amazing. But there's more! c) Open again the file you crated in case 'a', don't truncate it... randomize again 10000 positions and act as before. You're fast as before, about 0,03" to write 10000 times. This sounds Ok... try another step. d) Now open the file you created in case 'b', don't truncate it... randomize again 10000 positions and act as before. You're slow again and again, but the time is reduced to... 45"!! Maybe, trying again, the time will reduce. I actually wonder why... Any Idea? The following is part of the code I used to test what I told in previuos cases (you'll have to change someting in order to have a clean compilation, I just cut & paste from some source code, sorry). The sample can read and write, in random, ordered or reverse ordered mode, but write only in random order is the clearest test. We tryed using std::fstream but also using directly CreateFile(), WriteFile() and so on the results are the same (even if std::fstream is actually a little slower). Parameters for case 'a' = -f_tempdir_\casea.dat -n10000 -t -p -w Parameters for case 'b' = -f_tempdir_\caseb.dat -n10000 -t -v -w Parameters for case 'b.1' = -f_tempdir_\caseb.dat -n10000 -t -w Parameters for case 'c' = -f_tempdir_\casea.dat -n10000 -w Parameters for case 'd' = -f_tempdir_\caseb.dat -n10000 -w Run the test (and even others) and see... // iotest.cpp : Defines the entry point for the console application. // #include <windows.h> #include <iostream> #include <set> #include <vector> #include "stdafx.h" double RealTime_Microsecs() { LARGE_INTEGER fr = {0, 0}; LARGE_INTEGER ti = {0, 0}; double time = 0.0; QueryPerformanceCounter(&ti); QueryPerformanceFrequency(&fr); time = (double) ti.QuadPart / (double) fr.QuadPart; return time; } int main(int argc, char* argv[]) { std::string sFileName ; size_t stSize, stTimes, stBytes ; int retval = 0 ; char *p = NULL ; char *pPattern = NULL ; char *pReadBuf = NULL ; try { // Default stSize = 1<<30 ; // 1Gb stTimes = 1000 ; stBytes = 50 ; bool bTruncate = false ; bool bPre = false ; bool bPreFast = false ; bool bOrdered = false ; bool bReverse = false ; bool bWriteOnly = false ; // Comsumo i parametri for(int index=1; index < argc; ++index) { if ( '-' != argv[index][0] ) throw ; switch(argv[index][1]) { case 'f': sFileName = argv[index]+2 ; break ; case 's': stSize = xw::str::strtol(argv[index]+2) ; break ; case 'n': stTimes = xw::str::strtol(argv[index]+2) ; break ; case 'b':stBytes = xw::str::strtol(argv[index]+2) ; break ; case 't': bTruncate = true ; break ; case 'p' : bPre = true, bPreFast = false ; break ; case 'v' : bPreFast = true, bPre = false ; break ; case 'o' : bOrdered = true, bReverse = false ; break ; case 'r' : bReverse = true, bOrdered = false ; break ; case 'w' : bWriteOnly = true ; break ; default: throw ; break ; } } if ( sFileName.empty() ) { std::cout << "Usage: -f<File Name> -s<File Size> -n<Number of Reads and Writes> -b<Bytes per Read and Write> -t -p -v -o -r -w" << std::endl ; std::cout << "-t truncates the file, -p pre load the file, -v pre load 'veloce', -o writes in order mode, -r write in reverse order mode, -w Write Only" << std::endl ; std::cout << "Default: 1Gb, 1000 times, 50 bytes" << std::endl ; throw ; } if ( !stSize || !stTimes || !stBytes ) { std::cout << "Invalid Parameters" << std::endl ; return -1 ; } size_t stBestSize = 0x00100000 ; std::fstream fFile ; fFile.open(sFileName.c_str(), std::ios_base::binary|std::ios_base::out|std::ios_base::in|(bTruncate?std::ios_base::trunc:0)) ; p = new char[stBestSize] ; pPattern = new char[stBytes] ; pReadBuf = new char[stBytes] ; memset(p, 0, stBestSize) ; memset(pPattern, (int)(stBytes&0x000000ff), stBytes) ; double dTime = RealTime_Microsecs() ; size_t stCopySize, stSizeToCopy = stSize ; if ( bPre ) { do { stCopySize = std::min(stSizeToCopy, stBestSize) ; fFile.write(p, stCopySize) ; stSizeToCopy -= stCopySize ; } while (stSizeToCopy) ; std::cout << "Creating time is: " << xw::str::itoa(RealTime_Microsecs()-dTime, 5, 'f') << std::endl ; } else if ( bPreFast ) { fFile.seekp(stSize-1) ; fFile.write(p, 1) ; std::cout << "Creating Fast time is: " << xw::str::itoa(RealTime_Microsecs()-dTime, 5, 'f') << std::endl ; } size_t stPos ; ::srand((unsigned int)dTime) ; double dReadTime, dWriteTime ; stCopySize = stTimes ; std::vector<size_t> inVect ; std::vector<size_t> outVect ; std::set<size_t> outSet ; std::set<size_t> inSet ; // Prepare vector and set do { stPos = (size_t)(::rand()<<16) % stSize ; outVect.push_back(stPos) ; outSet.insert(stPos) ; stPos = (size_t)(::rand()<<16) % stSize ; inVect.push_back(stPos) ; inSet.insert(stPos) ; } while (--stCopySize) ; // Write & read using vectors if ( !bReverse && !bOrdered ) { std::vector<size_t>::iterator outI, inI ; outI = outVect.begin() ; inI = inVect.begin() ; stCopySize = stTimes ; dReadTime = 0.0 ; dWriteTime = 0.0 ; do { dTime = RealTime_Microsecs() ; fFile.seekp(*outI) ; fFile.write(pPattern, stBytes) ; dWriteTime += RealTime_Microsecs() - dTime ; ++outI ; if ( !bWriteOnly ) { dTime = RealTime_Microsecs() ; fFile.seekg(*inI) ; fFile.read(pReadBuf, stBytes) ; dReadTime += RealTime_Microsecs() - dTime ; ++inI ; } } while (--stCopySize) ; std::cout << "Write time is " << xw::str::itoa(dWriteTime, 5, 'f') << " (Ave: " << xw::str::itoa(dWriteTime/stTimes, 10, 'f') << ")" << std::endl ; if ( !bWriteOnly ) { std::cout << "Read time is " << xw::str::itoa(dReadTime, 5, 'f') << " (Ave: " << xw::str::itoa(dReadTime/stTimes, 10, 'f') << ")" << std::endl ; } } // End // Write in order if ( bOrdered ) { std::set<size_t>::iterator i = outSet.begin() ; dWriteTime = 0.0 ; stCopySize = 0 ; for(; i != outSet.end(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekp(stPos) ; fFile.write(pPattern, stBytes) ; dWriteTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Ordered Write time is " << xw::str::itoa(dWriteTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dWriteTime/stCopySize, 10, 'f') << ")" << std::endl ; if ( !bWriteOnly ) { i = inSet.begin() ; dReadTime = 0.0 ; stCopySize = 0 ; for(; i != inSet.end(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekg(stPos) ; fFile.read(pReadBuf, stBytes) ; dReadTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Ordered Read time is " << xw::str::itoa(dReadTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dReadTime/stCopySize, 10, 'f') << ")" << std::endl ; } }// End // Write in reverse order if ( bReverse ) { std::set<size_t>::reverse_iterator i = outSet.rbegin() ; dWriteTime = 0.0 ; stCopySize = 0 ; for(; i != outSet.rend(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekp(stPos) ; fFile.write(pPattern, stBytes) ; dWriteTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Reverse ordered Write time is " << xw::str::itoa(dWriteTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dWriteTime/stCopySize, 10, 'f') << ")" << std::endl ; if ( !bWriteOnly ) { i = inSet.rbegin() ; dReadTime = 0.0 ; stCopySize = 0 ; for(; i != inSet.rend(); ++i) { stPos = *i ; dTime = RealTime_Microsecs() ; fFile.seekg(stPos) ; fFile.read(pReadBuf, stBytes) ; dReadTime += RealTime_Microsecs() - dTime ; ++stCopySize ; } std::cout << "Reverse ordered Read time is " << xw::str::itoa(dReadTime, 5, 'f') << " in " << xw::str::itoa(stCopySize) << " (Ave: " << xw::str::itoa(dReadTime/stCopySize, 10, 'f') << ")" << std::endl ; } }// End dTime = RealTime_Microsecs() ; fFile.close() ; std::cout << "Flush/Close Time is " << xw::str::itoa(RealTime_Microsecs()-dTime, 5, 'f') << std::endl ; std::cout << "Program Terminated" << std::endl ; } catch(...) { std::cout << "Something wrong or wrong parameters" << std::endl ; retval = -1 ; } if ( p ) delete []p ; if ( pPattern ) delete []pPattern ; if ( pReadBuf ) delete []pReadBuf ; return retval ; }

    Read the article

  • Fake ISAPI Handler to serve static files with extention that are rewritted by url rewriter

    - by developerit
    Introduction I often map html extention to the asp.net dll in order to use url rewritter with .html extentions. Recently, in the new version of www.nouvelair.ca, we renamed all urls to end with .html. This works great, but failed when we used FCK Editor. Static html files would not get serve because we mapped the html extension to the .NET Framework. We can we do to to use .html extension with our rewritter but still want to use IIS behavior with static html files. Analysis I thought that this could be resolve with a simple HTTP handler. We would map urls of static files in our rewriter to this handler that would read the static file and serve it, just as IIS would do. Implementation This is how I coded the class. Note that this may not be bullet proof. I only tested it once and I am sure that the logic behind IIS is more complicated that this. If you find errors or think of possible improvements, let me know. Imports System.Web Imports System.Web.Services ' Author: Nicolas Brassard ' For: Solutions Nitriques inc. http://www.nitriques.com ' Date Created: April 18, 2009 ' Last Modified: April 18, 2009 ' License: CPOL (http://www.codeproject.com/info/cpol10.aspx) ' Files: ISAPIDotNetHandler.ashx ' ISAPIDotNetHandler.ashx.vb ' Class: ISAPIDotNetHandler ' Description: Fake ISAPI handler to serve static files. ' Usefull when you want to serve static file that has a rewrited extention. ' Example: It often map html extention to the asp.net dll in order to use url rewritter with .html. ' If you want to still serve static html file, add a rewritter rule to redirect html files to this handler Public Class ISAPIDotNetHandler Implements System.Web.IHttpHandler Sub ProcessRequest(ByVal context As HttpContext) Implements IHttpHandler.ProcessRequest ' Since we are doing the job IIS normally does with html files, ' we set the content type to match html. ' You may want to customize this with your own logic, if you want to serve ' txt or xml or any other text file context.Response.ContentType = "text/html" ' We begin a try here. Any error that occurs will result in a 404 Page Not Found error. ' We replicate the behavior of IIS when it doesn't find the correspoding file. Try ' Declare a local variable containing the value of the query string Dim uri As String = context.Request("fileUri") ' If the value in the query string is null, ' throw an error to generate a 404 If String.IsNullOrEmpty(uri) Then Throw New ApplicationException("No fileUri") End If ' If the value in the query string doesn't end with .html, then block the acces ' This is a HUGE security hole since it could permit full read access to .aspx, .config, etc. If Not uri.ToLower.EndsWith(".html") Then ' throw an error to generate a 404 Throw New ApplicationException("Extention not allowed") End If ' Map the file on the server. ' If the file doesn't exists on the server, it will throw an exception and generate a 404. Dim fullPath As String = context.Server.MapPath(uri) ' Read the actual file Dim stream As IO.StreamReader = FileIO.FileSystem.OpenTextFileReader(fullPath) ' Write the file into the response context.Response.Output.Write(stream.ReadToEnd) ' Close and Dipose the stream stream.Close() stream.Dispose() stream = Nothing Catch ex As Exception ' Set the Status Code of the response context.Response.StatusCode = 404 'Page not found ' For testing and bebugging only ! This may cause a security leak ' context.Response.Output.Write(ex.Message) Finally ' In all cases, flush and end the response context.Response.Flush() context.Response.End() End Try End Sub ' Automaticly generated by Visual Studio ReadOnly Property IsReusable() As Boolean Implements IHttpHandler.IsReusable Get Return False End Get End Property End Class Conclusion As you see, with our static files map to this handler using query string (ex.: /ISAPIDotNetHandler.ashx?fileUri=index.html) you will have the same behavior as if you ask for the uri /index.html. Finally, test this only in IIS with the html extension map to aspnet_isapi.dll. Url rewritting will work in Casini (Internal Web Server shipped with Visual Studio) but it’s not the same as with IIS since EVERY request is handle by .NET. Versions First release

    Read the article

  • Smart Help with UPK

    - by [email protected]
    A short lesson on how awesome Smart Help is. In Oracle UPK speak, there are targeted and non-targeted applications. Targeted applications are Oracle EBS, PeopleSoft, Siebel, JD Edwards, SAP and a few others. Non-targeted applications are either custom built or other third party off the shelf applications. For most targeted applications you'll see better object recognition (during recording) and also Help Integration for that application. Help integration means that someone technical modifies the help link in your application to call up the UPK content that has been created. If you have seen this presented before, this is usually where the term context sensitive help is mentioned and the Do It mode shows off. The fact that UPK builds context sensitive help for its targeted applications automatically is awesome enough, but there is a whole new world out there and it's called "custom and\or third party apps." For the purposes of Smart Help and this discussion, I'm talking about the browser based applications. How does UPK support these apps? It used to be that you had to have your vendor try to modify the Help link to point to UPK or if your company had control over the applications configuration menus, then you get someone on your team to modify this for you. But as you start to use UPK for more than one, two or three applications, the administration of this starts to become daunting. Multiple administrators, multiple player packages, multiple call points, multiple break points, help doesn't always work the same way for every application (picture the black white infomercial with an IT person trying to configure a bunch of wires or something funny like that). Introducing Smart Help! (in color of course, new IT person, probably wearing a blue shirt and smiling). Smart help eliminates the need to configure multiple browser help integration points, and adds a icon to the users browser itself. You're using your browser to read this now correct? Look up at the icons on your browser, you have the home link icon, print icon, maybe an RSS feed icon. Smart Help is icon that gets added to the users browser just like the others. When you click it, it first recognizes which application you're in and then finds the UPK created material for you and returns the best possible match, for (hold on to your seat now) both targeted and non-targeted applications (browser based applications). But wait, there's more. It does this automatically! You don't have to do anything! All you have to do is record content, UPK and Smart Help do the rest! This technology is not new. There are customers out there today that use this for as many as six applications! The real hero here is SMART MATCH. Smart match is the technology that's used to determine which application you're in and where you are when you click on Smart Help. We'll save that for a one-on-one conversation. Like most other awesome features of UPK, it ships with the product. All you have to do is turn it on. To learn more about Smart Help, Smart Match, Targeted and Non-Targeted applications, contact your UPK Sales Consultant or me directly at [email protected]

    Read the article

  • Routes for IIS Classic and Integrated Mode

    - by imran_ku07
         Introduction:             ASP.NET MVC Routing feature makes it very easy to provide clean URLs. You just need to configure routes in global.asax file to create an application with clean URLs. In most cases you define routes works in IIS 6, IIS 7 (or IIS 7.5) Classic and Integrated mode. But in some cases your routes may only works in IIS 7 Integrated mode, like in the case of using extension less URLs in IIS 6 without a wildcard extension map. So in this article I will show you how to create different routes which works in IIS 6 and IIS 7 Classic and Integrated mode.       Description:             Let's say that you need to create an application which must work both in Classic and Integrated mode. Also you have no control to setup a wildcard extension map in IIS. So you need to create two routes. One with extension less URL for Integrated mode and one with a URL with an extension for Classic Mode.   routes.MapRoute( "DefaultClassic", // Route name "{controller}.aspx/{action}/{id}", // URL with parameters new { controller = "Home", action = "Index", id = UrlParameter.Optional } // Parameter defaults ); routes.MapRoute( "DefaultIntegrated", // Route name "{controller}/{action}/{id}", // URL with parameters new { controller = "Home", action = "Index", id = UrlParameter.Optional } // Parameter defaults );               Now you have set up two routes, one for Integrated mode and one for Classic mode. Now you only need to ensure that Integrated mode route should only match if the application is running in Integrated mode and Classic mode route should only match if the application is running in Classic mode. For making this work you need to create two custom constraint for Integrated and Classic mode. So replace the above routes with these routes,     routes.MapRoute( "DefaultClassic", // Route name "{controller}.aspx/{action}/{id}", // URL with parameters new { controller = "Home", action = "Index", id = UrlParameter.Optional }, // Parameter defaults new { mode = new ClassicModeConstraint() }// Constraints ); routes.MapRoute( "DefaultIntegrated", // Route name "{controller}/{action}/{id}", // URL with parameters new { controller = "Home", action = "Index", id = UrlParameter.Optional }, // Parameter defaults new { mode = new IntegratedModeConstraint() }// Constraints );            The first route which is for Classic mode adds a ClassicModeConstraint and second route which is for Integrated mode adds a IntegratedModeConstraint. Next you need to add the implementation of these constraint classes.     public class ClassicModeConstraint : IRouteConstraint { public bool Match(HttpContextBase httpContext, Route route, string parameterName, RouteValueDictionary values, RouteDirection routeDirection) { return !HttpRuntime.UsingIntegratedPipeline; } } public class IntegratedModeConstraint : IRouteConstraint { public bool Match(HttpContextBase httpContext, Route route, string parameterName, RouteValueDictionary values, RouteDirection routeDirection) { return HttpRuntime.UsingIntegratedPipeline; } }             HttpRuntime.UsingIntegratedPipeline returns true if the application is running on Integrated mode; otherwise, it returns false. So routes for Integrated mode only matched when the application is running on Integrated mode and routes for Classic mode only matched when the application is not running on Integrated mode.       Summary:             During developing applications, sometimes developers are not sure that whether this application will be host on IIS 6 or IIS 7 (or IIS 7.5) Integrated mode or Classic mode. So it's a good idea to create separate routes for both Classic and Integrated mode so that your application will use extension less URLs where possible and use URLs with an extension where it is not possible to use extension less URLs. In this article I showed you how to create separate routes for IIS Integrated and Classic mode. Hope you will enjoy this article too.   SyntaxHighlighter.all()

    Read the article

  • SQL SERVER – Guest Post – Glenn Berry – Wait Type – Day 26 of 28

    - by pinaldave
    Glenn Berry works as a Database Architect at NewsGator Technologies in Denver, CO. He is a SQL Server MVP, and has a whole collection of Microsoft certifications, including MCITP, MCDBA, MCSE, MCSD, MCAD, and MCTS. He is also an Adjunct Faculty member at University College – University of Denver, where he has been teaching since 2000. He is one wonderful blogger and often blogs at here. I am big fan of the Dynamic Management Views (DMV) scripts of Glenn. His script are extremely popular and the reality is that he has inspired me to start this series with his famous DMV which I have mentioned in very first  wait stats blog post (I had forgot to request his permission to re-use the script but when asked later on his whole hearty approved it). Here is is his excellent blog post on this subject of wait stats: Analyzing cumulative wait stats in SQL Server 2005 and above has become a popular and effective technique for diagnosing performance issues and further focusing your troubleshooting and diagnostic  efforts.  Rather than just guessing about what resource(s) that SQL Server is waiting on, you can actually find out by running a relatively simple DMV query. Once you know what resources that SQL Server is spending the most time waiting on, you can run more specific queries that focus on that resource to get a better idea what is causing the problem. I do want to throw out a few caveats about using wait stats as a diagnostic tool. First, they are most useful when your SQL Server instance is experiencing performance problems. If your instance is running well, with no indication of any resource pressure from other sources, then you should not worry that much about what the top wait types are. SQL Server will always be waiting on some resource, but many wait types are quite benign, and can be safely ignored. In spite of this, I quite often see experienced DBAs obsessing over the top wait type, even when their SQL Server instance is running extremely well. Second, I often see DBAs jump to the wrong conclusion based on seeing a particular well-known wait type. A good example is CXPACKET waits. People typically jump to the conclusion that high CXPACKET waits means that they should immediately change their instance-level MADOP setting to 1. This is not always the best solution. You need to consider your workload type, and look carefully for any important “missing” indexes that might be causing the query optimizer to use a parallel plan to compensate for the missing index. In this case, correcting the index problem is usually a better solution than changing MAXDOP, since you are curing the disease rather than just treating the symptom. Finally, you should get in the habit of clearing out your cumulative wait stats with the  DBCC SQLPERF(‘sys.dm_os_wait_stats’, CLEAR); command. This is especially important if you have made an configuration or index changes, or if your workload has changed recently. Otherwise, your cumulative wait stats will be polluted with the old stats from weeks or months ago (since the last time SQL Server was started or the stats were cleared).  If you make a change to your SQL Server instance, or add an index, you should clear out your wait stats, and then wait a while to see what your new top wait stats are. At any rate, enjoy Pinal Dave’s series on Wait Stats. This blog post has been written by Glenn Berry (Twitter | Blog) Read all the post in the Wait Types and Queue series. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, Readers Contribution, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, SQL Wait Stats, SQL Wait Types, T SQL, Technology

    Read the article

  • Web.config WordPress rewrite rules next to Magento

    - by Flo
    I've installed Magento on IIS in folder: E:\mydomain\wwwroot (I already have it all running correctly). I have no deeper folder magento, I placed all files directly in the wwwroot folder, so: wwwroot\app wwwroot\downloader wwwroot\errors wwwroot\includes etc... UPDATE: since I'm on IIS my .htaccess is ignored completely and my web.config rules are used instead. Here's my web.config in folder e:\mydomain\wwwroot: <?xml version="1.0" encoding="UTF-8"?> <configuration> <system.webServer> <rewrite> <rules> <rule name="Magento SEO: remove index.php from URL"> <match url="^(?!index.php)([^?#]*)(\\?([^#]*))?(#(.*))?" /> <conditions> <add input="{URL}" pattern="^/(media|skin|js)/" ignoreCase="false" negate="true" /> <add input="{REQUEST_FILENAME}" matchType="IsFile" ignoreCase="false" negate="true" /> <add input="{REQUEST_FILENAME}" matchType="IsDirectory" ignoreCase="false" negate="true" /> </conditions> <action type="Rewrite" url="index.php/{R:0}" /> </rule> </rules> </rewrite> </system.webServer> </configuration> Next, I wanted to install WordPress. I unzipped all files in folder e:\mydomain\wwwroot\wordpress Browsed to www.mydomain.com/wordpress/wp-admin/install.php, where I configured everything for my database. Everything was installed correctly. I then navigate to http://www.mydomain.com/wordpress/wp-login.php where I type my credentials. I seem to be logged in and am redirected to http://www.mydomain.com/wordpress/wp-admin/ But there I receive an empty page. I enabled detailed error message in IIS following this article: http://www.iis.net/learn/troubleshoot/diagnosing-http-errors/how-to-use-http-detailed-errors-in-iis I also checkec with Fiddler and see that I receive a 500 error: GET /wordpress/wp-admin/ HTTP/1.1 Host: www.mydomain.com Connection: keep-alive Cache-Control: max-age=0 Accept: text/html,application/xhtml+xml,application/xml;q=0.9,/;q=0.8 User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.76 Safari/537.36 Referer: http://www.mydomain.com/wordpress/wp-login.php Accept-Encoding: gzip,deflate,sdch Accept-Language: en-US,en;q=0.8,nl;q=0.6 Cookie: wordpress_fabec4083cf12d8de89c98e8aef4b7e3=floran%7C1381236774%7C2d8edb4fc6618f290fadb49b035cad31; wordpress_test_cookie=WP+Cookie+check; wordpress_logged_in_fabec4083cf12d8de89c98e8aef4b7e3=floran%7C1381236774%7Cbf822163926b8b8df16d0f1fefb6e02e HTTP/1.1 500 Internal Server Error Content-Type: text/html Server: Microsoft-IIS/7.5 X-Powered-By: PHP/5.4.14 X-Powered-By: ASP.NET Date: Sun, 06 Oct 2013 12:56:03 GMT Content-Length: 0 My WordPress web.config in folder e:\mydomain\wwwroot\wordpress contains: <?xml version="1.0" encoding="UTF-8"?> <configuration> <system.webServer> <rewrite> <rules> <rule name="wordpress" patternSyntax="Wildcard"> <match url="*"/> <conditions> <add input="{REQUEST_FILENAME}" matchType="IsFile" negate="true"/> <add input="{REQUEST_FILENAME}" matchType="IsDirectory" negate="true"/> </conditions> <action type="Rewrite" url="index.php"/> </rule></rules> </rewrite> </system.webServer> </configuration> I also want my WordPress articles to be available on www.mydomain.com/blog instead of www.mydomain.com/wordpress Ofcourse my admin links for Magento and Wordpress should also work. How can I configure my web.config files to achieve the above?

    Read the article

  • SQL SERVER – ASYNC_IO_COMPLETION – Wait Type – Day 11 of 28

    - by pinaldave
    For any good system three things are vital: CPU, Memory and IO (disk). Among these three, IO is the most crucial factor of SQL Server. Looking at real-world cases, I do not see IT people upgrading CPU and Memory frequently. However, the disk is often upgraded for either improving the space, speed or throughput. Today we will look at another IO-related wait type. From Book On-Line: Occurs when a task is waiting for I/Os to finish. ASYNC_IO_COMPLETION Explanation: Any tasks are waiting for I/O to finish. If by any means your application that’s connected to SQL Server is processing the data very slowly, this type of wait can occur. Several long-running database operations like BACKUP, CREATE DATABASE, ALTER DATABASE or other operations can also create this wait type. Reducing ASYNC_IO_COMPLETION wait: When it is an issue related to IO, one should check for the following things associated to IO subsystem: Look at the programming and see if there is any application code which processes the data slowly (like inefficient loop, etc.). Note that it should be re-written to avoid this  wait type. Proper placing of the files is very important. We should check the file system for proper placement of the files – LDF and MDF on separate drive, TempDB on another separate drive, hot spot tables on separate filegroup (and on separate disk), etc. Check the File Statistics and see if there is a higher IO Read and IO Write Stall SQL SERVER – Get File Statistics Using fn_virtualfilestats. Check event log and error log for any errors or warnings related to IO. If you are using SAN (Storage Area Network), check the throughput of the SAN system as well as configuration of the HBA Queue Depth. In one of my recent projects, the SAN was performing really badly and so the SAN administrator did not accept it. After some investigations, he agreed to change the HBA Queue Depth on the development setup (test environment). As soon as we changed the HBA Queue Depth to quite a higher value, there was a sudden big improvement in the performance. It is very likely to happen that there are no proper indexes on the system and yet there are lots of table scans and heap scans. Creating proper index can reduce the IO bandwidth considerably. If SQL Server can use appropriate cover index instead of clustered index, it can effectively reduce lots of CPU, Memory and IO (considering cover index has lesser columns than cluster table and all other; it depends upon the situation). You can refer to the following two articles I wrote that talk about how to optimize indexes: Create Missing Indexes Drop Unused Indexes Checking Memory Related Perfmon Counters SQLServer: Memory Manager\Memory Grants Pending (Consistent higher value than 0-2) SQLServer: Memory Manager\Memory Grants Outstanding (Consistent higher value, Benchmark) SQLServer: Buffer Manager\Buffer Hit Cache Ratio (Higher is better, greater than 90% for usually smooth running system) SQLServer: Buffer Manager\Page Life Expectancy (Consistent lower value than 300 seconds) Memory: Available Mbytes (Information only) Memory: Page Faults/sec (Benchmark only) Memory: Pages/sec (Benchmark only) Checking Disk Related Perfmon Counters Average Disk sec/Read (Consistent higher value than 4-8 millisecond is not good) Average Disk sec/Write (Consistent higher value than 4-8 millisecond is not good) Average Disk Read/Write Queue Length (Consistent higher value than benchmark is not good) Read all the post in the Wait Types and Queue series. Note: The information presented here is from my experience and there is no way that I claim it to be accurate. I suggest reading Book OnLine for further clarification. All the discussions of Wait Stats in this blog are generic and vary from system to system. It is recommended that you test this on a development server before implementing it to a production server. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, SQL Wait Stats, SQL Wait Types, T SQL, Technology

    Read the article

  • Filtering data in LINQ with the help of where clause

    - by vik20000in
     LINQ has bought with itself a super power of querying Objects, Database, XML, SharePoint and nearly any other data structure. The power of LINQ lies in the fact that it is managed code that lets you write SQL type code to fetch data.  Whenever working with data we always need a way to filter out the data based on different condition. In this post we will look at some of the different ways in which we can filter data in LINQ with the help of where clause. Simple Filter for an array. Let’s say we have an array of number and we want to filter out data based on some condition. Below is an example int[] numbers = { 5, 4, 1, 3, 9, 8, 6, 7, 2, 0 }; var lowNums =                 from num in numbers                 where num < 5                 select num;   Filter based on one of the property in the class. With the help of LINQ we can also filer out data from a list based on value of some property. var soldOutProducts =                 from prod in products                 where prod.UnitsInStock == 0                 select prod; Filter based on Multiple of the property in the class. var expensiveInStockProducts =         from prod in products         where prod.UnitsInStock > 0 && prod.UnitPrice > 3.00M         select prod; Filter based on the index of the Item in the list.In the below example we can see that we are able to filter data based on the index of the item in the list. string[] digits = { "zero", "one", "two", "three", "four", "five", "six"}; var shortDigits = digits.Where((digit, index) => digit.Length < index); There are many other way in which we can filter out data in LINQ. In the above post I have tried and shown few ways using the LINQ. Vikram

    Read the article

  • SQL SERVER – IO_COMPLETION – Wait Type – Day 10 of 28

    - by pinaldave
    For any good system three things are vital: CPU, Memory and IO (disk). Among these three, IO is the most crucial factor of SQL Server. Looking at real-world cases, I do not see IT people upgrading CPU and Memory frequently. However, the disk is often upgraded for either improving the space, speed or throughput. Today we will look at an IO-related wait types. From Book On-Line: Occurs while waiting for I/O operations to complete. This wait type generally represents non-data page I/Os. Data page I/O completion waits appear as PAGEIOLATCH_* waits. IO_COMPLETION Explanation: Any tasks are waiting for I/O to finish. This is a good indication that IO needs to be looked over here. Reducing IO_COMPLETION wait: When it is an issue concerning the IO, one should look at the following things related to IO subsystem: Proper placing of the files is very important. We should check the file system for proper placement of files – LDF and MDF on a separate drive, TempDB on another separate drive, hot spot tables on separate filegroup (and on separate disk),etc. Check the File Statistics and see if there is higher IO Read and IO Write Stall SQL SERVER – Get File Statistics Using fn_virtualfilestats. Check event log and error log for any errors or warnings related to IO. If you are using SAN (Storage Area Network), check the throughput of the SAN system as well as the configuration of the HBA Queue Depth. In one of my recent projects, the SAN was performing really badly so the SAN administrator did not accept it. After some investigations, he agreed to change the HBA Queue Depth on development (test environment) set up and as soon as we changed the HBA Queue Depth to quite a higher value, there was a sudden big improvement in the performance. It is very possible that there are no proper indexes in the system and there are lots of table scans and heap scans. Creating proper index can reduce the IO bandwidth considerably. If SQL Server can use appropriate cover index instead of clustered index, it can effectively reduce lots of CPU, Memory and IO (considering cover index has lesser columns than cluster table and all other; it depends upon the situation). You can refer to the two articles that I wrote; they are about how to optimize indexes: Create Missing Indexes Drop Unused Indexes Checking Memory Related Perfmon Counters SQLServer: Memory Manager\Memory Grants Pending (Consistent higher value than 0-2) SQLServer: Memory Manager\Memory Grants Outstanding (Consistent higher value, Benchmark) SQLServer: Buffer Manager\Buffer Hit Cache Ratio (Higher is better, greater than 90% for usually smooth running system) SQLServer: Buffer Manager\Page Life Expectancy (Consistent lower value than 300 seconds) Memory: Available Mbytes (Information only) Memory: Page Faults/sec (Benchmark only) Memory: Pages/sec (Benchmark only) Checking Disk Related Perfmon Counters Average Disk sec/Read (Consistent higher value than 4-8 millisecond is not good) Average Disk sec/Write (Consistent higher value than 4-8 millisecond is not good) Average Disk Read/Write Queue Length (Consistent higher value than benchmark is not good) Note: The information presented here is from my experience and there is no way that I claim it to be accurate. I suggest reading Book OnLine for further clarification. All the discussions of Wait Stats in this blog are generic and vary from system to system. It is recommended that you test this on a development server before implementing it to a production server. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, SQL Wait Types, SQL White Papers, T SQL, Technology

    Read the article

  • E-Business Suite 12.1.3 Data Masking Certified with Enterprise Manager 12c

    - by Elke Phelps (Oracle Development)
    Following up on our prior announcement for EM 11g, we're pleased to announce the certification of the E-Business Suite 12.1.3 Data Masking Template for the Data Masking Pack with Enterprise Manager Cloud Control 12c. You can use the Oracle Data Masking Pack with Oracle Enterprise Manager Grid Control 12c to scramble sensitive data in cloned E-Business Suite environments.  Due to data dependencies, scrambling E-Business Suite data is not a trivial task.  The data needs to be scrubbed in such a way that allows the application to continue to function.  You may scramble data in E-Business Suite cloned environments with EM12c using the following template: E-Business Suite 12.1.3 Data Masking Template for Data Masking Pack with EM12c (Patch 14407414) What does data masking do in E-Business Suite environments? Application data masking does the following: De-identify the data:  Scramble identifiers of individuals, also known as personally identifiable information or PII.  Examples include information such as name, account, address, location, and driver's license number. Mask sensitive data:  Mask data that, if associated with personally identifiable information (PII), would cause privacy concerns.  Examples include compensation, health and employment information.   Maintain data validity:  Provide a fully functional application. How can EBS customers use data masking? The Oracle E-Business Suite Template for Data Masking Pack can be used in situations where confidential or regulated data needs to be shared with other non-production users who need access to some of the original data, but not necessarily every table.  Examples of non-production users include internal application developers or external business partners such as offshore testing companies, suppliers or customers.  The template works with the Oracle Data Masking Pack and Oracle Enterprise Manager to obscure sensitive E-Business Suite information that is copied from production to non-production environments. The Oracle E-Business Suite Template for Data Masking Pack is applied to a non-production environment with the Enterprise Manager Grid Control Data Masking Pack.  When applied, the Oracle E-Business Suite Template for Data Masking Pack will create an irreversibly scrambled version of your production database for development and testing.  What's new with EM 12c? Some of the execution steps may also be performed with EM Command Line Interface (EM CLI).  Support of EM CLI is a new feature with the E-Business Suite Release 12.1.3 template for EM 12c.  Is there a charge for this? Yes. You must purchase licenses for the Oracle Data Masking Pack plug-in. The Oracle E-Business Suite 12.1.3 Template for the Data Masking Pack is included with the Oracle Data Masking Pack license.  You can contact your Oracle account manager for more details about licensing. References Additional details and requirements are provided in the following My Oracle Support Note: Using Oracle E-Business Suite Release 12.1.3 Template for the Data Masking Pack with Oracle Enterprise Manager 12.1.0.2 Data Masking Tool (Note 1481916.1) Masking Sensitive Data in the Oracle Database Real Application Testing User's Guide 11g Release 2 (11.2) Related Articles Scrambling Sensitive Data in E-Business Suite

    Read the article

  • C++ problem with assimp 3D model loader

    - by Brendan Webster
    In my game I have model loading functions for Assimp model loading library. I can load the model and render it, but the model displays incorrectly. The models load in as if they were using a seperate projection matrix. I have looked over my code over and over again, but I probably keep on missing the obvious reason why this is happening. Here is an image of my game: It's simply a 6 sided cube, but it's off big time! Here are my code snippets for rendering the cube to the screen: void C_MediaLoader::display(void) { float tmp; glTranslatef(0,0,0); // rotate it around the y axis glRotatef(angle,0.f,0.f,1.f); glColor4f(1,1,1,1); // scale the whole asset to fit into our view frustum tmp = scene_max.x-scene_min.x; tmp = aisgl_max(scene_max.y - scene_min.y,tmp); tmp = aisgl_max(scene_max.z - scene_min.z,tmp); tmp = (1.f / tmp); glScalef(tmp/5, tmp/5, tmp/5); // center the model //glTranslatef( -scene_center.x, -scene_center.y, -scene_center.z ); // if the display list has not been made yet, create a new one and // fill it with scene contents if(scene_list == 0) { scene_list = glGenLists(1); glNewList(scene_list, GL_COMPILE); // now begin at the root node of the imported data and traverse // the scenegraph by multiplying subsequent local transforms // together on GL's matrix stack. recursive_render(scene, scene->mRootNode); glEndList(); } glCallList(scene_list); } void C_MediaLoader::recursive_render (const struct aiScene *sc, const struct aiNode* nd) { unsigned int i; unsigned int n = 0, t; struct aiMatrix4x4 m = nd->mTransformation; // update transform aiTransposeMatrix4(&m); glPushMatrix(); glMultMatrixf((float*)&m); // draw all meshes assigned to this node for (; n < nd->mNumMeshes; ++n) { const struct aiMesh* mesh = scene->mMeshes[nd->mMeshes[n]]; apply_material(sc->mMaterials[mesh->mMaterialIndex]); if(mesh->mNormals == NULL) { glDisable(GL_LIGHTING); } else { glEnable(GL_LIGHTING); } for (t = 0; t < mesh->mNumFaces; ++t) { const struct aiFace* face = &mesh->mFaces[t]; GLenum face_mode; switch(face->mNumIndices) { case 1: face_mode = GL_POINTS; break; case 2: face_mode = GL_LINES; break; case 3: face_mode = GL_TRIANGLES; break; default: face_mode = GL_POLYGON; break; } glBegin(face_mode); for(i = 0; i < face->mNumIndices; i++) { int index = face->mIndices[i]; if(mesh->mColors[0] != NULL) glColor4fv((GLfloat*)&mesh->mColors[0][index]); if(mesh->mNormals != NULL) glNormal3fv(&mesh->mNormals[index].x); glVertex3fv(&mesh->mVertices[index].x); } glEnd(); } } // draw all children for (n = 0; n < nd->mNumChildren; ++n) { recursive_render(sc, nd->mChildren[n]); } glPopMatrix(); } Sorry there is so much code to look through, but I really cannot find the problem, and I would love to have help.

    Read the article

  • Sprite animation in openGL - Some frames are being skipped

    - by Sid
    Earlier, I was facing problems on implementing sprite animation in openGL ES. Now its being sorted up. But the problem that i am facing now is that some of my frames are being skipped when a bullet(a circle) strikes on it. What I need : A sprite animation should stop at the last frame without skipping any frame. What I did : Collision Detection function and working properly. PS : Everything is working fine but i want to implement the animation in OPENGL ONLY. Canvas won't work in my case. ------------------------ EDIT----------------------- My sprite sheet. Consider the animation from Left to right and then from top to bottom Here is an image for a better understanding. My spritesheet ... class FragileSquare{ FloatBuffer fVertexBuffer, mTextureBuffer; ByteBuffer mColorBuff; ByteBuffer mIndexBuff; int[] textures = new int[1]; public boolean beingHitFromBall = false; int numberSprites = 20; int columnInt = 4; //number of columns as int float columnFloat = 4.0f; //number of columns as float float rowFloat = 5.0f; int oldIdx; public FragileSquare() { // TODO Auto-generated constructor stub float vertices [] = {-1.0f,1.0f, //byte index 0 1.0f, 1.0f, //byte index 1 //byte index 2 -1.0f, -1.0f, 1.0f,-1.0f}; //byte index 3 float textureCoord[] = { 0.0f,0.0f, 0.25f,0.0f, 0.0f,0.20f, 0.25f,0.20f }; byte indices[] = {0, 1, 2, 1, 2, 3 }; ByteBuffer byteBuffer = ByteBuffer.allocateDirect(4*2 * 4); // 4 vertices, 2 co-ordinates(x,y) 4 for converting in float byteBuffer.order(ByteOrder.nativeOrder()); fVertexBuffer = byteBuffer.asFloatBuffer(); fVertexBuffer.put(vertices); fVertexBuffer.position(0); ByteBuffer byteBuffer2 = ByteBuffer.allocateDirect(textureCoord.length * 4); byteBuffer2.order(ByteOrder.nativeOrder()); mTextureBuffer = byteBuffer2.asFloatBuffer(); mTextureBuffer.put(textureCoord); mTextureBuffer.position(0); } public void draw(GL10 gl){ gl.glFrontFace(GL11.GL_CW); gl.glEnableClientState(GL10.GL_VERTEX_ARRAY); gl.glVertexPointer(1,GL10.GL_FLOAT, 0, fVertexBuffer); gl.glEnable(GL10.GL_TEXTURE_2D); if(MyRender.flag2==1){ /** Collision has taken place*/ int idx = oldIdx==(numberSprites-1) ? (numberSprites-1) : (int)((System.currentTimeMillis()%(200*numberSprites))/200); gl.glMatrixMode(GL10.GL_TEXTURE); gl.glTranslatef((idx%columnInt)/columnFloat, (idx/columnInt)/rowFloat, 0); gl.glMatrixMode(GL10.GL_MODELVIEW); oldIdx = idx; } gl.glEnable(GL10.GL_BLEND); gl.glBlendFunc(GL10.GL_SRC_ALPHA, GL10.GL_ONE_MINUS_SRC_ALPHA); gl.glBindTexture(GL10.GL_TEXTURE_2D, textures[0]); //4 gl.glTexCoordPointer(2, GL10.GL_FLOAT,0, mTextureBuffer); //5 gl.glEnableClientState(GL10.GL_TEXTURE_COORD_ARRAY); gl.glDrawArrays(GL10.GL_TRIANGLE_STRIP, 0, 4); //7 gl.glFrontFace(GL11.GL_CCW); gl.glDisableClientState(GL10.GL_VERTEX_ARRAY); gl.glDisableClientState(GL10.GL_TEXTURE_COORD_ARRAY); gl.glMatrixMode(GL10.GL_TEXTURE); gl.glLoadIdentity(); gl.glMatrixMode(GL10.GL_MODELVIEW); } public void loadFragileTexture(GL10 gl, Context context, int resource) { Bitmap bitmap = BitmapFactory.decodeResource(context.getResources(), resource); gl.glGenTextures(1, textures, 0); gl.glBindTexture(GL10.GL_TEXTURE_2D, textures[0]); gl.glTexParameterf(GL10.GL_TEXTURE_2D, GL10.GL_TEXTURE_MIN_FILTER, GL10.GL_LINEAR); gl.glTexParameterf(GL10.GL_TEXTURE_2D, GL10.GL_TEXTURE_MAG_FILTER, GL10.GL_LINEAR); gl.glTexParameterf(GL10.GL_TEXTURE_2D, GL10.GL_TEXTURE_WRAP_S, GL10.GL_REPEAT); gl.glTexParameterf(GL10.GL_TEXTURE_2D, GL10.GL_TEXTURE_WRAP_T, GL10.GL_REPEAT); GLUtils.texImage2D(GL10.GL_TEXTURE_2D, 0, bitmap, 0); bitmap.recycle(); }

    Read the article

  • New Rapid Install StartCD 12.2.0.48 for EBS 12.2 Now Available

    - by Max Arderius
    A new Rapid Install startCD (Patch 18086193) for Oracle E-Business Suite Release 12.2 is now available. We recommend that all EBS customers installing or upgrading to EBS 12.2 use this latest update. The startCD updates are distributed to customers via My Oracle Support Patch which can be uncompressed on top of any previous 12.2 startCD under the main staging area. This patch replaces any previous startCDs. What's New in This Update? This new startCD version 12.2.0.48 includes important fixes for multi-node Installs, RAC, pre-install checks, platform specific issues, and upgrade scenario failures: 18703814 - QREP:122:RI:ISSUE WITH CHECKOS.CMD 18689527 - QREP:122:RI:ISSUE WITH FNDCORE.DLL SHIPPED AS PART OF R122 PACKAGE 18548485 - QREP1224:4:JAR SIGNER ISSUE DUE TO THE RI UPGRADE AUTOCONFIG CHANGES 18535812 - QREP:1220.48_4: 12.2.0 UPGRADE FILE SYSTEM LAY OUT IS AFFECTING THE DB TABLES 18507545 - WIN: UNABLE TO LAY DOWN FS PRIOR TO 12.2 UPGRADE WITHOUT AFFECTING RUNNING DB 18476041 - UNABLE TO LAY DOWN FS PRIOR TO 12.2 UPGRADE WITHOUT AFFECTING PRODUCTION DB 18459887 - R12.2 INSTALLATION FAILURE - OPMNCTL: NOT FOUND 18436053 - START CD 48_4 - ISSUES WITH TEMP SPACE CHECK 18424747 - QREP1224.3:ADD SERVER BROWSE BUTTON NOT WORKING 18421132 - *RW-50010: ERROR: - SCRIPT HAS RETURNED AN ERROR: 1 18403700 - QREP122.48:RI:UPGRADE RI PRECHECK HUNG IN SPLIT TIER APPS NODE ( NO SILENT ) 18383075 - ADD VERBOSE OPTION TO RAC VALIDATION 18363584 - UPTAKE INSTALL SCRIPTS FOR XB48_4 18336093 - QREP:122:RI:PATCH FS ADMIN SERVICE RUNNING AFTER RI UPGRADE CONFIGURE MODE 18320278 - QREP:1224.3:PLATFORM SPECIFIC SYNTAX ERRORS WITH DATE COMMAND IN DB CHECKER 18314643 - DISABLE SID=DB_NAME FOR RI UPGRADE FLOW IN RAC 18298977 - RI: EXCEPTION WHILE CLICKING RAC NODES BUTTON ON A NON-RAC SERVER 18286816 - QREP122:STARTCD48_3:TRAVERSING FROM VISION PASSW SCREEN TO PROD 18286371 - QREP122:STARTCD48_3:AMBIGUOUS MESSAGE DURING STAGE AREA CHECK ON HP 18275403 - QREP122:48:RI UPGRADE WITH EOH POST CHECKS HANGS IN SPLIT TIER DB NODE 18270631 - QREP122.48:MULTI-NODE RI USING NON-DEFAULT PASSWORDS NOT WORKING 18266046 - QREP122:48:RI NOT ALLOWING TO IGNORE THE RAC PRE-CHECK FAILURE 18242201 - UPTAKE TXK INSTALL SCRIPTS AND PLATFORMS.ZIP INTO STARTCD XB48_3 18236428 - QREP122.47:RI UPGRADE EXISTING OH FOR NON-DEFAULT APPS PASSWORD NOT WORKING 18220640 - INCONSISTENT DATABASE PORTS DURING EBS 12.2 INSTALLATION FOR STARTCD 12.2.0.47 18138796 - QREP122:47:RI 10.1.2 TECHSTACK NOT WORKING IF WE RUN RI FROM NEW STARTCD LOC 18138396 - TST1220: CONTROL FILE NAMING IN RAPID INSTALL SEEMS TO HAVE ISSUES 18124144 - IMPROVE HANDLING ERRORS FOUND IN CLUVFY LOG DURING PREINSTALL CHECKS 18111361 - VALIDATE ASM DB DATA FILES PATH AS +<DATA GROUP>/<PATH> 18102504 - QREP1220.47_5: UNZIP PANEL DOES NOT CREATE THE CORRECT STAGE 18083342 - 12.2 UPGRADE JAVA.NET.BINDEXCEPTION: CANNOT ASSIGN REQUESTED ADDRESS 18082140 - QREP122:47:RAC DB VALIDATION IS FAILS WITH EXIT STATUS IS 6 18062350 - 12.2.3 UPG: 12.2.0 INSTALLATION LOGS 18050840 - RI: UPGRADE WITH EXISTING RAC OH:SECONDARY DB NODE NAME IS BLANK 18049813 - RAC LOV DEFAULTS NOT SAVED UNLESS "SELECT" IS CLICKED 18003592 - TST1220:ADDITIONAL FREE SPACE CHECK FOR RI NEEDS TO BE CHECKED 17981471 - REMOVE ASM SPACE CHECK FROM RACVALIDATIONS.SH 17942179 - R12.2 INSTALL FAILING AT ADRUN11G.SH WITH ERRORS RW-50004 & RW-50010 17893583 - QREP1220.47:VALIDATION OF O.S IN RAPIDWIZ IN THE DB NODE CONFIGURATION SCREEN 17886258 - CLEANUP FND_NODES DURING UPGRADE FLOW 17858010 - RI POST INSTALL CHECKS (SSH VERIFICATION) STEP IS FAILING 17799807 - GEOHR: 12.2.0 - ERRORS IN RAPIDWIZ AND ADCONFIG LOGS 17786162 - QREP1223.4:RI:SERVICE_NAMES IS PRINTED AS SERVICE_NAME IN RI SCREEN 17782455 - RI: CONFIRM DEFAULT APPS PASSWORD IN SILENT MODE KICKOFF 17778130 - RI:ADMIN SERVER TO BE UP ON PRIMARY MID-TIER IN MULTI-NODE UPGRADE FS CREATION 17773989 - UN-SUPPORTED PLATFORM SHOWS 32 BIT AS HARD-CODED 17772655 - RELEVANT MESSAGE DURING THE RAPDIWIZ -TECHSTACK 17759279 - VERIFICATION PANEL DOES NOT EXPAND TECHNOLOGY STACK 17759183 - BUILDSTAGE SCRIPT MENU NEEDS TO BE ADJUSTED 17737186 - DATABASE PRE-REQ CHECK INCORRECTLY REPORTS SUCCESS ON AIX 17708082 - 12.2 INSTALLATION - OS PRE-REQUISITES CHECK 17701676 - TST122: GENERATE WRONG S_DBSID FOR PATCH FILE SYSTEM AT PHASE PREPARE 17630972 - /TMP PRE-REQ INSTALLATION CHECK 17617245 - 12.2 VISION INSTALL FAILS ON AIX 17603342 - OMCS: DB STAGING COMPLAINS WHILE MOVING IT TO FINAL LOCATION 17591171 - OMCS: DB STAGING FAILS WITH FRESH INSTALL R12.2 17588765 - CHECKER VERSION AND PLUGIN VERSION 17561747 - BUILDSTAGE.SH FAILS WITH ERROR WHEN STAGE HOSTED ON 32BIT LINUX 17539198 - RAPID INSTALL NEEDS TO IGNORE NON-REQUIRED STAGE ELEMENTS 17272808 - APPS USERS THAT HAVE DEFAULT PASSWORD AFTER 12.2 RAPID INSTALL References 12.2 Documentation Library 1581299.1 : EBS 12.2 Product Information Center 1320300.1 : Oracle E-Business Suite Release Notes, Release 12.2 1606170.1 : Oracle E-Business Suite Technology Stack and Applications DBA Release Notes for Release 12.2.3 1624423.1 : Oracle E-Business Suite Technology Stack and Applications DBA Release Notes for R12.TXK.C.Delta.4 and R12.AD.C.Delta.4 1594274.1 : Oracle E-Business Suite Release 12.2: Consolidated List of Patches and Technology Bug Fixes Related Articles Oracle E-Business Suite 12.2 Now Available startCD options to install Oracle E-Business Suite Release 12.2

    Read the article

  • MVC Portable Areas &ndash; Static Files as Embedded Resources

    - by Steve Michelotti
    This is the third post in a series related to build and deployment considerations as I’ve been exploring MVC Portable Areas: #1 – Using Web Application Project to build portable areas #2 – Conventions for deploying portable area static files #3 – Portable area static files as embedded resources In the last post, I walked through a convention for managing static files.  In this post I’ll discuss another approach to manage static files (e.g., images, css, js, etc.).  With this approach, you *also* compile the static files as embedded resources into the assembly similar to the *.aspx pages. Once again, you can set this to happen automatically by simply modifying your *.csproj file to include the desired extensions so you don’t have to remember every time you add a file: 1: <Target Name="BeforeBuild"> 2: <ItemGroup> 3: <EmbeddedResource Include="**\*.aspx;**\*.ascx;**\*.gif;**\*.css;**\*.js" /> 4: </ItemGroup> 5: </Target> We now need a reliable way to serve up these static files that are embedded in the assembly. There are a couple of ways to do this but one way is to simply create a Resource controller whose job is dedicated to doing this. 1: public class ResourceController : Controller 2: { 3: public ActionResult Index(string resourceName) 4: { 5: var contentType = GetContentType(resourceName); 6: var resourceStream = Assembly.GetExecutingAssembly().GetManifestResourceStream(resourceName); 7:   8: return this.File(resourceStream, contentType); 9: return View(); 10: } 11:   12: private static string GetContentType(string resourceName) 13: { 14: var extention = resourceName.Substring(resourceName.LastIndexOf('.')).ToLower(); 15: switch (extention) 16: { 17: case ".gif": 18: return "image/gif"; 19: case ".js": 20: return "text/javascript"; 21: case ".css": 22: return "text/css"; 23: default: 24: return "text/html"; 25: } 26: } 27: } In order to use this controller, we need to make sure we’ve registered the route in our portable area registration (shown in lines 5-6): 1: public class WidgetAreaRegistration : PortableAreaRegistration 2: { 3: public override void RegisterArea(System.Web.Mvc.AreaRegistrationContext context, IApplicationBus bus) 4: { 5: context.MapRoute("ResourceRoute", "widget1/resource/{resourceName}", 6: new { controller = "Resource", action = "Index" }); 7:   8: context.MapRoute("Widget1", "widget1/{controller}/{action}", new 9: { 10: controller = "Home", 11: action = "Index" 12: }); 13:   14: RegisterTheViewsInTheEmbeddedViewEngine(GetType()); 15: } 16:   17: public override string AreaName 18: { 19: get { return "Widget1"; } 20: } 21: } In my previous post, we relied on a custom Url helper method to find the actual physical path to the static file like this: 1: <img src="<%: Url.AreaContent("/images/arrow.gif") %>" /> Hello World! However, since we are now embedding the files inside the assembly, we no longer have to worry about the physical path. We can change this line of code to this: 1: <img src="<%: Url.Resource("Widget1.images.arrow.gif") %>" /> Hello World! Note that I had to fully quality the resource name (with namespace and physical location) since that is how .NET assemblies store embedded resources. I also created my own Url helper method called Resource which looks like this: 1: public static string Resource(this UrlHelper urlHelper, string resourceName) 2: { 3: var areaName = (string)urlHelper.RequestContext.RouteData.DataTokens["area"]; 4: return urlHelper.Action("Index", "Resource", new { resourceName = resourceName, area = areaName }); 5: } This method gives us the convenience of not having to know how to construct the URL – but just allowing us to refer to the resource name. The resulting html for the image tag is: 1: <img src="/widget1/resource/Widget1.images.arrow.gif" /> so we can always request any image from the browser directly. This is almost analogous to the WebResource.axd file but for MVC. What is interesting though is that we can encapsulate each one of these so that each area can have it’s own set of resources and they are easily distinguished because the area name is the first segment of the route. This makes me wonder if something like this ResourceController should be baked into portable areas itself. I’m definitely interested in anyone has any opinions on it or have taken alternative approaches.

    Read the article

  • Functional/nonfunctional requirements VS design ideas

    - by Nicholas Chow
    Problem domain Functional requirements defines what a system does. Non-Functional requirements defines quality attributes of what the system does as a whole.(performance, security, reliability, volume, useability, etc.) Constraints limits the design space, they restrict designers to certain types of solutions. Solution domain Design ideas , defines how the system does it. For example a stakeholder need might be we want to increase our sales, therefore we must improve the usability of our webshop so more customers will purchase, a requirement can be written for this. (problem domain) Design takes this further into the solution domain by saying "therefore we want to offer credit card payments in addition to the current prepayment option". My problem is that the transition phase from requirement to design seems really vague, therefore when writing requirements I am often confused whether or not I incorporated design ideas in my requirements, that would make my requirement wrong. Another problem is that I often write functional requirements as what a system does, and then I also specify in what timeframe it must be done. But is this correct? Is it then a still a functional requirement or a non functional one? Is it better to seperate it into two distinct requirements? Here are a few requirements I wrote: FR1 Registration of Organizer FR1 describes the registration of an Organizer on CrowdFundum FR1.1 The system shall display a registration form on the website. FR1.2 The system shall require a Name, Username, Document number passport/ID card, Address, Zip code, City, Email address, Telephone number, Bank account, Captcha code on the registration form when a user registers. FR1.4 The system shall display an error message containing: “Registration could not be completed” to the subscriber within 1 seconds after the system check of the registration form was unsuccessful. FR1.5 The system shall send a verification email containing a verification link to the subscriber within 30 seconds after the system check of the registration form was successful. FR1.6 The system shall add the newly registered Organizer to the user base within 5 seconds after the verification link was accessed. FR2 Organizer submits a Project FR2 describes the submission of a Project by an Organizer on CrowdFundum - FR2 The system shall display a submit Project form to the Organizer accounts on the website.< - FR2.3 The system shall check for completeness the Name of the Project, 1-3 Photo’s, Keywords of the Project, Punch line, Minimum and maximum amount of people, Funding threshold, One or more reward tiers, Schedule of when what will be organized, Budget plan, 300-800 Words of additional information about the Project, Contact details within 1 secondin after an Organizer submits the submit Project form. - FR2.8 The system shall add to the homepage in the new Projects category the Project link within 30 seconds after the system made a Project webpage - FR2.9 The system shall include in the Project link for the homepage : Name of the Project, 1 Photo, Punch line within 30 seconds after the system made a Project webpage. Questions: FR 1.1 : Have I incorporated a design idea here, would " the system shall have a registration form" be a better functional requirement? F1.2 ,2.3 : Is this not singular? Would the conditions be better written for each its own separate requirement FR 1.4: Is this a design idea? Is this a correct functional requirement or have I incorporated non functional(performance) in it? Would it be better if I written it like this: FR1 The system shall display an error message when check is unsuccessful. NFR: The system will respond to unsuccesful registration form checks within 1 seconds. Same question with FR 2.8 and 2.9. FR2.3: The system shall check for "completeness", is completeness here used ambigiously? Should I rephrase it? FR1.2: I added that the system shall require a "Captcha code" is this a functional requirement or does it belong to the "security aspect" of a non functional requirement. I am eagerly waiting for your response. Thanks!

    Read the article

  • Problem with Assimp 3D model loader

    - by Brendan Webster
    In my game I have model loading functions for Assimp model loading library. I can load the model and render it, but the model displays incorrectly. The models load in as if they were using a seperate projection matrix. I have looked over my code over and over again, but I probably keep on missing the obvious reason why this is happening. Here is an image of my game: It's simply a 6 sided cube, but it's off big time! Here are my code snippets for rendering the cube to the screen: void C_MediaLoader::display(void) { float tmp; glTranslatef(0,0,0); // rotate it around the y axis glRotatef(angle,0.f,0.f,1.f); glColor4f(1,1,1,1); // scale the whole asset to fit into our view frustum tmp = scene_max.x-scene_min.x; tmp = aisgl_max(scene_max.y - scene_min.y,tmp); tmp = aisgl_max(scene_max.z - scene_min.z,tmp); tmp = (1.f / tmp); glScalef(tmp/5, tmp/5, tmp/5); // center the model //glTranslatef( -scene_center.x, -scene_center.y, -scene_center.z ); // if the display list has not been made yet, create a new one and // fill it with scene contents if(scene_list == 0) { scene_list = glGenLists(1); glNewList(scene_list, GL_COMPILE); // now begin at the root node of the imported data and traverse // the scenegraph by multiplying subsequent local transforms // together on GL's matrix stack. recursive_render(scene, scene->mRootNode); glEndList(); } glCallList(scene_list); } void C_MediaLoader::recursive_render (const struct aiScene *sc, const struct aiNode* nd) { unsigned int i; unsigned int n = 0, t; struct aiMatrix4x4 m = nd->mTransformation; // update transform aiTransposeMatrix4(&m); glPushMatrix(); glMultMatrixf((float*)&m); // draw all meshes assigned to this node for (; n < nd->mNumMeshes; ++n) { const struct aiMesh* mesh = scene->mMeshes[nd->mMeshes[n]]; apply_material(sc->mMaterials[mesh->mMaterialIndex]); if(mesh->mNormals == NULL) { glDisable(GL_LIGHTING); } else { glEnable(GL_LIGHTING); } for (t = 0; t < mesh->mNumFaces; ++t) { const struct aiFace* face = &mesh->mFaces[t]; GLenum face_mode; switch(face->mNumIndices) { case 1: face_mode = GL_POINTS; break; case 2: face_mode = GL_LINES; break; case 3: face_mode = GL_TRIANGLES; break; default: face_mode = GL_POLYGON; break; } glBegin(face_mode); for(i = 0; i < face->mNumIndices; i++) { int index = face->mIndices[i]; if(mesh->mColors[0] != NULL) glColor4fv((GLfloat*)&mesh->mColors[0][index]); if(mesh->mNormals != NULL) glNormal3fv(&mesh->mNormals[index].x); glVertex3fv(&mesh->mVertices[index].x); } glEnd(); } } // draw all children for (n = 0; n < nd->mNumChildren; ++n) { recursive_render(sc, nd->mChildren[n]); } glPopMatrix(); } Sorry there is so much code to look through, but I really cannot find the problem, and I would love to have help.

    Read the article

  • RewriteRule not working for certain URLs

    - by keiki
    There are a few domains pointing towards the same server, and of course I need them all redirect to only one of them. Redirects work, but only for certain URLs. What works: http://www.domain.com, http://domain.com, domain.com/index.html, domain.com/index.php, , domain.com/nonExistentDirectory, and if I click in the menu the following URLs are also redirected correctly: domain.com/foo/bar, domain.com/foo/bar.html or .php or other extension. What doesn't work: domain.com/existentDirectory, domain.com/foo/bar (if I type the URL in the address bar). If anyone will have the time and skill and will to tell me where's the mistake, I'll be deeply grateful. Here's my .htaccess file: AddHandler x-httpd-php .html .htm <ifModule mod_gzip.c> mod_gzip_on Yes mod_gzip_dechunk Yes mod_gzip_item_include file \.(html?|txt|css|js|php|pl)$ mod_gzip_item_include handler ^cgi-script$ mod_gzip_item_include mime ^text/.* mod_gzip_item_include mime ^application/x-javascript.* mod_gzip_item_exclude mime ^image/.* mod_gzip_item_exclude rspheader ^Content-Encoding:.*gzip.* </ifModule> <ifModule mod_expires.c> ExpiresActive On ExpiresDefault "access plus 1 seconds" ExpiresByType text/html "access plus 1 seconds" ExpiresByType image/gif "access plus 2592000 seconds" ExpiresByType image/jpeg "access plus 2592000 seconds" ExpiresByType image/png "access plus 2592000 seconds" ExpiresByType text/css "access plus 2592000 seconds" ExpiresByType text/javascript "access plus 216000 seconds" ExpiresByType application/x-javascript "access plus 216000 seconds" </ifModule> <ifModule mod_headers.c> <filesMatch "\\.(ico|pdf|flv|jpg|jpeg|png|gif|swf)$"> Header set Cache-Control "max-age=2592000, public" </filesMatch> <filesMatch "\\.(css)$"> Header set Cache-Control "max-age=2592000, public" </filesMatch> <filesMatch "\\.(js)$"> Header set Cache-Control "max-age=216000, private" </filesMatch> <filesMatch "\\.(xml|txt)$"> Header set Cache-Control "max-age=216000, public, must-revalidate" </filesMatch> <filesMatch "\\.(html|htm|php)$"> Header set Cache-Control "max-age=1, private, must-revalidate" </filesMatch> </ifModule> <ifModule mod_headers.c> Header unset ETag </ifModule> FileETag None <ifModule mod_headers.c> Header unset Last-Modified </ifModule> # BEGIN WordPress <IfModule mod_rewrite.c> RewriteEngine On RewriteBase / RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule . /index.php [L] </IfModule> # END WordPress RewriteCond %{HTTP_HOST} ^foo\.com$ [OR] RewriteCond %{HTTP_HOST} ^www\.foo\.com$ RewriteRule (.*) http://domain.com/$1 [R=301,L,QSA] RewriteCond %{HTTP_HOST} ^foo1\.com$ [OR] RewriteCond %{HTTP_HOST} ^www\.foo1\.com$ RewriteRule (.*) http://domain.com/$1 [R=301,L,QSA] RewriteCond %{HTTP_HOST} ^foo2\.com$ [OR] RewriteCond %{HTTP_HOST} ^www\.foo2\.com$ RewriteRule (.*) http://domain.com/$1 [R=301,L,QSA] RewriteCond %{HTTP_HOST} ^foo3\.com$ [OR] RewriteCond %{HTTP_HOST} ^www\.foo3\.com$ RewriteRule (.*) http://domain.com/$1 [R=301,L,QSA] RewriteCond %{HTTP_HOST} ^foo8\.com$ [OR] RewriteCond %{HTTP_HOST} ^www\.foo8\.com$ RewriteRule (.*) http://domain.com/$1 [R=301,L,QSA] Thinking that the above version was overkill, I've also tried to redirect all the requests for domains different than the main on to be redirected to it like this: RewriteCond %{HTTP_HOST} !^domain\.com$ [NC] RewriteRule ^(.*)$ http://domain.com [L,R=301] Is it also wrong? Because it doesn't work either! P.S. @Sodved I've tried that and it doesn't help (I comment here because I can't seem to be able to comment your answer.) Removing the following piece of code didn't solve the issue either, so the problem must be somewhere else: # BEGIN WordPress <IfModule mod_rewrite.c> RewriteEngine On RewriteBase / RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule . /index.php [L] </IfModule> # END WordPress New details: using this tool for checking the redirects I got the following results for the URLs that are not redirected: Checked link: http://domain.com/aDirectory/ Type of link: direct link (note the trailing slash above) and: Checked link: http://domain.com/aDirectory Type of redirect: 301 Moved Permanently Redirected to: http://domain.com/aDirectory/ (no trailing slash here) I hope/suspect I'm getting closer to the cause of this behavior.

    Read the article

  • Reducing Deadlocks - not a DBA issue ?

    - by steveh99999
     As a DBA, I'm involved on an almost daily basis troubleshooting 'SQL Server' performance issues. Often, this troubleshooting soon veers away from a 'its a SQL Server issue' to instead become a wider application/database design/coding issue.One common perception with SQL Server is that deadlocking is an application design issue - and is fixed by recoding...  I see this reinforced by MCP-type questions/scenarios where the answer to prevent deadlocking is simply to change the order in code in which tables are accessed....Whilst this is correct, I do think this has led to a situation where many 'operational' or 'production support' DBAs, when faced with a deadlock, are happy to throw the issue over to developers without analysing the issue further....A couple of 'war stories' on deadlocks which I think are interesting :- Case One , I had an issue recently on a third-party application that I support on SQL 2008.  This particular third-party application has an unusual support agreement where the customer is allowed to change the index design on the third-party provided database.  However, we are not allowed to alter application code or modify table structure..This third-party application is also known to encounter occasional deadlocks – indeed, I have documentation from the vendor that up to 50 deadlocks per day is not unusual !So, as a DBA I have to support an application which in my opinion has too many deadlocks - but, I cannot influence the design of the tables or stored procedures for the application. This should be the classic - blame the third-party developers scenario, and hope this issue gets addressed in a future application release - ie we could wait years for this to be resolved and implemented in our production environment...But, as DBAs  can change the index layout, is there anything I could do still to reduce the deadlocks in the application ?I initially used SQL traceflag 1222 to write deadlock detection output to the SQL Errorlog – using this I was able to identify one table heavily involved in the deadlocks.When I examined the table definition, I was surprised to see it was a heap – ie no clustered index existed on the table.Using SQL profiler to see locking behaviour and plan for the query involved in the deadlock, I was able to confirm a table scan was being performed.By creating an appropriate clustered index - it was possible to produce a more efficient plan and locking behaviour.So, less locks, held for less time = less possibility of deadlocks. I'm still unhappy about the overall number of deadlocks on this system - but that's something to be discussed further with the vendor.Case Two,  a system which hadn't changed for months suddenly started seeing deadlocks on a regular basis. I love the 'nothing's changed' scenario, as it gives me the opportunity to appear wise and say 'nothings changed on this system, except the data'.. This particular deadlock occurred on a table which had been growing rapidly. By using DBCC SHOW_STATISTICS - the DBA team were able to see that the deadlocks seemed to be occurring shortly after auto-update stats had regenerated the table statistics using it's default sampling behaviour.As a quick fix, we were able to schedule a nightly UPDATE STATISTICS WITH FULLSCAN on the table involved in the deadlock - thus, greatly reducing the potential for stats to be updated via auto_update_stats, consequently reducing the potential for a bad plan to be generated based on an unrepresentative sample of the data. This reduced the possibility of a deadlock occurring.  Not a perfect solution by any means, but quick, easy to implement, and needed no application code changes. This fix gave us some 'breathing space'  to properly fix the code during the next scheduled application release.   The moral of this post - don't dismiss deadlocks as issues that can only be fixed by developers...

    Read the article

  • Stale statistics on a newly created temporary table in a stored procedure can lead to poor performance

    - by sqlworkshops
    When you create a temporary table you expect a new table with no past history (statistics based on past existence), this is not true if you have less than 6 updates to the temporary table. This might lead to poor performance of queries which are sensitive to the content of temporary tables.I was optimizing SQL Server Performance at one of my customers who provides search functionality on their website. They use stored procedure with temporary table for the search. The performance of the search depended on who searched what in the past, option (recompile) by itself had no effect. Sometimes a simple search led to timeout because of non-optimal plan usage due to this behavior. This is not a plan caching issue rather temporary table statistics caching issue, which was part of the temporary object caching feature that was introduced in SQL Server 2005 and is also present in SQL Server 2008 and SQL Server 2012. In this customer case we implemented a workaround to avoid this issue (see below for example for workarounds).When temporary tables are cached, the statistics are not newly created rather cached from the past and updated based on automatic update statistics threshold. Caching temporary tables/objects is good for performance, but caching stale statistics from the past is not optimal.We can work around this issue by disabling temporary table caching by explicitly executing a DDL statement on the temporary table. One possibility is to execute an alter table statement, but this can lead to duplicate constraint name error on concurrent stored procedure execution. The other way to work around this is to create an index.I think there might be many customers in such a situation without knowing that stale statistics are being cached along with temporary table leading to poor performance.Ideal solution is to have more aggressive statistics update when the temporary table has less number of rows when temporary table caching is used. I will open a connect item to report this issue.Meanwhile you can mitigate the issue by creating an index on the temporary table. You can monitor active temporary tables using Windows Server Performance Monitor counter: SQL Server: General Statistics->Active Temp Tables. The script to understand the issue and the workaround is listed below:set nocount onset statistics time offset statistics io offdrop table tab7gocreate table tab7 (c1 int primary key clustered, c2 int, c3 char(200))gocreate index test on tab7(c2, c1, c3)gobegin trandeclare @i intset @i = 1while @i <= 50000begininsert into tab7 values (@i, 1, ‘a’)set @i = @i + 1endcommit trangoinsert into tab7 values (50001, 1, ‘a’)gocheckpointgodrop proc test_slowgocreate proc test_slow @i intasbegindeclare @j intcreate table #temp1 (c1 int primary key)insert into #temp1 (c1) select @iselect @j = t7.c1 from tab7 t7 inner join #temp1 t on (t7.c2 = t.c1)endgodbcc dropcleanbuffersset statistics time onset statistics io ongo–high reads as expected for parameter ’1'exec test_slow 1godbcc dropcleanbuffersgo–high reads that are not expected for parameter ’2'exec test_slow 2godrop proc test_with_recompilegocreate proc test_with_recompile @i intasbegindeclare @j intcreate table #temp1 (c1 int primary key)insert into #temp1 (c1) select @iselect @j = t7.c1 from tab7 t7 inner join #temp1 t on (t7.c2 = t.c1)option (recompile)endgodbcc dropcleanbuffersset statistics time onset statistics io ongo–high reads as expected for parameter ’1'exec test_with_recompile 1godbcc dropcleanbuffersgo–high reads that are not expected for parameter ’2'–low reads on 3rd execution as expected for parameter ’2'exec test_with_recompile 2godrop proc test_with_alter_table_recompilegocreate proc test_with_alter_table_recompile @i intasbegindeclare @j intcreate table #temp1 (c1 int primary key)–to avoid caching of temporary tables one can create a constraint–but this might lead to duplicate constraint name error on concurrent usagealter table #temp1 add constraint test123 unique(c1)insert into #temp1 (c1) select @iselect @j = t7.c1 from tab7 t7 inner join #temp1 t on (t7.c2 = t.c1)option (recompile)endgodbcc dropcleanbuffersset statistics time onset statistics io ongo–high reads as expected for parameter ’1'exec test_with_alter_table_recompile 1godbcc dropcleanbuffersgo–low reads as expected for parameter ’2'exec test_with_alter_table_recompile 2godrop proc test_with_index_recompilegocreate proc test_with_index_recompile @i intasbegindeclare @j intcreate table #temp1 (c1 int primary key)–to avoid caching of temporary tables one can create an indexcreate index test on #temp1(c1)insert into #temp1 (c1) select @iselect @j = t7.c1 from tab7 t7 inner join #temp1 t on (t7.c2 = t.c1)option (recompile)endgoset statistics time onset statistics io ondbcc dropcleanbuffersgo–high reads as expected for parameter ’1'exec test_with_index_recompile 1godbcc dropcleanbuffersgo–low reads as expected for parameter ’2'exec test_with_index_recompile 2go

    Read the article

  • Generate texture for a heightmap

    - by James
    I've recently been trying to blend multiple textures based on the height at different points in a heightmap. However i've been getting poor results. I decided to backtrack and just attempt to recreate one single texture from an SDL_Surface (i'm using SDL) and just send that into opengl. I'll put my code for creating the texture and reading the colour values. It is a 24bit TGA i'm loading, and i've confirmed that the rest of my code works because i was able to send the surfaces pixels directly to my createTextureFromData function and it drew fine. struct RGBColour { RGBColour() : r(0), g(0), b(0) {} RGBColour(unsigned char red, unsigned char green, unsigned char blue) : r(red), g(green), b(blue) {} unsigned char r; unsigned char g; unsigned char b; }; // main loading code SDLSurfaceReader* reader = new SDLSurfaceReader(m_renderer); reader->readSurface("images/grass.tga"); // new texture unsigned char* newTexture = new unsigned char[reader->m_surface->w * reader->m_surface->h * 3 * reader->m_surface->w]; for (int y = 0; y < reader->m_surface->h; y++) { for (int x = 0; x < reader->m_surface->w; x += 3) { int index = (y * reader->m_surface->w) + x; RGBColour colour = reader->getColourAt(x, y); newTexture[index] = colour.r; newTexture[index + 1] = colour.g; newTexture[index + 2] = colour.b; } } unsigned int id = m_renderer->createTextureFromData(newTexture, reader->m_surface->w, reader->m_surface->h, RGB); // functions for reading pixels RGBColour SDLSurfaceReader::getColourAt(int x, int y) { Uint32 pixel; Uint8 red, green, blue; RGBColour rgb; pixel = getPixel(m_surface, x, y); SDL_LockSurface(m_surface); SDL_GetRGB(pixel, m_surface->format, &red, &green, &blue); SDL_UnlockSurface(m_surface); rgb.r = red; rgb.b = blue; rgb.g = green; return rgb; } // this function taken from SDL documentation // http://www.libsdl.org/cgi/docwiki.cgi/Introduction_to_SDL_Video#getpixel Uint32 SDLSurfaceReader::getPixel(SDL_Surface* surface, int x, int y) { int bpp = m_surface->format->BytesPerPixel; Uint8 *p = (Uint8*)m_surface->pixels + y * m_surface->pitch + x * bpp; switch (bpp) { case 1: return *p; case 2: return *(Uint16*)p; case 3: if (SDL_BYTEORDER == SDL_BIG_ENDIAN) return p[0] << 16 | p[1] << 8 | p[2]; else return p[0] | p[1] << 8 | p[2] << 16; case 4: return *(Uint32*)p; default: return 0; } } I've been stumped at this, and I need help badly! Thanks so much for any advice.

    Read the article

  • Retrofit Certification

    - by Bill Evjen
    Impact of Regulations on Cabin Systems Installation John Courtright, Structural Integrity Engineering There are “heightened” FAA attention to technical issues related to IFE and Wi-Fi Systems Installations The Aging Aircraft Safety Rule – EWIS & Damage Tolerance Analysis The Challenge: Maximize Flight Safety While Minimizing Costs Issue Papers & Testing, Testing, Testing The role of Airworthiness Directives (ADs) on the design of many IFE systems and all antenna systems. Goal is safety AND cost-effective maintenance intervals and inspection techniques The STC Process Briefly Stated Type Certifications (TC) Supplemental Type Certifications (STC) The STC Process Project Specific Certification Plan (PSCP) Managed by FAA Aircraft Certification Office (ACO) Type of Project (Electrical/Mechanical Systems or Structural) Specific Type of Aircraft Being Modified Schedule Design & Installation Location What does the STC Plan (PSCP) Cover? System Description – What does the system do? System qualification – Are the components qualified? Certification requirements – What FARs are applicable? Installation detail – what is being modified? Prototype installation – What is new? Functional hazard Assessment (FHA) – is it safe? EZAP-EWIS Requirements – Any aging aircraft issues? Certification Data – How is compliance achieved? Delegation and FAA involvement – Who is doing the work? Proposed certification schedule – When is the installation? Certification documentation – What the FAA Expects to see Cabin Systems Certification Concerns In addition to meeting the requirements for DO-160, Cabin System Certification needs to address issues related to: Power management: Generally, IFE and Wi-Fi Systems are classified as “Non-Essential Equipment” from a certification viewpoint. Connected to “non-essential” power buses Must be able to shed IFE & Wi-Fi Systems in a smoke/fire event or Other electrical emergency (FAA Policy 00-111-160) FAA is more relaxed with testing wi-fi. It used to be that you had to have 150 seats with laptops running wi-fi, but now it is down to around 50. Aging aircraft concerns – electrical and structural Issue papers addressing technical concerns involving: “Structural Certification Criteria for Large Antenna Installations” Antenna “Vibration/Buffeting Compliance Criteria” DO-160 : Environmental Test Procedures DO 160 – “Environmental Conditions and Test Procedures for Airborne Equipment”, Issued by RTCA Provides guidance to equipment manufacturers as to testing requirements Temperature: –40C to +55C Vibration and Shock Contaminant susceptibility – fluids and dust Electro-magnetic Interference Cabin systems are generally classified as “non-essential” Swissair 111 crashed (in part) due to non-standard wiring practices. EWIS Design Implications Installation design must take EWIS Requirements into account. This generally means: Aircraft surveys are needed to identify proper wire routing Ensure existing wiring diagrams are correct Identify primary/Secondary/Tertiary bus locations Verify proper separation of wire bundles exist Required separation from fuel quantity indicator system (FQIS) to prevent fuel tang ignition Enhanced Zonal Analysis Procedure (EZAP) Performed EZAP was developed by the Aging Transport Systems Rulemaking Advisory Committee (ATSRAC) EZAP is the method for analyzing airplane zones with an emphasis on evaluating wiring systems and the existence of combustibles  in the cabin. Certification Considerations for Wi-Fi Systems Electrical – All existing DO 160 testing required Issue papers required Onboard EMI testing – any interference with aircraft systems when multiple wi-fi users are logged on? Vibration/Buffeting compliance criteria – what is the effect of the antenna on aircraft flight characteristics? Structural certification criteria – what are the stress loads on the aircraft at the antenna location and what is the impact on maintenance inspection criteria for the airline? Damage tolerance analysis required Goal – minimize maintenance inspection intervals

    Read the article

  • FBX Importer - Texture Name

    - by CmasterG
    I have a problem with the FBX SDK. I read in the data for the vertex position and the uv coordinates. It works fine, but now I want to read for each polygon to which texture it belongs, so that I can have models with multiple textures. Can anyone tell me how I can get the texture name (file name) for my polygon. My code to read in vertex position and uv coordinates is the following: int i, j, lPolygonCount = pMesh->GetPolygonCount(); FbxVector4* lControlPoints = pMesh->GetControlPoints(); int vertexId = 0; for (i = 0; i < lPolygonCount; i++) { int lPolygonSize = pMesh->GetPolygonSize(i); for (j = 0; j < lPolygonSize; j++) { int lControlPointIndex = pMesh->GetPolygonVertex(i, j); FbxVector4 pos = lControlPoints[lControlPointIndex]; current_model[vertex_index].x = pos.mData[0] - pivot_offset[0]; current_model[vertex_index].y = pos.mData[1] - pivot_offset[1]; current_model[vertex_index].z = pos.mData[2]- pivot_offset[2]; FbxVector4 vertex_normal; pMesh->GetPolygonVertexNormal(i,j, vertex_normal); current_model[vertex_index].nx = vertex_normal.mData[0]; current_model[vertex_index].ny = vertex_normal.mData[1]; current_model[vertex_index].nz = vertex_normal.mData[2]; //read in UV data FbxStringList lUVSetNameList; pMesh->GetUVSetNames(lUVSetNameList); //get lUVSetIndex-th uv set const char* lUVSetName = lUVSetNameList.GetStringAt(0); const FbxGeometryElementUV* lUVElement = pMesh->GetElementUV(lUVSetName); if(!lUVElement) continue; // only support mapping mode eByPolygonVertex and eByControlPoint if( lUVElement->GetMappingMode() != FbxGeometryElement::eByPolygonVertex && lUVElement->GetMappingMode() != FbxGeometryElement::eByControlPoint ) return; //index array, where holds the index referenced to the uv data const bool lUseIndex = lUVElement->GetReferenceMode() != FbxGeometryElement::eDirect; const int lIndexCount= (lUseIndex) ? lUVElement->GetIndexArray().GetCount() : 0; FbxVector2 lUVValue; //get the index of the current vertex in control points array int lPolyVertIndex = pMesh->GetPolygonVertex(i,j); //the UV index depends on the reference mode //int lUVIndex = lUseIndex ? lUVElement->GetIndexArray().GetAt(lPolyVertIndex) : lPolyVertIndex; int lUVIndex = pMesh->GetTextureUVIndex(i, j); lUVValue = lUVElement->GetDirectArray().GetAt(lUVIndex); current_model[vertex_index].tu = (float)lUVValue.mData[0]; current_model[vertex_index].tv = (float)lUVValue.mData[1]; vertex_index ++; } } float v1[3], v2[3], v3[3]; v1[0] = current_model[vertex_index - 3].x; v1[1] = current_model[vertex_index - 3].y; v1[2] = current_model[vertex_index - 3].z; v2[0] = current_model[vertex_index - 2].x; v2[1] = current_model[vertex_index - 2].y; v2[2] = current_model[vertex_index - 2].z; v3[0] = current_model[vertex_index - 1].x; v3[1] = current_model[vertex_index - 1].y; v3[2] = current_model[vertex_index - 1].z; collision_model->addTriangle(v1,v2,v3);

    Read the article

  • NHibernate Pitfalls: Custom Types and Detecting Changes

    - by Ricardo Peres
    This is part of a series of posts about NHibernate Pitfalls. See the entire collection here. NHibernate supports the declaration of properties of user-defined types, that is, not entities, collections or primitive types. These are used for mapping a database columns, of any type, into a different type, which may not even be an entity; think, for example, of a custom user type that converts a BLOB column into an Image. User types must implement interface NHibernate.UserTypes.IUserType. This interface specifies an Equals method that is used for comparing two instances of the user type. If this method returns false, the entity is marked as dirty, and, when the session is flushed, will trigger an UPDATE. So, in your custom user type, you must implement this carefully so that it is not mistakenly considered changed. For example, you can cache the original column value inside of it, and compare it with the one in the other instance. Let’s see an example implementation of a custom user type that converts a Byte[] from a BLOB column into an Image: 1: [Serializable] 2: public sealed class ImageUserType : IUserType 3: { 4: private Byte[] data = null; 5: 6: public ImageUserType() 7: { 8: this.ImageFormat = ImageFormat.Png; 9: } 10: 11: public ImageFormat ImageFormat 12: { 13: get; 14: set; 15: } 16: 17: public Boolean IsMutable 18: { 19: get 20: { 21: return (true); 22: } 23: } 24: 25: public Object Assemble(Object cached, Object owner) 26: { 27: return (cached); 28: } 29: 30: public Object DeepCopy(Object value) 31: { 32: return (value); 33: } 34: 35: public Object Disassemble(Object value) 36: { 37: return (value); 38: } 39: 40: public new Boolean Equals(Object x, Object y) 41: { 42: return (Object.Equals(x, y)); 43: } 44: 45: public Int32 GetHashCode(Object x) 46: { 47: return ((x != null) ? x.GetHashCode() : 0); 48: } 49: 50: public override Int32 GetHashCode() 51: { 52: return ((this.data != null) ? this.data.GetHashCode() : 0); 53: } 54: 55: public override Boolean Equals(Object obj) 56: { 57: ImageUserType other = obj as ImageUserType; 58: 59: if (other == null) 60: { 61: return (false); 62: } 63: 64: if (Object.ReferenceEquals(this, other) == true) 65: { 66: return (true); 67: } 68: 69: return (this.data.SequenceEqual(other.data)); 70: } 71: 72: public Object NullSafeGet(IDataReader rs, String[] names, Object owner) 73: { 74: Int32 index = rs.GetOrdinal(names[0]); 75: Byte[] data = rs.GetValue(index) as Byte[]; 76: 77: this.data = data as Byte[]; 78: 79: if (data == null) 80: { 81: return (null); 82: } 83: 84: using (MemoryStream stream = new MemoryStream(this.data ?? new Byte[0])) 85: { 86: return (Image.FromStream(stream)); 87: } 88: } 89: 90: public void NullSafeSet(IDbCommand cmd, Object value, Int32 index) 91: { 92: if (value != null) 93: { 94: Image data = value as Image; 95: 96: using (MemoryStream stream = new MemoryStream()) 97: { 98: data.Save(stream, this.ImageFormat); 99: value = stream.ToArray(); 100: } 101: } 102: 103: (cmd.Parameters[index] as DbParameter).Value = value ?? DBNull.Value; 104: } 105: 106: public Object Replace(Object original, Object target, Object owner) 107: { 108: return (original); 109: } 110: 111: public Type ReturnedType 112: { 113: get 114: { 115: return (typeof(Image)); 116: } 117: } 118: 119: public SqlType[] SqlTypes 120: { 121: get 122: { 123: return (new SqlType[] { new SqlType(DbType.Binary) }); 124: } 125: } 126: } In this case, we need to cache the original Byte[] data because it’s not easy to compare two Image instances, unless, of course, they are the same.

    Read the article

  • Pro/con of using Angular directives for complex form validation/ GUI manipulation

    - by tengen
    I am building a new SPA front end to replace an existing enterprise's legacy hodgepodge of systems that are outdated and in need of updating. I am new to angular, and wanted to see if the community could give me some perspective. I'll state my problem, and then ask my question. I have to generate several series of check boxes based on data from a .js include, with data like this: $scope.fieldMappings.investmentObjectiveMap = [ {'id':"CAPITAL PRESERVATION", 'name':"Capital Preservation"}, {'id':"STABLE", 'name':"Moderate"}, {'id':"BALANCED", 'name':"Moderate Growth"}, // etc {'id':"NONE", 'name':"None"} ]; The checkboxes are created using an ng-repeat, like this: <div ng-repeat="investmentObjective in fieldMappings.investmentObjectiveMap"> ... </div> However, I needed the values represented by the checkboxes to map to a different model (not just 2-way-bound to the fieldmappings object). To accomplish this, I created a directive, which accepts a destination array destarray which is eventually mapped to the model. I also know I need to handle some very specific gui controls, such as unchecking "None" if anything else gets checked, or checking "None" if everything else gets unchecked. Also, "None" won't be an option in every group of checkboxes, so the directive needs to be generic enough to accept a validation function that can fiddle with the checked state of the checkbox group's inputs based on what's already clicked, but smart enough not to break if there is no option called "NONE". I started to do that by adding an ng-click which invoked a function in the controller, but in looking around Stack Overflow, I read people saying that its bad to put DOM manipulation code inside your controller - it should go in directives. So do I need another directive? So far: (html): <input my-checkbox-group type="checkbox" fieldobj="investmentObjective" ng-click="validationfunc()" validationfunc="clearOnNone()" destarray="investor.investmentObjective" /> Directive code: .directive("myCheckboxGroup", function () { return { restrict: "A", scope: { destarray: "=", // the source of all the checkbox values fieldobj: "=", // the array the values came from validationfunc: "&" // the function to be called for validation (optional) }, link: function (scope, elem, attrs) { if (scope.destarray.indexOf(scope.fieldobj.id) !== -1) { elem[0].checked = true; } elem.bind('click', function () { var index = scope.destarray.indexOf(scope.fieldobj.id); if (elem[0].checked) { if (index === -1) { scope.destarray.push(scope.fieldobj.id); } } else { if (index !== -1) { scope.destarray.splice(index, 1); } } }); } }; }) .js controller snippet: .controller( 'SuitabilityCtrl', ['$scope', function ( $scope ) { $scope.clearOnNone = function() { // naughty jQuery DOM manipulation code that // looks at checkboxes and checks/unchecks as needed }; The above code is done and works fine, except the naughty jquery code in clearOnNone(), which is why I wrote this question. And here is my question: after ALL this, I think to myself - I could be done already if I just manually handled all this GUI logic and validation junk with jQuery written in my controller. At what point does it become foolish to write these complicated directives that future developers will have to puzzle over more than if I had just written jQuery code that 99% of us would understand with a glance? How do other developers draw the line? I see this all over Stack Overflow. For example, this question seems like it could be answered with a dozen lines of straightforward jQuery, yet he has opted to do it the angular way, with a directive and a partial... it seems like a lot of work for a simple problem. Specifically, I suppose I would like to know: how SHOULD I be writing the code that checks whether "None" has been selected (if it exists as an option in this group of checkboxes), and then check/uncheck the other boxes accordingly? A more complex directive? I can't believe I'm the only developer that is having to implement code that is more complex than needed just to satisfy an opinionated framework.

    Read the article

< Previous Page | 210 211 212 213 214 215 216 217 218 219 220 221  | Next Page >