Search Results

Search found 22900 results on 916 pages for 'pascal case'.

Page 111/916 | < Previous Page | 107 108 109 110 111 112 113 114 115 116 117 118  | Next Page >

• C++ Dynamic object construction

  - by Rajesh Subramanian

  I have a base class, class Msg { ParseMsg() { ParseMsgData(); ParseTrailer(); } virtual void ParseMsgData() = 0; ParseTrailer(); }; and derived classes, class InitiateMsg { void ParseMsgData() { ... } }; class ReadOperationMsg { void ParseMsgData() { ... } }; class WriteOperationMsg { void ParseMsgData() { ... } }; and the scenario is below, void UsageFunction(string data) { Msg* msg = ParseHeader(data); ParseMsg } Msg* ParseHeader(string data) { Msg *msg = NULL; .... switch() { case 1: msg = new InitiateMsg(); break; case 2: msg = new ReadOperationMsg{(); break; case 3: msg = new WriteOperationMsg{(); break; .... } return msg; } based on the data ParseHeader method will decide which object has to be created, So I have implemented ParseHeader function outside the class where I am using. How can I make the ParseHeader function inside the Msg class and then use it? In C# the same is achieved by defining ParseHeader method as static with in class and use it from outside,

  Read the article

• In Xlib, How can I animate until an event occurs?

  - by Animateur

  Hi, I've been trying to animate in a C program using Xlib and I wanna do something when an event occurs, otherwise I wanna keep animating. Here's an example code snippet of what I am doing currently: while( 1 ) { // If an event occurs, stop and do whatever is needed. // If no event occurs, skip this if statement. if ( XEventsQueued( display, QueuedAlready ) > 0 ) { XNextEvent( display, &event ) switch ( event.type ) { // Don't do anything case Expose: while ( event.xexpose.count != 0 ) break; // Do something, when a button is pressed case ButtonPress: ... break; // Do something, when a key is pressed case KeyPress: ... break; } } animate(); // Do animation step i.e. change any drawings... repaint(); // Paint again with the new changes from animation... } So basically, I wanna keep looping if the user hasn't clicked the mouse OR pressed a key in the keyboard yet. When the user presses a key OR clicks the mouse, I wanna stop and do a specific action. The problem in my above code is that, it doesnt stop whenever I do an action. If I remove the if statement, the animation blocks until an event occurs, however I do not want this. It's a simple problem, but I'm kinda new to Xlib/animations so any help would be highly appreciated. Thanks.

  Read the article

• How to reset keyboard for an entry field?

  - by David.Chu.ca

  I am using tag field as a flag for text fields text view fields for auto-jumping to the next field: - (BOOL)findNextEntryFieldAsResponder:(UIControl *)field { BOOL retVal = NO; for (UIView* aView in mEntryFields) { if (aView.tag == (field.tag + 1)) { [aView becomeFirstResponder]; retVal = YES; break; } } return retVal; } It works fine in terms of auto-jumping to the next field when Next key is pressed. However, my case is that the keyboards are different some fields. For example, one fields is numeric & punctuation, and the next one is default (alphabetic keys). For the numeric & punctuation keyboard is OK, but the next field will stay as the same layout. It requires user to press 123 to go back ABC keyboard. I am not sure if there is any way to reset the keyboard for a field as its keyboard defined in xib? Not sure if there is any APIs available? I guess I have to do something is the following delegate? -(void)textFieldDidBegingEditing:(UITextField*) textField { // reset to the keyboard to request specific keyboard view? .... } OK. I found a solution close to my case by slatvik: -(void) textFieldDidBeginEditing:(UITextField*) textField { textField.keyboardType = UIKeybardTypeAlphabet; } However, in the case of the previous text fields is numeric, the keyboard stays numeric when auto-jumped to the next field. Is there any way to set keyboard to alphabet mode?

  Read the article

• MySQL subquery and bracketing

  - by text

  Here are my tables respondents: field sample value respondentid : 1 age : 2 gender : male survey_questions: id : 1 question : Q1 answer : sample answer answers: respondentid : 1 question : Q1 answer : 1 --id of survey question I want to display all respondents who answered the certain survey, display all answers and total all the answer and group them according to the age bracket. I tried using this query: SELECT res.Age, res.Gender, answer.id, answer.respondentid, SUM(CASE WHEN res.Gender='Male' THEN 1 else 0 END) AS males, SUM(CASE WHEN res.Gender='Female' THEN 1 else 0 END) AS females, CASE WHEN res.Age < 1 THEN 'age1' WHEN res.Age BETWEEN 1 AND 4 THEN 'age2' WHEN res.Age BETWEEN 4 AND 9 THEN 'age3' WHEN res.Age BETWEEN 10 AND 14 THEN 'age4' WHEN res.Age BETWEEN 15 AND 19 THEN 'age5' WHEN res.Age BETWEEN 20 AND 29 THEN 'age6' WHEN res.Age BETWEEN 30 AND 39 THEN 'age7' WHEN res.Age BETWEEN 40 AND 49 THEN 'age8' ELSE 'age9' END AS ageband FROM Respondents AS res INNER JOIN Answers as answer ON answer.respondentid=res.respondentid INNER JOIN Questions as question ON answer.Answer=question.id WHERE answer.Question='Q1' GROUP BY ageband ORDER BY res.Age ASC I was able to get the data but the listing of all answers are not present. Do I have to subquery SELECT into my current SELECT statement to show the answers? I want to produce something like this: ex: # of Respondents is 3 ages: 2,3 and 6 Question: what are your favorite subjects? Ages 1-4: subject 1: 1 subject 2: 2 subject 3: 2 total respondents for ages 1-4 : 2 Ages 5-10: subject 1: 1 subject 2: 1 subject 3: 0 total respondents for ages 5-10 : 1

  Read the article

• Response Redirect - Open Link in New Window

  - by bacis09

  First, I've taken the time to review this question which seems to be the most similar, however, the solution that seems to have been selected will not work in my scenario.Not to mention I worry about some of the comments claiming it to be brittle or an inadequate solution. http://stackoverflow.com/questions/104601/asp-net-response-redirect-to-new-window -We have an XML document which basically contains all of the information for a Side menu. -We have numerous URLS which are stored in a constants class. -One of the elements in a string of XML (well call it label) is used to determine if the menu item is created as a LinkButton or a Label. -Links use a custom user control that is used standard for all links across the application (why suggestion on similar thread doesn't work - I don't want all links to open in a new window - just one) -One of the elements in a string of XML (well call it function) is used in a Switch statement to generate our links using Response Redirect. It may look something like this. switch (function) { case goto 1: string url; if (user_group == 1) { url = Constants.CONSTANT1 } else { url = Constants.CONSTANT2 } Response.Redirect(url) case goto 2: ...... default: ...... break; } Given this scenario, I'm trying to find the best way to quickly open a New Window, when a specific case in this switch statement is met. Can it be done with Response Redirect (this seems to be arguable - people say no it can't, yet other people say they have made it work)? If not, what alternative can work here?

  Read the article

• Display TableViews corresponding to segmentedControl in a single tableview without pushing a new view

  - by user1727927

  I have a tableViewController where I have used the Interface Builder to insert a Segmented Controller having two segments. Since by default, first segment is always selected, I am not facing any problem in displaying the tableview corresponding to first segment. However, when I click on the second segment, I want to display another tableView. Here goes the problem. I am calling newTableViewController class on clicking the second segment. Hence, this view is getting pushed instead. Please suggest me a method to have these two tableViews in the main tableView upon clicking the segments. I have used switch case for switching between the segments. Here's the relevant part of the code: This method is in the FirstTableViewController since first segment is by default selected. -(IBAction) segmentedControlChanged { switch(segmentedControl.selectedSegmnentIndex) { case 0: //default first index selected. [tableView setHidden:NO]; break; case 1: NewViewController *controller=[[NewViewController alloc] initWithNibName:@"NewViewController" bundle:nil]; self.navigationController pushViewController:controller animated:YES]; [controller release]; break; default: break; } }

  Read the article

• What is the n in O(n) when comparing sorting algorithms?

  - by Mumfi

  The question is rather simple, but I just can't find a good enough answer. I've taken a look at the most upvoted question regarding the Big-Oh notation, namely this: Plain English explanation of Big O It says there that: For example, sorting algorithms are typically compared based on comparison operations (comparing two nodes to determine their relative ordering). Now let's consider the simple bubble sort algorithm: for (int i = arr.length - 1; i > 0 ; i--) { for (int j = 0; j<i; j++) { if (arr[j] > arr[j+1]) { switchPlaces(...) } } } I know that worst case is O(n^2) and best case is O(n), but what is n exactly? If we attempt to sort an already sorted algorithm (best case), we would end up doing nothing, so why is it still O(n)? We are looping through 2 for-loops still, so if anything it should be O(n^2). n can't be the number of comparison operations, because we still compare all the elements, right? This confuses me, and I appreciate if someone could help me.

  Read the article

• Need help: input int from console and pass it into method in different class and do math

  - by christophe

  i'm a beginner, Need help, Please!!! I want to read optional number "a" from console and then store it in variable to use as passing to a different class (different .java file). and pint the sum separetely by optional inputting. How do i code the 2 classes? thanks /* * DemoApp.java */ public class DemoApp { public static void main(String[] args) { Scanner input = new Scanner(System.in); int a; System.out.println("Input one of the following 3 numbers: 100, 200, 300"); System.out.print("Enter: "); a = input.nextInt(); TestApplication testapp = new TestApplication(); testapp.test(a); } } /* * TestApplication.java * */ public class TestApplication { private int a; public void test(int a) { this.a = a; // TODO: where to get the "a"? (entered by users from console) System.out.println("The number_a was passed in: "+a); } protected void printNum() throws Exception { int num; switch (a) { case 100: num = num + 10; break; case 200: num = num + 20; break; case 300: num = num + 30; break; default: // TODO: unexpected number input. throw(); break; } System.out.println("I got a sum number"+num); } }

  Read the article

• C++: Switch statement within while loop?

  - by Jason

  I just started C++ but have some prior knowledge to other languages (vb awhile back unfortunately), but have an odd predicament. I disliked using so many IF statements and wanted to use switch/cases as it seemed cleaner, and I wanted to get in the practice.. But.. Lets say I have the following scenario (theorietical code): while(1) { //Loop can be conditional or 1, I use it alot, for example in my game char something; std::cout << "Enter something\n -->"; std::cin >> something; //Switch to read "something" switch(something) { case 'a': cout << "You entered A, which is correct"; break; case 'b': cout << "..."; break; } } And that's my problem. Lets say I wanted to exit the WHILE loop, It'd require two break statements? This obviously looks wrong: case 'a': cout << "You entered A, which is correct"; break; break; So can I only do an IF statement on the 'a' to use break;? Am I missing something really simple? This would solve a lot of my problems that I have right now.

  Read the article

• Something wrong on my very first LINQ to SQL c # code

  - by user334813

  using System; using System.Collections.Generic; using System.ComponentModel; using System.Data; using System.Drawing; using System.Linq; using System.Text; using System.Windows.Forms; namespace Advanced_LinQ_Query { public partial class Form1 : Form { public Form1() { InitializeComponent(); } private DataClasses1DataContext database = new DataClasses1DataContext(); private void Form1_Load(object sender, EventArgs e) { database.Log= Console.Out; comboBox.SelectedIndex=0; } private void titleBindingNavigatorSaveItem_Click(object sender, EventArgs e) { Validate(); titleBindingSource.EndEdit(); database.SubmitChanges(); comboBox.SelectedIndex=0; } private void comboBox_SelectedIndexChanged(object sender, EventArgs e) { switch (comboBox.SelectedIndex) { case 0: titleBindingSource.DataSource = from Title in database.Titles orderby Title.BookTitle select Title; break; case 1: titleBindingSource.DataSource = from Title in database.Titles where Title.Copyright == "2008" orderby Title.BookTitle select Title; break; case 2: titleBindingSource.DataSource = from Title in database.Titles where Title.BookTitle.EndsWith("How to Program") orderby Title.BookTitle select Title; break; } titleBindingSource.MoveFirst(); } } } no connection seems to built after debugging between Title table in my database (book.mdf) and titleBindingSource! Where is the problem?

  Read the article

• How grouping and totaling are done into three tables using JOIN

  - by text

  Here are my tables respondents: field sample value respondentid : 1 age : 2 gender : male survey_questions: id : 1 question : Q1 answer : sample answer answers: respondentid : 1 question : Q1 answer : 1 --id of survey question I want to display all respondents who answered the certain survey, display all answers and total all the answer and group them according to the age bracket. I tried using this query: $sql = "SELECT res.Age, res.Gender, answer.id, answer.respondentid, SUM(CASE WHEN res.Gender='Male' THEN 1 else 0 END) AS males, SUM(CASE WHEN res.Gender='Female' THEN 1 else 0 END) AS females, CASE WHEN res.Age < 1 THEN 'age1' WHEN res.Age BETWEEN 1 AND 4 THEN 'age2' WHEN res.Age BETWEEN 4 AND 9 THEN 'age3' WHEN res.Age BETWEEN 10 AND 14 THEN 'age4' WHEN res.Age BETWEEN 15 AND 19 THEN 'age5' WHEN res.Age BETWEEN 20 AND 29 THEN 'age6' WHEN res.Age BETWEEN 30 AND 39 THEN 'age7' WHEN res.Age BETWEEN 40 AND 49 THEN 'age8' ELSE 'age9' END AS ageband FROM Respondents AS res INNER JOIN Answers as answer ON answer.respondentid=res.respondentid INNER JOIN Questions as question ON answer.Answer=question.id WHERE answer.Question='Q1' GROUP BY ageband ORDER BY res.Age ASC"; I was able to get the data but the listing of all answers are not present. What's wrong with my query. I want to produce something like this: ex: # of Respondents is 3 ages: 2,3 and 6 Question: what are your favorite subjects? Ages 1-4: subject 1: 1 subject 2: 2 subject 3: 2 total respondents for ages 1-4 : 2 Ages 5-10: subject 1: 1 subject 2: 1 subject 3: 0 total respondents for ages 5-10 : 1

  Read the article

• Associate activity with database ID

  - by Mohit Deshpande

  I have a main ListView that is based on an adapter from my database. Each database id is "assigned" to an Activity via the ListView. And in my AndroidManifest, each activity has an intent filter with a custom action. Now with this, I have had to create this class: public final class ActivityLauncher { private ActivityLauncher() { } public static void launch(Context c, int id) { switch(id) { case 1: Intent intent = new Intent(); intent.setAction(SomeActivity.ACTION_SOMEACTIVITY); c.startActivity(intent); break; case 2: ... break; ... } } private static void st(Context context, String action) { Intent intent = new Intent(); intent.setAction(action); context.startActivity(intent); } } So I have to manually create a new case for the switch statement. This would get troublesome if I have to rearrange or delete an id. Is there any way to get around this?

  Read the article

• Passing a string to a function in C++

  - by Chef Flambe

  I want to pass a string like "Celcius" into a function that I have but I keep getting errors tossed back at me from the Function. System::Console::WriteLine' : none of the 19 overloads could convert all the argument types I figure I just have something simple wrong. Can someone point out my mistake please? Using MS Visual C++ 2010 I've posted the offending code. The other functions (not posted) work fine. void PrintResult( double result, std::string sType ); // Print result and string // to the console //============================================================================================= // start of main //============================================================================================= void main( void ) { ConsoleKeyInfo CFM; // Program Title and Description ProgramDescription(); // Menu Selection and calls to data retrieval/calculation/result Print CFM=ChooseFromMenu(); switch(CFM.KeyChar) // ************************************************************ { //* case '1' : PrintResult(F2C(GetTemperature()),"Celsius"); //* break; //* //* case '2' : PrintResult(C2F(GetTemperature()),"Fahrenheit"); //* break; //* //* default : Console::Write("\n\nSwitch : Case !!!FAILURE!!!"); //* } //************************************************************ system("pause"); return; } //Function void PrintResult( double result, std::string sType ) { Console::WriteLine("\n\nThe converted temperature is {0:F2} degrees {1}\n\n",result,sType); return; }

  Read the article

• NHibernate Native SQL multiple joins

  - by Chris

  Hi all, I"m having some problems with Nhibernate and native sql. I've got an entity with alot of collections and I am doing an SQL Fulltext search on it. So when returning 100 or so entities, I dont want all collections be lazy loaded. For this I changed my SQL query: SELECT Query.* FROM (SELECT {spr.*}, {adr.*}, {adrt.*}, {cty.*}, {com.*}, {comt.*}, spft.[Rank] AS [Rak], Row_number() OVER(ORDER BY spft.[Rank] DESC) AS rownum FROM customer spr INNER JOIN CONTAINSTABLE ( customerfulltext , computedfulltextindex , '" + parsedSearchTerm + @"' ) AS spft ON spr.customerid = spft.[Key] LEFT JOIN [Address] adr ON adr.customerid = spr.customerid INNER JOIN [AddressType] adrt ON adrt.addresstypeid = adr.addresstypeid INNER JOIN [City] cty ON cty.cityid = adr.cityid LEFT JOIN [Communication] com ON com.customerid = spr.customerid INNER JOIN [CommunicationType] comt ON comt.communicationtypeid = com.communicationtypeid) as Query ORDER BY Query.[Rank] DESC This is how I setup the query: var items = GetCurrentSession() .CreateSQLQuery(query) .AddEntity("spr", typeof(Customer)) .AddJoin("adr", "spr.addresses") .AddJoin("adrt", "adr.Type") .AddJoin("cty", "adr.City") .AddJoin("com", "spr.communicationItems") .AddJoin("comt", "com.Type") .List<Customer>(); What happens now is, that the query returns customers twice (or more), I assume this is because of the joins since for each customer address, communicationItem (e.g. phone, email), a new sql row is returned. In this case I thought I could use the DistinctRootEntityResultTransformer. var items = GetCurrentSession() .CreateSQLQuery(query) .AddEntity("spr", typeof(Customer)) .AddJoin("adr", "spr.addresses") .AddJoin("adrt", "adr.Type") .AddJoin("cty", "adr.City") .AddJoin("com", "spr.communicationItems") .AddJoin("comt", "com.Type") .SetResultTransformer(new DistinctRootEntityResultTransformer()) .List<Customer>(); Doing so an exception is thrown. This is because I try to list customers .List<Customer>() but the transformer returns only entities of the last join added. E.g. in the case above, the entity with alias "comt" is returned when doing .List() instead of .List(). If I would switch last join with the join alias "cty", then the transformer returns a list of cities only... Anyone knows how I can return a clean list of customers in this case?

  Read the article

• SDK2 query for counting: which is more efficient?

  - by user1195996

  I have an app that is displaying metrics about defects in a project. I have the option of making one query that returns all the defects, and from that I can break out about four different metrics (How many defects escaped QA in 90 days, 180 days, and then the same metrics again but only counting sev1/sev2 defects). I could make four queries and limit the results to one so that I just get a count for each. Or I could make one query that encompass them all (all defects that escaped QA in 180 days) and then count up the difference. I'm figuring worst case, the number of defects that escaped QA in the last six months will generally be less than 100, certainly less 500 worst case. Which would you do-- four queryies with one result each, or one single query that on average might return 50, perhaps worst case 500? And I guess the key question is-- where are the inflections points? Perhaps I have more metrics tomorrow (who knows, 8?) and a different average defect counts. Is there a rule of thumb I could use to help choose which approach?

  Read the article

• Help with enum values in registry c++

  - by vBx

  DWORD type = REG_NONE; int i = 0; size = sizeof(ValueName); size2 = sizeof(ValueData); BOOL bContinue = TRUE; do { lRet = RegEnumValue(Hkey , i , ValueName , &size , 0 , &type , ValueData , &size2); switch(lRet) { case ERROR_SUCCESS: print_values(ValueName , type , ValueData , size2); i++; size = sizeof(ValueName); size2 = sizeof(ValueData); break; case ERROR_MORE_DATA: size2 = sizeof(ValueData); if(NULL != ValueData) delete [] ValueData; ValueData = new BYTE[size2]; break; case ERROR_NO_MORE_ITEMS: bContinue = false; break; default: cout << "Unexpected error: " << GetLastError() << endl; bContinue = false; break; } }while(bContinue); it always goes to ERROR_NO_MORE_DATA ,why is that ? :-/

  Read the article

• Parallelizing for loop

  - by vman049

  I have MATLAB code which I'm trying to parallelize with a simple change from "for" to "parfor." I'm unable to do so because of an error I'm receiving on the variable "votes" which states: Valid indices for 'votes' are restricted in PARFOR loops. Explanation: For MATLAB to execute parfor loops efficiently, the amount of data sent to the MATLAB workers must be minimal. One of the ways MATLAB achieves this is by restricting the way variables can be indexed in parfor iterations. The indicated variable is indexed in a way that is incompatible with parfor. Suggested Action: Fix the indexing. For a description of the indexing restrictions, see “Sliced Variables” in the Parallel Computing Toolbox documentation. Below is my code: votes = zeros(num_layers, size(spikes, 1), size(SVMs_layer1, 1)); predDir = zeros(size(spikes, 1), 1); chronProb = zeros([num_layers, size(chronDists)]); for i = 1:num_layers switch i case 1 B = B1; k_elem_temp = k_elem1; rest_elem_temp = rest_elem1; case 2 B = B2; k_elem_temp = k_elem2; rest_elem_temp = rest_elem2; case 3 B = B3; k_elem_temp = k_elem3; rest_elem_temp = rest_elem3; end for j = 1:length(chronPred) if chronDists(i, j, :) ~= 0 parfor k = 1:8 chronProb(i, j, k) = logistic(B{k}(1) + chronDists(i, j, k).*(B{k}(2))); votes(i, j, k_elem_temp(k, :)) = votes(i, j, k_elem_temp(k, :)) + chronProb(i, j, k)/num_k(i)/num_layers; votes(i, j, rest_elem_temp(k, :)) = votes(i, j, rest_elem_temp(k, :)) + (1 - chronProb(i, j, k))/num_rest(i)/num_layers; end end end end Do you have any suggestions as to how I could adjust my code so that it runs in parallel? Thank you!

  Read the article

• How can I combine a LINQ query with an IQueryable<Guid>

  - by John

  I have a LINQ query that uses 1 table + a large number of views. I'd like to be able to write something like this: IQueryable<Guid> mostViewedWriters; switch (datePicker) { case DatePicker.Last12Hours: mostViewedWriters = from x in context.tempMostViewed12Hours select x.GuidId; break; case DatePicker.Last24Hours: mostViewedWriters = from x in context.tempMostViewed12Hours select x.GuidId; break; case DatePicker.Last36Hours: mostViewedWriters = from x in context.tempMostViewed12Hours select x.GuidId; break; } var query = from x1 in context.Articles join x2 in context.Authors on x1.AuthorId == x2.AuthorId join x3 in mostViewedWriters on x2.AuthorId == x3.Id select new { x2.AuthorName, x1.ArticleId, x1.ArticleTitle }; The above C# is pseudo-code written to protect the innocent (me). The gist of the question is this: I have a query that is related to the results of a view. That view, however, could be one of many different views. All the views return the same data type. I thought that I might be able to create an IQueryable that would contain the Ids that I need and use that query. Alas, that effort has stalled.

  Read the article

• difficulties in javascript coding [on hold]

  - by user3718986

  Question is, It takes me 3 hours to fly from NY to CA. How much it will take me if I fly directly from NY to Florida. suppose that distance from CA to Flordia is 8 hours by air..Rule is if you suppose to travel from NY to Florida you will have to pass through CA. I did the quesiton in JavaScript below but coding isn't correct. Can someone fixed this issue for me please? var destination = prompt('Please enter your destinations. We are currently flying to NY,CA and FL'); var locatioon = prompt("specify your current location. "); switch (destination) { case 'NY': { distanceTeller(locatioon); break; } break; case 'CA': { distanceTeller(locatioon); break; } break; case 'FL': { alert("11 HR"); } default: alert('dont look at me'); break; } function distanceTeller(locatioon) { if (locatioon == 'CA') { alert('it will take you 3 hours'); } else if (locatioon == 'FL') { alert('it will take you 8 hours'); } else alert('it will take you 11 hours to reach NY'); }

  Read the article

• Using FiddlerCore to capture HTTP Requests with .NET

  - by Rick Strahl

  Over the last few weeks I’ve been working on my Web load testing utility West Wind WebSurge. One of the key components of a load testing tool is the ability to capture URLs effectively so that you can play them back later under load. One of the options in WebSurge for capturing URLs is to use its built-in capture tool which acts as an HTTP proxy to capture any HTTP and HTTPS traffic from most Windows HTTP clients, including Web Browsers as well as standalone Windows applications and services. To make this happen, I used Eric Lawrence’s awesome FiddlerCore library, which provides most of the functionality of his desktop Fiddler application, all rolled into an easy to use library that you can plug into your own applications. FiddlerCore makes it almost too easy to capture HTTP content! For WebSurge I needed to capture all HTTP traffic in order to capture the full HTTP request – URL, headers and any content posted by the client. The result of what I ended up creating is this semi-generic capture form: In this post I’m going to demonstrate how easy it is to use FiddlerCore to build this HTTP Capture Form.  If you want to jump right in here are the links to get Telerik’s Fiddler Core and the code for the demo provided here. FiddlerCore Download FiddlerCore on NuGet Show me the Code (WebSurge Integration code from GitHub) Download the WinForms Sample Form West Wind Web Surge (example implementation in live app) Note that FiddlerCore is bound by a license for commercial usage – see license.txt in the FiddlerCore distribution for details. Integrating FiddlerCore FiddlerCore is a library that simply plugs into your application. You can download it from the Telerik site and manually add the assemblies to your project, or you can simply install the NuGet package via:       PM> Install-Package FiddlerCore The library consists of the FiddlerCore.dll as well as a couple of support libraries (CertMaker.dll and BCMakeCert.dll) that are used for installing SSL certificates. I’ll have more on SSL captures and certificate installation later in this post. But first let’s see how easy it is to use FiddlerCore to capture HTTP content by looking at how to build the above capture form. Capturing HTTP Content Once the library is installed it’s super easy to hook up Fiddler functionality. Fiddler includes a number of static class methods on the FiddlerApplication object that can be called to hook up callback events as well as actual start monitoring HTTP URLs. In the following code directly lifted from WebSurge, I configure a few filter options on Form level object, from the user inputs shown on the form by assigning it to a capture options object. In the live application these settings are persisted configuration values, but in the demo they are one time values initialized and set on the form. Once these options are set, I hook up the AfterSessionComplete event to capture every URL that passes through the proxy after the request is completed and start up the Proxy service:void Start() { if (tbIgnoreResources.Checked) CaptureConfiguration.IgnoreResources = true; else CaptureConfiguration.IgnoreResources = false; string strProcId = txtProcessId.Text; if (strProcId.Contains('-')) strProcId = strProcId.Substring(strProcId.IndexOf('-') + 1).Trim(); strProcId = strProcId.Trim(); int procId = 0; if (!string.IsNullOrEmpty(strProcId)) { if (!int.TryParse(strProcId, out procId)) procId = 0; } CaptureConfiguration.ProcessId = procId; CaptureConfiguration.CaptureDomain = txtCaptureDomain.Text; FiddlerApplication.AfterSessionComplete += FiddlerApplication_AfterSessionComplete; FiddlerApplication.Startup(8888, true, true, true); } The key lines for FiddlerCore are just the last two lines of code that include the event hookup code as well as the Startup() method call. Here I only hook up to the AfterSessionComplete event but there are a number of other events that hook various stages of the HTTP request cycle you can also hook into. Other events include BeforeRequest, BeforeResponse, RequestHeadersAvailable, ResponseHeadersAvailable and so on. In my case I want to capture the request data and I actually have several options to capture this data. AfterSessionComplete is the last event that fires in the request sequence and it’s the most common choice to capture all request and response data. I could have used several other events, but AfterSessionComplete is one place where you can look both at the request and response data, so this will be the most common place to hook into if you’re capturing content. The implementation of AfterSessionComplete is responsible for capturing all HTTP request headers and it looks something like this:private void FiddlerApplication_AfterSessionComplete(Session sess) { // Ignore HTTPS connect requests if (sess.RequestMethod == "CONNECT") return; if (CaptureConfiguration.ProcessId > 0) { if (sess.LocalProcessID != 0 && sess.LocalProcessID != CaptureConfiguration.ProcessId) return; } if (!string.IsNullOrEmpty(CaptureConfiguration.CaptureDomain)) { if (sess.hostname.ToLower() != CaptureConfiguration.CaptureDomain.Trim().ToLower()) return; } if (CaptureConfiguration.IgnoreResources) { string url = sess.fullUrl.ToLower(); var extensions = CaptureConfiguration.ExtensionFilterExclusions; foreach (var ext in extensions) { if (url.Contains(ext)) return; } var filters = CaptureConfiguration.UrlFilterExclusions; foreach (var urlFilter in filters) { if (url.Contains(urlFilter)) return; } } if (sess == null || sess.oRequest == null || sess.oRequest.headers == null) return; string headers = sess.oRequest.headers.ToString(); var reqBody = sess.GetRequestBodyAsString(); // if you wanted to capture the response //string respHeaders = session.oResponse.headers.ToString(); //var respBody = session.GetResponseBodyAsString(); // replace the HTTP line to inject full URL string firstLine = sess.RequestMethod + " " + sess.fullUrl + " " + sess.oRequest.headers.HTTPVersion; int at = headers.IndexOf("\r\n"); if (at < 0) return; headers = firstLine + "\r\n" + headers.Substring(at + 1); string output = headers + "\r\n" + (!string.IsNullOrEmpty(reqBody) ? reqBody + "\r\n" : string.Empty) + Separator + "\r\n\r\n"; BeginInvoke(new Action<string>((text) => { txtCapture.AppendText(text); UpdateButtonStatus(); }), output); } The code starts by filtering out some requests based on the CaptureOptions I set before the capture is started. These options/filters are applied when requests actually come in. This is very useful to help narrow down the requests that are captured for playback based on options the user picked. I find it useful to limit requests to a certain domain for captures, as well as filtering out some request types like static resources – images, css, scripts etc. This is of course optional, but I think it’s a common scenario and WebSurge makes good use of this feature. AfterSessionComplete like other FiddlerCore events, provides a Session object parameter which contains all the request and response details. There are oRequest and oResponse objects to hold their respective data. In my case I’m interested in the raw request headers and body only, as you can see in the commented code you can also retrieve the response headers and body. Here the code captures the request headers and body and simply appends the output to the textbox on the screen. Note that the Fiddler events are asynchronous, so in order to display the content in the UI they have to be marshaled back the UI thread with BeginInvoke, which here simply takes the generated headers and appends it to the existing textbox test on the form. As each request is processed, the headers are captured and appended to the bottom of the textbox resulting in a Session HTTP capture in the format that Web Surge internally supports, which is basically raw request headers with a customized 1st HTTP Header line that includes the full URL rather than a server relative URL. When the capture is done the user can either copy the raw HTTP session to the clipboard, or directly save it to file. This raw capture format is the same format WebSurge and also Fiddler use to import/export request data. While this code is application specific, it demonstrates the kind of logic that you can easily apply to the request capture process, which is one of the reasonsof why FiddlerCore is so powerful. You get to choose what content you want to look up as part of your own application logic and you can then decide how to capture or use that data as part of your application. The actual captured data in this case is only a string. The user can edit the data by hand or in the the case of WebSurge, save it to disk and automatically open the captured session as a new load test. Stopping the FiddlerCore Proxy Finally to stop capturing requests you simply disconnect the event handler and call the FiddlerApplication.ShutDown() method:void Stop() { FiddlerApplication.AfterSessionComplete -= FiddlerApplication_AfterSessionComplete; if (FiddlerApplication.IsStarted()) FiddlerApplication.Shutdown(); } As you can see, adding HTTP capture functionality to an application is very straight forward. FiddlerCore offers tons of features I’m not even touching on here – I suspect basic captures are the most common scenario, but a lot of different things can be done with FiddlerCore’s simple API interface. Sky’s the limit! The source code for this sample capture form (WinForms) is provided as part of this article. Adding Fiddler Certificates with FiddlerCore One of the sticking points in West Wind WebSurge has been that if you wanted to capture HTTPS/SSL traffic, you needed to have the full version of Fiddler and have HTTPS decryption enabled. Essentially you had to use Fiddler to configure HTTPS decryption and the associated installation of the Fiddler local client certificate that is used for local decryption of incoming SSL traffic. While this works just fine, requiring to have Fiddler installed and then using a separate application to configure the SSL functionality isn’t ideal. Fortunately FiddlerCore actually includes the tools to register the Fiddler Certificate directly using FiddlerCore. Why does Fiddler need a Certificate in the first Place? Fiddler and FiddlerCore are essentially HTTP proxies which means they inject themselves into the HTTP conversation by re-routing HTTP traffic to a special HTTP port (8888 by default for Fiddler) and then forward the HTTP data to the original client. Fiddler injects itself as the system proxy in using the WinInet Windows settings  which are the same settings that Internet Explorer uses and that are configured in the Windows and Internet Explorer Internet Settings dialog. Most HTTP clients running on Windows pick up and apply these system level Proxy settings before establishing new HTTP connections and that’s why most clients automatically work once Fiddler – or FiddlerCore/WebSurge are running. For plain HTTP requests this just works – Fiddler intercepts the HTTP requests on the proxy port and then forwards them to the original port (80 for HTTP and 443 for SSL typically but it could be any port). For SSL however, this is not quite as simple – Fiddler can easily act as an HTTPS/SSL client to capture inbound requests from the server, but when it forwards the request to the client it has to also act as an SSL server and provide a certificate that the client trusts. This won’t be the original certificate from the remote site, but rather a custom local certificate that effectively simulates an SSL connection between the proxy and the client. If there is no custom certificate configured for Fiddler the SSL request fails with a certificate validation error. The key for this to work is that a custom certificate has to be installed that the HTTPS client trusts on the local machine. For a much more detailed description of the process you can check out Eric Lawrence’s blog post on Certificates. If you’re using the desktop version of Fiddler you can install a local certificate into the Windows certificate store. Fiddler proper does this from the Options menu: This operation does several things: It installs the Fiddler Root Certificate It sets trust to this Root Certificate A new client certificate is generated for each HTTPS site monitored Certificate Installation with FiddlerCore You can also provide this same functionality using FiddlerCore which includes a CertMaker class. Using CertMaker is straight forward to use and it provides an easy way to create some simple helpers that can install and uninstall a Fiddler Root certificate:public static bool InstallCertificate() { if (!CertMaker.rootCertExists()) { if (!CertMaker.createRootCert()) return false; if (!CertMaker.trustRootCert()) return false; } return true; } public static bool UninstallCertificate() { if (CertMaker.rootCertExists()) { if (!CertMaker.removeFiddlerGeneratedCerts(true)) return false; } return true; } InstallCertificate() works by first checking whether the root certificate is already installed and if it isn’t goes ahead and creates a new one. The process of creating the certificate is a two step process – first the actual certificate is created and then it’s moved into the certificate store to become trusted. I’m not sure why you’d ever split these operations up since a cert created without trust isn’t going to be of much value, but there are two distinct steps. When you trigger the trustRootCert() method, a message box will pop up on the desktop that lets you know that you’re about to trust a local private certificate. This is a security feature to ensure that you really want to trust the Fiddler root since you are essentially installing a man in the middle certificate. It’s quite safe to use this generated root certificate, because it’s been specifically generated for your machine and thus is not usable from external sources, the only way to use this certificate in a trusted way is from the local machine. IOW, unless somebody has physical access to your machine, there’s no useful way to hijack this certificate and use it for nefarious purposes (see Eric’s post for more details). Once the Root certificate has been installed, FiddlerCore/Fiddler create new certificates for each site that is connected to with HTTPS. You can end up with quite a few temporary certificates in your certificate store. To uninstall you can either use Fiddler and simply uncheck the Decrypt HTTPS traffic option followed by the remove Fiddler certificates button, or you can use FiddlerCore’s CertMaker.removeFiddlerGeneratedCerts() which removes the root cert and any of the intermediary certificates Fiddler created. Keep in mind that when you uninstall you uninstall the certificate for both FiddlerCore and Fiddler, so use UninstallCertificate() with care and realize that you might affect the Fiddler application’s operation by doing so as well. When to check for an installed Certificate Note that the check to see if the root certificate exists is pretty fast, while the actual process of installing the certificate is a relatively slow operation that even on a fast machine takes a few seconds. Further the trust operation pops up a message box so you probably don’t want to install the certificate repeatedly. Since the check for the root certificate is fast, you can easily put a call to InstallCertificate() in any capture startup code – in which case the certificate installation only triggers when a certificate is in fact not installed. Personally I like to make certificate installation explicit – just like Fiddler does, so in WebSurge I use a small drop down option on the menu to install or uninstall the SSL certificate:   This code calls the InstallCertificate and UnInstallCertificate functions respectively – the experience with this is similar to what you get in Fiddler with the extra dialog box popping up to prompt confirmation for installation of the root certificate. Once the cert is installed you can then capture SSL requests. There’s a gotcha however… Gotcha: FiddlerCore Certificates don’t stick by Default When I originally tried to use the Fiddler certificate installation I ran into an odd problem. I was able to install the certificate and immediately after installation was able to capture HTTPS requests. Then I would exit the application and come back in and try the same HTTPS capture again and it would fail due to a missing certificate. CertMaker.rootCertExists() would return false after every restart and if re-installed the certificate a new certificate would get added to the certificate store resulting in a bunch of duplicated root certificates with different keys. What the heck? CertMaker and BcMakeCert create non-sticky CertificatesI turns out that FiddlerCore by default uses different components from what the full version of Fiddler uses. Fiddler uses a Windows utility called MakeCert.exe to create the Fiddler Root certificate. FiddlerCore however installs the CertMaker.dll and BCMakeCert.dll assemblies, which use a different crypto library (Bouncy Castle) for certificate creation than MakeCert.exe which uses the Windows Crypto API. The assemblies provide support for non-windows operation for Fiddler under Mono, as well as support for some non-Windows certificate platforms like iOS and Android for decryption. The bottom line is that the FiddlerCore provided bouncy castle assemblies are not sticky by default as the certificates created with them are not cached as they are in Fiddler proper. To get certificates to ‘stick’ you have to explicitly cache the certificates in Fiddler’s internal preferences. A cache aware version of InstallCertificate looks something like this:public static bool InstallCertificate() { if (!CertMaker.rootCertExists()) { if (!CertMaker.createRootCert()) return false; if (!CertMaker.trustRootCert()) return false; App.Configuration.UrlCapture.Cert = FiddlerApplication.Prefs.GetStringPref("fiddler.certmaker.bc.cert", null); App.Configuration.UrlCapture.Key = FiddlerApplication.Prefs.GetStringPref("fiddler.certmaker.bc.key", null); } return true; } public static bool UninstallCertificate() { if (CertMaker.rootCertExists()) { if (!CertMaker.removeFiddlerGeneratedCerts(true)) return false; } App.Configuration.UrlCapture.Cert = null; App.Configuration.UrlCapture.Key = null; return true; } In this code I store the Fiddler cert and private key in an application configuration settings that’s stored with the application settings (App.Configuration.UrlCapture object). These settings automatically persist when WebSurge is shut down. The values are read out of Fiddler’s internal preferences store which is set after a new certificate has been created. Likewise I clear out the configuration settings when the certificate is uninstalled. In order for these setting to be used you have to also load the configuration settings into the Fiddler preferences *before* a call to rootCertExists() is made. I do this in the capture form’s constructor:public FiddlerCapture(StressTestForm form) { InitializeComponent(); CaptureConfiguration = App.Configuration.UrlCapture; MainForm = form; if (!string.IsNullOrEmpty(App.Configuration.UrlCapture.Cert)) { FiddlerApplication.Prefs.SetStringPref("fiddler.certmaker.bc.key", App.Configuration.UrlCapture.Key); FiddlerApplication.Prefs.SetStringPref("fiddler.certmaker.bc.cert", App.Configuration.UrlCapture.Cert); }} This is kind of a drag to do and not documented anywhere that I could find, so hopefully this will save you some grief if you want to work with the stock certificate logic that installs with FiddlerCore. MakeCert provides sticky Certificates and the same functionality as Fiddler But there’s actually an easier way. If you want to skip the above Fiddler preference configuration code in your application you can choose to distribute MakeCert.exe instead of certmaker.dll and bcmakecert.dll. When you use MakeCert.exe, the certificates settings are stored in Windows so they are available without any custom configuration inside of your application. It’s easier to integrate and as long as you run on Windows and you don’t need to support iOS or Android devices is simply easier to deal with. To integrate into your project, you can remove the reference to CertMaker.dll (and the BcMakeCert.dll assembly) from your project. Instead copy MakeCert.exe into your output folder. To make sure MakeCert.exe gets pushed out, include MakeCert.exe in your project and set the Build Action to None, and Copy to Output Directory to Copy if newer. Note that the CertMaker.dll reference in the project has been removed and on disk the files for Certmaker.dll, as well as the BCMakeCert.dll files on disk. Keep in mind that these DLLs are resources of the FiddlerCore NuGet package, so updating the package may end up pushing those files back into your project. Once MakeCert.exe is distributed FiddlerCore checks for it first before using the assemblies so as long as MakeCert.exe exists it’ll be used for certificate creation (at least on Windows). Summary FiddlerCore is a pretty sweet tool, and it’s absolutely awesome that we get to plug in most of the functionality of Fiddler right into our own applications. A few years back I tried to build this sort of functionality myself for an app and ended up giving up because it’s a big job to get HTTP right – especially if you need to support SSL. FiddlerCore now provides that functionality as a turnkey solution that can be plugged into your own apps easily. The only downside is FiddlerCore’s documentation for more advanced features like certificate installation which is pretty sketchy. While for the most part FiddlerCore’s feature set is easy to work with without any documentation, advanced features are often not intuitive to gleam by just using Intellisense or the FiddlerCore help file reference (which is not terribly useful). While Eric Lawrence is very responsive on his forum and on Twitter, there simply isn’t much useful documentation on Fiddler/FiddlerCore available online. If you run into trouble the forum is probably the first place to look and then ask a question if you can’t find the answer. The best documentation you can find is Eric’s Fiddler Book which covers a ton of functionality of Fiddler and FiddlerCore. The book is a great reference to Fiddler’s feature set as well as providing great insights into the HTTP protocol. The second half of the book that gets into the innards of HTTP is an excellent read for anybody who wants to know more about some of the more arcane aspects and special behaviors of HTTP – it’s well worth the read. While the book has tons of information in a very readable format, it’s unfortunately not a great reference as it’s hard to find things in the book and because it’s not available online you can’t electronically search for the great content in it. But it’s hard to complain about any of this given the obvious effort and love that’s gone into this awesome product for all of these years. A mighty big thanks to Eric Lawrence  for having created this useful tool that so many of us use all the time, and also to Telerik for picking up Fiddler/FiddlerCore and providing Eric the resources to support and improve this wonderful tool full time and keeping it free for all. Kudos! Resources FiddlerCore Download FiddlerCore NuGet Fiddler Capture Sample Form Fiddler Capture Form in West Wind WebSurge (GitHub) Eric Lawrence’s Fiddler Book© Rick Strahl, West Wind Technologies, 2005-2014Posted in .NET  HTTP   Tweet !function(d,s,id){var js,fjs=d.getElementsByTagName(s)[0];if(!d.getElementById(id)){js=d.createElement(s);js.id=id;js.src="//platform.twitter.com/widgets.js";fjs.parentNode.insertBefore(js,fjs);}}(document,"script","twitter-wjs"); (function() { var po = document.createElement('script'); po.type = 'text/javascript'; po.async = true; po.src = 'https://apis.google.com/js/plusone.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(po, s); })();

  Read the article

• Improving Partitioned Table Join Performance

  - by Paul White

  The query optimizer does not always choose an optimal strategy when joining partitioned tables. This post looks at an example, showing how a manual rewrite of the query can almost double performance, while reducing the memory grant to almost nothing. Test Data The two tables in this example use a common partitioning partition scheme. The partition function uses 41 equal-size partitions: CREATE PARTITION FUNCTION PFT (integer) AS RANGE RIGHT FOR VALUES ( 125000, 250000, 375000, 500000, 625000, 750000, 875000, 1000000, 1125000, 1250000, 1375000, 1500000, 1625000, 1750000, 1875000, 2000000, 2125000, 2250000, 2375000, 2500000, 2625000, 2750000, 2875000, 3000000, 3125000, 3250000, 3375000, 3500000, 3625000, 3750000, 3875000, 4000000, 4125000, 4250000, 4375000, 4500000, 4625000, 4750000, 4875000, 5000000 ); GO CREATE PARTITION SCHEME PST AS PARTITION PFT ALL TO ([PRIMARY]); There two tables are: CREATE TABLE dbo.T1 ( TID integer NOT NULL IDENTITY(0,1), Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T1 PRIMARY KEY CLUSTERED (TID) ON PST (TID) );   CREATE TABLE dbo.T2 ( TID integer NOT NULL, Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T2 PRIMARY KEY CLUSTERED (TID, Column1) ON PST (TID) ); The next script loads 5 million rows into T1 with a pseudo-random value between 1 and 5 for Column1. The table is partitioned on the IDENTITY column TID: INSERT dbo.T1 WITH (TABLOCKX) (Column1) SELECT (ABS(CHECKSUM(NEWID())) % 5) + 1 FROM dbo.Numbers AS N WHERE n BETWEEN 1 AND 5000000; In case you don’t already have an auxiliary table of numbers lying around, here’s a script to create one with 10 million rows: CREATE TABLE dbo.Numbers (n bigint PRIMARY KEY);   WITH L0 AS(SELECT 1 AS c UNION ALL SELECT 1), L1 AS(SELECT 1 AS c FROM L0 AS A CROSS JOIN L0 AS B), L2 AS(SELECT 1 AS c FROM L1 AS A CROSS JOIN L1 AS B), L3 AS(SELECT 1 AS c FROM L2 AS A CROSS JOIN L2 AS B), L4 AS(SELECT 1 AS c FROM L3 AS A CROSS JOIN L3 AS B), L5 AS(SELECT 1 AS c FROM L4 AS A CROSS JOIN L4 AS B), Nums AS(SELECT ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) AS n FROM L5) INSERT dbo.Numbers WITH (TABLOCKX) SELECT TOP (10000000) n FROM Nums ORDER BY n OPTION (MAXDOP 1); Table T1 contains data like this: Next we load data into table T2. The relationship between the two tables is that table 2 contains ‘n’ rows for each row in table 1, where ‘n’ is determined by the value in Column1 of table T1. There is nothing particularly special about the data or distribution, by the way. INSERT dbo.T2 WITH (TABLOCKX) (TID, Column1) SELECT T.TID, N.n FROM dbo.T1 AS T JOIN dbo.Numbers AS N ON N.n >= 1 AND N.n <= T.Column1; Table T2 ends up containing about 15 million rows: The primary key for table T2 is a combination of TID and Column1. The data is partitioned according to the value in column TID alone. Partition Distribution The following query shows the number of rows in each partition of table T1: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T1 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are 40 partitions containing 125,000 rows (40 * 125k = 5m rows). The rightmost partition remains empty. The next query shows the distribution for table 2: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T2 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are roughly 375,000 rows in each partition (the rightmost partition is also empty): Ok, that’s the test data done. Test Query and Execution Plan The task is to count the rows resulting from joining tables 1 and 2 on the TID column: SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; The optimizer chooses a plan using parallel hash join, and partial aggregation: The Plan Explorer plan tree view shows accurate cardinality estimates and an even distribution of rows across threads (click to enlarge the image): With a warm data cache, the STATISTICS IO output shows that no physical I/O was needed, and all 41 partitions were touched: Running the query without actual execution plan or STATISTICS IO information for maximum performance, the query returns in around 2600ms. Execution Plan Analysis The first step toward improving on the execution plan produced by the query optimizer is to understand how it works, at least in outline. The two parallel Clustered Index Scans use multiple threads to read rows from tables T1 and T2. Parallel scan uses a demand-based scheme where threads are given page(s) to scan from the table as needed. This arrangement has certain important advantages, but does result in an unpredictable distribution of rows amongst threads. The point is that multiple threads cooperate to scan the whole table, but it is impossible to predict which rows end up on which threads. For correct results from the parallel hash join, the execution plan has to ensure that rows from T1 and T2 that might join are processed on the same thread. For example, if a row from T1 with join key value ‘1234’ is placed in thread 5’s hash table, the execution plan must guarantee that any rows from T2 that also have join key value ‘1234’ probe thread 5’s hash table for matches. The way this guarantee is enforced in this parallel hash join plan is by repartitioning rows to threads after each parallel scan. The two repartitioning exchanges route rows to threads using a hash function over the hash join keys. The two repartitioning exchanges use the same hash function so rows from T1 and T2 with the same join key must end up on the same hash join thread. Expensive Exchanges This business of repartitioning rows between threads can be very expensive, especially if a large number of rows is involved. The execution plan selected by the optimizer moves 5 million rows through one repartitioning exchange and around 15 million across the other. As a first step toward removing these exchanges, consider the execution plan selected by the optimizer if we join just one partition from each table, disallowing parallelism: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = 1 AND $PARTITION.PFT(T2.TID) = 1 OPTION (MAXDOP 1); The optimizer has chosen a (one-to-many) merge join instead of a hash join. The single-partition query completes in around 100ms. If everything scaled linearly, we would expect that extending this strategy to all 40 populated partitions would result in an execution time around 4000ms. Using parallelism could reduce that further, perhaps to be competitive with the parallel hash join chosen by the optimizer. This raises a question. If the most efficient way to join one partition from each of the tables is to use a merge join, why does the optimizer not choose a merge join for the full query? Forcing a Merge Join Let’s force the optimizer to use a merge join on the test query using a hint: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN); This is the execution plan selected by the optimizer: This plan results in the same number of logical reads reported previously, but instead of 2600ms the query takes 5000ms. The natural explanation for this drop in performance is that the merge join plan is only using a single thread, whereas the parallel hash join plan could use multiple threads. Parallel Merge Join We can get a parallel merge join plan using the same query hint as before, and adding trace flag 8649: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN, QUERYTRACEON 8649); The execution plan is: This looks promising. It uses a similar strategy to distribute work across threads as seen for the parallel hash join. In practice though, performance is disappointing. On a typical run, the parallel merge plan runs for around 8400ms; slower than the single-threaded merge join plan (5000ms) and much worse than the 2600ms for the parallel hash join. We seem to be going backwards! The logical reads for the parallel merge are still exactly the same as before, with no physical IOs. The cardinality estimates and thread distribution are also still very good (click to enlarge): A big clue to the reason for the poor performance is shown in the wait statistics (captured by Plan Explorer Pro): CXPACKET waits require careful interpretation, and are most often benign, but in this case excessive waiting occurs at the repartitioning exchanges. Unlike the parallel hash join, the repartitioning exchanges in this plan are order-preserving ‘merging’ exchanges (because merge join requires ordered inputs): Parallelism works best when threads can just grab any available unit of work and get on with processing it. Preserving order introduces inter-thread dependencies that can easily lead to significant waits occurring. In extreme cases, these dependencies can result in an intra-query deadlock, though the details of that will have to wait for another time to explore in detail. The potential for waits and deadlocks leads the query optimizer to cost parallel merge join relatively highly, especially as the degree of parallelism (DOP) increases. This high costing resulted in the optimizer choosing a serial merge join rather than parallel in this case. The test results certainly confirm its reasoning. Collocated Joins In SQL Server 2008 and later, the optimizer has another available strategy when joining tables that share a common partition scheme. This strategy is a collocated join, also known as as a per-partition join. It can be applied in both serial and parallel execution plans, though it is limited to 2-way joins in the current optimizer. Whether the optimizer chooses a collocated join or not depends on cost estimation. The primary benefits of a collocated join are that it eliminates an exchange and requires less memory, as we will see next. Costing and Plan Selection The query optimizer did consider a collocated join for our original query, but it was rejected on cost grounds. The parallel hash join with repartitioning exchanges appeared to be a cheaper option. There is no query hint to force a collocated join, so we have to mess with the costing framework to produce one for our test query. Pretending that IOs cost 50 times more than usual is enough to convince the optimizer to use collocated join with our test query: -- Pretend IOs are 50x cost temporarily DBCC SETIOWEIGHT(50);   -- Co-located hash join SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (RECOMPILE);   -- Reset IO costing DBCC SETIOWEIGHT(1); Collocated Join Plan The estimated execution plan for the collocated join is: The Constant Scan contains one row for each partition of the shared partitioning scheme, from 1 to 41. The hash repartitioning exchanges seen previously are replaced by a single Distribute Streams exchange using Demand partitioning. Demand partitioning means that the next partition id is given to the next parallel thread that asks for one. My test machine has eight logical processors, and all are available for SQL Server to use. As a result, there are eight threads in the single parallel branch in this plan, each processing one partition from each table at a time. Once a thread finishes processing a partition, it grabs a new partition number from the Distribute Streams exchange…and so on until all partitions have been processed. It is important to understand that the parallel scans in this plan are different from the parallel hash join plan. Although the scans have the same parallelism icon, tables T1 and T2 are not being co-operatively scanned by multiple threads in the same way. Each thread reads a single partition of T1 and performs a hash match join with the same partition from table T2. The properties of the two Clustered Index Scans show a Seek Predicate (unusual for a scan!) limiting the rows to a single partition: The crucial point is that the join between T1 and T2 is on TID, and TID is the partitioning column for both tables. A thread that processes partition ‘n’ is guaranteed to see all rows that can possibly join on TID for that partition. In addition, no other thread will see rows from that partition, so this removes the need for repartitioning exchanges. CPU and Memory Efficiency Improvements The collocated join has removed two expensive repartitioning exchanges and added a single exchange processing 41 rows (one for each partition id). Remember, the parallel hash join plan exchanges had to process 5 million and 15 million rows. The amount of processor time spent on exchanges will be much lower in the collocated join plan. In addition, the collocated join plan has a maximum of 8 threads processing single partitions at any one time. The 41 partitions will all be processed eventually, but a new partition is not started until a thread asks for it. Threads can reuse hash table memory for the new partition. The parallel hash join plan also had 8 hash tables, but with all 5,000,000 build rows loaded at the same time. The collocated plan needs memory for only 8 * 125,000 = 1,000,000 rows at any one time. Collocated Hash Join Performance The collated join plan has disappointing performance in this case. The query runs for around 25,300ms despite the same IO statistics as usual. This is much the worst result so far, so what went wrong? It turns out that cardinality estimation for the single partition scans of table T1 is slightly low. The properties of the Clustered Index Scan of T1 (graphic immediately above) show the estimation was for 121,951 rows. This is a small shortfall compared with the 125,000 rows actually encountered, but it was enough to cause the hash join to spill to physical tempdb: A level 1 spill doesn’t sound too bad, until you realize that the spill to tempdb probably occurs for each of the 41 partitions. As a side note, the cardinality estimation error is a little surprising because the system tables accurately show there are 125,000 rows in every partition of T1. Unfortunately, the optimizer uses regular column and index statistics to derive cardinality estimates here rather than system table information (e.g. sys.partitions). Collocated Merge Join We will never know how well the collocated parallel hash join plan might have worked without the cardinality estimation error (and the resulting 41 spills to tempdb) but we do know: Merge join does not require a memory grant; and Merge join was the optimizer’s preferred join option for a single partition join Putting this all together, what we would really like to see is the same collocated join strategy, but using merge join instead of hash join. Unfortunately, the current query optimizer cannot produce a collocated merge join; it only knows how to do collocated hash join. So where does this leave us? CROSS APPLY sys.partitions We can try to write our own collocated join query. We can use sys.partitions to find the partition numbers, and CROSS APPLY to get a count per partition, with a final step to sum the partial counts. The following query implements this idea: SELECT row_count = SUM(Subtotals.cnt) FROM ( -- Partition numbers SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1 ) AS P CROSS APPLY ( -- Count per collocated join SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals; The estimated plan is: The cardinality estimates aren’t all that good here, especially the estimate for the scan of the system table underlying the sys.partitions view. Nevertheless, the plan shape is heading toward where we would like to be. Each partition number from the system table results in a per-partition scan of T1 and T2, a one-to-many Merge Join, and a Stream Aggregate to compute the partial counts. The final Stream Aggregate just sums the partial counts. Execution time for this query is around 3,500ms, with the same IO statistics as always. This compares favourably with 5,000ms for the serial plan produced by the optimizer with the OPTION (MERGE JOIN) hint. This is another case of the sum of the parts being less than the whole – summing 41 partial counts from 41 single-partition merge joins is faster than a single merge join and count over all partitions. Even so, this single-threaded collocated merge join is not as quick as the original parallel hash join plan, which executed in 2,600ms. On the positive side, our collocated merge join uses only one logical processor and requires no memory grant. The parallel hash join plan used 16 threads and reserved 569 MB of memory:   Using a Temporary Table Our collocated merge join plan should benefit from parallelism. The reason parallelism is not being used is that the query references a system table. We can work around that by writing the partition numbers to a temporary table (or table variable): SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   CREATE TABLE #P ( partition_number integer PRIMARY KEY);   INSERT #P (partition_number) SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1;   SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals;   DROP TABLE #P;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; Using the temporary table adds a few logical reads, but the overall execution time is still around 3500ms, indistinguishable from the same query without the temporary table. The problem is that the query optimizer still doesn’t choose a parallel plan for this query, though the removal of the system table reference means that it could if it chose to: In fact the optimizer did enter the parallel plan phase of query optimization (running search 1 for a second time): Unfortunately, the parallel plan found seemed to be more expensive than the serial plan. This is a crazy result, caused by the optimizer’s cost model not reducing operator CPU costs on the inner side of a nested loops join. Don’t get me started on that, we’ll be here all night. In this plan, everything expensive happens on the inner side of a nested loops join. Without a CPU cost reduction to compensate for the added cost of exchange operators, candidate parallel plans always look more expensive to the optimizer than the equivalent serial plan. Parallel Collocated Merge Join We can produce the desired parallel plan using trace flag 8649 again: SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: One difference between this plan and the collocated hash join plan is that a Repartition Streams exchange operator is used instead of Distribute Streams. The effect is similar, though not quite identical. The Repartition uses round-robin partitioning, meaning the next partition id is pushed to the next thread in sequence. The Distribute Streams exchange seen earlier used Demand partitioning, meaning the next partition id is pulled across the exchange by the next thread that is ready for more work. There are subtle performance implications for each partitioning option, but going into that would again take us too far off the main point of this post. Performance The important thing is the performance of this parallel collocated merge join – just 1350ms on a typical run. The list below shows all the alternatives from this post (all timings include creation, population, and deletion of the temporary table where appropriate) from quickest to slowest: Collocated parallel merge join: 1350ms Parallel hash join: 2600ms Collocated serial merge join: 3500ms Serial merge join: 5000ms Parallel merge join: 8400ms Collated parallel hash join: 25,300ms (hash spill per partition) The parallel collocated merge join requires no memory grant (aside from a paltry 1.2MB used for exchange buffers). This plan uses 16 threads at DOP 8; but 8 of those are (rather pointlessly) allocated to the parallel scan of the temporary table. These are minor concerns, but it turns out there is a way to address them if it bothers you. Parallel Collocated Merge Join with Demand Partitioning This final tweak replaces the temporary table with a hard-coded list of partition ids (dynamic SQL could be used to generate this query from sys.partitions): SELECT row_count = SUM(Subtotals.cnt) FROM ( VALUES (1),(2),(3),(4),(5),(6),(7),(8),(9),(10), (11),(12),(13),(14),(15),(16),(17),(18),(19),(20), (21),(22),(23),(24),(25),(26),(27),(28),(29),(30), (31),(32),(33),(34),(35),(36),(37),(38),(39),(40),(41) ) AS P (partition_number) CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: The parallel collocated hash join plan is reproduced below for comparison: The manual rewrite has another advantage that has not been mentioned so far: the partial counts (per partition) can be computed earlier than the partial counts (per thread) in the optimizer’s collocated join plan. The earlier aggregation is performed by the extra Stream Aggregate under the nested loops join. The performance of the parallel collocated merge join is unchanged at around 1350ms. Final Words It is a shame that the current query optimizer does not consider a collocated merge join (Connect item closed as Won’t Fix). The example used in this post showed an improvement in execution time from 2600ms to 1350ms using a modestly-sized data set and limited parallelism. In addition, the memory requirement for the query was almost completely eliminated  – down from 569MB to 1.2MB. The problem with the parallel hash join selected by the optimizer is that it attempts to process the full data set all at once (albeit using eight threads). It requires a large memory grant to hold all 5 million rows from table T1 across the eight hash tables, and does not take advantage of the divide-and-conquer opportunity offered by the common partitioning. The great thing about the collocated join strategies is that each parallel thread works on a single partition from both tables, reading rows, performing the join, and computing a per-partition subtotal, before moving on to a new partition. From a thread’s point of view… If you have trouble visualizing what is happening from just looking at the parallel collocated merge join execution plan, let’s look at it again, but from the point of view of just one thread operating between the two Parallelism (exchange) operators. Our thread picks up a single partition id from the Distribute Streams exchange, and starts a merge join using ordered rows from partition 1 of table T1 and partition 1 of table T2. By definition, this is all happening on a single thread. As rows join, they are added to a (per-partition) count in the Stream Aggregate immediately above the Merge Join. Eventually, either T1 (partition 1) or T2 (partition 1) runs out of rows and the merge join stops. The per-partition count from the aggregate passes on through the Nested Loops join to another Stream Aggregate, which is maintaining a per-thread subtotal. Our same thread now picks up a new partition id from the exchange (say it gets id 9 this time). The count in the per-partition aggregate is reset to zero, and the processing of partition 9 of both tables proceeds just as it did for partition 1, and on the same thread. Each thread picks up a single partition id and processes all the data for that partition, completely independently from other threads working on other partitions. One thread might eventually process partitions (1, 9, 17, 25, 33, 41) while another is concurrently processing partitions (2, 10, 18, 26, 34) and so on for the other six threads at DOP 8. The point is that all 8 threads can execute independently and concurrently, continuing to process new partitions until the wider job (of which the thread has no knowledge!) is done. This divide-and-conquer technique can be much more efficient than simply splitting the entire workload across eight threads all at once. Related Reading Understanding and Using Parallelism in SQL Server Parallel Execution Plans Suck © 2013 Paul White – All Rights Reserved Twitter: @SQL_Kiwi

  Read the article

• Nagging As A Strategy For Better Linking: -z guidance

  - by user9154181

  The link-editor (ld) in Solaris 11 has a new feature that we call guidance that is intended to help you build better objects. The basic idea behind guidance is that if (and only if) you request it, the link-editor will issue messages suggesting better options and other changes you might make to your ld command to get better results. You can choose to take the advice, or you can disable specific types of guidance while acting on others. In some ways, this works like an experienced friend leaning over your shoulder and giving you advice — you're free to take it or leave it as you see fit, but you get nudged to do a better job than you might have otherwise. We use guidance to build the core Solaris OS, and it has proven to be useful, both in improving our objects, and in making sure that regressions don't creep back in later. In this article, I'm going to describe the evolution in thinking and design that led to the implementation of the -z guidance option, as well as give a brief description of how it works. The guidance feature issues non-fatal warnings. However, experience shows that once developers get used to ignoring warnings, it is inevitable that real problems will be lost in the noise and ignored or missed. This is why we have a zero tolerance policy against build noise in the core Solaris OS. In order to get maximum benefit from -z guidance while maintaining this policy, I added the -z fatal-warnings option at the same time. Much of the material presented here is adapted from the arc case: PSARC 2010/312 Link-editor guidance The History Of Unfortunate Link-Editor Defaults The Solaris link-editor is one of the oldest Unix commands. It stands to reason that this would be true — in order to write an operating system, you need the ability to compile and link code. The original link-editor (ld) had defaults that made sense at the time. As new features were needed, command line option switches were added to let the user use them, while maintaining backward compatibility for those who didn't. Backward compatibility is always a concern in system design, but is particularly important in the case of the tool chain (compilers, linker, and related tools), since it is a basic building block for the entire system. Over the years, applications have grown in size and complexity. Important concepts like dynamic linking that didn't exist in the original Unix system were invented. Object file formats changed. In the case of System V Release 4 Unix derivatives like Solaris, the ELF (Extensible Linking Format) was adopted. Since then, the ELF system has evolved to provide tools needed to manage today's larger and more complex environments. Features such as lazy loading, and direct bindings have been added. In an ideal world, many of these options would be defaults, with rarely used options that allow the user to turn them off. However, the reality is exactly the reverse: For backward compatibility, these features are all options that must be explicitly turned on by the user. This has led to a situation in which most applications do not take advantage of the many improvements that have been made in linking over the last 20 years. If their code seems to link and run without issue, what motivation does a developer have to read a complex manpage, absorb the information provided, choose the features that matter for their application, and apply them? Experience shows that only the most motivated and diligent programmers will make that effort. We know that most programs would be improved if we could just get you to use the various whizzy features that we provide, but the defaults conspire against us. We have long wanted to do something to make it easier for our users to use the linkers more effectively. There have been many conversations over the years regarding this issue, and how to address it. They always break down along the following lines: Change ld Defaults Since the world would be a better place the newer ld features were the defaults, why not change things to make it so? This idea is simple, elegant, and impossible. Doing so would break a large number of existing applications, including those of ISVs, big customers, and a plethora of existing open source packages. In each case, the owner of that code may choose to follow our lead and fix their code, or they may view it as an invitation to reconsider their commitment to our platform. Backward compatibility, and our installed base of working software, is one of our greatest assets, and not something to be lightly put at risk. Breaking backward compatibility at this level of the system is likely to do more harm than good. But, it sure is tempting. New Link-Editor One might create a new linker command, not called 'ld', leaving the old command as it is. The new one could use the same code as ld, but would offer only modern options, with the proper defaults for features such as direct binding. The resulting link-editor would be a pleasure to use. However, the approach is doomed to niche status. There is a vast pile of exiting code in the world built around the existing ld command, that reaches back to the 1970's. ld use is embedded in large and unknown numbers of makefiles, and is used by name by compilers that execute it. A Unix link-editor that is not named ld will not find a majority audience no matter how good it might be. Finally, a new linker command will eventually cease to be new, and will accumulate its own burden of backward compatibility issues. An Option To Make ld Do The Right Things Automatically This line of reasoning is best summarized by a CR filed in 2005, entitled 6239804 make it easier for ld(1) to do what's best The idea is to have a '-z best' option that unchains ld from its backward compatibility commitment, and allows it to turn on the "best" set of features, as determined by the authors of ld. The specific set of features enabled by -z best would be subject to change over time, as requirements change. This idea is more realistic than the other two, but was never implemented because it has some important issues that we could never answer to our satisfaction: The -z best proposal assumes that the user can turn it on, and trust it to select good options without the user needing to be aware of the options being applied. This is a fallacy. Features such as direct bindings require the user to do some analysis to ensure that the resulting program will still operate properly. A user who is willing to do the work to verify that what -z best does will be OK for their application is capable of turning on those features directly, and therefore gains little added benefit from -z best. The intent is that when a user opts into -z best, that they understand that z best is subject to sometimes incompatible evolution. Experience teaches us that this won't work. People will use this feature, the meaning of -z best will change, code that used to build will fail, and then there will be complaints and demands to retract the change. When (not if) this occurs, we will of course defend our actions, and point at the disclaimer. We'll win some of those debates, and lose others. Ultimately, we'll end up with -z best2 (-z better), or other compromises, and our goal of simplifying the world will have failed. The -z best idea rolls up a set of features that may or may not be related to each other into a unit that must be taken wholesale, or not at all. It could be that only a subset of what it does is compatible with a given application, in which case the user is expected to abandon -z best and instead set the options that apply to their application directly. In doing so, they lose one of the benefits of -z best, that if you use it, future versions of ld may choose a different set of options, and automatically improve the object through the act of rebuilding it. I drew two conclusions from the above history: For a link-editor, backward compatibility is vital. If a given command line linked your application 10 years ago, you have every reason to expect that it will link today, assuming that the libraries you're linking against are still available and compatible with their previous interfaces. For an application of any size or complexity, there is no substitute for the work involved in examining the code and determining which linker options apply and which do not. These options are largely orthogonal to each other, and it can be reasonable not to use any or all of them, depending on the situation, even in modern applications. It is a mistake to tie them together. The idea for -z guidance came from consideration of these points. By decoupling the advice from the act of taking the advice, we can retain the good aspects of -z best while avoiding its pitfalls: -z guidance gives advice, but the decision to take that advice remains with the user who must evaluate its merit and make a decision to take it or not. As such, we are free to change the specific guidance given in future releases of ld, without breaking existing applications. The only fallout from this will be some new warnings in the build output, which can be ignored or dealt with at the user's convenience. It does not couple the various features given into a single "take it or leave it" option, meaning that there will never be a need to offer "-zguidance2", or other such variants as things change over time. Guidance has the potential to be our final word on this subject. The user is given the flexibility to disable specific categories of guidance without losing the benefit of others, including those that might be added to future versions of the system. Although -z fatal-warnings stands on its own as a useful feature, it is of particular interest in combination with -z guidance. Used together, the guidance turns from advice to hard requirement: The user must either make the suggested change, or explicitly reject the advice by specifying a guidance exception token, in order to get a build. This is valuable in environments with high coding standards. ld Command Line Options The guidance effort resulted in new link-editor options for guidance and for turning warnings into fatal errors. Before I reproduce that text here, I'd like to highlight the strategic decisions embedded in the guidance feature: In order to get guidance, you have to opt in. We hope you will opt in, and believe you'll get better objects if you do, but our default mode of operation will continue as it always has, with full backward compatibility, and without judgement. Guidance suggestions always offers specific advice, and not vague generalizations. You can disable some guidance without turning off the entire feature. When you get guidance warnings, you can choose to take the advice, or you can specify a keyword to disable guidance for just that category. This allows you to get guidance for things that are useful to you, without being bothered about things that you've already considered and dismissed. As the world changes, we will add new guidance to steer you in the right direction. All such new guidance will come with a keyword that let's you turn it off. In order to facilitate building your code on different versions of Solaris, we quietly ignore any guidance keywords we don't recognize, assuming that they are intended for newer versions of the link-editor. If you want to see what guidance tokens ld does and does not recognize on your system, you can use the ld debugging feature as follows: % ld -Dargs -z guidance=foo,nodefs debug: debug: Solaris Linkers: 5.11-1.2275 debug: debug: arg[1] option=-D: option-argument: args debug: arg[2] option=-z: option-argument: guidance=foo,nodefs debug: warning: unrecognized -z guidance item: foo The -z fatal-warning option is straightforward, and generally useful in environments with strict coding standards. Note that the GNU ld already had this feature, and we accept their option names as synonyms: -z fatal-warnings | nofatal-warnings --fatal-warnings | --no-fatal-warnings The -z fatal-warnings and the --fatal-warnings option cause the link-editor to treat warnings as fatal errors. The -z nofatal-warnings and the --no-fatal-warnings option cause the link-editor to treat warnings as non-fatal. This is the default behavior. The -z guidance option is defined as follows: -z guidance[=item1,item2,...] Provide guidance messages to suggest ld options that can improve the quality of the resulting object, or which are otherwise considered to be beneficial. The specific guidance offered is subject to change over time as the system evolves. Obsolete guidance offered by older versions of ld may be dropped in new versions. Similarly, new guidance may be added to new versions of ld. Guidance therefore always represents current best practices. It is possible to enable guidance, while preventing specific guidance messages, by providing a list of item tokens, representing the class of guidance to be suppressed. In this way, unwanted advice can be suppressed without losing the benefit of other guidance. Unrecognized item tokens are quietly ignored by ld, allowing a given ld command line to be executed on a variety of older or newer versions of Solaris. The guidance offered by the current version of ld, and the item tokens used to disable these messages, are as follows. Specify Required Dependencies Dynamic executables and shared objects should explicitly define all of the dependencies they require. Guidance recommends the use of the -z defs option, should any symbol references remain unsatisfied when building dynamic objects. This guidance can be disabled with -z guidance=nodefs. Do Not Specify Non-Required Dependencies Dynamic executables and shared objects should not define any dependencies that do not satisfy the symbol references made by the dynamic object. Guidance recommends that unused dependencies be removed. This guidance can be disabled with -z guidance=nounused. Lazy Loading Dependencies should be identified for lazy loading. Guidance recommends the use of the -z lazyload option should any dependency be processed before either a -z lazyload or -z nolazyload option is encountered. This guidance can be disabled with -z guidance=nolazyload. Direct Bindings Dependencies should be referenced with direct bindings. Guidance recommends the use of the -B direct, or -z direct options should any dependency be processed before either of these options, or the -z nodirect option is encountered. This guidance can be disabled with -z guidance=nodirect. Pure Text Segment Dynamic objects should not contain relocations to non-writable, allocable sections. Guidance recommends compiling objects with Position Independent Code (PIC) should any relocations against the text segment remain, and neither the -z textwarn or -z textoff options are encountered. This guidance can be disabled with -z guidance=notext. Mapfile Syntax All mapfiles should use the version 2 mapfile syntax. Guidance recommends the use of the version 2 syntax should any mapfiles be encountered that use the version 1 syntax. This guidance can be disabled with -z guidance=nomapfile. Library Search Path Inappropriate dependencies that are encountered by ld are quietly ignored. For example, a 32-bit dependency that is encountered when generating a 64-bit object is ignored. These dependencies can result from incorrect search path settings, such as supplying an incorrect -L option. Although benign, this dependency processing is wasteful, and might hide a build problem that should be solved. Guidance recommends the removal of any inappropriate dependencies. This guidance can be disabled with -z guidance=nolibpath. In addition, -z guidance=noall can be used to entirely disable the guidance feature. See Chapter 7, Link-Editor Quick Reference, in the Linker and Libraries Guide for more information on guidance and advice for building better objects. Example The following example demonstrates how the guidance feature is intended to work. We will build a shared object that has a variety of shortcomings: Does not specify all it's dependencies Specifies dependencies it does not use Does not use direct bindings Uses a version 1 mapfile Contains relocations to the readonly allocable text (not PIC) This scenario is sadly very common — many shared objects have one or more of these issues. % cat hello.c #include <stdio.h> #include <unistd.h> void hello(void) { printf("hello user %d\n", getpid()); } % cat mapfile.v1 # This version 1 mapfile will trigger a guidance message % cc hello.c -o hello.so -G -M mapfile.v1 -lelf As you can see, the operation completes without error, resulting in a usable object. However, turning on guidance reveals a number of things that could be better: % cc hello.c -o hello.so -G -M mapfile.v1 -lelf -zguidance ld: guidance: version 2 mapfile syntax recommended: mapfile.v1 ld: guidance: -z lazyload option recommended before first dependency ld: guidance: -B direct or -z direct option recommended before first dependency Undefined first referenced symbol in file getpid hello.o (symbol belongs to implicit dependency /lib/libc.so.1) printf hello.o (symbol belongs to implicit dependency /lib/libc.so.1) ld: warning: symbol referencing errors ld: guidance: -z defs option recommended for shared objects ld: guidance: removal of unused dependency recommended: libelf.so.1 warning: Text relocation remains referenced against symbol offset in file .rodata1 (section) 0xa hello.o getpid 0x4 hello.o printf 0xf hello.o ld: guidance: position independent (PIC) code recommended for shared objects ld: guidance: see ld(1) -z guidance for more information Given the explicit advice in the above guidance messages, it is relatively easy to modify the example to do the right things: % cat mapfile.v2 # This version 2 mapfile will not trigger a guidance message $mapfile_version 2 % cc hello.c -o hello.so -Kpic -G -Bdirect -M mapfile.v2 -lc -zguidance There are situations in which the guidance does not fit the object being built. For instance, you want to build an object without direct bindings: % cc -Kpic hello.c -o hello.so -G -M mapfile.v2 -lc -zguidance ld: guidance: -B direct or -z direct option recommended before first dependency ld: guidance: see ld(1) -z guidance for more information It is easy to disable that specific guidance warning without losing the overall benefit from allowing the remainder of the guidance feature to operate: % cc -Kpic hello.c -o hello.so -G -M mapfile.v2 -lc -zguidance=nodirect Conclusions The linking guidelines enforced by the ld guidance feature correspond rather directly to our standards for building the core Solaris OS. I'm sure that comes as no surprise. It only makes sense that we would want to build our own product as well as we know how. Solaris is usually the first significant test for any new linker feature. We now enable guidance by default for all builds, and the effect has been very positive. Guidance helps us find suboptimal objects more quickly. Programmers get concrete advice for what to change instead of vague generalities. Even in the cases where we override the guidance, the makefile rules to do so serve as documentation of the fact. Deciding to use guidance is likely to cause some up front work for most code, as it forces you to consider using new features such as direct bindings. Such investigation is worthwhile, but does not come for free. However, the guidance suggestions offer a structured and straightforward way to tackle modernizing your objects, and once that work is done, for keeping them that way. The investment is often worth it, and will replay you in terms of better performance and fewer problems. I hope that you find guidance to be as useful as we have.

  Read the article

• problem with sIFR 3 not displaying in IE just getting XXX

  - by user288306

  I am having a problem with sIFR 3 not displaying in IE. I get 3 larges black XXX in IE yet it displays fine in Firefox. I have checked i do have the most recent version of flash installed correctly. Here is the code on the page <div id="features"> <div id="mainmessage_advertisers"><h2>Advertisers</h2><br /><br /><h3><a href="">Reach your customers where they browse. Buy directly from top web publishers.</a></h3><br /><br /><br /><a href=""><img src="img/buyads.gif" border="0"></a></div> <div id="mainmessage_publishers"><h2>Publishers</h2><br /><br /><h3>Take control of your ad space and start generating more revenue than <u>ever before</u>.</h3><br /><br /><br /><a href=""><img src="img/sellads.gif" border="0"></a></div> </div>` Here is the code from my global.css #mainmessage_advertisers { width: 395px; height: 200px; padding: 90px 50px; border: 1px; float: left; } #mainmessage_publishers { width: 395px; height: 200px; padding: 90px 50px; float: right; } and here is what i have in my sifr.js /*********************************************************************** SIFR 3.0 (BETA 1) FUNCTIONS ************************************************************************/ var parseSelector=(function(){var _1=/\s*,\s*/;var _2=/\s*([\s>+~(),]|^|$)\s*/g;var _3=/([\s>+~,]|[^(]\+|^)([#.:@])/g;var _4=/^[^\s>+~]/;var _5=/[\s#.:>+~()@]|[^\s#.:>+~()@]+/g;function parseSelector(_6,_7){_7=_7||document.documentElement;var _8=_6.split(_1),_9=[];for(var i=0;i<_8.length;i++){var _b=[_7],_c=toStream(_8[i]);for(var j=0;j<_c.length;){var _e=_c[j++],_f=_c[j++],_10="";if(_c[j]=="("){while(_c[j++]!=")"&&j<_c.length){_10+=_c[j]}_10=_10.slice(0,-1)}_b=select(_b,_e,_f,_10)}_9=_9.concat(_b)}return _9}function toStream(_11){var _12=_11.replace(_2,"$1").replace(_3,"$1*$2");if(_4.test(_12)){_12=" "+_12}return _12.match(_5)||[]}function select(_13,_14,_15,_16){return (_17[_14])?_17[_14](_13,_15,_16):[]}var _18={toArray:function(_19){var a=[];for(var i=0;i<_19.length;i++){a.push(_19[i])}return a}};var dom={isTag:function(_1d,tag){return (tag=="*")||(tag.toLowerCase()==_1d.nodeName.toLowerCase())},previousSiblingElement:function(_1f){do{_1f=_1f.previousSibling}while(_1f&&_1f.nodeType!=1);return _1f},nextSiblingElement:function(_20){do{_20=_20.nextSibling}while(_20&&_20.nodeType!=1);return _20},hasClass:function(_21,_22){return (_22.className||"").match("(^|\\s)"+_21+"(\\s|$)")},getByTag:function(tag,_24){return _24.getElementsByTagName(tag)}};var _17={"#":function(_25,_26){for(var i=0;i<_25.length;i++){if(_25[i].getAttribute("id")==_26){return [_25[i]]}}return []}," ":function(_28,_29){var _2a=[];for(var i=0;i<_28.length;i++){_2a=_2a.concat(_18.toArray(dom.getByTag(_29,_28[i])))}return _2a},">":function(_2c,_2d){var _2e=[];for(var i=0,_30;i<_2c.length;i++){_30=_2c[i];for(var j=0,_32;j<_30.childNodes.length;j++){_32=_30.childNodes[j];if(_32.nodeType==1&&dom.isTag(_32,_2d)){_2e.push(_32)}}}return _2e},".":function(_33,_34){var _35=[];for(var i=0,_37;i<_33.length;i++){_37=_33[i];if(dom.hasClass([_34],_37)){_35.push(_37)}}return _35},":":function(_38,_39,_3a){return (pseudoClasses[_39])?pseudoClasses[_39](_38,_3a):[]}};parseSelector.selectors=_17;parseSelector.pseudoClasses={};parseSelector.util=_18;parseSelector.dom=dom;return parseSelector})(); var sIFR=new function(){var _3b=this;var _3c="sIFR-active";var _3d="sIFR-replaced";var _3e="sIFR-flash";var _3f="sIFR-ignore";var _40="sIFR-alternate";var _41="sIFR-class";var _42="sIFR-layout";var _43="http://www.w3.org/1999/xhtml";var _44=6;var _45=126;var _46=8;var _47="SIFR-PREFETCHED";var _48=" ";this.isActive=false;this.isEnabled=true;this.hideElements=true;this.replaceNonDisplayed=false;this.preserveSingleWhitespace=false;this.fixWrap=true;this.registerEvents=true;this.setPrefetchCookie=true;this.cookiePath="/";this.domains=[];this.fromLocal=true;this.forceClear=false;this.forceWidth=true;this.fitExactly=false;this.forceTextTransform=true;this.useDomContentLoaded=true;this.debugMode=false;this.hasFlashClassSet=false;var _49=0;var _4a=false,_4b=false;var dom=new function(){this.getBody=function(){var _4d=document.getElementsByTagName("body");if(_4d.length==1){return _4d[0]}return null};this.addClass=function(_4e,_4f){if(_4f){_4f.className=((_4f.className||"")==""?"":_4f.className+" ")+_4e}};this.removeClass=function(_50,_51){if(_51){_51.className=_51.className.replace(new RegExp("(^|\\s)"+_50+"(\\s|$)"),"").replace(/^\s+|(\s)\s+/g,"$1")}};this.hasClass=function(_52,_53){return new RegExp("(^|\\s)"+_52+"(\\s|$)").test(_53.className)};this.create=function(_54){if(document.createElementNS){return document.createElementNS(_43,_54)}return document.createElement(_54)};this.setInnerHtml=function(_55,_56){if(ua.innerHtmlSupport){_55.innerHTML=_56}else{if(ua.xhtmlSupport){_56=["<root xmlns=\"",_43,"\">",_56,"</root>"].join("");var xml=(new DOMParser()).parseFromString(_56,"text/xml");xml=document.importNode(xml.documentElement,true);while(_55.firstChild){_55.removeChild(_55.firstChild)}while(xml.firstChild){_55.appendChild(xml.firstChild)}}}};this.getComputedStyle=function(_58,_59){var _5a;if(document.defaultView&&document.defaultView.getComputedStyle){_5a=document.defaultView.getComputedStyle(_58,null)[_59]}else{if(_58.currentStyle){_5a=_58.currentStyle[_59]}}return _5a||""};this.getStyleAsInt=function(_5b,_5c,_5d){var _5e=this.getComputedStyle(_5b,_5c);if(_5d&&!/px$/.test(_5e)){return 0}_5e=parseInt(_5e);return isNaN(_5e)?0:_5e};this.getZoom=function(){return _5f.zoom.getLatest()}};this.dom=dom;var ua=new function(){var ua=navigator.userAgent.toLowerCase();var _62=(navigator.product||"").toLowerCase();this.macintosh=ua.indexOf("mac")>-1;this.windows=ua.indexOf("windows")>-1;this.quicktime=false;this.opera=ua.indexOf("opera")>-1;this.konqueror=_62.indexOf("konqueror")>-1;this.ie=false/*@cc_on || true @*/;this.ieSupported=this.ie&&!/ppc|smartphone|iemobile|msie\s5\.5/.test(ua)/*@cc_on && @_jscript_version >= 5.5 @*/;this.ieWin=this.ie&&this.windows/*@cc_on && @_jscript_version >= 5.1 @*/;this.windows=this.windows&&(!this.ie||this.ieWin);this.ieMac=this.ie&&this.macintosh/*@cc_on && @_jscript_version < 5.1 @*/;this.macintosh=this.macintosh&&(!this.ie||this.ieMac);this.safari=ua.indexOf("safari")>-1;this.webkit=ua.indexOf("applewebkit")>-1&&!this.konqueror;this.khtml=this.webkit||this.konqueror;this.gecko=!this.webkit&&_62=="gecko";this.operaVersion=this.opera&&/.*opera(\s|\/)(\d+\.\d+)/.exec(ua)?parseInt(RegExp.$2):0;this.webkitVersion=this.webkit&&/.*applewebkit\/(\d+).*/.exec(ua)?parseInt(RegExp.$1):0;this.geckoBuildDate=this.gecko&&/.*gecko\/(\d{8}).*/.exec(ua)?parseInt(RegExp.$1):0;this.konquerorVersion=this.konqueror&&/.*konqueror\/(\d\.\d).*/.exec(ua)?parseInt(RegExp.$1):0;this.flashVersion=0;if(this.ieWin){var axo;var _64=false;try{axo=new ActiveXObject("ShockwaveFlash.ShockwaveFlash.7")}catch(e){try{axo=new ActiveXObject("ShockwaveFlash.ShockwaveFlash.6");this.flashVersion=6;axo.AllowScriptAccess="always"}catch(e){_64=this.flashVersion==6}if(!_64){try{axo=new ActiveXObject("ShockwaveFlash.ShockwaveFlash")}catch(e){}}}if(!_64&&axo){this.flashVersion=parseFloat(/([\d,?]+)/.exec(axo.GetVariable("$version"))[1].replace(/,/g,"."))}}else{if(navigator.plugins&&navigator.plugins["Shockwave Flash"]){var _65=navigator.plugins["Shockwave Flash"];this.flashVersion=parseFloat(/(\d+\.?\d*)/.exec(_65.description)[1]);var i=0;while(this.flashVersion>=_46&&i<navigator.mimeTypes.length){var _67=navigator.mimeTypes[i];if(_67.type=="application/x-shockwave-flash"&&_67.enabledPlugin.description.toLowerCase().indexOf("quicktime")>-1){this.flashVersion=0;this.quicktime=true}i++}}}this.flash=this.flashVersion>=_46;this.transparencySupport=this.macintosh||this.windows;this.computedStyleSupport=this.ie||document.defaultView&&document.defaultView.getComputedStyle&&(!this.gecko||this.geckoBuildDate>=20030624);this.css=true;if(this.computedStyleSupport){try{var _68=document.getElementsByTagName("head")[0];_68.style.backgroundColor="#FF0000";var _69=dom.getComputedStyle(_68,"backgroundColor");this.css=!_69||/\#F{2}0{4}|rgb\(255,\s?0,\s?0\)/i.test(_69);_68=null}catch(e){}}this.xhtmlSupport=!!window.DOMParser&&!!document.importNode;this.innerHtmlSupport;try{var n=dom.create("span");if(!this.ieMac){n.innerHTML="x"}this.innerHtmlSupport=n.innerHTML=="x"}catch(e){this.innerHtmlSupport=false}this.zoomSupport=!!(this.opera&&document.documentElement);this.geckoXml=this.gecko&&(document.contentType||"").indexOf("xml")>-1;this.requiresPrefetch=this.ieWin||this.khtml;this.verifiedKonqueror=false;this.supported=this.flash&&this.css&&(!this.ie||this.ieSupported)&&(!this.opera||this.operaVersion>=8)&&(!this.webkit||this.webkitVersion>=412)&&(!this.konqueror||this.konquerorVersion>3.5)&&this.computedStyleSupport&&(this.innerHtmlSupport||!this.khtml&&this.xhtmlSupport)};this.ua=ua;var _6b=new function(){function capitalize($){return $.toUpperCase()}this.normalize=function(str){if(_3b.preserveSingleWhitespace){return str.replace(/\s/g,_48)}return str.replace(/(\s)\s+/g,"$1")};this.textTransform=function(_6e,str){switch(_6e){case "uppercase":str=str.toUpperCase();break;case "lowercase":str=str.toLowerCase();break;case "capitalize":var _70=str;str=str.replace(/^\w|\s\w/g,capitalize);if(str.indexOf("function capitalize")!=-1){var _71=_70.replace(/(^|\s)(\w)/g,"$1$1$2$2").split(/^\w|\s\w/g);str="";for(var i=0;i<_71.length;i++){str+=_71[i].charAt(0).toUpperCase()+_71[i].substring(1)}}break}return str};this.toHexString=function(str){if(typeof (str)!="string"||!str.charAt(0)=="#"||str.length!=4&&str.length!=7){return str}str=str.replace(/#/,"");if(str.length==3){str=str.replace(/(.)(.)(.)/,"$1$1$2$2$3$3")}return "0x"+str};this.toJson=function(obj){var _75="";switch(typeof (obj)){case "string":_75="\""+obj+"\"";break;case "number":case "boolean":_75=obj.toString();break;case "object":_75=[];for(var _76 in obj){if(obj[_76]==Object.prototype[_76]){continue}_75.push("\""+_76+"\":"+_6b.toJson(obj[_76]))}_75="{"+_75.join(",")+"}";break}return _75};this.convertCssArg=function(arg){if(!arg){return {}}if(typeof (arg)=="object"){if(arg.constructor==Array){arg=arg.join("")}else{return arg}}var obj={};var _79=arg.split("}");for(var i=0;i<_79.length;i++){var $=_79[i].match(/([^\s{]+)\s*\{(.+)\s*;?\s*/);if(!$||$.length!=3){continue}if(!obj[$[1]]){obj[$[1]]={}}var _7c=$[2].split(";");for(var j=0;j<_7c.length;j++){var $2=_7c[j].match(/\s*([^:\s]+)\s*\:\s*([^\s;]+)/);if(!$2||$2.length!=3){continue}obj[$[1]][$2[1]]=$2[2]}}return obj};this.extractFromCss=function(css,_80,_81,_82){var _83=null;if(css&&css[_80]&&css[_80][_81]){_83=css[_80][_81];if(_82){delete css[_80][_81]}}return _83};this.cssToString=function(arg){var css=[];for(var _86 in arg){var _87=arg[_86];if(_87==Object.prototype[_86]){continue}css.push(_86,"{");for(var _88 in _87){if(_87[_88]==Object.prototype[_88]){continue}css.push(_88,":",_87[_88],";")}css.push("}")}return escape(css.join(""))}};this.util=_6b;var _5f={};_5f.fragmentIdentifier=new function(){this.fix=true;var _89;this.cache=function(){_89=document.title};function doFix(){document.title=_89}this.restore=function(){if(this.fix){setTimeout(doFix,0)}}};_5f.synchronizer=new function(){this.isBlocked=false;this.block=function(){this.isBlocked=true};this.unblock=function(){this.isBlocked=false;_8a.replaceAll()}};_5f.zoom=new function(){var _8b=100;this.getLatest=function(){return _8b};if(ua.zoomSupport&&ua.opera){var _8c=document.createElement("div");_8c.style.position="fixed";_8c.style.left="-65536px";_8c.style.top="0";_8c.style.height="100%";_8c.style.width="1px";_8c.style.zIndex="-32";document.documentElement.appendChild(_8c);function updateZoom(){if(!_8c){return}var _8d=window.innerHeight/_8c.offsetHeight;var _8e=Math.round(_8d*100)%10;if(_8e>5){_8d=Math.round(_8d*100)+10-_8e}else{_8d=Math.round(_8d*100)-_8e}_8b=isNaN(_8d)?100:_8d;_5f.synchronizer.unblock();document.documentElement.removeChild(_8c);_8c=null}_5f.synchronizer.block();setTimeout(updateZoom,54)}};this.hacks=_5f;var _8f={kwargs:[],replaceAll:function(){for(var i=0;i<this.kwargs.length;i++){_3b.replace(this.kwargs[i])}this.kwargs=[]}};var _8a={kwargs:[],replaceAll:_8f.replaceAll};function isValidDomain(){if(_3b.domains.length==0){return true}var _91="";try{_91=document.domain}catch(e){}if(_3b.fromLocal&&sIFR.domains[0]!="localhost"){sIFR.domains.unshift("localhost")}for(var i=0;i<_3b.domains.length;i++){if(_3b.domains[i]=="*"||_3b.domains[i]==_91){return true}}return false}this.activate=function(){if(!ua.supported||!this.isEnabled||this.isActive||!isValidDomain()){return}this.isActive=true;if(this.hideElements){this.setFlashClass()}if(ua.ieWin&&_5f.fragmentIdentifier.fix&&window.location.hash!=""){_5f.fragmentIdentifier.cache()}else{_5f.fragmentIdentifier.fix=false}if(!this.registerEvents){return}function handler(evt){_3b.initialize();if(evt&&evt.type=="load"){if(document.removeEventListener){document.removeEventListener("DOMContentLoaded",handler,false);document.removeEventListener("load",handler,false)}if(window.removeEventListener){window.removeEventListener("load",handler,false)}}}if(window.addEventListener){if(_3b.useDomContentLoaded&&ua.gecko){document.addEventListener("DOMContentLoaded",handler,false)}window.addEventListener("load",handler,false)}else{if(ua.ieWin){if(_3b.useDomContentLoaded&&!_4a){document.write("<scr"+"ipt id=__sifr_ie_onload defer src=//:></script>");document.getElementById("__sifr_ie_onload").onreadystatechange=function(){if(this.readyState=="complete"){handler();this.removeNode()}}}window.attachEvent("onload",handler)}}};this.setFlashClass=function(){if(this.hasFlashClassSet){return}dom.addClass(_3c,dom.getBody()||document.documentElement);this.hasFlashClassSet=true};this.removeFlashClass=function(){if(!this.hasFlashClassSet){return}dom.removeClass(_3c,dom.getBody());dom.removeClass(_3c,document.documentElement);this.hasFlashClassSet=false};this.initialize=function(){if(_4b||!this.isActive||!this.isEnabled){return}_4b=true;_8f.replaceAll();clearPrefetch()};function getSource(src){if(typeof (src)!="string"){if(src.src){src=src.src}if(typeof (src)!="string"){var _95=[];for(var _96 in src){if(src[_96]!=Object.prototype[_96]){_95.push(_96)}}_95.sort().reverse();var _97="";var i=-1;while(!_97&&++i<_95.length){if(parseFloat(_95[i])<=ua.flashVersion){_97=src[_95[i]]}}src=_97}}if(!src&&_3b.debugMode){throw new Error("sIFR: Could not determine appropriate source")}if(ua.ie&&src.charAt(0)=="/"){src=window.location.toString().replace(/([^:]+)(:\/?\/?)([^\/]+).*/,"$1$2$3")+src}return src}this.prefetch=function(){if(!ua.requiresPrefetch||!ua.supported||!this.isEnabled||!isValidDomain()){return}if(this.setPrefetchCookie&&new RegExp(";?"+_47+"=true;?").test(document.cookie)){return}try{_4a=true;if(ua.ieWin){prefetchIexplore(arguments)}else{prefetchLight(arguments)}if(this.setPrefetchCookie){document.cookie=_47+"=true;path="+this.cookiePath}}catch(e){if(_3b.debugMode){throw e}}};function prefetchIexplore(_99){for(var i=0;i<_99.length;i++){document.write("<embed src=\""+getSource(_99[i])+"\" sIFR-prefetch=\"true\" style=\"display:none;\">")}}function prefetchLight(_9b){for(var i=0;i<_9b.length;i++){new Image().src=getSource(_9b[i])}}function clearPrefetch(){if(!ua.ieWin||!_4a){return}try{var _9d=document.getElementsByTagName("embed");for(var i=_9d.length-1;i>=0;i--){var _9f=_9d[i];if(_9f.getAttribute("sIFR-prefetch")=="true"){_9f.parentNode.removeChild(_9f)}}}catch(e){}}function getRatio(_a0){if(_a0<=10){return 1.55}if(_a0<=19){return 1.45}if(_a0<=32){return 1.35}if(_a0<=71){return 1.3}return 1.25}function getFilters(obj){var _a2=[];for(var _a3 in obj){if(obj[_a3]==Object.prototype[_a3]){continue}var _a4=obj[_a3];_a3=[_a3.replace(/filter/i,"")+"Filter"];for(var _a5 in _a4){if(_a4[_a5]==Object.prototype[_a5]){continue}_a3.push(_a5+":"+escape(_6b.toJson(_6b.toHexString(_a4[_a5]))))}_a2.push(_a3.join(","))}return _a2.join(";")}this.replace=function(_a6,_a7){if(!ua.supported){return}if(_a7){for(var _a8 in _a6){if(typeof (_a7[_a8])=="undefined"){_a7[_a8]=_a6[_a8]}}_a6=_a7}if(!_4b){return _8f.kwargs.push(_a6)}if(_5f.synchronizer.isBlocked){return _8a.kwargs.push(_a6)}var _a9=_a6.elements;if(!_a9&&parseSelector){_a9=parseSelector(_a6.selector)}if(_a9.length==0){return}this.setFlashClass();var src=getSource(_a6.src);var css=_6b.convertCssArg(_a6.css);var _ac=getFilters(_a6.filters);var _ad=(_a6.forceClear==null)?_3b.forceClear:_a6.forceClear;var _ae=(_a6.fitExactly==null)?_3b.fitExactly:_a6.fitExactly;var _af=_ae||(_a6.forceWidth==null?_3b.forceWidth:_a6.forceWidth);var _b0=parseInt(_6b.extractFromCss(css,".sIFR-root","leading"))||0;var _b1=_6b.extractFromCss(css,".sIFR-root","background-color",true)||"#FFFFFF";var _b2=_6b.extractFromCss(css,".sIFR-root","opacity",true)||"100";if(parseFloat(_b2)<1){_b2=100*parseFloat(_b2)}var _b3=_6b.extractFromCss(css,".sIFR-root","kerning",true)||"";var _b4=_a6.gridFitType||_6b.extractFromCss(css,".sIFR-root","text-align")=="right"?"subpixel":"pixel";var _b5=_3b.forceTextTransform?_6b.extractFromCss(css,".sIFR-root","text-transform",true)||"none":"none";var _b6="";if(_ae){_6b.extractFromCss(css,".sIFR-root","text-align",true)}if(!_a6.modifyCss){_b6=_6b.cssToString(css)}var _b7=_a6.wmode||"";if(_b7=="transparent"){if(!ua.transparencySupport){_b7="opaque"}else{_b1="transparent"}}for(var i=0;i<_a9.length;i++){var _b9=_a9[i];if(!ua.verifiedKonqueror){if(dom.getComputedStyle(_b9,"lineHeight").match(/e\+08px/)){ua.supported=_3b.isEnabled=false;this.removeFlashClass();return}ua.verifiedKonqueror=true}if(dom.hasClass(_3d,_b9)||dom.hasClass(_3f,_b9)){continue}var _ba=false;if(!_b9.offsetHeight||!_b9.offsetWidth){if(!_3b.replaceNonDisplayed){continue}_b9.style.display="block";if(!_b9.offsetHeight||!_b9.offsetWidth){_b9.style.display="";continue}_ba=true}if(_ad&&ua.gecko){_b9.style.clear="both"}var _bb=null;if(_3b.fixWrap&&ua.ie&&dom.getComputedStyle(_b9,"display")=="block"){_bb=_b9.innerHTML;dom.setInnerHtml(_b9,"X")}var _bc=dom.getStyleAsInt(_b9,"width",ua.ie);if(ua.ie&&_bc==0){var _bd=dom.getStyleAsInt(_b9,"paddingRight",true);var _be=dom.getStyleAsInt(_b9,"paddingLeft",true);var _bf=dom.getStyleAsInt(_b9,"borderRightWidth",true);var _c0=dom.getStyleAsInt(_b9,"borderLeftWidth",true);_bc=_b9.offsetWidth-_be-_bd-_c0-_bf}if(_bb&&_3b.fixWrap&&ua.ie){dom.setInnerHtml(_b9,_bb)}var _c1,_c2;if(!ua.ie){_c1=dom.getStyleAsInt(_b9,"lineHeight");_c2=Math.floor(dom.getStyleAsInt(_b9,"height")/_c1)}else{if(ua.ie){var _bb=_b9.innerHTML;_b9.style.visibility="visible";_b9.style.overflow="visible";_b9.style.position="static";_b9.style.zoom="normal";_b9.style.writingMode="lr-tb";_b9.style.width=_b9.style.height="auto";_b9.style.maxWidth=_b9.style.maxHeight=_b9.style.styleFloat="none";var _c3=_b9;var _c4=_b9.currentStyle.hasLayout;if(_c4){dom.setInnerHtml(_b9,"<div class=\""+_42+"\">X<br />X<br />X</div>");_c3=_b9.firstChild}else{dom.setInnerHtml(_b9,"X<br />X<br />X")}var _c5=_c3.getClientRects();_c1=_c5[1].bottom-_c5[1].top;_c1=Math.ceil(_c1*0.8);if(_c4){dom.setInnerHtml(_b9,"<div class=\""+_42+"\">"+_bb+"</div>");_c3=_b9.firstChild}else{dom.setInnerHtml(_b9,_bb)}_c5=_c3.getClientRects();_c2=_c5.length;if(_c4){dom.setInnerHtml(_b9,_bb)}_b9.style.visibility=_b9.style.width=_b9.style.height=_b9.style.maxWidth=_b9.style.maxHeight=_b9.style.overflow=_b9.style.styleFloat=_b9.style.position=_b9.style.zoom=_b9.style.writingMode=""}}if(_ba){_b9.style.display=""}if(_ad&&ua.gecko){_b9.style.clear=""}_c1=Math.max(_44,_c1);_c1=Math.min(_45,_c1);if(isNaN(_c2)||!isFinite(_c2)){_c2=1}var _c6=Math.round(_c2*_c1);if(_c2>1&&_b0){_c6+=Math.round((_c2-1)*_b0)}var _c7=dom.create("span");_c7.className=_40;var _c8=_b9.cloneNode(true);for(var j=0,l=_c8.childNodes.length;j<l;j++){_c7.appendChild(_c8.childNodes[j].cloneNode(true))}if(_a6.modifyContent){_a6.modifyContent(_c8,_a6.selector)}if(_a6.modifyCss){_b6=_a6.modifyCss(css,_c8,_a6.selector)}var _cb=handleContent(_c8,_b5);if(_a6.modifyContentString){_cb=_a6.modifyContentString(_cb,_a6.selector)}if(_cb==""){continue}var _cc=["content="+_cb.replace(/\</g,"&lt;").replace(/>/g,"&gt;"),"width="+_bc,"height="+_c6,"fitexactly="+(_ae?"true":""),"tunewidth="+(_a6.tuneWidth||""),"tuneheight="+(_a6.tuneHeight||""),"offsetleft="+(_a6.offsetLeft||""),"offsettop="+(_a6.offsetTop||""),"thickness="+(_a6.thickness||""),"sharpness="+(_a6.sharpness||""),"kerning="+_b3,"gridfittype="+_b4,"zoomsupport="+ua.zoomSupport,"filters="+_ac,"opacity="+_b2,"blendmode="+(_a6.blendMode||""),"size="+_c1,"zoom="+dom.getZoom(),"css="+_b6];_cc=encodeURI(_cc.join("&amp;"));var _cd="sIFR_callback_"+_49++;var _ce={flashNode:null};window[_cd+"_DoFSCommand"]=(function(_cf){return function(_d0,arg){if(/(FSCommand\:)?resize/.test(_d0)){var $=arg.split(":");_cf.flashNode.setAttribute($[0],$[1]);if(ua.khtml){_cf.flashNode.innerHTML+=""}}}})(_ce);_c6=Math.round(_c2*getRatio(_c1)*_c1);var _d3=_af?_bc:"100%";var _d4;if(ua.ie){_d4=["<object classid=\"clsid:D27CDB6E-AE6D-11cf-96B8-444553540000\" id=\"",_cd,"\" sifr=\"true\" width=\"",_d3,"\" height=\"",_c6,"\" class=\"",_3e,"\">","<param name=\"movie\" value=\"",src,"\"></param>","<param name=\"flashvars\" value=\"",_cc,"\"></param>","<param name=\"allowScriptAccess\" value=\"always\"></param>","<param name=\"quality\" value=\"best\"></param>","<param name=\"wmode\" value=\"",_b7,"\"></param>","<param name=\"bgcolor\" value=\"",_b1,"\"></param>","<param name=\"name\" value=\"",_cd,"\"></param>","</object>","<scr","ipt event=FSCommand(info,args) for=",_cd,">",_cd,"_DoFSCommand(info, args);","</","script>"].join("")}else{_d4=["<embed class=\"",_3e,"\" type=\"application/x-shockwave-flash\" src=\"",src,"\" quality=\"best\" flashvars=\"",_cc,"\" width=\"",_d3,"\" height=\"",_c6,"\" wmode=\"",_b7,"\" bgcolor=\"",_b1,"\" name=\"",_cd,"\" allowScriptAccess=\"always\" sifr=\"true\"></embed>"].join("")}dom.setInnerHtml(_b9,_d4);_ce.flashNode=_b9.firstChild;_b9.appendChild(_c7);dom.addClass(_3d,_b9);if(_a6.onReplacement){_a6.onReplacement(_ce.flashNode)}}_5f.fragmentIdentifier.restore()};function handleContent(_d5,_d6){var _d7=[],_d8=[];var _d9=_d5.childNodes;var i=0;while(i<_d9.length){var _db=_d9[i];if(_db.nodeType==3){var _dc=_6b.normalize(_db.nodeValue);_dc=_6b.textTransform(_d6,_dc);_d8.push(_dc.replace(/\%/g,"%25").replace(/\&/g,"%26").replace(/\,/g,"%2C").replace(/\+/g,"%2B"))}if(_db.nodeType==1){var _dd=[];var _de=_db.nodeName.toLowerCase();var _df=_db.className||"";if(/\s+/.test(_df)){if(_df.indexOf(_41)){_df=_df.match("(\\s|^)"+_41+"-([^\\s$]*)(\\s|$)")[2]}else{_df=_df.match(/^([^\s]+)/)[1]}}if(_df!=""){_dd.push("class=\""+_df+"\"")}if(_de=="a"){var _e0=_db.getAttribute("href")||"";var _e1=_db.getAttribute("target")||"";_dd.push("href=\""+_e0+"\"","target=\""+_e1+"\"")}_d8.push("<"+_de+(_dd.length>0?" ":"")+escape(_dd.join(" "))+">");if(_db.hasChildNodes()){_d7.push(i);i=0;_d9=_db.childNodes;continue}else{if(!/^(br|img)$/i.test(_db.nodeName)){_d8.push("</",_db.nodeName.toLowerCase(),">")}}}if(_d7.length>0&&!_db.nextSibling){do{i=_d7.pop();_d9=_db.parentNode.parentNode.childNodes;_db=_d9[i];if(_db){_d8.push("</",_db.nodeName.toLowerCase(),">")}}while(i<_d9.length&&_d7.length>0)}i++}return _d8.join("").replace(/\n|\r/g,"")}}; sIFR.prefetch({ src: 'swf/sifr/helvetica.swf' }); sIFR.activate(); sIFR.replace({ selector: 'h2, h3', src: 'swf/sifr/helvetica.swf', wmode: 'transparent', css: { '.sIFR-root' : { 'color': '#000000', 'font-weight': 'bold', 'letter-spacing': '-1' }, 'a': { 'text-decoration': 'none' }, 'a:link': { 'color': '#000000' }, 'a:hover': { 'color': '#000000' }, '.span': { 'color': '#979797' }, 'label': { 'color': '#E11818' } } }); sIFR.replace({ selector: 'h4', src: 'swf/sifr/helvetica.swf', wmode: 'transparent', css: { '.sIFR-root' : { 'color': '#7E7E7E', 'font-weight': 'bold', 'letter-spacing': '-0.8' }, 'a': { 'text-decoration': 'none' }, 'a:link': { 'color': '#7E7E7E' }, 'a:hover': { 'color': '#7E7E7E' }, 'label': { 'color': '#E11818' } } }); sIFR.replace({ selector: '#cart p', src: 'swf/sifr/helvetica-lt.swf', wmode: 'transparent', css: { '.sIFR-root' : { 'color': '#979797', 'font-weight': 'bold', 'letter-spacing': '-0.8' }, 'a': { 'text-decoration': 'none' }, 'a:link': { 'color': '#979797' }, 'a:hover': { 'color': '#000000' }, 'label': { 'color': '#979797' } } }); Thank you in advance for your help!

  Read the article

• SQL SERVER – Subquery or Join – Various Options – SQL Server Engine knows the Best

  - by pinaldave

  This is followup post of my earlier article SQL SERVER – Convert IN to EXISTS – Performance Talk, after reading all the comments I have received I felt that I could write more on the same subject to clear few things out. First let us run following four queries, all of them are giving exactly same resultset. USE AdventureWorks GO -- use of = SELECT * FROM HumanResources.Employee E WHERE E.EmployeeID = ( SELECT EA.EmployeeID FROM HumanResources.EmployeeAddress EA WHERE EA.EmployeeID = E.EmployeeID) GO -- use of in SELECT * FROM HumanResources.Employee E WHERE E.EmployeeID IN ( SELECT EA.EmployeeID FROM HumanResources.EmployeeAddress EA WHERE EA.EmployeeID = E.EmployeeID) GO -- use of exists SELECT * FROM HumanResources.Employee E WHERE EXISTS ( SELECT EA.EmployeeID FROM HumanResources.EmployeeAddress EA WHERE EA.EmployeeID = E.EmployeeID) GO -- Use of Join SELECT * FROM HumanResources.Employee E INNER JOIN HumanResources.EmployeeAddress EA ON E.EmployeeID = EA.EmployeeID GO Let us compare the execution plan of the queries listed above. Click on image to see larger image. It is quite clear from the execution plan that in case of IN, EXISTS and JOIN SQL Server Engines is smart enough to figure out what is the best optimal plan of Merge Join for the same query and execute the same. However, in the case of use of Equal (=) Operator, SQL Server is forced to use Nested Loop and test each result of the inner query and compare to outer query, leading to cut the performance. Please note that here I no mean suggesting that Nested Loop is bad or Merge Join is better. This can very well vary on your machine and amount of resources available on your computer. When I see Equal (=) operator used in query like above, I usually recommend to see if user can use IN or EXISTS or JOIN. As I said, this can very much vary on different system. What is your take in above query? I believe SQL Server Engines is usually pretty smart to figure out what is ideal execution plan and use it. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Joins, SQL Optimization, SQL Performance, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

  Read the article

• Password Security: Short and Complex versus ‘Short or Lengthy’ and Less Complex

  - by Akemi Iwaya

  Creating secure passwords for our online accounts is a necessary evil due to the huge increase in database and account hacking that occurs these days. The problem though is that no two companies have a similar policy for complex and secure password creation, then factor in the continued creation of insecure passwords or multi-site use of the same password and trouble is just waiting to happen. Ars Technica decided to take a look at multiple password types, how users fared with them, and how well those password types held up to cracking attempts in their latest study. The password types that Ars Technica looked at were comprehensive8, basic8, and basic16. The comprehensive type required a variety of upper-case, lower-case, digits, and symbols with no dictionary words allowed. The only restriction on the two basic types was the number of characters used. Which type do you think was easier for users to adopt and did better in the two password cracking tests? You can learn more about how well users did with the three password types and the results of the tests by visiting the article linked below. What are your thoughts on the matter? Are shorter, more complex passwords better or worse than using short or long, but less complex passwords? What methods do you feel work best since most passwords are limited to approximately 16 characters in length? Perhaps you use a service like LastPass or keep a dedicated list/notebook to manage your passwords. Let us know in the comments!    

  Read the article

< Previous Page | 107 108 109 110 111 112 113 114 115 116 117 118  | Next Page >