Search Results

Search found 10312 results on 413 pages for 'compiler bug'.

Page 49/413 | < Previous Page | 45 46 47 48 49 50 51 52 53 54 55 56  | Next Page >

  • Bizarre C++ compiler problem

    - by Yassin
    Hi, I have the following C++ code: typedef istream_iterator<string> isi; // (1) vector<string> lineas(isi(cin), isi()); // (2) //vector<string> lineas; //copy(isi(cin), isi(), back_inserter(lineas)); typedef vector<string>::iterator vci; for (vci it = lineas.begin(); it != lineas.end(); ++it) cout &lt;&lt; *it &lt;&lt; endl; However, I get the error while compiling: test.cpp: In function 'int main(int, char**)': test.cpp:16: error: request for member 'begin' in 'lineas', which is of non-class type 'std::vector<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::basic_string<char, std::char_traits<char>, std::allocator<char> > > >(main(int, char**)::isi, main(int, char**)::isi (*)())' test.cpp:16: error: request for member 'end' in 'lineas', which is of non-class type 'std::vector<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::basic_string<char, std::char_traits<char>, std::allocator<char> > > >(main(int, char**)::isi, main(int, char**)::isi (*)())' However, if I replace (1) by (2), it compiles. I'm using g++ 4.4.0 What's wrong?

    Read the article

  • Firebird 2.1 + EXISTS = query bug?

    - by Atlas
    Using Delphi 2009 + Firebird 2.1.3. Database is ODS 11.1, default char set is UTF8. My prepared query is as follows: SELECT a.po_id, a.po_no FROM purchase_order a WHERE EXISTS (SELECT 1 FROM sales_order_item z1 JOIN purchase_order_item z2 ON z2.so_item_id = z1.so_item_id AND z2.po_id = a.po_id WHERE z1.so_id = :soid) ORDER BY a.po_no Now when I loop this say 1000 times because I have 1000 x so_id, the CPU usage get at 100% for FBSERVER.EXE Anyone encountered this problem?

    Read the article

  • sprintf bug with php & apache in windows?

    - by potatoe
    I've run into a strange problem on a WAMP server setup (PHP version 5.3.0, Apache 2.2.11). When using sprintf to output a number, I occasionally get erroneous characters in the output string. Example: (not trimmed from anything, this is the only code in the script) $dt1 = new DateTime('now'); $dt2 = new DateTime('now - 10 min'); $interval = $dt1->diff($dt2); $number = 10.0; $string = sprintf("%.1f", $number); echo "number: $number, string: $string\n"; If I run this at the command prompt with PHP CLI, I get the expected output: number: 10, string: 10.0 However, if I serve it using Apache, in the browser I get number: 10, string: :.0 with a colon where '10' should be. (Note that ':' is the next ascii character in sequence after '9', if $number is 0-9, everything works. Numbers greater than 10 appear to use ascii equivalents - so 11 is ';', 12 is '<', etc.) The strangest part is that the first four lines in the above code sample seem to affect the results. Logically, those statements should have no impact, but if I comment them out or remove them the problem goes away. Any ideas? Anyone else able to replicate this?

    Read the article

  • CString a = "Hello " + "World!"; Is it possible?

    - by Sanctus2099
    I'm making my own string class and I was wondering if there's any way to replace the compiler behaviour around " characters. As I said in the title I'd like to make it so that CString a = "Hello " + "World!"; would actually work and not give a compiler error telling that it can't add 2 pointers. My string class automatically converts to char* when needed and thus writing printf(a) would not break the code. I'm sure this is a rather weird question but if it's possible I'd really like to know how to do it. Thank you very much

    Read the article

  • Cannot understand the behaviour of C# compiler while instantiating a class thru interface

    - by Newbie
    I have a class that implements an interface. The interface is public interface IRiskFactory { void StartService(); void StopService(); } The class that implements the interface is public class RiskFactoryService : IRiskFactory { } Now I have a console application and one window service. From the console application if I write the following code static void Main(string[] args) { IRiskFactory objIRiskFactory = new RiskFactoryService(); objIRiskFactory.StartService(); Console.ReadLine(); objIRiskFactory.StopService(); } It is working fine. However, when I mwrite the same piece of code in Window service public partial class RiskFactoryService : ServiceBase { IRiskFactory objIRiskFactory = null; public RiskFactoryService() { InitializeComponent(); objIRiskFactory = new RiskFactoryService(); <- ERROR } /// <summary> /// Starts the service /// </summary> /// <param name="args"></param> protected override void OnStart(string[] args) { objIRiskFactory.StartService(); } /// <summary> /// Stops the service /// </summary> protected override void OnStop() { objIRiskFactory.StopService(); } } It throws error: Cannot implicitly convert type 'RiskFactoryService' to 'IRiskFactory'. An explicit conversion exists (are you missing a cast?) When I type cast to the interface type, it started working objIRiskFactory = (IRiskFactory)new RiskFactoryService(); My question is why so?

    Read the article

  • Python - creating a list with 2 characteristics bug

    - by user2733911
    The goal is to create a list of 99 elements. All elements must be 1s or 0s. The first element must be a 1. There must be 7 1s in total. import random import math import time # constants determined through testing generation_constant = 0.96 def generate_candidate(): coin_vector = [] coin_vector.append(1) for i in range(0, 99): random_value = random.random() if (random_value > generation_constant): coin_vector.append(1) else: coin_vector.append(0) return coin_vector def validate_candidate(vector): vector_sum = sum(vector) sum_test = False if (vector_sum == 7): sum_test = True first_slot = vector[0] first_test = False if (first_slot == 1): first_test = True return (sum_test and first_test) vector1 = generate_candidate() while (validate_candidate(vector1) == False): vector1 = generate_candidate() print vector1, sum(vector1), validate_candidate(vector1) Most of the time, the output is correct, saying something like [1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0] 7 True but sometimes, the output is: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0] 2 False What exactly am I doing wrong?

    Read the article

  • endpoints (bug?) behavior in xts

    - by Alex
    If I try to get endpoints for every year, for example, and do the following: xts.data <- xts(1:10000, order.by=seq(from=as.Date("1970-01-01"), by=1, len=10000)) z <- endpoints(xts.data, on="months", k=12) The return value for this is: > z [1] 0 10000 Same with numbers higher than 12. Why would xts simply not return indices of every year or 13th month starting from the beginning. Is it limited by the number of periods in a year? That is if i did: z <- endpoints(xts.data, on="weeks", k=54) This seems to work fine.

    Read the article

  • JQuery input hidden bug

    - by Abude
    this is the code: Jsfiddle when you clear the url filed and leave it empty the input is hidden and disappear , need to return to the input tag wit h display if the value is empty by click or tab. i have a form with inputs the url input is edited by clicking on the link double click or click next to the link that means in the div area when it's done editing it converts the text to link the Problem: when you leave the input empty it make the attribute of the code and the input attribute hidden so no info is show neither can type an info. how can i make if that input with the id url0/url1 is empty to return to the input option to make it visible and can type?

    Read the article

  • Perl, waitpid() exit code returning wrong value?

    - by Mike
    Consder this trivial example of fork()ing then waiting for a child to die in Perl #!/usr/bin/perl use strict; use warnings; if (fork() == 0) { exit(1); } waitpid(-1,0); print $?; $perl test.pl 256 I suspect the values of are being shifted upwards because when I do exit(2) in the child, the output becomes 512 I can't seem to find this documented in perl's waitpid. Is this a bug on my system or am I doing something wrong? (btw, my OS is solaris 10)

    Read the article

  • IE6 Hacks: Getting jQuery tools expose to work properly on my site

    - by Wild Thing
    Hi, I am trying to get the jQuery tools 'expose' function to work properly in IE6. The page is http://204.51.246.10:12123/ You will need IE6 to reproduce the bug (obviously :)). If you click 'Sign In' on the top-right corner (below the 'Join Today!' button), you will notice that the sign in panel (fieldset#login-controls) is getting covered by the background of its parent element (.signinPanel), which is very weird. It is not on the top of z-index, and I can't figure out why. Please note that in order to get fieldset#login-controls positioned correctly, I have applied the following hack for IE6: $("#header, #header div, #header table, #header td, #header tr").css("position", "static"); $("fieldset#login-controls *").css("position", "static"); This might be the reason for the above. I am pretty lost at this point, as I can't figure what's going on - any advise would be appreciated. Cheers, Wild Thing

    Read the article

  • Is the valid state domain of a program a regular language?

    - by BCS
    If you look at the call stack of a program and treat each return pointer as a token, what kind of automata is needed to build a recognizer for the valid states of the program? As a corollary, what kind of automata is needed to build a recognizer for a specific bug state? My thought is that if these form regular languages than some interesting tools could be built around that. E.g. given a set of crash/failure dumps, automatically group them and generate a recognizer to identify new instances of know bugs.

    Read the article

  • Comparison of Code Review Tools/Systems

    - by SytS
    There are a number of tools/systems available aimed at streamlining and enhancing the code review process, including: CodeStriker Review Board, code review system in use at VMWare Code Collaborator, commercial product by SmartBear Rietveld, based on Modrian, the code review system in use at Google Crucible, commercial product by Atlassian These systems all have varying feature sets, and differ in degrees of maturity and polish; the selection is a little bewildering for someone who is evaluating code review systems for the frist time. Some of these tools have already been mentioned in other questions/answers on StackOverflow, but I would like to see a more comprehensive comparison of the more popular systems, especially with respect to: integration with source control systems integration with bug tracking systems supported workflow (reviews pre/post commit, review or contiguous/non-contigous revision ranges, etc) deployment/maintenance requirements

    Read the article

  • Delphi 2009 RAD Studio - Desktop Selector not Working

    - by Andreas Rejbrand
    In the Delphi 2009 RAD Studio, there is a drop-down combo box where you can select the desktop configuration, i.e. "Default Layout", "Debug Layout, "Classic Undocked", etc. This has stopped working for me, i.e. when I select a desktop, nothing happens. I think I once "customized" the toolbar, by removing the drop-down combo box, and then I added it again, and ever since it has been dead. Is this a known bug? Is there a cure?

    Read the article

  • Stop IE6 overwriting Button text when setting value.

    - by RoToRa
    I've got a button which contains a span I need for styling: <button id="example" name="example" value="Example"><span>Example</span></button> IE6 has a bug that it submits the contents of the button instead of it's value, but I've already worked around that. Now I need to change the value of the button and the text in the span in JavaScript/jQuery: jQuery("#example").val("Changed").children("span").text("Changed"); However IE6 also changes the contents of the button when setting the value per JavaScript/jQuery, so that the span is lost. Does any one know of a way other that identifying IE6 with jQuery.browser that I could use to avoid that?

    Read the article

  • (this == null) in C#!

    - by SLaks
    Due to a bug that was fixed in C# 4, the following program prints true. (Try it in LINQPad) void Main() { new Derived(); } class Base { public Base(Func<string> valueMaker) { Console.WriteLine(valueMaker()); } } class Derived : Base { string CheckNull() { return "Am I null? " + (this == null); } public Derived() : base(() => CheckNull()) { } } In VS2008 in Release mode, in throws an InvalidProgramException. (In Debug mode, it works fine) In VS2010 Beta 2, it doesn't compile (I didn't try Beta 1); I learned that the hard way Is there any other way to make this == null in pure C#?

    Read the article

  • Is WinForms ListView in VirtualMode limited to 100,000,000 rows?

    - by damageboy
    I have some grid scenario with 500,000,000 rows I would like to display in ListView. If I artificially limit my ListView to display 100,000,000: _listView.VirtualListSize = _data.Count; if (_listView.VirtualListSize > 100000000) _listView.VirtualListSize = 100000000; Everything works fine (In VirtualMode naturally). When I change my code to: _listView.VirtualListSize = _data.Count; if (_listView.VirtualListSize > 100000001) _listView.VirtualListSize = 100000001; The ListView display an empty grid... Is this a Microsoft Bug? Where is this coming from? Is this a Win32 ListView limitation? Most importantly, why is this not documented?

    Read the article

  • asp.net could a half submitted web page be processed?

    - by c00ke
    Having a weird bug in production and just wondering if it's possible for a half submitted web page to processed by the server? The page has no view state just using plain old html controls and accessing data displayed in repeater on the back end via Request.Form[name] etc. Is it possible for a request to be truncated perhaps due to lost internet connection and the page still processed by the server. Therefore if field not part of the request Request.Form[name] could result in null? I know can use fiddler to modify request but unfortunately we are not allowed to change group policy and change the proxy! Many Thanks

    Read the article

  • MediaWiki: workaround for watchlist restricted to 7 days?

    - by Mark Robinson
    I'm running MediaWiki 1.13.2. When I go to my watchlist, I'm limited to only viewing the last 7 days (though MediaWiki should permit me to view the last 30 days). I've tried: Clicking the all button Changing the URL Changing the settings under MY PREFERENCES-WATCHLIST. When I save (e.g. 30 days) it resets back to 7 days! Googling for an answer... Anyone heard of this bug before? Is there a workaround? I can't see a setting anywhere for this to be a maximum.

    Read the article

  • Cannot convert []string to []interface {}

    - by karlrh
    I'm writing some code, and I need it to catch the arguments and pass them through fmt.Println (I want its default behaviour, to write arguments separated by spaces and followed by a newline). However it takes []interface {} but flag.Args() returns a []string. Here's the code example package main import ( "fmt" "flag" ) func main() { flag.Parse() fmt.Println(flag.Args()...) } This returns the following error: ./example.go:10: cannot use args (type []string) as type []interface {} in function argument Is this a bug? Shouldn't fmt.Println take any array? By the way, I've also tried to do this: var args = []interface{}(flag.Args()) but I get the following error: cannot convert flag.Args() (type []string) to type []interface {} Is there a "Go" way to workaround this?

    Read the article

  • How can I get the following compiled on UVA?

    - by Michael Tsang
    Note the comment below. It cannot compiled on UVA because of a bug in GCC. #include <cstdio> #include <cstring> #include <cctype> #include <map> #include <stdexcept> class Board { public: bool read(FILE *); enum Colour {none, white, black}; Colour check() const; private: struct Index { size_t x; size_t y; Index &operator+=(const Index &) throw(std::range_error); Index operator+(const Index &) const throw(std::range_error); }; const static std::size_t size = 8; char data[size][size]; // Cannot be compiled on GCC 4.1.2 due to GCC bug 29993 // http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29993 typedef bool CheckFunction(Colour, const Index &) const; CheckFunction pawn, knight, bishop, king, rook; bool queen(const Colour c, const Index &location) const { return rook(c, location) || bishop(c, location); } static char get_king(Colour c) { return c == white ? 'k' : 'K'; } template<std::size_t n> bool check_consecutive(Colour c, const Index &location, const Index (&offsets)[n]) const { for(const Index *p = offsets; p != (&offsets)[1]; ++p) { try { Index target = location + *p; for(; data[target.x][target.y] == '.'; target += *p) { } if(data[target.x][target.y] == get_king(c)) return true; } catch(std::range_error &) { } } return false; } template<std::size_t n> bool check_distinct(Colour c, const Index &location, const Index (&offsets)[n]) const { for(const Index *p = offsets; p != (&offsets)[1]; ++p) { try { Index target = location + *p; if(data[target.x][target.y] == get_king(c)) return true; } catch(std::range_error &) { } } return false; } }; int main() { Board board; for(int d = 1; board.read(stdin); ++d) { Board::Colour c = board.check(); const char *sp; switch(c) { case Board::black: sp = "white"; break; case Board::white: sp = "black"; break; case Board::none: sp = "no"; break; } std::printf("Game #%d: %s king is in check.\n", d, sp); std::getchar(); // discard empty line } } bool Board::read(FILE *f) { static const char empty[] = "........" "........" "........" "........" "........" "........" "........" "........"; // 64 dots for(char (*p)[size] = data; p != (&data)[1]; ++p) { std::fread(*p, size, 1, f); std::fgetc(f); // discard new-line } return std::memcmp(empty, data, sizeof data); } Board::Colour Board::check() const { std::map<char, CheckFunction Board::*> fp; fp['P'] = &Board::pawn; fp['N'] = &Board::knight; fp['B'] = &Board::bishop; fp['Q'] = &Board::queen; fp['K'] = &Board::king; fp['R'] = &Board::rook; for(std::size_t i = 0; i != size; ++i) { for(std::size_t j = 0; j != size; ++j) { CheckFunction Board::* p = fp[std::toupper(data[i][j])]; if(p) { Colour ret; if(std::isupper(data[i][j])) ret = white; else ret = black; if((this->*p)(ret, (Index){i, j}/* C99 extension */)) return ret; } } } return none; } bool Board::pawn(const Colour c, const Index &location) const { const std::ptrdiff_t sh = c == white ? -1 : 1; const Index offsets[] = { {sh, 1}, {sh, -1} }; return check_distinct(c, location, offsets); } bool Board::knight(const Colour c, const Index &location) const { static const Index offsets[] = { {1, 2}, {2, 1}, {2, -1}, {1, -2}, {-1, -2}, {-2, -1}, {-2, 1}, {-1, 2} }; return check_distinct(c, location, offsets); } bool Board::bishop(const Colour c, const Index &location) const { static const Index offsets[] = { {1, 1}, {1, -1}, {-1, -1}, {-1, 1} }; return check_consecutive(c, location, offsets); } bool Board::rook(const Colour c, const Index &location) const { static const Index offsets[] = { {1, 0}, {0, -1}, {0, 1}, {-1, 0} }; return check_consecutive(c, location, offsets); } bool Board::king(const Colour c, const Index &location) const { static const Index offsets[] = { {-1, -1}, {-1, 0}, {-1, 1}, {0, 1}, {1, 1}, {1, 0}, {1, -1}, {0, -1} }; return check_distinct(c, location, offsets); } Board::Index &Board::Index::operator+=(const Index &rhs) throw(std::range_error) { if(x + rhs.x >= size || y + rhs.y >= size) throw std::range_error("result is larger than size"); x += rhs.x; y += rhs.y; return *this; } Board::Index Board::Index::operator+(const Index &rhs) const throw(std::range_error) { Index ret = *this; return ret += rhs; }

    Read the article

  • How to configure bugzilla to not advance to next bug when updating?

    - by WilliamKF
    By default, when you apply changes to a Bugzilla entry, the web interface advances to the next bug in your list. I would like to disable this feature since it is almost never what I desire, planning to make further updates later. Further, I often update the wrong bug subsequently due to its changing the current bug without my noticing. How do I configure Bugzilla to not advance like this?

    Read the article

  • Custom DataType in DataTemplate breaks WPF designer

    - by PRINCESS FLUFF
    Why does the DataTemplate line break the WPF designer in Visual Studio 2008? The program compiles and runs properly. The DataTemplate is applied as it should. However the entire DataTemplate block of code is underlined in red, and when I simply "build" the program without running, I get the error "Type reference cannot find public type named 'Character'" How come it can't find it in the designer yet the program applies the template properly? <UserControl x:Class="WPF_Tests.Tests.TwoCollecViews.TwoViews" xmlns="http://schemas.microsoft.com/winfx/2006/xaml/presentation" xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml" xmlns:DetailsPane="clr-namespace:WPF_Tests.Tests.DetailsPane" > <UserControl.Resources> <DataTemplate DataType="{x:Type DetailsPane:Character}"> <StackPanel Orientation="Horizontal"> <TextBlock Text="{Binding Path=Name}"></TextBlock> </StackPanel> </DataTemplate> </UserControl.Resources> <Grid> <ListBox ItemsSource="{Binding Path=Characters}" /> </Grid> </UserControl> EDIT: I am being told that this may be a bug in Visual Studio 2008, as it worked correctly in 2010. You can download the code here: http://www.mediafire.com/?z1myytvwm4n - The Test/TwoCollec xaml file's designer will break with this code.

    Read the article

  • What headaches should I expect from using Trac?

    - by Dolph Mathews
    No tool is perfect, and I'm about to start several long-term projects using Trac, and wanted a heads up of the kinds of problems I may or may not experience with it. In other words, Trac meets my needs in the short term, and I've already made the decision to use it, but I want to know what to expect down the road. I am not looking for: "Use product X instead of Trac because..." answers. "Trac is great because..." answers. A comparison to any other specific system. "Trac doesn't support Feature X" answers. I can read the feature list too, thank you very much. I am looking for: "Feature X does not behave as expected..." "Trac behaves oddly when..." "Trac doesn't fully support..." "Trac itself has a known bug that will likely never be fixed..." And especially "Trac can't handle..." etc So, what Trac-induced headaches do I have to look forward to? For future reference, this question was asked while Trac v0.11 was the latest stable release.

    Read the article

  • Why am I having this InstantiationException in Java when accessing final local variables?

    - by Oscar Reyes
    I was playing with some code to make a "closure like" construct ( not working btw ) Everything looked fine but when I tried to access a final local variable in the code, the exception InstantiationException is thrown. If I remove the access to the local variable either by removing it altogether or by making it class attribute instead, no exception happens. The doc says: InstantiationException Thrown when an application tries to create an instance of a class using the newInstance method in class Class, but the specified class object cannot be instantiated. The instantiation can fail for a variety of reasons including but not limited to: - the class object represents an abstract class, an interface, an array class, a primitive type, or void - the class has no nullary constructor What other reason could have caused this problem? Here's the code. comment/uncomment the class attribute / local variable to see the effect (lines:5 and 10 ). import javax.swing.*; import java.awt.event.*; import java.awt.*; class InstantiationExceptionDemo { //static JTextField field = new JTextField();// works if uncommented public static void main( String [] args ) { JFrame frame = new JFrame(); JButton button = new JButton("Click"); final JTextField field = new JTextField();// fails if uncommented button.addActionListener( new _(){{ System.out.println("click " + field.getText()); }}); frame.add( field ); frame.add( button, BorderLayout.SOUTH ); frame.pack();frame.setVisible( true ); } } class _ implements ActionListener { public void actionPerformed( ActionEvent e ){ try { this.getClass().newInstance(); } catch( InstantiationException ie ){ throw new RuntimeException( ie ); } catch( IllegalAccessException ie ){ throw new RuntimeException( ie ); } } } Is this a bug in Java? edit Oh, I forgot, the stacktrace ( when thrown ) is: Caused by: java.lang.InstantiationException: InstantiationExceptionDemo$1 at java.lang.Class.newInstance0(Class.java:340) at java.lang.Class.newInstance(Class.java:308) at _.actionPerformed(InstantiationExceptionDemo.java:25)

    Read the article

  • Why would the IE Developer Toolbar claim a style is applied, yet that supposed fact is not reflected

    - by Deane
    I have a situation where IE7 is simply not applying styles, even though it claims it is. I have an element on my page. In the CSS, I have defined a rule that should apply "display: none" to it, so it should not be displayed. It's still displaying. I downloaded the IE Developer Toolbar, and found the element in the DOM selector. I right-clicked and selected "Applied Styles." Right there, IE claims that it is applying my "display: none" rule. In fact, the "Applied Styles" dialog confirms everything I think I know about my CSS and how it should be applied. Yet the element remains. Now, I'm not asking anyone to debug my CSS here. I'm asking, if the IE Developer Toolbar claims/confirms this element should be gone, but it's still there...what does that mean, exactly? Since the Toolbar is on my side, I think my CSS is fine. Is there some IE7 bug I'm not considering? Edit: One thing that might be relevant: the LINK elements that load the stylesheets are applied to the page in Javascript, via "document.write". I'm starting to suspect that has something to do with it.

    Read the article

< Previous Page | 45 46 47 48 49 50 51 52 53 54 55 56  | Next Page >