| Author |
 Topic  |
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
 Posted - Jan 18 2008 : 3:05:11 PM
|
How quickly is something supposed to show up? I have done something to most of these functions (97 out of 116) and still no sign of anything.
I have installed the DirectX SDK Aug2006 since I had that one downloaded all ready, onto a winXP test sysem using VS2005 and VA 1624.
Is VA supposed to have, or required to have, some knowledge of the DirextX SDK? I am not seeing any reference to it in your VA about information.
Since I am not sure what you mean by / require by implementing the functions I have been converting the lines from:
STDMETHOD(SetVertexShaderConstantB)(THIS_ UINT StartRegister,CONST BOOL* pConstantData,UINT BoolCount) PURE;
into:
STDMETHOD(SetVertexShaderConstantB)(THIS_ UINT StartRegister,CONST BOOL* pConstantData,UINT BoolCount)
{
}
And nothing is happening, other than me concluding that this needs doing with a macro, or some form of search and replace, since it is terribly boring. |
zen is the art of being at one with the two'ness |
 |
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 18 2008 : 3:14:03 PM
|
quote:
Originally posted by feline
How quickly is something supposed to show up? I have done something to most of these functions (97 out of 116) and still no sign of anything.
Well, it's supposed to show up faster than I would normally type it right? That is, I don't want to sit around waiting for VAX to recognize a symbol. I want it to work when I start typing basically.
quote:
Originally posted by feline
Is VA supposed to have, or required to have, some knowledge of the DirextX SDK? I am not seeing any reference to it in your VA about information.
My experience shows that VAX can find the correct headers by using the IDE-defined PATH (which is where I've defined the path to the DirectX SDK at the moment. I'm able to use the Alt+G feature just fine, which leads me to believe VAX has found and parsed the header, right?
quote:
Originally posted by feline
Since I am not sure what you mean by / require by implementing the functions I have been converting the lines from:
STDMETHOD(SetVertexShaderConstantB)(THIS_ UINT StartRegister,CONST BOOL* pConstantData,UINT BoolCount) PURE;
into:
STDMETHOD(SetVertexShaderConstantB)(THIS_ UINT StartRegister,CONST BOOL* pConstantData,UINT BoolCount)
{
}
And nothing is happening, other than me concluding that this needs doing with a macro, or some form of search and replace, since it is terribly boring.
In the example I posted a link to above, you'll notice that each overridden method calls the "real" DX interface, like so:
STDMETHOD(SetVertexShaderConstantB)(THIS_ UINT StartRegister,CONST BOOL* pConstantData,UINT BoolCount)
{
return m_device->SetVertexShaderConstantB(StartRegister, pConstantData, BoolCount);
}
The problems I'm seeing happen when I want the list of functions for m_device, as well as auto-complete for the local variables (i.e. StartRegister, pConstantData, and BoolCount). |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 18 2008 : 3:38:23 PM
|
Apologies, I am not making myself clear at all. How many functions did you have to implement before you started seeing a problem? 2? 20? All 100+?
If the problem should show up after 5 functions I don't want to sit here and type in all 100+ of them without seeing any problems.
When you say "IDE-defined PATH" do you mean the projects C++ additional include directories?
Or do you mean:
IDE tools menu -> Options -> Projects and Solutions -> VC++ Directories -> Show directories for = Include files
There was no sign in your VA about information that you had used the second method.
Have you modified the code in this project? The file:
c:\\src_2005\\Direct3D-Hook\\Direct3D-Hook\\MyDirect3DDevice9.h
does not contain any #include lines, so on my system "IDirect3DDevice9" is not coloured as a type. VA does not know about this class. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 18 2008 : 3:48:50 PM
|
quote:
Originally posted by feline
Apologies, I am not making myself clear at all. How many functions did you have to implement before you started seeing a problem? 2? 20? All 100+?
It appeared quite fast, like after 2 or 5. Then it happened for pretty much all the methods (although sometimes it would work perfectly for e.g. 5 functions in a row).
quote:
Originally posted by feline
When you say "IDE-defined PATH" do you mean the projects C++ additional include directories?
Or do you mean:
IDE tools menu -> Options -> Projects and Solutions -> VC++ Directories -> Show directories for = Include files
There was no sign in your VA about information that you had used the second method.
The latter. I apologize too, though; the DirectX SDK is only a recent addition to my solution (I added it a few days ago), which is why the paths don't show up in the VA information I pasted earlier.
quote:
Originally posted by feline
Have you modified the code in this project? The file:
c:\\src_2005\\Direct3D-Hook\\Direct3D-Hook\\MyDirect3DDevice9.h
does not contain any #include lines, so on my system "IDirect3DDevice9" is not coloured as a type. VA does not know about this class.
Well, you're right the file doesn't contain any #include lines. Nevertheless, it compiles fine and VAX is able to deduce the type information on my end at least. Furthermore, in my own project, which has a similar class, I had an #include line at the top of the file and I still noticed the mysterious VAX behavior. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 18 2008 : 8:29:46 PM
|
I have added:
C:\\Program Files\\Microsoft DirectX SDK (August 2006)\\Include
to the IDE's list of include directories and restarted the IDE. This directory is now showing up in VA's list of stable include directories.
I reverted back to the clean solution, as downloaded, and re-ran this test. I have now overloaded the first 20 functions and I am not seeing any problem.
How "extreme" is this problem? Is it likely that I am going to miss it while typing at a slow / medium speed? I am having to pay close attention to what I am doing, but I am always getting and accepting listboxes, and rarely typing more than a 5 or 6 characters to trigger anything, function or parameter names. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 19 2008 : 04:34:00 AM
|
| If you're interested, I'm up for letting you remote control my system so you can see for yourself what happens. It sounds like you can't reproduce it. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 19 2008 : 1:53:54 PM
|
That may be worth trying eventually. For now the basic problem with this is that even if I see the problem it probably will not offer any further clues to help me reproduce it here.
Given this test project, and an easy way to reproduce the problem, can you try the following two tests please.
Firstly press the button:
VA Options -> Performance -> Rebuild symbol databases
then restart the IDE, and simply load this test project. When VA has finished parsing, and has applied its enhanced syntax highlighting to the code can you re-test and see if you still see the problem?
Second test, can you disable the IDE's intellisense by renaming "feacp.dll" as explained here:
http://docs.wholetomato.com?W133
Do either of these make any difference? |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 25 2008 : 06:36:07 AM
|
I apologize for the delayed response, feline, but I've been swamped with work.
When I first opened the solution, I went to the MyDirect3DDevice9 class, deleted all the overridden methods, went to d3d9.h and copied all the pure virtual functions from line 399 into MyDirect3DDevice9. Then I started overriding the methods by calling the real implementation on m_device. The first and second functions didn't work. IntellliSense was unable to deduce the type of m_device until after a while, and VAX didn't suggest any of the functions' parameters unless I waited.
Does the IntelliSense disabling you mention in http://docs.wholetomato.com?W133 work for VS2008 too?
Over the past week, I've noticed more and more cases where VAX doesn't work. Usually, as you know, I can't get any symbols on my objects which is really frustrating since I'll have to manually go to the class to copy the function name if I can't remember its exact spelling.
I have multiple projects that have files of the same name (the files are totally unrelated since they're in different projects, but they share a common name). However, when I use Alt+O to switch between the .h and .cpp files, VAX gets confused and suggests multiple files, i.e. not just the file in the current project but also the ones in the other projects. This is quite annoying, since it requires several extra keypresses (e.g. I need to select the right file from the suggestion box and click Enter).
Once I had rebuilt the entire VAX database I carried out the same scenario as above with the same results.
It's worth noting, that if I stay in the same view (i.e. don't scroll up or down) then it's like VAX gets better to recognize the symbols (it's as if it gets time to analyze it). However, once I scroll I'm back to square one. In fact, after I've scrolled and I insert an opening bracket ({) then VAX fails to insert the closing bracket.
Once again, I'm sorry for the slow reply. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 25 2008 : 08:28:24 AM
|
VS2008 intellisense, I am seeing the file:
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\vcpackages\\feacp.dll
which is such a good match for the VS2005 file that I renamed it, and the IDE loaded quite happily. The IDE loads quite happily with this renamed, but I am not sure how much difference it made.
Have you changed IDE's? I was doing my testing in VS2005.
Alt-o, you are describing a well known problem. This happens when VA cannot pair up the files. This normally happens when only some of the files have been added to the solution. Are all of the files actually listed in solution explorer? If so can you add the missing files to the solution. This should fix the problem.
The problems you are seeing with the virtual functions, I am not sure what is going on. Do you have time to see how quickly and easily you can reproduce this in the "test" project you referenced, that I downloaded? |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 25 2008 : 11:48:46 AM
|
quote:
Originally posted by feline
VS2008 intellisense, I am seeing the file:
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\vcpackages\\feacp.dll
which is such a good match for the VS2005 file that I renamed it, and the IDE loaded quite happily. The IDE loads quite happily with this renamed, but I am not sure how much difference it made.
Disabling IntelliSense by following the method outlined in your forum post did not fix the issues. Instead, VAX became less useful since it now behaves like it would if I disable the "Get content from default Intellisense". That is, VAX does not recognize the symbols instantly, it has problems with STL containers, it doesn't insert ending brackets, etc, etc.
quote:
Originally posted by feline
Have you changed IDE's? I was doing my testing in VS2005.
I've been using VS2008 all along.
quote:
Originally posted by feline
Alt-o, you are describing a well known problem. This happens when VA cannot pair up the files. This normally happens when only some of the files have been added to the solution. Are all of the files actually listed in solution explorer? If so can you add the missing files to the solution. This should fix the problem.
Yes, all files are in the solution, albeit in different projects.
quote:
Originally posted by feline
The problems you are seeing with the virtual functions, I am not sure what is going on. Do you have time to see how quickly and easily you can reproduce this in the "test" project you referenced, that I downloaded?
I'm not sure if I can reproduce it easily. I'll see what I can do |
Edited by - dreijer on Jan 25 2008 11:49:03 AM |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 25 2008 : 12:18:14 PM
|
Here's another thing I've noticed. Take a look at the following piece of code:
WinCountsCont::iterator nextUserEntry = winCounts.find( (*nextUserItor).GetUserId() );
...
nextUserEntry->second.second++;
When I have the "Get content from default Intellisense" feature enabled, I get auto-completion when I type the last line; that is, I get a suggestion for 'nextUserEntry' when I start typing the variable name, I get the list box of functions when I type 'nextUserEntry->', and I get another list of functions when I type 'nextUserEntry->second.' However, if I move the cursor away from the line to another location and wait for a few seconds, the '->second' gets underlined by VAX in red. If I hover the mouse cursor over the symbol, I can get the type of the variable just fine, and clicking 'Go To Definition' from the right-click shortcut menu takes me to the right spot. This is all IntelliSense though (the stuff that works  ). The question is why VAX has problems with it. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 25 2008 : 1:41:10 PM
|
Alt-o first, can you paste the full path names of the problem files please? I can then try and setup a similar test here.
clearly VA is having problems with STL code in your project, but I am not sure why. What you describe with the underlining of "second" makes sense if VA does not understand what "nextUserEntry" is.
How is "WinCountsCont" defined?
If you place the caret into this and look at VA's context and definition fields do they show the correct information?
Hopefully we can make a bit more progress with this underlining problem, since that is more clear cut. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 25 2008 : 1:49:59 PM
|
quote:
Originally posted by feline
Alt-o first, can you paste the full path names of the problem files please? I can then try and setup a similar test here.
Here are two of the problem files:
c:\\Users\\dreijer\\Documents\\DTU\\Final Project\\src\\Evolve\\src\\Shared\\DataEntities\\Game.h
c:\\Users\\dreijer\\Documents\\DTU\\Final Project\\src\\Evolve\\src\\SDK\\Game.cpp
quote:
Originally posted by feline
How is "WinCountsCont" defined?
typedef map<unsigned long, pair<User, unsigned long> > WinCountsCont;
Both std::map and std::pair are defined at the top of the source file with 'using std::XX;'
quote:
Originally posted by feline
If you place the caret into this and look at VA's context and definition fields do they show the correct information?
I guess the context and definition fields are the one above the source window, right? No information is shown for the variable. However, correct information is shown for both 'nextUserEntry' and '.second'.
Just for the record, whenever VAX fails to pop up the suggestion list box or recognize a symbol, the context and definition fields are empty (as I've mentioned earlier).
I hope we can make some progress soon too. I've spent quite a bit of time on this already, heh. :)
Oh, and also, I expect that VA has issues with the virtual base classes since IntelliSense has problems parsing it as well (I make extensive use of boost::shared_ptr's and other similar constructs to hold the pointers). |
Edited by - dreijer on Jan 25 2008 1:54:24 PM |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 25 2008 : 2:18:59 PM
|
For the alt-o problem files, are there any other files called "Game"? Or only those two?
I am seeing something with the iterator here, currently trying to pin exactly when the problem starts. Can you try changing the typedef from:
typedef map<unsigned long, pair<User, unsigned long> > WinCountsCont;
to:
typedef std::map<unsigned long, std::pair<User, unsigned long> > WinCountsCont;
and see if this makes any difference? |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 25 2008 : 2:24:19 PM
|
quote:
Originally posted by feline
For the alt-o problem files, are there any other files called "Game"? Or only those two?
Only those two.
quote:
Originally posted by feline
I am seeing something with the iterator here, currently trying to pin exactly when the problem starts. Can you try changing the typedef from:
typedef map<unsigned long, pair<User, unsigned long> > WinCountsCont;
to:
typedef std::map<unsigned long, std::pair<User, unsigned long> > WinCountsCont;
and see if this makes any difference?
That sort of fixed it. Now the first '->second' is recognized, but '.second' isn't. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 25 2008 : 2:52:59 PM
|
I have "reduced" the typedef problem down to this "simple" test case:
#include <map>
#include <utility>
using std::map;
using std::pair;
typedef map<long, long> typedefMapViaUsing;
typedef std::map<long, long> typedefMapFullName;
typedef map<long, pair<long, long> > typedefComplexMapViaUsing;
typedef std::map<long, std::pair<long, long> > typedefComplexMapFullName;
static void testTemplateUnderliningAndListboxes()
{
typedefMapViaUsing::iterator itTypeViaUsing;
itTypeViaUsing->second; // understood correctly
itTypeViaUsing; // spontaneous listbox when typing ->
typedefMapFullName::iterator itTypeFullName;
itTypeFullName->second; // understood correctly
itTypeFullName; // spontaneous listbox when typing ->
typedefComplexMapViaUsing::iterator itTypeComplexViaUsing;
itTypeComplexViaUsing->second; // "second" underlined
itTypeComplexViaUsing; // WRONG spontaneous listbox when typing ->
typedefComplexMapFullName::iterator itTypeComplexFullName;
itTypeComplexFullName->second; // understood correctly
itTypeComplexFullName->second.second; // understood correctly
itTypeComplexFullName; // spontaneous listbox when typing ->
itTypeComplexFullName->second; // NO spontaneous listbox when typing Dot
map<long, pair<long, long> >::iterator itComplexViaUsing;
itComplexViaUsing->second; // no underlining, definition field BLANK for "second"
itComplexViaUsing->second.second; // no underlining, definition field and alt-g work for final "second"
itComplexViaUsing; // NO spontaneous listbox when typing ->
itComplexViaUsing->second; // spontaneous listbox when typing Dot
std::map<long, std::pair<long, long> >::iterator itComplexFullName;
itComplexFullName->second; // understood correctly
itComplexFullName->second.second; // no underlining, definition field BLANK, alt-g does nothing for final "second"
itComplexFullName; // spontaneous listbox when typing ->
itComplexFullName->second; // NO spontaneous listbox when typing Dot
}
I don't see any obvious work arounds at this point. Even removing the typedef completely does not seem to be an ideal solution.
You might have some success adding a #define version of your typedef to VA's "StdAfx.h" file as explained in this FAQ entry:
http://docs.wholetomato.com?W302
perhaps:
#define WinCountsCont std::map<unsigned long, std::pair<User, unsigned long> >
at the bottom. However I have not tested this, and I am not sure how many typedef's you have.
This file is used to help VA's parser with difficult code, and can be used to work around odd effects. After modifying this file you need to rebuild the VA symbol database for the changes to take effect:
VA Options -> Performance -> General -> Rebuild symbol databases |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 25 2008 : 3:07:17 PM
|
Adding each and everyone of my typedefs to that file is just not a viable solution. I have way too many containers in my code to add all of them to StdAfx.h so that VA recognizes the symbol. The required rebuilding of the database makes it an even less attractive solution.
I just tried swapping my typedef with the real actual std::map definition. This did *not* fix the issue. Now the '->second' isn't highlighted in red but it still doesn't show up in the VA context and definition bars.
The following also gives problems (I've highlighted the items in red that VA highlights in red):
multimap< unsigned long, User > sortedUsers;
for (map< unsigned long, pair<User, unsigned long> >::const_iterator curUserItor
= winCounts.begin(); curUserItor != winCounts.end(); ++curUserItor)
{
sortedUsers.insert( std::make_pair( curUserItor->second.second, curUserItor->second.first ) );
}
So, where does this leave us? Do you have any ETA on when this bug might be fixed? Are we assuming that this issue is what's causing the other problems I'm seeing? I'm seriously considering downgrading to an older version just to have a VA that actually aids me in programming rather than requiring me to play nurse while typing.
*edited by feline to make this narrower* |
Edited by - feline on Jan 28 2008 1:19:38 PM |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 25 2008 : 3:54:57 PM
|
I have put a bug report in for that test code:
case=11842
Note that using that test code I get identical results, identical failures, in VA 1561.
The bug report is down as urgent, but unfortunately I have no estimate on when it will be done. I am not sure how easy this will be.
The fact that I get the same results on this code in VA 1561 is interesting. Do you remember what version of VA you were using before?
I still need to get to the bottom of why VA is getting worse for you.
Alt-o, I have just created the required directories. I now have the same two files as you, in the same directories. I have added both files to a VS2005 solution. They are in the same project inside the solution. Having just added the files alt-o is working perfectly.
Restart the IDE, alt-o is still working perfectly. Open File dialog, filtered on "game", only lists these two files.
I assume my test situation is different to the situation you have. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 26 2008 : 09:46:48 AM
|
quote:
Originally posted by feline
The fact that I get the same results on this code in VA 1561 is interesting. Do you remember what version of VA you were using before?
I think it was 1549 or 1541.
Don't you just love bugs that can't be reproduced in-house? |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 26 2008 : 3:42:55 PM
|
Since you ask, no!
Do you feel that the problems with these STL iterators is a new bug in VA 16xx? There seem to be several different things going on, and going wrong here.
win2k, VS2005, VA 1559, the same test case, I am seeing the same results.
win2k, VS2005, VA 1549, the same test case, I am seeing slightly different results. I would say VA is working slightly worse in this version. Certainly it is not significantly better.
win2k, VS2005, VA 1541, the same test case, this is significantly better. In one case I am getting the wrong listbox content in 1541 where I get the correct listbox content with VA 1626, so it is not an improvement in all areas. However this does now qualify as a regression.
Finally getting somewhere!
Since this is a regression I have bumped the bug priority one notch up to the top group.
This does not help with the alt-o problem, but it does hopefully help with some / most of the problems you are seeing, or will when this is fixed. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 26 2008 : 4:09:13 PM
|
quote:
Originally posted by feline
Do you feel that the problems with these STL iterators is a new bug in VA 16xx? There seem to be several different things going on, and going wrong here.
If I recall correctly, I remember seeing red underlined words in the previous version I used too (although I'm not sure they were in the same location as our previous examples). In previous versions, VA worked as expected, though, with correct list box suggestions and recognized symbols. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 28 2008 : 08:03:10 AM
|
Are you seeing problems with listboxes and underlining on Non STL code?
The test case I have, and the bug report, will hopefully address quite a few of the problems you are seeing with STL code.
Which raises the question, how many of the problems you are seeing are actually with STL code? |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 28 2008 : 09:29:14 AM
|
| The DX project we were looking at earlier didn't contain any STL code, and VA had just as many problems with those symbols. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 28 2008 : 1:23:43 PM
|
This is true. This feeds back to my confusion over IDE versions, although that probably does not matter. VA 1614 was the first public build to support VS2008 Beta 2. So in the VA 15xx builds you must have been using VS2005, or earlier, if you were working with VA support.
Do you have time to try a fairly quick DX project test please?
If you use the button:
VA Options -> Performance -> Rebuild symbol databases
then restart the IDE, but only load the DX test project you sent me. Wait for VA to finish parsing and apply syntax highlighting to the current file. If you now implement the virtual functions do you run into the same problem? Or does this work?
I am wondering if you are seeing this problem because something in your main solution is corrupting VA's symbol database somehow.
If you see the problem quickly and easily under these conditions then we just need to find the differences between my system and yours, and try to eliminate them one by one. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 28 2008 : 2:09:45 PM
|
| Rebuilding the symbol database and opening only the DX project didn't fix anything. It was broken instantly when I started overriding an arbitrary method in the class. |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 28 2008 : 2:32:36 PM
|
Thank you. That is surprisingly clear. Did you copy / paste in a load of files from another file? Or just set to work in the test project? I want to make sure I am running a sensible / valid test here, otherwise I may end up testing the wrong thing.
Can you please go to:
VA Options -> System Info -> Copy Info
and paste the details (from the clipboard) into this thread. This will give us the basic information about your setup. I will then try to get as close as possible to that information here. |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 28 2008 : 3:35:19 PM
|
Only the files that were originally in the solution when I downloaded it (plus d3d9.h as described earlier) are open in that test project.
VA_X.dll file version 10.4.1624.0 built 2007.12.07
DevEnv.exe version 9.0.21022.8
msenv.dll version 9.0.21022.8
Font: Courier New 13(Pixels)
Comctl32.dll version 6.10.6000.16386
Windows Vista 6.0 Build 6000
2 processors
Platform: Win32
Stable Includes:
C:\\Program Files\\Microsoft DirectX SDK (November 2007)\\Include;
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\include;
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\atlmfc\\include;
;
C:\\Program Files\\Microsoft SDKs\\Windows\\v6.0A\\include;
Other Includes:
Stable Source Directories:
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\atlmfc\\src\\mfc;
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\atlmfc\\src\\mfcm;
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\atlmfc\\src\\atl;
C:\\Program Files\\Microsoft Visual Studio 9.0\\VC\\crt\\src; |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 28 2008 : 5:34:07 PM
|
| I am currently downloading the DirectX November 2007 SDK. Once I have that I will set up a test and see what happens. |
zen is the art of being at one with the two'ness |
|
|
|
feline
Whole Tomato Software
United Kingdom
18939 Posts |
Posted - Jan 29 2008 : 1:55:49 PM
|
I have taken the following steps.
* Start with a Vista machine with VS2008 and DirectX SDK, IDE include directories matching yours
* open the directX test project in VS2008, it needs converting, so do that
* close and re-load the IDE to make sure the project opens happily
* open the IDE, and open the two files:
c:\\code 2008\\Direct3D-Hook\\Direct3D-Hook\\MyDirect3DDevice9.h
C:\\Program Files\\Microsoft DirectX SDK (November 2007)\\Include\\d3d9.h
so that they will be open after the symbol rebuild.
* tell VA to rebuild its symbol database, and restart VS2008
* wait for VA to finish parsing and to apply its syntax highlighting. At this point the code in the test project has not been modified at all.
* sitting in the SDK header file select from line 399 down to line 519 and copy this block of code to line 645 in the test project header file
I have worked down this block, implementing functions, as far as:
STDMETHOD_(void, SetCursorPosition)(THIS_ int X,int Y,DWORD Flags)
From your description I should be seeing something by now. I am not, everything is working perfectly.
Unless there is some key fact / step I am missing log files are the next step.
Can you please rebuild your symbol database and reproduce the problem on the test project, to try and keep things simple, with logging enabled.
Please see this FAQ for details of turning on VA's logging, and sending us the log files
http://docs.wholetomato.com?W305 |
zen is the art of being at one with the two'ness |
|
|
|
dreijer
Ketchup Master
62 Posts |
Posted - Jan 30 2008 : 03:02:23 AM
|
| I've submitted the log files. |
|
|
|
Topic |
|
My issues with 1624
