|
Hi,
I use VC 6 with SP4. CFile is nice but part of this project I can't use MFC. I also don't want to do fopen(), so I attempt C++ stream's way of life. But the unnecessary pain is ... to me as a 'iostream' newbie.
Two problems hurt most:
(1) I am working on a Dialog based project (with MFC support but my work is on a non-MFC utility type helper). Everything is going fine until I try to use the C++ file class to open a file for binary i/o. When I added the line #include "fstream.h" in the, say, MyDlgApp.cpp all hell breaks loose. The compiler gives me no end of complaints, over 100s, shouting error C2872: 'streambuf' : ambiguous symbol or 'ostream' : ambiguous symbol or 'ios' : ambiguous symbol etc. etc. That is before I even declare an open file stream object yet! This came just when I included the header!!
Not understand why, I created a new test Dialog application and entered #include <fstream.h> as I did before. This time everything is OK and I can even 'really' open a file. It seems something in my original app have broken. But what? and how to fix it?. Can someone help. It is driving me crazy.
(2) It is a real misery to me, I coded
ifstream in;
in.open("test.txt", ios::nocreate | ios::binary);
if (!in) { return 0;}
then the compiler shouted
error C2039: 'nocreate' : is not a member of 'basic_ios<char,struct std::char_traits<char=""> >'
error C2065: 'nocreate' : undeclared identifier
The misery to me is when I took away ios::nocreate (leaving ios::binary) then everything is OK again. I see both of these are defined as an enum in ios.h, so why one is accepted and the other not? 
Am I doing something stupid here, or I should include something else as well. Your suggestion is greatly appreciated. ![Rose | [Rose]](https://www.codeproject.com/script/Forums/Images/rose.gif)
Will
|
|
|
|
|
> When I added the line #include "fstream.h" in the, say, MyDlgApp.cpp all hell breaks loose.
Try using #include <fstream> instead. Add "using namespace std;" statement after #includes.
The .h headers are for the old iostream library that isn't compatible with the Standard C++ library. See the page "Differences in iostream Implementation" in the Visual C++ Programmers Guide (under Port to the Standard C++ Library)
> The misery to me is when I took away ios::nocreate (leaving ios::binary) then everything is OK again.
ios::nocreate isn't in the Standard C++ library, it's in the old iostream library.
|
|
|
|
|
Thank you for pointing me the direction. I lookup MSDN and it causes more confusion regarding your comment about ios::nocreate isn't in the Std C++ library, It appears you are right as the compiler also has the same idea; I cut & paste part of MSDN on ifstream constructor ref, what confuses me is here it still says the 2nd paramater "nMode" can be ios::nocreate. May be it comes down to old and new lib again. I use this flag just because I don't want to create an empty file if the one I specify doesn't exist. Am I looking at the old stuff?. I don't really care but just want to use the fstream class instead of fopen(). This should not be an unreasonable wish,
////From MSDN
ifstream::ifstream
ifstream();
ifstream( const char* szName, int nMode = ios::in, int nProt = filebuf::openprot );
ifstream( filedesc fd );
ifstream( filedesc fd, char* pch, int nLength );
Parameters
szName
The name of the file to be opened during construction.
nMode
An integer that contains mode bits defined as ios enumerators that can be combined with the bitwise OR ( | ) operator. The nMode parameter must have one of the following values:
ios::in The file is opened for input (default).
ios::nocreate If the file does not already exist, the function fails.
ios::binary Opens the file in binary mode (the default is text mode).
Note that the ios::nocreate flag is necessary if you intend to test for the file’s existence (the usual case).
////End MSDN
I found from ios.h (really no idea about the old and new thing) that ios::nocreate is 0x20. I then tried "file.open("file.txt", 0x20 | ios:binary)" and it works??? but I should not need to do this. don't you agree?
I have included the angle brack on "fstream" but this paser just cut it out, so I used double quote here. Again, nothing works
Any idea??
Will
|
|
|
|
|
You're looking at the documentation for the old iostream library. Instead, look at basic_fstream. fstream is just a typedef for basic_fstream<char, char_traits<char=""> > which specializes the stream for character read/writes.
|
|
|
|
|
Hi,
I need to know if there are APIs available to partition HD and getting the info of partitions on a hard disk?
Ammar
|
|
|
|
|
I don't know why this error happens...~~~ㅠ.ㅠ
wrong page absence in GIF89.DLL module
-----------------------------------------------------------------------
Registers:
EAX=00000000 CS=0187 EIP=10007bce EFLGS=00010246
EBX=00000400 SS=018f ESP=0192f7e8 EBP=0192f804
ECX=0192f814 DS=018f ESI=00000000 FS=2cff
EDX=00000002 ES=018f EDI=10009938 GS=0000
CS:EIP's byte number:
8b 0e 50 50 50 8d 45 f0 50 6a 04 53 57 ff 75 0c
Stcak Dump:
0192f8ac 01210720 1000b28c 0192f824 0192f814 00000001 00000001 0192f840
10007644 00000000 00000000 fffffffd 00802e30 0192fa20 00000000 0000000b
|
|
|
|
|
i assume the gif89.dll module is a third party library you are using and i would further assume you either found a bug in their code or the gif file you are trying to do stuff with has a corrupted or unknown tag in it ... they should handle such things more gracefully but either way you should contact the vendor of the library and ask them what gives
prolly send them the code snippet and the file that causes the fault
the stuff you are seeing (if you don't know) is a hex dump of the cpu registers and stack contents
errr ... i should add that of course it could also be something you are doing that causes the crash (like a null ptr or suchlike) ... heh
---
"every year we invent better idiot proof systems and every year they invent better idiots"
|
|
|
|
|
Does anyone know how, or have a link to a place that has info on how to do realtime mixing of wav files?
I've gotten the mixing down, and it works fine, just have channels and "mix" them by adding bytes of the wav. My problem is that I'm doing this for a game, and because the output buffer is 1024 bytes, sometimes you can notice a delay between when the event (an explosion in this case) occurs and when the sound plays. But making the buffer short enough so that the delay isn't noticable is so short that my computer (pIII-450) can't keep up. Is there a way to access the data you've already passed to waveOutWrite? Or is there some other way I might be able to get rid of this stupid delay. (I know directsound can probably handle it, but I'm interested in a non-directx solution, if one exists.)
-wP
|
|
|
|
|
Hi,
have you tried using more but shorter buffers ? Ok this will cause a higher use of the write functions - but shorter buffers have of course shorter delay in change.
But wait - You are using 1024 Sample Buffers ? I guess for a game you are using 44,1k/16bit/mono ? That means the total play-time of the buffer is about 23ms. Therefor a change needs at least this time to appear. On the other side this time can only appear as a short chorusing effect because the time is to short for human hearing to notify 2 seperate sounds. I think you should not notify a delay. Maybe you have a great preformance loss in your wave-mixer ? You should check it out.
To raise up speed you should not use the wave-event (don`t know if you did). Better use the WaveOutProc-CALLBACK.
Then you should put the output routine in a seperate thread.
Hope it helps
Mario ///
----------------------
www.klangwerker.de
mario@klangwerker.de
----------------------
|
|
|
|
|
Hi,
I would recommend having a quick look at the FMOD sound library. It's been around for a while, so is quite mature and stable. Have a look at www.fmod.org for further details.
Hope this helps,
Regards,
Craig.
|
|
|
|
|
am just curious ... is there a reason for not using directx? after all they have spent the past 6 years getting it kinda right i understand
---
"every year we invent better idiot proof systems and every year they invent better idiots"
|
|
|
|
|
Hi! I'm coding a dialog based application, and when the user take some actions
(an specific button, or keyword) my application must show a document based in some data that I take from a database. this Document must have a toolbar and menu resources. I have the datbase connection, but I don't know how show this data in the sdi?
|
|
|
|
|
If I was writing this, I would start with an SDI application, but show dialogs until I needed to bring up a view, then hide/show the view as needed.
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Hello, the codegurus around the world. 
As long as we focus on the dialog based application, we had better call
another exe file with SDI inside OnInitDiaog.
However, if we want to get the data from the dialog in SDI, we need additional code.
So, as the other poster told us, it is easy to create SDI and call the dialog before
SDI application shows.
However, if this dialog deals with hiding and showing SDI at any time,
I have to examine the code. 
Have a nice day!
-Masaaki Onishi-
|
|
|
|
|
Basically your CWinApp derived class can hold a pointer to the SDI main frame, which can be hidden, shown and moved using ShowWindow, SetWindowPos and MoveWindow. It can be accessed from a dialog class.
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Hello, the codegurus around the world.
I doubt a little about your approach.
Even though we put on the creation of SDI on CWinApp::InitInstance(),
CMyDialog object already gets the address of m_pMainWnd.
At run time, we can't change the main object (Dialog or SDI) to m_pMainWnd.
Or, you're talking about another approach?
I mean, there are several approaches to satisfy with the requirement.
So, I am simply curious about your idea.
Have a nice day!
-Masaaki Onishi-
|
|
|
|
|
I'm not sure what you think I meant, but I am talking about starting a new SDI application, and moving the dialog based code into it. On reflection, I'd start the dialog in CMainFrame, not in CMyApp, but the core of the idea is that the app is SDI, but the user does not see the SDI window until it is needed, instead they see the dialog boxes. I don't see any other way to get the Doc/View architecture working for one part of a dialog based app. An SDI app can easily lauch dialog boxes, but I don't know of a mechanism for a dialog app to lauch SDI windows, and dialog boxes most certainly cannot have a Windows toolbar.
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
mmmm I'm gonna prove what u say, the problem it's the data that must be shared between SDI and Dialog. Well I've got to probe it and I'll tell u. Thanks, anyway I'll really apreciate if u have information on that 
|
|
|
|
|
newbie question created a dialog program but as you get deeper into the prog you end up with a lot of old dialog boxes in the background. the question is can i/how do you get rid of them i am using msvisual c++ 6.0
thanks for the help
don lamica
fubarsmut@aol.com
|
|
|
|
|
mydialog.DestroyWindow
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Hello, the codegurus around the world.
In fact, I don't understand the question very well, but.
1) Model dialog - use EndDialog() or call CDialog::OnOK() or OnCancel().
2) Modeless dialog - use DestoryWindow().
Humm. If you want to destroy the dialog called by the other application,
you need CWnd of this dialog........?
Have a nice day!
-Masaaki Onishi-
|
|
|
|
|
It is very easy to enumerate all open windows in order to get the CWnd of an open window.
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
I've posted this problem to the board yesterday, but I still don't know the solution.
Now I have a TouchScreen. When my finger touches the screen, a message will come out. I have to write a MFC program to access this message. For example, when the user touches the screen suddenly, the program will pop up a message box.
How can my MFC program checks this message and do the action suddenly and immediately?
|
|
|
|
|
> Now I have a TouchScreen. When my finger touches the screen, a message will come out.
Most Touch-Screens that I know of have drivers that make them simulate mouse movements with finger traces and clicks with finger taps.
> I have to write a MFC program to access this message. For example, when the user
> touches the screen suddenly, the program will pop up a message box. How can my
> MFC program checks this message and do the action suddenly and immediately?
If your Touch-Screen works as above, them your app can just lie in wait of a mouse message, and react to it. If not, what kind of drivers do you have for the Touch-Screen? A mouse-like driver? A driver that gives you grid/coord. information?
Peace!
-=- James.
|
|
|
|
|
Yeah.....I know I can make it as a "virtual" mouse message, but I don't want to do it in this way because there is indeed a "touchscreen" message. This message will give me coord. and touch or untouch or error touch message......
Thanks for your help.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
