I think that I made my choice : LibXML++
It can be usefull that I explain :
- Xerces/Xalan : Not maintained in the same way, static, leaks...
- TinyXML/TinyXPath : No XSD validation
- MSXML : COM,no support for XPath 2.0
- Arabica : Seemed to be ok, but finally : "provides uniform SAX2 wrappers for the Expat parser, Xerces, Libxml2 and, on Windows, for the Microsoft XML parser",so,not fully standalone.Moreover,as I am lazy,I have seen that the compilation gives only static libraries (not good for me).
- LibXML++: Installation is a bit "crappy", cause need glib & glibmm.But seems to be more standalone.
after playing a bit with LibXML & LibXMl++,I have to say that it is incomplete.
First,the LibXML++ lacks of some functionalities,so I need to call the base C LibXML instead.Anyway,this is not the main problem.Apperently,LibXML has an incomplete support for XSD schemas,so that's not ok for me.
I will have a deeper look on Arabica right now,even if it uses MSXML (by default) on windows,but anyway it is portable,so even if it is not my priority it is better.
Update : I tried Arabica, but first bad impression : If I try to parse a document that does not exists,no error is reported.Moreover, XSD does not seams to be supported....I am lost....Xerces/Xalan or MSXML are much more complete I have to recognize.
If I disable the edit control, the control is not painted correctly.
It is painted with text in someother edit control or someother part of my form view.
Before solving this problem, I wanted temporary fix by avoiding focus or something else.
The problem was inside OnCtlColor() where I was doing some coloring with form control.
It is fixed. Thanks..
Are there functions available in the windows api that will paint windows controls to a certain rectangle using GDI. Something similar to the ControlPaint class in .NET. If there are, can someone point me to the documentation that details how to use them.
I don't have an instance of a control. All I want to do, is call some routine that will paint me a button to a specified DC, like the ButtonRenderer.DrawButton method for the .NET framework. This allows you to pass an arbitrary Graphics object and some state information and it will draw a button using windows styles to it without needing an instance of Button.
Have a look at OpenThemeData[^], DrawThemeBackground[^] and related methods, but note that these only work under Windows XP (and i supose above) with themes enabled, so it might not be enough for your needs.
> The problem with computers is that they do what you tell them to do and not what you want them to do. < > Life: great graphics, but the gameplay sux. <
To translate an applicaion to different languages, App.rc is needed to translate the text in it. I would like someone to give me an idea or suggestion of what I might need to do with the hardcoded strings for messages through out the application( that are not part of .rc file - have like 400 of them), is there a way I can put all these messages in a file may be and use them as needed, I am just trying to find a way to make it easier later on to add more such strings that are not part of the .rc file in case of translation or anything else... what's the best way of doing this?
MessageId = 1004
Severity = Error
SymbolicName = RAD_FCE_DRVNOTINSTALLED
Language = English
Allied Vision firewire message: Kernel mode driver not installed
which when compiled gives me a .rc and a .h
I recall that at some point I built a .dll using the resource that had one method: GetString(...)
and could just recompile the .dll when (in theory) a new language was added; which sounds like what you are looking for. I've never had cause to actually add another language though so can't advise on that.
I dropped the .dll idea eventually and now just compile the .h and .rc straight into my main code as it's easier for me.
Though a .dll could export such a function instead of needing a COM object. Within that function handler I loaded the string, though looking now I see I used a string table rather than a message table. The info on message-tables I'm sure I found on Code-Project. I actually used the CComBSTR within the function as it had a LoadString method. (All this was done in VS6).
I created one file suppose "abc.txt". I denied all access to all the users and groups for that file. (I did it by going to Security tab of property dialog of that file.). When I programmatically try to open this file by using API NtOpenFile() with access mask READ_CONTROL in user mode it succeeds. In this case I am opening the file in the user context who is creator owner of that file.
When I try to open the file in the context of other user who is not creator of that file NtOpenFile() fails with error "Access Denied". It may fails due to that user doesn’t have permissions.
Now when I try to open same file in native mode, NtOpenFile() fails with error "Access Denied". What will be the reason? How I can access the file having all permissions denied to all users in native mode?
Can anybody tell who is the user in native mode? Is it System?
It's impolite to repeat your question less than eight hours after first posting it. Don't do it again.
Anyway - an answer:
There is no user in native mode. Windows *DOES NOT EXIST* when in native mode. That means there's no concept of users as Windows understands them, no security sub-system etc etc etc. More importantly, I don't believe there is any concept of ACLs in native-mode, which means that READ_CONTROL is meaningless in native mode. This is reinforced by looking at the NtOpenFile[^] (and then ZwCreateFile[^]) documentation - READ_CONTROL is not a valid access mask for either of those functions.
Hi, I'm planning to program following:
2 connected computers (one windows one linux) would communicate between each other via UDP (User Data Protocol). Both sided will receive and send data based on input.
The question is: in UDP do I have to always configure client/server side? Can a server send data to client based on user input and the same on client side? If yes then I need to define just one IP and one port (am I right?).