|
We are currently scoping a new (smallish) project, and need to make decision as to whether we develop it in .Net or C++.
Does anyone know of any statistics that detail the distribution of the .Net framework on desktop systems?
|
|
|
|
|
Why not just have your applications installation package install the .NET Framework. It is freely redistributable.
EuroCPian Spring 2004 Get Together[^]
"You can have everything in life you want if you will just help enough other people get what they want." --Zig Ziglar
"Get in touch with your Inner Capitalist - I wish you much success!" -- Christopher Duncan, Lounge 9-Feb-2004
|
|
|
|
|
I've been in the same situation. If your software is distributed over the web to end users you have no control over, .NET is a HUGE requirement. I've asked this question in other forums and to some MS people; it's really strange how some people have a hard time understanding that .NET is indeed a very difficult pill for end users to swallow, thanks to it being over 21 MB.
My thought is this:
.NET will be everywhere by 2007. Until then, it's a guessing game; while a vast majority of desktops have XP, only some of them will have .NET. If your product is small, and will be finished within a year, and will be distributed over the web, I'd say use C or C++.
If your product is fairly large project that will take a year or more to develop, or will be distributed on a physical media, then yes, write it in a .NET language for productivity's sake. By the time your product is released, many more users will have the .NET framework installed already, and you'd be even closer to Longhorn, when everyone will have the framework installed.
My rant is that MS needs to get .NET to the masses sooner than Longhorn. If they would just bundle it with Office or IE, questions like these wouldn't arise and they'd get more developers writing .NET-based software.
The graveyards are filled with indispensible men.
|
|
|
|
|
Hi
i have a strange problem:
i use code like this:
ODBCConnection cnn = new ODBCConnection("... cnn string ...");
ODBCDataAdapter da = new ODBCDataAdapter("select * from tbl left join tbl2
on tbl.refid = tbl2.id", cnn);
DataTable dt = new DataTable;
DateTime t = Now();
da.Fill(dt);
TimeSpan ts = Now.Substruct(t);
Trace....... Show the ts.
.....
i use this code in 2 programs:
(the remoting server is a console app.)
position:
Time
1. remoting server program on a marshaled object
in static method.
8.7 seconds.
2. the same server, but in the main file.
5.4 seconds
3. the client, in the remoting object (the same
method as 1, but i know here it is not the same
remoting object).
4.2 seconds.
4. different console program( that not using any
remoting, built to the test.)
0.5 second !
u see a huge changes between these methods.
I operate the 2 programs (remoting and test) concurrently and the results
was the same +-.
so i assume it is the remoting that affect the Fill method.
but why ??
what could be the reason ??
if any1 want i will send him the remoting configuration section in the
config file so he might watch it...
|
|
|
|
|
Hello,
Does anyone know about How to write a code for custom controls which the user can do drag and drop feature while browsing the page. For Example i am having a login control in my let hand side, at the time of browsing the user can change the location to right hand side by just doing drag and drop facility.
Is it possible to do? Any documents for this or any procedures?
Thanks in advance.
Best regds,
Kandasamy.
" BE extremely passionate about what you do. Combine it with patience, confidence in your abilities and a dedicated team of professionals. Sucess is inevitable "
|
|
|
|
|
Hello All,
I need a technical help, which is related to Performance and Application Design issue.
I have a function, which is accessible from a page alone. Now the question is which is the best approach - To keep the function in my Code-behind file itself or keeping it in a class module and call it.
Which is the best way to do. For this if you got any supportive documents also pls do send me ASAP.
Thanks in Advance.
Regds,
Kandasamy.
|
|
|
|
|
There are several related questions that point in one direction or the other. Primarily, where and how often do you intend to use a function, and what information does the function act upon? For example, if a function returns a SqlDataReader that lists a customer's recent orders by customer ID, the function probably belongs in a class like "OrdersData"... thus making it available to any method within the entire application. If, on the other hand, it is some kind of static/shared function, for example, a function that takes a datetime value and returns a formatted, human-readable date & time string, and you expect it to be called several times in several different classes, you could have a class library of "helper functions". Or, if it is a function used only by one page for one reason, leave it as a member of the page class.
But then again, I could be wrong.
What a piece of work is man, how noble in reason, how infinite in faculties, in form and moving how express and admirable . . . and yet to me, what is this quintessence of dust? -- Hamlet, Act II, Scene ii.
|
|
|
|
|
My windows service is strong names by putting
<assembly: assemblykeyfileattribute("c:\mykeyfile.snk")=""> in assemblyinfor.vb
But when I write a setup and deployment package for it and set it up I get the error :
Unable to create an instance of (my installer class) installer type. ->Exception has been thrown by the target of an invocation.-> Security Error
Any one got any idea how to get round this?
MOO!!
|
|
|
|
|
Don't know if this is a known issue, but Application.EnableVisualStyles does not work on vertical scrollbars (I assume it won't work on horizontal scrollbars either). Oddly, if I use a create .manifest file for my .exe, the vertical scrollbars will indeed get the XP Visual Style.
The graveyards are filled with indispensible men.
|
|
|
|
|
Hi,
Does anyone know where i can find some info about compiled .exe/.dll files' format? In other words, how can I extract IL assembly instructions from a .exe/.dll file? (well, i can use the 'ildasm' to get the IL instructions, but I wanna know how it does that since I'm gonna have a project dealing with those binary files).
Any answer, book or link is appreciated!
|
|
|
|
|
download reflector or anakrino
|
|
|
|
|
.NET executables are PE/COFF files (Portable Executable / Common Object File Format). The instructions are stored in the "text" section, the class and method tables are stored as meta data.
|
|
|
|
|
Hi,
First of all, thank for the answer.
Do you know how i can idenfify the "text" section? Is it started or ended with a special character or what? When a open an execuatble file with a hexdump program, i see alot of junks, although I can see different sections of junk.
Also, I assume each instruction is equivalent to a hex opcode, but i can't find a reference to all of the opcodes. Anyone knows?
|
|
|
|
|
|
Hello again,
I just tried to download the ofiicial Microsoft PE/COFF specification. The .doc and .pdf versions don't exist, you'll only get a "Sorry, the page you requested.........".
But the compressed .doc (pecoffw.exe) is still there
|
|
|
|
|
|
|
http://www.dotnetspider.com/technology/
This site offers C# tutorials for beginners. C# and .Net explained in simple ENglish and ste by step.
|
|
|
|
|
Well I had code break on a compiled version of a Visual Basic .net application that was in use. Here is the story.
Environment: Windows network using Windows 2000 advanced Servers and Windows 2000 workstation Service Pack 4. .net Applications written with .net framework 1.1
The other day we had a series of security upgrades deployed on our network. As soon as that happened a Visual Basic .net application in use for some time developed numerous problems and was generating error messages all over the place. On workstations that had earlier versions of the .net framework, those versions had been removed and only the latest version which had been installed during the security upgrades remained.
I looked at the source code of two applications written in vb.net and found the following problem. All Catch statements that worked previously had an error in the source code when I examined it. The basic Catch statement Catch ex as exception no longer was valid. the squiggly blue line appeared under the ex. To rectify the problem I was required to type:
Catch ex as System.exception
Obviously the reference for System had always been present in the application. Just as an experiment I added Imports System to some of the class modules, but I was still required to type system.exception where ever a Catch statement was used in the apps that existed prior to the network upgrades. If I create a new project however, catch ex as exception statement doesn't have an error and the code compiles with no need to use the expanded system.exception. But in the .net apps that were written before the upgrade that code errors 100% of the time on catch statements and I must now type system.exception in all my catch statements.
I have tried deleting the references and re-entering them and all sorts of other tricks. I am pretty sure that those older apps are looking for a different version of the framework that is no longer there.
I could just leave all the syntax changed to Catch ex as system.exception but I am worried about the implications of what has happened to cause this code to break, and would sure like to know how to fix the problem without having to change code that shouldn't be broken.
Any thoughts or suggestions would be greatly appreciated.
|
|
|
|
|
I have a large enterprise application. It involves 50 - 100 seperate .exes and several thousand .dlls. We want to group the .exes in their own directories with the .dlls they use. There are about 100 base .dlls that are commonly used. We would like to keep these in a single directory and let all the other apps use them.
This does not appear to be possible. We can place them in the Global Assembly cache, but they have to be strongly named. If I understand correctly, using strong names will require me to rebuild all the apps that reference them each time we change one of the base .dlls.
The ony other alternative seems to be to put a copy of each .dll used by an application in the same directory as the application. This causes us to have many locations to keep synchronized, or keep all the .dlls and .exes in the sam directory.
Is there a better way?? Am I missing something here?
Thanks for any help I receive.
|
|
|
|
|
You might be able to load all those dlls manually at start up.
Just use Assemlby.Load, or you might register event hanlder for AppDomain.AssemblyResolve and then load the assembly if it can't be found thru normal lookup.
|
|
|
|
|
You can also setup different paths for AppDomain.RelativeSearchPath or AppDomain.DynamicDirectory
|
|
|
|
|
Hey
I have a .NET remote server application I want to access from a smart device(.NET compact framework). Is there any other way, than webservices to access this remote server? I can't find the .NET remote for the compact framework.....
best regards
Rasmus
|
|
|
|
|
Hi All
I have a MFC client wich is a part of big 3-tier application. Recently, I've begun to convert some parts of it into managed code.
By that, I mean that I rewrote some components of the app using C# and imported them into my app as a COM components (including the communication with the server wich I changed from COM+ to .Net remoring).
It works fine, but when I close the application I get exception in call to corExitProcess().
Does anybody know something about that?
Do I have to shut the CLR in some way?
thanks!
|
|
|
|
|
I suspect you're violating some condition that the CLR is relying on.
I suggest running your code using CLR SPY[^] and working from there.
|
|
|
|