Forum Moderators: open
Is there a way to do codebehind using c++?
With a wrapper perhaps?
The *main* reason is to obtain a dll that is more difficult to reverse engineer than the .net style dll/runtime model.
tia,
plumsauce
Microsoft is unlikely to completely abandon classic ASP. However, it is unlikely that there will be any enhancements to the language, environment, or development tools.
The DLL is never accessible directly from the web, so there is no way that it can be decompiled unless the server is hacked. So C++ has no advantage over C#.
I agree that this is the case from the perspective of end user visibility and access.
The problem arises when the application has high value, and the potential client asks to install a copy on their server instead of using a remotely hosted service. For an example of the value, the current single server license of my competitors who are using open source 'LAMP' platforms is in the range of 100k to 500k USD depending on the country served. There is huge incentive at these levels to reverse engineer competitors products to gain an edge.
I have missed sales opportunities because I require that the software be hosted on my servers.
This is why I am casting about looking for a safe alternative.
There is *some* safety in that much of the logic is in
encrypted stored procedures, but unencryption of tsql is somewhat trivial.
Is c++.net usable as the codebehind .dll?
And does it use the CLR in the same way?
I have searched for this many times, however *G* resolves "c++" as being equal to "c#" due to dropping the non-alpha characters.
Can I tie c++ classic, ie. VC 6.0, dlls to aspx pages and thus avoid the CLR?
Anyone here ever tried C++ Server Pages from micronovae?
++
There is no magic in compiling programs. If you can lay your hands on the executable, in any form, then the logic expressed in that executable can be extracted with enough effort.
In my opinion, people get overly hung up on protecting code. I don't need source code to reverse engineer products. I just need to see what the product does and how it reacts. There are very few algorithms that can't be reverse engineered if you can supply inputs and see the outputs, especially in real time.
Look at how much effort is placed here on reverse engineering Google's algorithms, with different degrees of success--and Google has huge incentive not to let their algorithms be known. If there were real-time changes to Google's results when you made a change to a web page, then the algorithm could be reverse engineered in a matter of days. The only reason it hasn't been completely reverse engineered is because Google only lets you see the results of changes infrequently, and changes the algorithm constantly.
I've frequently stated: There are six algorithms in the world worth protecting, and yours isn't one of them.
The proper technique is to make them sign a license agreement saying they won't reverse engineer the product. Specify liquidated damages in the contract that they will have cough up x thousand dollars should they violate the agreement.
I asked for certain clarifications about the interaction of the CLR environment. The considerations in your lecture are already known to me. However, there is some value in not making the job any easier.
The *flavour* of what I am reading is that there is a definite price to pay in trying to use C/C++ in the world of iis and asp.net. The cpu cycle cost is mostly in either marshalling or pinvoke. If I am wrong I would love to hear about it because that is the cost I have been looking to avoid for ages. A low cost method of transitioning from managed to unmanaged code would be ideal in these circumstances.
The framework looks very appealing. The problem, in addition to the CLR/idasm matter, is that isapi in C has been my environment of choice in the past. But for things that require very little in the way of html manipulation.
Now, I want the same advantages, but need the dynamic pages.
I will mention csp from micronovae again, and encourage everyone to look at it. I have *no* connection to the product other than being *very* curious about it. This poor guy slaving away in Greece somewhere has a concept that should have flown.
Much more full-featured than plain ol' ISAPI and should be fast and obfuscatable. All that template code should obfuscate it a fair amount anyway. ;)
All those curly braces sprinkled through the html templates make me dizzy. I *know* that sounds very silly, but past experience has taught me that if a particular syntax gives me the heaves, then I will eventually abandon it.
Maybe I just have to have a go at reformatting some of the example code.
