On Wed, Dec 15, 2010 at 6:25 PM, Peter Maddin <[email protected]> wrote: > Hi Greg > > > Did the Delphi authors do that?! > > Sorry no idea. The controls were purchased from a European site because I > could not get anything out of the States. The company, no longer exists so I > am glad I purchased source code or I would be screwed. I tried to purchase > TurboPower's LockBox but they would not supply it due to an export > moratorium on cryptographic cipher technology. This has been relaxed for > some time now but I am stuck with the legacy of went before. > > The code I purchased has been in use for many years and has been very > reliable (except for a memory leak that I managed to fix). I do not feel > competent to recode it.
FWIW, it's probably far easier then you imagine. For any popular algorithm you'll find examples in just popular languages (java/c#), and you can easily translate that. There will be test vectors provided so you can test your implementation. It's not a few hours, but 2 or 3 days and you should easily get it done (at least, I'd consider investigating it). You should know that SHA-1 is considered "not good" as a hash: http://valerieaurora.org/hash.html, so I'd avoid it's use if possible. > The idea of writing a COM object to encapsulate the > Delphi code seems like the optimal solution. I am not that well versed in > COM so I have some study to do. At least I have texts on the subject. I > thought that with .NET, COM was deprecated (even if COM+ is not). Oh well. > > I will give the Delphi to C# converter suggested by Ian, a go as well. > > My fall back is to use the framework's cryptographic support (RSA not > elliptic curve) but that would mean two parallel delivery streams, the old > and the new that are not compatible with each other. -- silky http://dnoondt.wordpress.com/ "Every morning when I wake up, I experience an exquisite joy — the joy of being this signature."
