If your only goal is to stop FxCop from screaming at you, then you have found the best practice.
The best practice for signing your assemblies is that it completely depends on your goals and needs. We will need additional information, for example, the planned deployment:
- For personal use.
- For use on a corporate PC network as a client application
- Work on a web server
- Running in SQL Server
- Uploaded over the Internet
- Sold on CD in shrink film
- Loaded directly into the cybernetic brain.
- Etc.
Typically, you use code signing to ensure that assemblies are sourced from a specific trusted source and have not been modified. Thus, each with the same key is excellent. Now, as this trust and identification is defined, this is another story.
UPDATE:. How beneficial is it to your end users when deploying over the Internet if you have received a certificate of certification from a certification authority . Then, when they download your assemblies, they can confirm that they came from Domenic Software Emporium and that they were not modified or damaged along the way. You will also want to sign the installer when it is downloaded. This prevents the warning that some browsers show that it was obtained from an unknown source.
Note. You will pay for a software signature certificate. What you get is a certification authority that has become a trusted third party that checks who you are. This works because of a trust network that accesses the root certificate installed on their operating system. There are several certification authorities to choose from, but you must make sure that they are supported by root certificates in the target operating system.
Jim McKeeth Aug 29 '08 at 21:56 2008-08-29 21:56
source share