.NET: How does the use of components in .NET differ to pre-.NET?

How does the use of components in .NET differ to pre-.NET?

Are the differences as follows?

• .NET components don't have to be centrally registered on a machine and can merely be invoked at run-time from a specified location
• if .NET components are 'registered' in the GAC, 开发者_Python百科problems associated with different versions of the same DLL ("DLL hell") are avoided because each version will have its own id/key which is known to the calling program

Is the above correct?

What other relevant differences are there?

---

In the era of COM, one could host the DLL as a DCOM service and could invoke the service hosted in one m/c from another. After web services gained popularity, DCOM has become obsolete.

---

The differences given in the question are the most relevant. Other differences relate only to ease of use with IDEs.