I've developed a few Firefox browser plugins. I love the ease of development that the Mozilla extension API gives you (once you figure out the file and directory structure). It allows an average programmer to create an extension with nothing more than notepad, and no compilers. Or if you want more power you can create a full Java compiled extension. And with the advent of Silverlight (and some creative hacking) you can create a C# extension. But the point is that almost anyone who has basic web dev experience can write a Firefox extension.
So recently I decided to write an IE plugin, and um...oh no. I dont want to write a C++ COM dll, I'm a .Net developer. Ok, I can write it in C#, but shoot...I have to port all the interfaces over to C#. Man...this is hard. Why Mozilla proved that it doesnt need to be hard.
So my question is this. What is the future, if anyone from the right MS team is reading this, if the IE plugin architecture. Will it continue to be a compilcated list of COM interfaces and registry hacks, or will IE move over to a much easier model, similar to what Mozilla has done
The reason is that there are thousands of really useful Firefox plugins (and 100 times that number of not-useful plugins). Almost any idea that I come up with for a useful Firefox plugin, almost asuradly has already been created. But the same can not be said of IE plugins, because it takes specialized and hard earned knowledge of C++ and COM. Its so compilcated and anoying to create one that there are very few.
If IE went to a more open, accessable model for their plugin architecture, IE users would have a plethora of plugin options.