Support MessagePack serealization/deserialization in nanoFramework

[RelaxSpirit]

I took the liberty to implement MassagePak support for nanoFramework solutions.
Might you consider adding this solution directly to nanoFramework, if appropriate and necessary?

https://github.com/RelaxSpirit/nanoFramework.MessagePack

[josesimoes]

Hello @RelaxSpirit! Nice work.
Interestingly I was talking to someone the other day that was asking for this.

Let me discuss this internally, but for sure we would like to host the repo and offer it as an official library.

[RelaxSpirit]

@josesimoes No, Discord is blocked in my country. Using a VPN is not convenient at all.
That leaves Telegram and WhatsApp.
Discord has never been popular with us, so no one noticed its blocking :)

[josesimoes]

@RelaxSpirit I've made the final adjustments in the source code, added the AZDO pipeline and enable testing in the pipeline.
First official nuget has been published.
Thank you again for your work on this library. 💯 👏🏻 👏🏻 👏🏻

[josesimoes]

I have a few requests for you, if you don't mind 😅 (nothing urgent):

1. Add a sample app demoing usage of MsgPack to our samples repo. Easy peasy. Just follow one of the other serialization samples there and follow the pattern.
2. Improve the readme in the repo to cover the section on how to add a serializer for a custom type. Nice to have this also covered in the samples.
3. Add a benchmark project to the repo with a few test cases comparing MsgPack with the other available serializers: Json and native Binary serializer. Time and size would be the two major comparisons to make.

[RelaxSpirit]

@josesimoes Thank you so much for working to bring the project to its first final release! I've already started using the new NuGet package! :)

By points:

1. It's probably better to post examples here https://github.com/nanoframework/Samples/tree/main/samples but I really would not like to do the mapping of this solution for myself. I'll think about where to add examples.
2. The Readme in the repo can be updated. I'll try to do this first.
3. I've been thinking about the benchmark project, too, until I get around to it, but it's a good and necessary idea.

I will try to finish everything if possible :)

[josesimoes]

Just a simple example is more than enough. Pretty much the code that there is now in README. The major benefit is that the Samples repo is the goto for developers searching for code sample. Therefore, this increases the library visibility.

On the benchmark, take a look at the benchmark project in Json repo, for example. We have a Benchmark library available so it should be pretty straightforward to implement. Again: start with something simple. Just a few comparisons. With time that can be improved with more comparisons. The general idea is to have ballpark number to compare the different serializations available out-of-the-box in nanoFramework.

[RelaxSpirit]

Thank you all for your participation, help and a positive ending to the question, the solution is included in nanoFramework.