Improve nullability about ydotnet HOT 3 CLOSED

Hi, @SebastianStehle!

I agree with your proposal to fix and improve the documentation here but I think it's part of a bigger problem that I, on purpose, didn't handle in the library so far: exception handling.

You probably noticed wherever something can go wrong in the library, if it happens there's no handling at all. So, I'd suggest renaming this issue to something like "Improve exception handling" and then we can include items like:

• Handling unexpected null results
• Handling invalid parameters (like reading beyond the valid indexes of an array)
• etc.

However, I don't intend to do that in a "local" way as it's suggested here or, for example, by adding a bunch of if statements throughout the code, I'd like to find a global approach to throw exceptions keeping the Rust stack trace and without crashing the process, for example.

So, if you agree, let's try to make this scope bigger and investigate how to handle this.

from ydotnet.

Also, please keep in mind that the priority of this kind of task is low for now because I'm still improving the features and memory management.

In the best case, for now, if there are no programming errors, the exceptions shouldn't happen very often.

from ydotnet.

Hi, @SebastianStehle!

Since there are many issues in a row that could be closed after #55 (I think this is the 4th or 5th already), I'll just go ahead and close them for now.

If you notice there's anything else pending to do, please feel free to re-open any of them or just open new issues so we can start discussions with a clean slate.

from ydotnet.