Pavel Koneski

Another additional trigger for the cleanup could be `gc.collect` (unless it is already happening somehow).

> I sort of dislike having the extra dependency on the core, but maybe we could break off the SQLite module into its own optional NuGet package Another possibility is...

[Originally](https://github.com/IronLanguages/ironpython3/issues/1208#issuecomment-834757250) I intended to remove `__classcell__` and replace it with an extra parameter to `PythonOps.MakeClass`, but since Python 3.6 introduces it and IronPython 3.4 already implements a number of Python...

I think that upgrading the parser to be able to parse Python 3.6 is a very good idea, even if it turns out that there are other blocking issues in...

I have never created a NuGet package, but if they have something like a post-install script, then it can do `chmod -x` on Posix systems.

Indeed, as @mwchase writes, this is a _NuGet_ problem and should be addressed there. As for a workaround here: since we have now a .NET tool package and a complete...

I am aware it has been resolved — sort of… There is no crash during compilation anymore, but the semantics/behaviour of this construct is different than CPython's. This is covered...

It also adds `_imp.create_dynamic` and `_imp.exec_dynamic` which seem to be the decomposed `_imp.load_dynamic` from 3.4. But ipy 3.4 doesn't even have `_imp.load_dynamic`. I think it would make sense to implement...

Thanks for the submission. This seems to be related to #21, which is only partially finished. What is still needed is a review of all modules and converting their API...

Yes, `from_buffer_copy` will be simple. I can pick it up since the int/long unification stuff is drying up. For the `from_buffer`, indeed #1297 needs to be done first.