Alex Gaynor

So the trick is figuring out how sqlalchemy itself uses the underlying connection. My inclination is that we'd have to replace more layers, and probably copy-paste-and-edit more code. I'm definitely...

Maybe. It's all about the relationship between all the pieces, and being able to propogate deferreds correctly.

`txpostgres` might not be the perfect approach, but I think the analysis of `sqlalchemy` that you did will apply, whatever connection library you use, because at the core you need...

I think this is related to "connection" pinning, which we need on database adapters that aren't thread safe. What we want, I think, is that there's N+1 queues, where N...

The answer to this is predicated on whether I can figure out what that interface should look like, is their prior art her (scala perhaps?)

So, some notes on my own PR: I'm not happy about an "assign" opcode, that seems pretty wrong. The genesis of that is that local variables are still represented as...

I was definitely planining to do two IRs already, one untyped, then one typed. Doing one with locals and one with SSA makes sense I think. On Wed Nov 12...

Ok, I think I'm reasonable happy with teh scope of this patch at this point. Right now it demonstrates: - locals - conditional branches - some arithmetic Is there anything...

Bagel is not going to be that function, chains of `->` will be uncommon.

You're asking what the "type" of a function is? On Mon, Oct 20, 2014 at 11:52 AM, Glyph [email protected] wrote: > This does raise the question of how one declares...