Paul Colomiets

Hm, I could catch this issue today. Although, I'm not sure how to reproduce it. And it looks like my hypothesis is right: ``` $ sudo lslocks --output-all COMMAND PID...

Well, in my opinion, internally "service" is a `Sink`. A disconnected Sink returns Error on `poll_complete`. Then there is a connection-pool layer, that removes errored Sink from the pool, but...

@carllerche I understand the current situation. At a glance in https://github.com/tokio-rs/tokio-proto/issues/75 the [protocol implementation is a Sink](https://github.com/tokio-rs/tokio-proto/pull/75/files#diff-be3e9d6fb37416cf1bdeb2740e38929bR64). So what I'm trying to say is: we don't need to hide that...

I also think that `Pipfile.lock` should contain original versions of the packages. Not just to read it from setup.py, but also to be able to check quickly if lockfile is...

Any updates on this issue? It should be pretty trivial to do and maintenance cost should be minimal. Should I make a PR?

> This feels to me like it is adding another quirk to python packaging. If we do this, setup.py will depend on Pipfile.lock which is very weird: now abstract dependencies...

The problem with using `HashMap` is that arguments can be of different types. So dynamic typing should be used. I.e. `edgedb_protocol::value::Value` type. Unforutately, we don't have a nice constructor for...

Usually I'm trying to avoid macros. Also I don't think we can have impl for some of listed types, which we don't want have in the database itself. (e.g. we...

> The query functions taking a single arbitrary `Value` looks weird as well, since it's always a list of key-value pairs. I'd have expected some kind of map or named...

> If it's a tuple, how does the query access it? Something like `$0` instead of `$name`? My preference would be using a smaller enum `QueryArgs` with only two options...