Henry

I wasn't sure how the sub-queries are handled internally, I assumed it would iterate the inner query and than pass the resulting slice to the outer `IN` predicate. From your...

Thanks for the clarification. > The ql (or any other) database/sql/driver implementation never sees the slice and cannot handle it. Yup, I came up to the same conclusion while doing...

Hrm.. seeing the documentation on `Value` I'm not so sure any more about my last suggestion. ``` Value is a value that drivers must be able to handle. It is...

correct me if I'm wrong but in postgres it's nice to have this on the query-language level because you reduce the amount of _context switching_ between database and application layer...

I can’t believe 4 years have passed since I pushed this... I just stumbled over this again after seeing #8 To be honest I never went beyond using this for...

One remedy @keks and I thought about is buffering unhandled requests and draining them on a worker pool of static or scalable size. The idea is to still drain the...

The above approach would stall on live querys. For this to be feasable we need some kind of switchboard that can accumulate live querys and serve multiple muxrpc sessions from...

Yup, i thought about a simple token bucket limiter for new calls as well, would be easier to add into the existing code for sure.

An increased `bufSize` just hides the problem in the garbage collector. https://github.com/cryptoscope/go-muxrpc/blob/b76bb2cd2c270b2ab6964926d167b97806586a5d/rpc.go#L54 This can be especially nasty in the case of `ssb-blobs` requests where whole files can hide in this...

a bit more background: there is a single read-loop (`rpc.Serve`) that reads new packets, spawns requests and pours packets to existing requests via `luigi.Pipe()`s. These have a buffer size, the...