Tim Blechmann

> They originally were non-const, but I noticed that they were const in the `{fmt}` API examples. I can revert them to non-const. no strong opinion from my side. in...

many thanks! i've tested it with my codebase and it works nicely!

my code looks like this: ``` #include namespace a::b::c { struct foo {}; using tests = std::tuple< foo >; TEMPLATE_LIST_TEST_CASE( "test", "", tests ) {} } // namespace a::b::c ```...

i've tried to reproduce it again. it seems to be triggered by using a clang icecc toolchain with local proprocessing. so probably the namespace was a red herring. fwiw, the...

> > Isn't there a trade off here in terms of optimization? > > It would be worth benchmarking this, say, by creating a SynthDef with a couple hundred different...

@rollbear, i was wondering on how to go forward with this?

i do like the approach of using an explicit `strong::partial_ordering`, it gives a bit more control over just relying on the type-inferred result of ``

template specializations do not compose well in this case. it would be great to have a more robust stringification mechanism, but from the top of my head i don't see...

there's a good reason to initialize delay ugens to the input sample rather than initializing it to zero: zero makes sense for "audio" signal, but not necessarily for "control" signals...

> I propose a non-conventional default: initVal: 0.0 when the UGen runs at audio rate, initVal: (in) when run at control rate. it's not necessarily about "audio rate" vs "control...