programming-language

We have both Symbol and Int getindex for a NamedTuple, but only Symbol setindex.

Is there a reason for this? its seems like we should have both in both cases.

There is the test/libsolidity/SolidityEndToEndTest.cpp file, which contains a lot of semantic tests. It is a large monolithic single C++ file, which takes a long time to compile and is hard to maintain.

We want to move as many test cases out to individual files under test/libsolidity/semanticTests. They should be grouped logically, by following the grouping in SolidityEndToEnd.

Have a

So currently, install in our "unix" Makefile isn't hooked into cmake so it can lead to errors.

in particular if you do

make configure arch=not-our-default
make build
make install

you will end up with a broken install because the arch override from configure is used during build but isn't used during install, instead a different arch (our default) is used and you end

Currently it displays:

  ⚠️
  
  I was expecting the namespace . to be empty for this operation, but it isn't.

which didn't feel super clear.

How about:

`pull-request.load` downloads the PR into the current namespace and I don't want to clutter up your current one, so please switch to an empty namespace and try again.

This likely goes away altogether with #2247,

Motivation

The Non-Realtime Synthesis guide (https://doc.sccode.org/Guides/Non-Realtime-Synthesis.html) currently doesn't contain much info or examples that make use of busses and synth mapping/lfo mapping.

Description of Proposed Feature

Add info ala the examples here:
https://scsynth.org/t/how-to-allocate-busses-in-nrt-mode/

i) offsetLeft / offsetRight
ii) layerX / layerY
iii) offsetX / offsetY
These three things really helps in DOM manipulation. It is very cumbersome to do things without these methods. It will bring developer happiness hopefully.