I was excited to see the introduction to modern Haskell as I find that it's very confusing to enter Haskell these days as many or most introductions doesn't actually match current practices (Text vs. String, cabal vs. plain GHC, etc). However, I was bewildered when the Cabal & Stack section didn't mention Stack. Is this merely waiting for someone to write it? At the very least, we should be

Hi,

I know I am reopening an old discussion that has been already discussed here (#1032 or #162), but this is the one thing that make miasm hard to use. So I wonder if the issue with it is :

1. Not enough time to write it ?
2. The API is changing too often too create a documentation that stays up to date ?

If it is 1, I think it should be possible to have some other people involved and

[ 70%] Built target gtest_main
/home/akochkov/tmp/gtirb/src/CFG.cpp:24:54: warning: statement should be inside braces [readability-braces-around-statements]
  if (auto it = IdTable.find(N); it != IdTable.end())
                                                     ^
                                                      {
/home/akochkov/tmp/gtirb/src/CFG.cpp:36:54: warning: statement shoul

I think it would be good to start enforcing coding style guidelines so the codebase is consistent. @rsetaluri would vote for https://google.github.io/styleguide/cppguide.html#Formatting. A simpler, automated option is to use https://clang.llvm.org/docs/ClangFormat.html.

I think we should at least run a tool like clang-format on the entire code base to get it into a consistent state. The benef

Yesterday @rgommers and I talked a bit about the thread he started on the NumPy mailing list (notes). We determined that one part of the discussion focused on how NumPy could define some of its API in terms of other parts and whether abstract base classes are the best way to get there.

It would be interesting to compare a metadsl approach to one of

we are missing a lot docs for constant gates, need to add them later when one has some more time

Currently we do not store the line numbers in the AST which means we cannot report errors with the line number information

In the documentation I've specified the versions of dependency components that were used to test and develop the Linux build and codegen. Those were the current versions on an older release of debian that I used to dev/test the code.

After upgrading to a more recent version of debian, I found that the generated code crashed early during initializing the runtime library.

I don't have this deb