Matthew White

Ok, python it is then! Pull request created. Gentoo GNU/Linux 3.17.7 x86_64 GNU Emacs 24.4.1

I've read the whole #228 : - without `process-adaptive-read-buffering` it takes "infinity+1" to complete: **total of 88.87s**, running and **capturing clang's output takes 86.62s**, the output **parsing takes just 2.16s**...

Today I've sent the signed form to the FSF, they should receive it in about 20 days. PS: I'm studying clang/llvm to understand how to filter by prefix outside emacs...

In the meantime I found a way to parse system headers' comments, solving the problem described in the [comment 71353325](https://github.com/company-mode/company-mode/issues/286#issuecomment-71353325): `-Xclang --no-system-header-prefix=` when no prefix is specified it means "any"...

We should time measure the case when there is of a **lot of nesting (headers)**, and when there is a **lot of volume (lines of code)**, yet we know that...

> Sounds nice, but... any drawbacks, like lower performance? From what I see in the manual, system headers enjoy suppression of warnings and precompilation: > > http://clang.llvm.org/docs/UsersManual.html#controlling-diagnostics-in-system-headers > http://clang.llvm.org/docs/UsersManual.html#precompiled-headers >...

> I've noticed that you never include a summary for the benchmark results. Why? Sometimes I prefer to be pedantic and write all the steps I follow. Are you suggesting...

@dpsutton I'm happy you enjoy the code! - I see what you mean by `ivy-buffer-list`... messages are replaced by the timed resize-window's status. What about making it **optional** via `defcustom`?...

> Absolutely! Happy to help and I really appreciate the work you have already done. Thanks! > About squashing, I'm not sure squashing brings any benefit. Your commits are atomic...

> we can just spinloop the notifications for some time to come back when the minibuffer is no longer active. That is really a good idea! Great thinking! > You...