Re: GStreamermm and GitHub

[Daniel Boles <dboles src gmail com>]

On 1 October 2017 at 11:48, Russel Winder <russel winder org uk> wrote:

Given that D and Rust can create a mostly automated system for generating
bindings, I am surprised to hear C++ would be difficult. Clearly girtod and
gir already exists, and it seems nothing for C++ does. I guess the question is
the a "market" for doing something.

I should clarify that what I said about thinking they're not compatible was purely based on the fact that no one has made them compatible yet... which is not very good reasoning, now that I think about it :) and yes, may be doable already.

I would be in the market as a user, but of course making this work could be distinctly non-trivial on the part of the maintainers, and it's possible they don't have sufficient time or interest.

 

I'm trying to give up all use of Autotools: Meson, SCons, and if necessary
CMake, FTW.

Yeah, I spent a few days converting one of my more 'toy' projects from Autotools to Meson, and while I still have to convert the tests (snore), it's obvious that Meson is infinitely easier to write and more maintainable.

I can't wait to see the git commit where I finally bin Autotools - I can only guess how many lines and bytes of bloat that'll be.

I appreciate what Autotools set out to do, but it kept doing it for far too long. The part of this I'll miss the most is Automake, as I kinda like its nonlinear syntax and find it quite expressive in some areas - but I think Meson wins in having files that are easier to reason about and not loaded with Autoconf baggage.