Re: GJS camelCase methods (at least)

[philip chimento gmail com]

Hi Andrea,

Don't forget to keep replies on the list — other people might read it and get inspired :-)

On Sun, Oct 15, 2017 at 6:26 AM Andrea Giammarchi <andrea giammarchi gmail com> wrote:

Hi Philips,

  as probably previously mentioned my strongest PL skills are on _javascript_, not cpp, and I don't see me helping in the best way on cpp or, at least, it might take quite some time.

What if you worked out how the edge cases should be handled and wrote the regression tests in JS? I think the amount of required C++ code is actually quite small.

However, I wonder if we can go for a migration pattern instead of a core change, maybe behind a flag, and see how much adoption it gets.

As example, I have also the whole loaded namespace lazy patched on demand, which has almost zero performance penalization on methods usage, and way better performance than jsgtk on bootstrap (it's lazy, only what's used gets patched once).

As summary: would a JS contribution behind a flag `GJS_CAMEL=true` be OK for the time being?

I'm not overly excited about it, I'd rather have it in core to begin with. If the goal is to see how much adoption it gets, I don't think the results would be much different than just seeing which projects copy your gist into their code. I'd be happy to "endorse" that gist as the way to get camel case for now, at least as far as my endorsement carries any weight.

Regards,

Philip C.