Apropos of nothing, some unstructured thoughts on C++ and the game industry (thread):
1
11
59
Conversation
I see a lot of frustration with C++ in the game industry, compile times and runtime performance of debug builds being the main complaints. I don't want to downplay these, I've experienced both first-hand and I know how much of a burden they can be. I get it.
1
11
But loudly whining about it on twitter achieves nothing but making us look like children. Yelling "thing bad, fix it now" and then refusing to interact constructively when people offer to help isn't going to achieve anything but a sense of animosity.
1
21
It's /completely/ unreasonable to take something for free, complain about the parts you don't like, and then snub the people responsible when they make a good-faith effort to engage and address it.
2
20
thank you.
but also:
Quote Tweet
If you use C++, and if you are currently under the mistaken impression that you want it to be improved (rather than frozen permanently), please spend some time thinking about this frame from youtube.com/watch?v=4_ODvZ.
Backward compatibility generally precludes proper design.
Show this thread
2
Wonder if it would just be better to have linearly checked, explicit `drop`. Kinda like in ATS?
1
Ahh interesting - must have been before I started using Rust (0.4). I'd still like to try it out again though.
wrote a pretty good writeup about the problems with linear types in Rust gankro.github.io/blah/linear-ru
1
4
ergonomics fools: what if Rc autocloned, since it's generally completely unininteresting
me, a clarity/orthogonality genius: what if every single place that decremented a reference count had to manually try to run the concrete destructor of its contents. Also screw type erasure.
1
3
Show replies