Damus

Recent Notes

David Chisnall (*Now with 50% more sarcasm!*) · 9w
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq0cg7txfs7vf3t603zmvelerwaa52sw2qvycz25eu559s4skcgwxq5g9smr nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqgqn4mglfftd6292303jp2r060qwy69m...
Marcel Weiher profile picture
@nprofile1q... @nprofile1q...

OK. So C++, even if not ideal, is one of the type systems you like. And so one that should result in fewer bugs.

And yet, C++ was in the study that showed no benefits for languages with static type systems.

That's the empirical evidence.

I am sorry that reality does not conform to your expectations.
1
David Chisnall (*Now with 50% more sarcasm!*) · 9w
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq0cg7txfs7vf3t603zmvelerwaa52sw2qvycz25eu559s4skcgwxq5g9smr nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqgqn4mglfftd6292303jp2r060qwy69mpgdfprlumsuhf68wft27svfzgz0 I have no interest in debating the strawmen that you keep bringing up....
note1v4c6a...
Marcel Weiher profile picture
@nprofile1q... @nprofile1q...

Good grief.

It illustrates the documentation benefits of static typing using a C/Objective-C snippet.

That's it.

In your original post, you talked about moving from C to C++.

Does C++ have the kind of type system you like? What about Typescript?

How does what you "like" impact the fact that there is no evidence for static typing actually reducing bugs?

Or are you just going to keep no-scotsmaning this?
1
David Chisnall (*Now with 50% more sarcasm!*) · 9w
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq0cg7txfs7vf3t603zmvelerwaa52sw2qvycz25eu559s4skcgwxq5g9smr nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqgqn4mglfftd6292303jp2r060qwy69mpgdfprlumsuhf68wft27svfzgz0 C++ has the kind of static type system that I like, where I can enforc...
note1v7728...
Marcel Weiher profile picture
@nprofile1q... @nprofile1q...

Once again: the post does not even mention either Objective-C or Smalltalk and is certainly not in the least bit about comparing those two.

Yet you claim that that is what the post is about. How is that possible?

In the comments there is also a reference to a paper published later that claimed proof of safety benefits, only to be utterly debunked shortly after (the paper).

https://2019.splashcon.org/details/splash-2019-oopsla/75/On-the-Impact-of-Programming-Languages-on-Code-Quality

It looks like you're "no-true-scotsman"-ing this.
David Chisnall (*Now with 50% more sarcasm!*) · 10w
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqgqn4mglfftd6292303jp2r060qwy69mpgdfprlumsuhf68wft27svfzgz0 That’s a fairly superficial analysis and doesn’t really look at the properties t...
Marcel Weiher profile picture
@nprofile1q... @nprofile1q...

The blog post in question does not look at Smalltalk or Objective-C at all, did you even read it at all? Heck, neither are even mentioned. Instead, it looks at the actual research.

And yes: the documentation effect is powerful and is actually well supported by the research. Unlike the safety effect, which is not just not supported, but actually consistently contradicted by the research.

Static typing is a convenience feature, not a safety feature.