C++ Logo

std-proposals

Advanced search

Re: [std-proposals] Proposal to extended dynamic polymorphism (to fix issues that std::variant fails at addressing)

From: Jason McKesson <jmckesson_at_[hidden]>
Date: Mon, 30 Mar 2026 11:13:32 -0400
On Mon, Mar 30, 2026 at 2:07 AM Muneem via Std-Proposals
<std-proposals_at_[hidden]> wrote:
>
> Recently, I faced some issues in writing code that relied heavily on repetitive branching, which forced me to use AI. This proposal proposes techniques to counter this, in cases where the overhead of single dispatch is highly undesirable.

Let's start off with the form of your proposal.

Using a text file is fine, in so far as being able to read it goes.
The problem is the formatting; it's bad. None of the code is properly
indented, which makes reading it a huge pain. Paragraphs are neither
indented nor double-spaced. Grammar is all over the place; sometimes
you capitalize `I` correctly; sometimes you don't. Your code contains
obvious errors, like a lack of `break` statements in switch/case.

Overall, purely from a presentation perspective, it is very difficult
to read and get any information out of it.

In terms of the content, it's not much better. You mention this type:

vector<vector<type_tag>>, deque<vector<type_tag>>

That's not how `std::vector` works. The second template argument needs
to be an allocator, which `std::deque` is not. So is this your own
`vector` type or is this just nonsense code.

The code you linked to does not seem to be valid C++. It features both
`if constexpr` and `constexpr if`. Only one of those is legal C++.

Overall, this seems to be slop. Whether AI generated or human
generated is irrelevant. Maybe there's a coherent idea in there
*somewhere*, but you need to do some actual work to present it in a
reasonable fashion.

Received on 2026-03-30 15:13:45