C++ Logo

std-proposals

Advanced search

Re: [std-proposals] Letter to the C++ community

From: Sebastian Wittmeier <wittmeier_at_[hidden]>
Date: Thu, 1 Aug 2024 16:51:30 +0200
Hi Nadir,   your proposal got on of the best treatments of any proposal on this mailing list: Many people read through the posts in your thread, thought through the ideas and (around 10!) gave multiple feedbacks to the proposal and helped you to understand, how it relates to the language and the implementations.   At the same time some better prepared proposals in July (3rd) like Esa's about contract checking got no feedback at all: https://esapulkkinen.github.io/cifl-math-library/C++/contracts.pdf   Nearly all answers to you were patient, went into great detail and tried to help you formulate a core idea better or understand, were you had wrong assumptions. They came from many different perspectives, trying to understand and improve your proposal directly - correctness of code or words/terms used, talking about math side, implementability or language philosophy behind it, be it the current or the future on.   I have to say: Some of the best answers you have not followed through or just rejected by saying, in this or similar words "one has to be bold, knowing the current C++ is a hindrance for progress and one should not keep the current C++ in mind, when developing it forward" without you working through the points of the posts themselves. After having been treated with so much advice that is rather ungrateful by you to not treating the answers with at least the same diligence as your posts and proposal has been. Also the most critical answers (those talking about theoretical or technical points) are often the best feedback in the tech world.   You are assuming that participants in the discussion have still not understood, what you want to achieve with the proposal.   Perhaps some or most have understood it, but reject it for inner discrepancies, wrong assumptions about compile-time and runtime type information, which make important parts of the proposal not work at all, mathematical reasons (cleanliness of the language philosophy), implementability problems (for compiler and runtime implementations) or just a lack of presented practical use cases.   If you ignore all of this feedback than the question is, what do you expect from this list except good advice, corrections and honest feedback?     Best, Sebastian   PS   You complain about the community. This mailing list is not about education or getting pats on your back for having a (seemingly) good idea. It is for first feedback for worked out proposals. The task of the people here is to point out shortcomings or discrepancies to improve proposals, including recommendations to better abandon them early on.   C++ is by far the most important system language, exists since four decades and is improved with a steady pace with a new standard every 3 years and typically more than a hundred proposals integrated each time, most of those have been prepared by a lot of people and discussed for many years. To fit all of this together, lots of care and diligence is used. It is a commercially very important language, but it is not owned by a single corporation, neither by an internet community.   There is a very well working ISO process for improving the language. And you have the opportunity to create variants by your own from open source compilers to try out features, as have done many before, even before they are accepted into the standard.

Received on 2024-08-01 14:51:33