On Wed, Aug 4, 2021 at 2:04 PM Phil Bouchard via Std-Proposals <std-proposals@lists.isocpp.org> wrote:


On 8/4/21 1:44 PM, Andrey Semashev via Std-Proposals wrote:
On 8/4/21 8:15 PM, Phil Bouchard via Std-Proposals wrote:

On 8/3/21 11:41 PM, Phil Bouchard via Std-Proposals wrote:


On 8/3/21 11:07 PM, Emile Cormier via Std-Proposals wrote:
Sorry, but who do you expect to get involved in a C++ language extensions encumbered by patents? Does your "C++ Superset" allow a patent-free, open-source implementation?

Part of what I mentioned is patent pending but I can certainly loosen restrictions, but the Root Pointer headers will remain patented.

BTW thank God software can now be patented. Here's my anecdote:

- I was working for Corel Linux back in 2000 until Microsoft dissolved it;

- I wrote my own Fornux Powercalc and proposed it to Microsoft but got silently embraced and extended my Microsoft Powertoys:

https://github.com/philippeb8/fcalc

https://upload.wikimedia.org/wikipedia/en/6/6f/Powercalc.PNG

- Herb Sutter from Microsoft almost embraced and extended the logic of Root Pointer:

https://github.com/hsutter/gcpp


So sorry for the patent implications but this is the only way to protect ourselves these days.

Patents are a good way to bury your software and ideas in the sand. Sane people will be very cautious about using patented stuff and some won't even touch it with a ten feet pole. IMO, a language (extension) that builds on top of a patented technology is DOA.

Well it depends on the importance of the problem that is being fixed. But in no way Microsoft will embrace and extend my efforts anymore.



If you want to make profit, that's fine, but creating new programming languages is not the way. You can create tools that make the language safer - compilers, static analyzers, instrumenters, etc. - and sell those. If you want to make the language itself better then you work on that in a way that makes as easy to employ and distribute across the widest possible variety of tools, developers, applications across the industry, and patents are a major blocker in this.

Obviously following ISO standards is the way to go on the long term and to have robust support from other commercial compilers is desirable at the same time.

Again:

- 70% of cybersecurity problems are memory safety issues;

- memory management is the leading cause of security vulnerabilities in Google Chrome;

So 1 + 1 = 2. If we do not solve these problems then we'll have a huge national security issue with legacy code.

So this is my contribution to the community at the same time.


Also Linus Torvalds said that "C++ solves the wrong problems" and perhaps he was right if he was referring to memory management, but not anymore. He can just run a C++ Superset compiler over his kernel and be done with memory leaks and segfaults.

I would be curious to see Linus' response to this idea. For some reason I don't think he would be very enthusiastic. Nor do I believe in claims about another magic tool that will fix everyone's code.

Well yeah other than minor compile time fixes that may be needed, if the kernel is recompiled from scratch including all modules then it's all good. It's way better this way than rewriting and testing everything in Rust and way more convenient on the developer's learning curves and work loads.



You should do that then, and show the results, and let pen-testers give it a go.
 

--

Phil Bouchard
Founder & CTO
C.: (819) 328-4743

Fornux Logo
--
Std-Proposals mailing list
Std-Proposals@lists.isocpp.org
https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals


--
Be seeing you,
Tony