Is Microsoft at war with C++ trying to hinder its development?

The graphical capabilities of C++ they will never be developed on a fairly simple level in the future. Microsoft has long sought to block this possible development path. They even tried to kill C++ language as a tool used outside of corporate circles replacing it with theirs impotent C# version, which supposedly offers some advantages over C++ which are actually weaknesses when it comes to performance.

Implementing graphics capabilities build-in C++ this is Microsoft's biggest nightmare. This would make the existence of their operating system Windows redundant. For a many years they had almost no desire to control over the illegal versions of Windows. Everyone used Windows for free - and they were happy with that fact because in this way they gradually imposed a monopoly on almost every computer in the world as well as its software which is closely related to Windows too. And this monopoly gave them the strength to break in its infancy any sacrilege idea C++ to have build-in Graphics because the only thing Windows has IS GUI.

Do you know that because of Windows almost all games are currently unable to take full advantage of the new fast hardware and next-generation processors because this operating system throttling them and there is no way to solve this problem?

Windows and Clouds, are simply products of the US desire to put its hand on the information of all companies around the world. It's even funny now how companies around the world are being forced to use cloud services. Every country should impose a ban on its private and public business to use such cloud services because this is in fact a threat to their state security. What happens if Microsoft decides to shut down its servers or one of these countries came into conflict with the United States and they impose a sanction on them by depriving them of access to their own company information? What if a small company applies management models that are much more innovative and efficient than those of large corporations - will Microsoft delve into the personal information of this company and steal it? Of course, these are rhetorical questions. They will be a reality very soon.

That's why I'm telling you C++/Build-In-Graphic doesn't have future. This language can eliminate the largest spyware in the world WINDOWS - and no one will allow it.

But what if someone tries to create a simple operating system whose sole purpose will be to maintain the new C++ language standard and this new C++ will have built-in graphic capabilities?

COS (C Operating System) - as I said, very simple operating system, which is only the shell of the new superpower C++ standard. :-) -i.e. this operating system will be designed to serve only this programming language and nothing else. Everything will be kept simple, logical and clear - just the opposite of what they are trying to do right now in programming. Every programmer wants to be able to create something from scratch and dependence on others makes him feel uncomfortable. Maybe it's time to fulfill this dream of programmers for self-dependence.

P.s. I would be happy to join such a open project.
Last edited on
The graphical capabilities of C++ they will never be developed on a fairly simple level in the future. Microsoft has long sought to block this possible development path.

Microsoft hasn't much to do with the failure of recent efforts.

I definitely do not want to see graphics capability added to C++'s standard library and would advocate against it. The results of such an effort would be yet another graphics tool-chain to be added to the enormous stack of well-developed and fast-evolving systems which already exist.

The committee's resources are not finite, and would be better spent resolving language defects, of which there are many, and finding ways to make C++ simpler for average users.

This language can eliminate the largest spyware in the world WINDOWS

C++ is not a competitor to Windows.

One is a tool for communicating with the computer, while the other is an operating system and software suite. Totally different products.

(C Operating System)

... Unix?
Last edited on
The graphical capabilities of C++ they will never be developed on a fairly simple level in the future.

MS has little to no say about how C++ is developed.

I'm with mbozzi, adding graphic capabilities to the C++ standard when there are many excellent 3rd party libraries available would be a major step back.

This language can eliminate the largest spyware in the world WINDOWS

C++ is a programming language, Windows is an operating system. There is ZERO competition between the two. You are comparing apples and asteroids.

*nix and Mac advocates have tried to wrest the crown from Windows for years, MS has lost some of its dominance in the PC OS market, but Windows will still live on. And on and on.
I definitely do not want to see graphics capability added to C++'s standard library and would advocate against it. The results of such an effort would be yet another graphics tool-chain to be added to the enormous stack of well-developed and fast-evolving systems which already exist.


Not if it will be standard library for C++ itself. The now ones are not well-developed and not fast-evolving systems. They are full junk and soon many of them will have trouble working with Windows, which Microsoft will not want to resolve this issue, and will not give them enough information to solve these problems themselves. Microsoft is trying to shut down its operating system for external developers and restrict their access to knowledge of basic processes - and the situation will not get rosier in the future.

The committee's resources are not finite, and would be better spent resolving language defects, of which there are many, and finding ways to make C++ simpler for average users.


C++ language defects?! This language is simple enough, even its memory management is not a problem that Microsoft is trying to suggest to everyone so that its fucking operating system can take full control of memory instead of C++. They just want a monopoly and get it with carefully planned deceptions that they spread among the society of programmers. These things about the complexity of C++ are myths and legends. The only thing that sets it apart from other more modern languages is that you need to clearly understand what you are doing while programming - and that's not a bad thing at all.

Any problems with C++ they come just when you try to make programs for users with GUI. Ie just as you start interacting with Windows and Microsoft. For example, Microsoft has neglected so much MFC that after my last session with it, I started finding bug after bug - and that was a year ago. These bugs are reported but I haven't checked to see if they've fixed them. So this yours
well-developed
don't appeal to me. If you're talking to me about a third parties GUI developers for C++, you will soon find out how their products will not start properly on Windows.

GUI capabilities in standard C++ library they should now be almost mandatory. Their absence makes this language completely meaningless. If this is the standard, these programs will run on any system that have "C Operating System" - and this language will not depend on Microsoft in any way anymore.

... Unix?


I don't talk about Operation System that can handle C/C++. I talk about operation system that which is solely intended to serve the standard C/C++ and the new graphics capabilities of this language. One operation system for only one language which can run on any platform without being dependent on Microsoft. Any changes to this new operating system should only be related to changes to the C++ standard itself. This will give complete independence to this language and will release the shackles with which it is nailed to different types of operating systems now.

Do you know that Windows 10, before the last update could not recognize an integrated graphics cards of Intel as "IntelĀ® HD Graphics 630" for example? I had to tell them that - because they don't have the brains to find their own problems. And here we are talking about Intel. Or that Windows 10 Insider Preview Build 20231 burn my smart HP AC Power Adapter because it force me to update my BIOS in a way that is not appropriate to do so on HP systems?

This is the only way to save C++ from this brainless f-ckers - new graphic standard and new operating system.

This is my opinion.

P.s. If this new operating system is created from scratch, it has every chance of being much more stable and secure than anything created so far. Like I said, operating systems like Windows has problems with new hardware which are far from repairable. Microsoft can't do it, because no one in it does not understand in detail how it works and what the weaknesses are. They are trying to replace it with a new one, imperceptibly - but this cannot happen. A new operating system of the type I'm describing will crush them if it's free. Even the great variety of software products which made dependent users of Windows it will not save them.
Last edited on
unix is not gonna happen. For 30+ years unix has failed to take the gaming market, and the gaming market IS the graphics market. They have made headway: a lot of games for windows now ALSO support unix, but there are exactly what, zero? One? major games that were made for unix that had to be ported over to windows. No one makes major games targeted to anything but windows (for computers, clearly there is the console and phone world) first, anything else a distant second. A new OS isnt going to solve it either. Its been a pipe dream to have a universal OS that everyone likes for as long as there have been OS and still we soldier on wishing in one hand.

whatever M$ does, they have to convince the game devs to USE it. Managed code is garbage already; I have not met anyone (outside of MS and tangential companies) who uses it for c++ for serious development. Its used to bridge multi-language projects, mostly, from what I know of it.

Direct?? family of graphics stuff is well loved; it is a good tool set. If they can add something this good, it will be used, whether its a library or a language monkeying (remember the lawsuits over java that lead to pound c language?). But if they try to call it c++ with language additions of this magnitude, they may be in for a fight with the committee and more. Are they dumb enough to do this AGAIN? Time will tell. But its certainly not going to be a major change in the reality of c++ for at least another decade. It may spawn another new language, because we don't have enough not quite c++ watered down copies of c++ out there already.

I am a user. If someone makes a good product, I will use it. If its garbage, I won't.

I will also argue that with millenials and younger, windows has LOST some market share. Phones have cut into the real computer market deeply already, and its growing worse as the micro tech cpus are starting to be able to do things.
Last edited on
C++ is not an OS, Windows is not a programming language.
Re graphics and C++. This was supposed to be the purpose of SG13 - but it seems to be dying by death of 1000 cuts. See https://www.reddit.com/r/cpp/comments/89q6wr/sg13_2d_graphics_why_it_failed/

The latest proposal re graphics in C++ (Oct 2019) is at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2019/p0267r10.pdf

For info, details of 'experimental' features of C++ can be found at https://en.cppreference.com/w/cpp/experimental

Last edited on
The now ones are not well-developed and not fast-evolving systems
The committee does not have the resources to compete with current efforts (e.g., those of Khronos group on the quickly-evolving Vulkan), etc.; the result would be a toy unsuitable for any real work. The committee's expertise is elsewhere, and their effort is needed elsewhere. They're already grappling with process issues.

C++ language defects?!
Here is the list of active core language issues:
http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_active.html#1875

This language is simple enough [...] These things about the complexity of C++ are myths and legends.
Hinnant, Orr, Stroustrup, Vandevoorde, and Wong have this to say in p0939r4:
p0939r4 wrote:
C++ is complicated, too complicated, yet we cannot remove significant facilities and changing them is very hard. Changing parts deemed insignificant can be risky (we need better analysis tools[Winter,2016]) and the potential gains would be insignificant. However, we badly need to simplify use of C++ [...]

https://wg21.link/p0939r4
Please note the authors. Do any of those names ring any bells to you?

And this language will not depend on Microsoft in any way anymore
Microsoft's influence is limited to their status as a major implementer. As the LLVM project now competes with Visual Studio on Windows, Microsoft has less influence than ever. Indeed, ever since 2011 or so, Microsoft's standard C++ support and the general quality of their compiler has increased substantially.

GUI capabilities in standard C++ library they should now be almost mandatory.
Maybe ask politely to join the SG13 mailing list and gently gauge opinions from the folks who are trying to get 2D graphics into the standard. Note the committee's closed lists are intolerant of bullshit.

Also read the stuff @seeplus linked you. That would be a good starting place to understand their motivation & the reasons the 2D Graphics TS didn't make progress.

This will give complete independence to this language and will release the shackles with which it is nailed to different types of operating systems now.
C++ is regularly used on bare metal. I don't understand how or why you feel Microsoft has anything to do with "shackling" C++.

So this "well-developed" doesn't appeal to me.
Industry projects can be demanding. If a dependency doesn't support a required feature, it cannot be selected. Half-baked toy projects aren't sufficient. This is the reality.

Unix
Unix is the "C operating system". It was written in C, from scratch and designed to support C, from the bottom up. That's why the libc is part of the operating system; the design motivated POSIX.

I talk about operation system that which is solely intended to serve the standard C/C++

It would never be used.
Last edited on
If this new operating system is created from scratch, it has every chance of being much more stable and secure than anything created so far. Like I said, operating systems like Windows has problems

Don't get me wrong, I sympathize with the idea, but it's not realistic. Migrating to an entirely novel system is just not a reasonable choice.

- No matter how good the system is, until it is production-tested, it's too dangerous to use in production.
- Until the system is in production, you'll be limited by the number of interested developers. Without developers, there's no new software, and hence no motivation to adopt the new system.
- With no software, there's no users, and with no users there's no reason to make software.

A lower-impedance approach is needed.
Last edited on
There are already windows-like os's that have been produced. eg Haiku, ReactOS. None has gained much traction AFAIK.
Hinnant, Orr, Stroustrup, Vandevoorde, and Wong have this to say in p0939r4:


C++ is complicated, too complicated, yet we cannot remove significant facilities and changing them is very hard. Changing parts deemed insignificant can be risky (we need better analysis tools[Winter,2016]) and the potential gains would be insignificant. However, we badly need to simplify use of C++ [...]


These are empty words, unfortunately from people who should be responsible for the development of C++ language. The possibility of restructuring some of C++ " significant facilities" just to convert C++ in a perverted variant of Python so to make language accessible to much of the low-intelligence consumers and thus make it popular among them - looks a bit like of attempting to create "Instagram in programming circles". If something meaningful needs to change in that NEW C++ language it will be changed - but not at the cost of losing productivity at the expense of search for popularity. Who said it should have backward compatibility with the OLD C++ standards?! Here we are talking about a new operating system, with a new programming language on base of C++ with increased graphics capabilities, adapted to the new realities of the hardware which is currently available and totally FREE for users and coders. This should be a symbiosis between the operating system and the programming language - which exist for each other. JAVA lessons must be learned already - and this possibility of subsequent change of license they must be removed at the outset of this new C++ standard.

Microsoft is time to leave. Too much power, too little quality brain material. This is the only way we to put end of their monopoly - as they are already a brake on the technological development for all of us. It would be funny if it wasn't tragic anymore.

https://developercommunity2.visualstudio.com/t/i-would-like-visual-studio-to-have-spell-checker-f/582265?from=email&viewtype=all#T-ND1230849

P.s. As I said, this is my opinion. I may or may not be right but that is. If there is such a project, I will join it on 100 % (emilenchev1978@yahoo.com). When I'm done with Microsoft, there will be nothing left of them.
Last edited on
Everything you've said in this thread reflects a severe misunderstanding of the engineering problems inherent in software and language development (as well as misunderstandings of basic concepts), and a lot of it seems informed by personal biases that don't seem to go deeper than "I don't like Microsoft and I wish they would fail", rather than by an actual desire to solve real problems.

It would serve you well not to speak so confidently about things you clearly don't know so much about.
Last edited on
These are empty words, unfortunately from people who should be responsible for the development of C++ language. The possibility of restructuring some of C++ " significant facilities" just to convert C++ in a perverted variant of Python so to make language accessible to much of the low-intelligence consumers
C++ could be simplified without violating its fundamental design principles.

It is not a perfect language. Engineering trade-offs were made in its fundamental design in order to increase the chance of getting the language adopted. For example, Stroustrup chose to extend C partly for this reason; the rationale is written down in D&E. C compatibility continues to be a major source of difficulty to this day.
As I said, this is my opinion. I may or may not be right but that is.

You are blindly opinionated to the degree that is doing you a great disservice.

Nothing wrong with having an opinion, but your opinion should be laced with some verifiable facts. "I hate MS because they are a big poopy-head company and evil" is not a fact.
Storm, meet Teacup.

But seriously, I'm trying to work out who is at the bottom of this obviously menacing conspiracy. So far it seems like a toss-up between Hunter Biden and George Soros.

@lastchance is amazingly quiet, maybe Fortran is the sleeper here.
Topic archived. No new replies allowed.