Good opportunity for a free pascal operating system...

[MarkMLl]

Aside from this event probably only being a footnote in the history of Linux, do you even have a remote idea of what amount of work goes into developing a general purpose operating system? Neither Linux nor Window nor macOS came into being as they are now in a short time, they've all been worked on for decades. At the company I work at we develop our own operating system (in C++) that is underpinning our products and it's definitely a more complete and general purpose operating system than many of the toy operating system around, but despite being twenty years old it still has some way to go.

Not sure about you, but I'd estimate that the basic bring-up for an arbitrary chip is 6 months for one person (i.e. to the point where it can trap exceptions without supplemental hardware debugging). Follow that with somebody working for another six months to get a basic process switcher working, and somebody else on the memory manager. Follow that with lots of people working for a week or so at a time for each basic peripheral... you're very rapidly getting into many man-years, and that's obviously with the benefit of "knowing how to do it" since many architectures have followed the same path over the last 50 years.

MarkMLl

[VisualLab]

If I had that level of proficiency I'd budget a couple of weeks to design and build a custom CPU to go with it.

Oh, that, that! I support it with both hands and feet In this project I will support you... spiritually Because unfortunately I can't design CPU and I don't have the right budget

I respect everyone's opinion and their political stance. But this is Lazarus/Pascal programming forum, if anyone wants to talk about politics please do it in other forums.

As for respecting other people's opinions, I find this a somewhat controversial approach But I fully support the view of not raising any political issues on a technical forum. Especially since even people with extensive technical knowledge and many years of experience can sometimes have „very unusual” points of view on political issues

(and politics is generally nasty).


Linux has its flaws and quirks (and is even a pain in the ass for most people who use desktop computers on a daily basis). During its development, various people came and went (e.g. Alan Cox). As others have written, you should by no means expect any "Linux collapse". It has too many user groups who are interested in its continuation (as well as its development). About "writing" or "creating" your own OS by a small group people, it's not even worth mentioning. These are delusions - a multi-page essay could be written on this topic (but there are better ways to waste time).

[440bx]

do you even have a remote idea of what amount of work goes into developing a general purpose operating system?

That's what I was thinking too.  Writing a minimally capable O/S is a daunting task. 

In addition to that, not much of compiler writing knowledge applies to writing an O/S.

[dbannon]

Its important to keep the idea of a Linus's Kernel and Linux separate. Linus does an amazing job of keeping the Kernel project moving. But I'd expect far more lines of code go into GNU, all the absolutely essential "extras" that make up an operating system. Then beyond that, there are several more layers, desktops, application etc.

I suggest that Nomad needs to fork the kernel, that way, the Nomadix OS can continue to use GNU and, perhaps, Gnome, sounds like he might be a Gnome user. Then, he can start re-coding in FPC, each library/utility/application, one by one.

I bet there is something on youtube about how to live a longer life ....

Davo

[nomad]

Hi
There's a *tiny* difference, between a compiler and an OS ...or didn't you know that?!?
@nomad: You just keep smoking that pipe of yours, dude   
Regards Benny

You can use a compiler to write an operating system.

[nomad]

I see a opportunity for the free pascal community to take the lead when linux community goes into a frenzy.

This is... wild, to say the least. First of all, you think that what happens on the Linux Foundation side (which doesn't deal with just Linux, but also projects like Ceph, Xen, Let's Encrypt, PyTorch, RISC-V etc.) is of any concern to Free Pascal. GCC is more relevant to Linux than Free Pascal, and believing otherwise is pure delusion. Not surprising from someone fear-mongering and finding the absolute worst source you could for Linux news (Lunduke).

Weather (sic!) you believe in it or not is irrelevant because everyone is allowed to have an idea unless you are a dictator.

Flat Earth or "Tomatoes are blue" are also ideas, but are they correct? Everyone is allowed to have an idea, yes, but because you posted this on a public forum it mustn't go unchallenged, because otherwise that's how you get misinformation.

I am just saying that this mess with the linux kernel could lead to fragmentation of the community. As far as I am concerned linux is going to fall apart.

And that's because... um, a couple of maintainers were kicked? Are you for real? More people probably left Linux when Linus added Rust to the kernel than because of this nothing burger. You and Steve Ballmer are on the same level of enthusiasm and delusion to believe Linux is going to fall apart. Linux didn't fall apart in the 90s, Linux didn't fall apart in the 00s and it certainly doesn't fall apart now. This is plain misinformation and ragebait. I suppose your fingers were bored and wanted some action after watching your favorite "unbiased" news. Or Fox News, either one works.

This is a good opportunity fore the free pascal community to do a operating system that learns from the mistakes of the linux foundation. Would love to hear what the community thinks about these events.

Let's assume for a moment you're right (which is a big stretch). What makes you think that other communities wouldn't do the same and why would we be more successful? You have plenty of kernels to choose from (FOSS, ignoring Linux and the BSDs). If you really want Free Pascal, you have a couple of options, like:

• https://github.com/rezgui/fpos
• https://github.com/necokeine/StreamOS
• https://torokernel.io/
• https://github.com/Laksen/fp-rtos
• https://ultibo.org/ (the RTOS part)
• https://bitbucket.org/elera0/elerais/src
• https://www.spexeah.com/index.php/asuro/

Why are you fragmenting the community by wanting to make yet another OS, when you've got so many options? For something a tad bit more useful, the ETH Zürich team would really like your help extending and improving Oberon. Even improving https://ironclad.nongnu.org/, which is a "formally verified, hard real-time capable kernel for general-purpose and embedded uses, written in SPARK and Ada", would be amazing, and it is quite unique among kernels.

The people that are the most willing to do an alternative OS to Linux is... the Rust people. https://www.redox-os.org/ is probably the most notable Rust OS, and they could have the man power to pull off an usable OS (I personally wish those man-years could be spent into Haiku or ReactOS or AROS, but that's just me). We don't have the man power necessary, not even if you bring all somewhat active people, to do something like this. But if you truly believe Linux is going to fail, you still have the BSDs, and none of them are concerned about what happens to Linux (maybe FreeBSD cares a little bit, maybe...). You can always go to Windows or macOS or some other proprietary OS if you so wish, but I would hope that goes against your beliefs. Lead the path to a better OS, make a GitHub/GitLab/... repo with your OS and your vision for how it should be, and other people will follow suit if you can display competence, ambition and perseverance. Rēs, nōn verba. Maybe you are somehow smarter than Linus and all contributors (that intelligence should've guided you towards avoiding fear-mongering and watching Lunduke, but I digress...) and your day is 72 hours long and you can be the one to do it and compete with the likes of Haiku, ReactOS, AROS, Redox, SerenityOS etc. But I doubt it.

What you call fear mongering is what I call reality. If things go this way then there will be a need. The fact that you showed links to free pascal operating systems just proves my point. I already know about them. My point is to keep an open mind as the world is going crazy.

[nomad]

I see a opportunity for the free pascal community to take the lead when linux community goes into a frenzy. Weather you believe in it or not is irrelevant because everyone is allowed to have an idea unless you are a dictator. I am just saying that this mess with the linux kernel could lead to fragmentation of the community. As far as I am concerned linux is going to fall apart. The free pascal community could take advantage of the gap that will be left and become a new platform. AS I said. Its just an idea.

Have you /ever/ been even minimally involved in the development or maintenance of an operating system, particularly one that had to deal with PC-style equipment?

I can assure you that even something apparently well-designed like GNU Hurd gets totally impossible, particularly when one starts measuring it against the Real World with questions such as "can we run a browser on it?".

Linux undoubtedly has massive problems, including certain geopolitical ones which potentially affect the FPC and Lazarus projects equally. And for that matter the current browser-based user environment has massive problems, starting with its reliance on extremely messy scripting most of which only exists to make life easy for Google et al.

But basically, we're stuck with them.

MarkMLl

Thats what I am afraid of. The mess of the linux situation.

[nomad]

Hi,

@nomad
before discussing the "end" of Linux, read the wikipedia.

a lot of servers, supercomputers are based on Linux, even at NASA.

And I agree, the Lazarus forum is not a place to talk politics.

https://en.wikipedia.org/wiki/Linux

B->

I am not talking politics. I am trying to put others on alert to when the linux mess gets out of control.

[nomad]

Aside from this event probably only being a footnote in the history of Linux, do you even have a remote idea of what amount of work goes into developing a general purpose operating system? Neither Linux nor Window nor macOS came into being as they are now in a short time, they've all been worked on for decades. At the company I work at we develop our own operating system (in C++) that is underpinning our products and it's definitely a more complete and general purpose operating system than many of the toy operating system around, but despite being twenty years old it still has some way to go.

Not sure about you, but I'd estimate that the basic bring-up for an arbitrary chip is 6 months for one person (i.e. to the point where it can trap exceptions without supplemental hardware debugging). Follow that with somebody working for another six months to get a basic process switcher working, and somebody else on the memory manager. Follow that with lots of people working for a week or so at a time for each basic peripheral... you're very rapidly getting into many man-years, and that's obviously with the benefit of "knowing how to do it" since many architectures have followed the same path over the last 50 years.

MarkMLl

We do not need to write many of the parts of the operating systems. We just need a system good enough to emulate the api's of other operating systems like windows and linux. i.e. win api and posix. Even have dosbox to run dos apps. kolibri uses dosbox to run dos apps. freebsd emulates the posix api's to run linux apps. So you just need a base system to map the api's.

[440bx]

We do not need to write many of the parts of the operating systems.

How do you figure that ?  O/Ss don't write themselves... are you suggesting re-using parts of an existing O/S ?

We just need a system good enough to emulate the api's of other operating systems like windows and linux. i.e. win api and posix.

You make it sound like that's an easy and straightforward task.  The ring-3 Windows API consists of roughly 13,000+ functions, a good number of them are either not documented or poorly documented.  That alone is a phenomenal amount of work.

Then comes the _real_ O/S which is all the ring-0 stuff.  I don't know the number of ring-0 functions Windows has but, it is in the many thousands and I do know that they are significantly harder to use correctly than ring-3 functions, which means coding them correctly is also much harder.

Look at the source of ReactOS.  That will give you an idea of the size and complexity of what you're talking about.

[nomad]

We do not need to write many of the parts of the operating systems.

How do you figure that ?  O/Ss don't write themselves... are you suggesting re-using parts of an existing O/S ?

We just need a system good enough to emulate the api's of other operating systems like windows and linux. i.e. win api and posix.

You make it sound like that's an easy and straightforward task.  The ring-3 Windows API consists of roughly 13,000+ functions, a good number of them are either not documented or poorly documented.  That alone is a phenomenal amount of work.

Then comes the _real_ O/S which is all the ring-0 stuff.  I don't know the number of ring-0 functions Windows has but, it is in the many thousands and I do know that they are significantly harder to use correctly than ring-3 functions, which means coding them correctly is also much harder.

Look at the source of ReactOS.  That will give you an idea of the size and complexity of what you're talking about.

I never said it was easy. I am showing you that using api emulation can allow you to run existing apps on a new operating system. That means you do not need to rewrite many applications. That saves a lot of development time. You just meed to work on a os parts (windowing system, sound, input etc) that you can map the various api's to. I did my research. There is alot of existing code online like from the OS dev etc. So what you are saying makes no sense. It is just a matter of will to do it. Here is a question for you. Have you ever written an operating system? Even a boot loader. 

[440bx]

It is just a matter of will to do it.

I think you forgot that it's also a matter of knowledge and time. 

Here is a question for you. Have you ever written an operating system? Even a boot loader. 

I've played with some O/S code.  I enjoyed "Developing Your Own 32-Bit Operating System".  Have also played with MINIX but I would never claim to have written an O/S (at least not something I would consider to be a useful, functional O/S.)  I was _peripherally_ involved in the porting of an existing O/S to a new CPU architecture but did not work on the O/S itself, only on some supporting programs.

It's hard to believe you have any concept of the amount of work, time and knowledge necessary to develop a usable, functional O/S.  If you did you wouldn't be so casual about writing something to replace Linux.
 

[Handoko]

I never said it was easy. I am showing you that using api emulation can allow you to run existing apps on a new operating system. That means you do not need to rewrite many applications. That saves a lot of development time. You just meed to work on a os parts (windowing system, sound, input etc) that you can map the various api's to. I did my research.

I like your confidence and enthusiasm to take on big challenges.

In my early days of learning Pascal, instead of studying how to use DirectX, I decided to write my own. Unfortunate the performance was very poor, less than 0.1 FPS. Not willing to use DBase, I wrote my own database using Pascal. It worked without any bug and being used by some users, unfortunately it did not support index so the performance wasn't good. Instead of using any available game builder/engine, I'm now working on my own. I like reinventing the wheel. Some say I'm stupid but that is me. If I were now in my early 20, I would consider to join forces with you to write a new OS.

Because my computer was frequently infected by Stoned and Brain viruses, I studied how diskette boot sector work and wrote a tool to replace the infected diskette' boot sector with my own.

[MarkMLl]

We do not need to write many of the parts of the operating systems. We just need a system good enough to emulate the api's of other operating systems like windows and linux. i.e. win api and posix. Even have dosbox to run dos apps. kolibri uses dosbox to run dos apps. freebsd emulates the posix api's to run linux apps. So you just need a base system to map the api's.

That doesn't work. Ever. The problem is that by the time that an API is 25+ years old there is so much duplication and so many "edge cases" and not-quite-understood undocumented interactions that producing a complete implementation results in an unending stream of support tickets as things that must run are found to not run. Believe me: I'd /love/ it if it were possible to clean things up, but we're too far down a hole of our own digging for that to ever happen.

I could usefully interpose a Vernor Vinge quote here about a timekeeping system in the distant future which, ultimately, appeared to be based on when Man first set foot on his planet's moon. But when one dug just a bit further one found a discrepancy ("roughy 13 megaseconds"), because in actual fact it was based on UNIX's zero-second epoch.

The /only/ way to fix this is to start off with a cleaned-up net interface layer which throws out much of the cruft and designs in security and authentication: ignore the screams of the activists who argue that they will be held responsible for promoting sedition. On top of that put cleaned-up transport layers equivalent to TCP, UDP and SCTP.

Then design a cleaned-up browser, with well-defined client facilities which eliminate the reliance on Javascript. Focus on the fact that banking, commerce and education are important, not advertising, music and porn.

Then design an OS kernel, properly implementing process migration etc. (i.e. instead of just hiding everything behind a hypervisor, which has implications on network endpoint description). Then salvage as many Linux and Windows device drivers as possible.

I've been in this game a long time, and turned my hand to a fair number of things: perhaps not in as much depth as some around here, but with fair OS, compiler etc. coverage. The above might possibly work, but realistically it would take a global apocalypse to make it happen (which would probably destroy the infrastructure which would make it achievable).

MarkMLl

[MarkMLl]

I am not talking politics. I am trying to put others on alert to when the linux mess gets out of control.

The mess has always been out of control, so there's no need to put anybody on alert. The main issue is balancing innovation against prudence: unix+net has traditionally leant heavily towards innovation (open standards, free inspection of others' protocols and content etc.) and ignored prudence (commercial security, protection of minors, prevention of subversion by hostile states etc.).

40 years ago I worked for a mainframe company, which from the 1960s onwards provided tree-structured datacomms offering the sort of end-user experience that TBL introduced with HTTP/HTML in the early '90s. The difference is that the mainframes could- in theory- provide full security etc., while HTTP+HTML allows an undirected graph of servers and clients with anybody who can get an IP address able to set himself up as a server.

So on the one side we've got robust, secure centrally-administered systems which are hostile to individual innovation, and on the other side we've got utter anarchy on top of which the banks etc. do their best to operate.

And without intending this to invite political discussion we've also got the interests of law enforcement and state security who attempt to underwrite commerce, protect minors, and prevent subversion in a completely anarchic environment.

And in inconspicuous corners of that environment there are projects like Linux (strictly, LGX: Linux+GNU+X11), FPC and RISC-V, which do their best to make progress with a population of developers and users which might easily fall foul of the countries hosting their servers.

The situation is profoundly broken, in ways that you apparently cannot begin to understand.

MarkMLl