One of the most important things is to have structs that say their size within, so later additions won't harm the old software that expects the old struct. Win32 API does this in a lot of places and I've seen them make major sweeping changes right under my nose, because my old software still compiled and worked just fine, and it was at my leisure to adopt the new changes, which were always easy because it all used the same API. This would make the dirent/name length dilemma a breeze to fix.
I didn't watch the whole thing and all previous videos, but the answer is: if you want to, you have to start. And start by doing the research on what is needed for an operating system and make a design and implement it. Start small with room for growth. The knowledge is out there, etc.
If Terry Davis could do it, surely we all could? I certainly have the same desire for the same reasons. M$ has already poisoned the well somewhat with securboot and whatnot. But if you're just focusing on yourself, that shouldn't be an issue. I think the old hobbyist microcomputer mindset like we had with Digital Research is boiling just under the surface waiting to be unleashed again. I'm currently going through old computers trying to build DOS from scratch trying to learn. But I do want to scale up and make something low level, but high powered, free of binaryblob blackboxes from gfx cards/wifi radios/etc. Please do what you have in mind. Any mistake can be fixed later. I feel like time is of the essence as the internet is being attacked along with our files and our rights. Godspeed.
in norway, tv channels have almost removed the option to turn off subtitling. The time of year when a program is produced is also removed from the EPG. almost no TV channels run 1080p but 720p Is technology going backwards??
That's weird. I just went onto the NRK website and subtitles are on by default but can easily be turned off in the settings. Are you sure that your device isn't malfunctioning?
@@Mackenway You can turn off subtitles on something, but it has become very common to have stuck-on text that you can't turn off. tv2 and many other channels. I saw a film on TV where you can't turn off subtitles. quite a few channels and programs where you cannot turn off text. the fact that you can't turn off text has become a bit too common. you NEVER se in the EPG. when the program is produced as you always did before. PS. don't think only NRK
I think it's one of those "if you (start) build(ing) it, they will come" type moments like with the new browser engine. Although a lot of people are weirdly attached to Linux, so I don't know if you'd get the same level of support that someone like the guys working on that new browser engine are getting. Also people simp hard for open source, so going closed source or source available might cause some issues as well.
Let's not forget how now ur forced to sign up for a Microsoft account now. Bye bye local accounts. I WON'T BE ONLINE JUST TO BE ABLE TO LOG IN TO WINDOWS. NOPE NOT HAPPENING
there are still many ways to bypass microsofts bull-ish. look up NTLite and Chris Titus Tool, both pretty awesome pieces oft software, should work with both 10 and 11 although i mainly use windows 10. most systems i tried this on utilize around 1.5~2.5gb ram in idle, windows folder is 15~23gb big (depending on the hardware i believe) and in most cases less than 100 processes on the cpu (also in idle, so no major programs open)
I contributed some code to SerenityOS before it split off into that Ladybird project. I still like the idea of SerenityOS though, but perhaps port it to run on top of the Linux kernel, but still have its own window manager, libraries, built in applications, etc. That was one major issue of Serenity was its lack of hardware support to make it widely available, it was mostly confined to running inside of a virtual machine. I think if you could build on top of an established kernel, but control everything above that it could be a decent OS, sort of like Mac OS X where Apple controls everything from the window manager, applications, compositor, etc for a unified experience.
I'm trying my best to develop one-- I have some interesting ideas around microkernels and legacy (e.g., 32-bit x86) hardware. I wish more people would give me the time of day, though and help out! lol
Firstly wee need not an OS but rather a standard , and this standards should be rock-solid and easy to comprehend. This documentation should be writen (on purpose) so that anyone could easy comprehend it for few days fully. And documentation should cover only fundamentals , things that are really meaningfull like OpenGL or Input/Output, That what it means to be fundamental , be essential. Bare minimum to do pretty much anything. Upon this fundamentals there could be built an OS (still as documentation, as a standard) And this standard should be a bare minimum as well , fully described reguarding its inner working, Nothing should be left unnatended. So you have nice system that have very structured small and consise description that have only an essential staff there. You could lern it for very short period of time, use to create pretty much anything Everything there could be created in the sam fasion. Everything to make it easy to memorize, easy to comprehend, easy to imagine and there should be no doubt how this thing would work(from top to bottom) , just rock-solid thing that you could learn in few days and use the rest of your life. This staff should be done on purpose so that is easy to understand , easy to use And you could know everything you need about this system to use and develope pretty much enything and everything in a few days. Such a description is a must. (from the people for the people) And this system should be hardware free (not tied to any particular hardware , something like java but with functionality of C++ but with ease of use of assembler - you get the idea) There should be only standard how this thing should work exactly on the hardware and software level(documentetion) because ties with hardware is a bad thing , and there might be a proprietary stuff but nevertheless this thin should have strictly determined behavior. (like OpenGL but better because there could be some loose things here and there, people should have confidence that this stuff would behave exactly this way) For such a system hardware could be built at any time, so noone could restrict it And it could be implemented in existing one if needed, x86 , arm , ... whatever , ... bring it on! And if there is no hardware (everything propriatary) we could built it ourselves , or use existing non propriatary stuff. === To implement the system we should have a bunch of Basic functions(essentials) , list of this staff and its exact behavior should be in documantation. (standard of behavior - like OpenGL) And this functions should be enough to built system upon them. Then you should implement only them and you essentially done porting such a system to a new platform. === For such a system could be writen its own compiler with strictly determined behavior There shouldnt be any question or ambiguity how it works. Comprehensible description no less important then the Thing itself. , it should be written on purpose to leave no blank spaces, ans so that you could learn it in a few days.(and use for life) Keep it simple to use , to learn , to develop easy to implement/port to any platform and simultaneously effective Simple principles ,and all of them should be explicitly stated , nothing should be left in assumption that people should know it, or should anderstand that, Description is like a number each statement could make picture more crisp(more digits after a comma) if you anderstand something it doesnt mean that everyone else do , and system it is for everyone, nothing should be left unattended.(thats the Basis of such a system , from the people for the people with care) Good description so important that without it make no sence , So you should team up with the people who could make such a description that people who prior dont even know what computer is could weild this stuff skillfuly in a few days. Be easy to use and comprehend is a main thing. === Thats my take on this matter
One idea I've had for a while is to have a distributed ledger to handle software publishing and dependencies, and have that be linked to a package manager. Of course in not actually hosting the software, that's a whole other issue, but at least in terms of bookkeeping it in git style. If there's one massive advantage a new OS should have over others is to never go down the route of dependency hell.
@@user-yx3wk7tc2t From what I've gleaned, Rust's only notable benefit is that it's supposed to be safe against buffer overflows and such. That isn't worth throwing out C to pull off. The "community" around Rust is slam full of SJW types. If someone recognizes your name and something you said that they don't like then your participation will yield cancel culture targeted at you. It's cancer and just not worth it.
Everything on the internet takes up more space. higher quality images and video. More advertising. the internet site itself is of a higher resolution and there is more going on. More security running in the background. All this takes up more space in memory and requires more CPU. power. But win 11 itself probably runs a lot of unnecessary junk in the background
Hay, it takes a lot of resources to maintain a viable surveillance state AND continuously scrape other peoples hard work off the internet to plagiarize using "A.I.".
I would love to see that! Wish I knew enough of programming to get on/start a project like that. I totally agree on what you said about functionality vs user interface ... damn, I'd rather have an OS that looks like Windows 95/98, but has way better functionality and offer me programs that could maybe, potentially replace mega corp shit like Adobe (but for real, not like KDE/Gnome bullshit). I really like the idea/point about parallel processes, opposed to what we currently have. I think the most important thing (after fully functional system is established potentially) would be making sure it would have completely proprietary and non-opensource kernel + components like WM or desktop environment, so we don't end up with same mess that Linux became now.
Wouldn't it be better to just take Linux and make an actually polished distro? Then you could at least partially use work done before. You could make the main focus for the distro usability for the average user. Also Rollin rambles are my favorite videos on the channel. Great for driving to work.
@@JodyBruchon 10-4. Just wondering would it be easier to completely customize the Linux kernel, and maybe some software, or create a new kernel from scratch? This is coming from someone with just a little coding experience.
We need to simplify the computing infrastructure so that complex operating systems are not needed. We should be able to do basic computing stuff using very primitive operating systems that don't take years for one person to write, or even completely without an operating system at all - just a bootable program should suffice for most use cases. We need to develop simple permacomputing-friendly common protocols and file formats, and not normalize bloated operating systems.
@@JodyBruchon yeah sorry i was only part way through when i posted that. you had mentioned some things about every system data harvesting and such, so i wondered if you had considered which options may be best to avoid that. i see you were also getting into more low level stuff. starting from the ground up makes sense and surely would be a major task. i wish i had some good insights or advice to help, but for now i will just wish you the best in whatever you do. :)
"We need a new OS" That's like wanting people to use a new currency, it's possible, but it'll take time for people to get used to. Even Linux (or how you say 😆) people are just giving it a chance now because of how 💩 Windows 11 is.
>we need a whole new OS and programs yeah sure but we won't get any, we might get a new OS but the problem would remain the same as it is with Linux ATM because the reality of the matter is that software development will never favor tiny market-shares. And any project that grows to any relevant size will be INEVITABLY coopted, maybe not in the lifetime of the owner, but after it for certain. You can not fix computing without fixing the underlying political issues.
well, have a look at reactos, it is not Linux or the others but is under GNU license and is still under heavy development yes but someone did what you want but the work is complex and very long . get a look at the source code or just try a nightly build to see what it is .
One of the most important things is to have structs that say their size within, so later additions won't harm the old software that expects the old struct. Win32 API does this in a lot of places and I've seen them make major sweeping changes right under my nose, because my old software still compiled and worked just fine, and it was at my leisure to adopt the new changes, which were always easy because it all used the same API. This would make the dirent/name length dilemma a breeze to fix.
@@BigRedTower Excellent insight!
I'm not sick of Rollin' Rambles; they are good second monitor content.
Remember, Microsoft bought Github FOR A REASON.
I didn't watch the whole thing and all previous videos, but the answer is: if you want to, you have to start. And start by doing the research on what is needed for an operating system and make a design and implement it. Start small with room for growth. The knowledge is out there, etc.
Jody modern IT is sad, that is the problem, you're reached the point where most of us have been since the early 2000s
If Terry Davis could do it, surely we all could? I certainly have the same desire for the same reasons. M$ has already poisoned the well somewhat with securboot and whatnot. But if you're just focusing on yourself, that shouldn't be an issue. I think the old hobbyist microcomputer mindset like we had with Digital Research is boiling just under the surface waiting to be unleashed again. I'm currently going through old computers trying to build DOS from scratch trying to learn. But I do want to scale up and make something low level, but high powered, free of binaryblob blackboxes from gfx cards/wifi radios/etc. Please do what you have in mind. Any mistake can be fixed later. I feel like time is of the essence as the internet is being attacked along with our files and our rights. Godspeed.
in norway, tv channels have almost removed the option to turn off subtitling.
The time of year when a program is produced is also removed from the EPG.
almost no TV channels run 1080p but 720p
Is technology going backwards??
That's weird. I just went onto the NRK website and subtitles are on by default but can easily be turned off in the settings. Are you sure that your device isn't malfunctioning?
@@Mackenway You can turn off subtitles on something, but it has become very common to have stuck-on text that you can't turn off. tv2 and many other channels. I saw a film on TV where you can't turn off subtitles. quite a few channels and programs where you cannot turn off text. the fact that you can't turn off text has become a bit too common. you NEVER se in the EPG. when the program is produced as you always did before. PS. don't think only NRK
Rollin Rambles are the best content on the channel. 🤷♂️
3:14 oh no he's not holding the steering wheel someone call the cops!!!
🤣
SKREEEEEEEEEEEEEBOOM
What I would give to have a browser like Netscape today.
KalibriOS is a good example of what can be achieved.
Saw some projects, but they never moved on from being just a hobby project. SerenityOS, Essence OS…
I think it's one of those "if you (start) build(ing) it, they will come" type moments like with the new browser engine. Although a lot of people are weirdly attached to Linux, so I don't know if you'd get the same level of support that someone like the guys working on that new browser engine are getting. Also people simp hard for open source, so going closed source or source available might cause some issues as well.
Let's not forget how now ur forced to sign up for a Microsoft account now. Bye bye local accounts. I WON'T BE ONLINE JUST TO BE ABLE TO LOG IN TO WINDOWS. NOPE NOT HAPPENING
there are still many ways to bypass microsofts bull-ish. look up NTLite and Chris Titus Tool, both pretty awesome pieces oft software, should work with both 10 and 11 although i mainly use windows 10. most systems i tried this on utilize around 1.5~2.5gb ram in idle, windows folder is 15~23gb big (depending on the hardware i believe) and in most cases less than 100 processes on the cpu (also in idle, so no major programs open)
If someone focuses on why can't write an OS then never will. Focus on how to do it.
The video is the latter, not the former.
I think we need mobile OS that could run normal binaries, not walled garden bs. Desktop OS is nowhere as intolerable as mobile OS is.
I contributed some code to SerenityOS before it split off into that Ladybird project. I still like the idea of SerenityOS though, but perhaps port it to run on top of the Linux kernel, but still have its own window manager, libraries, built in applications, etc. That was one major issue of Serenity was its lack of hardware support to make it widely available, it was mostly confined to running inside of a virtual machine. I think if you could build on top of an established kernel, but control everything above that it could be a decent OS, sort of like Mac OS X where Apple controls everything from the window manager, applications, compositor, etc for a unified experience.
I'm trying my best to develop one-- I have some interesting ideas around microkernels and legacy (e.g., 32-bit x86) hardware. I wish more people would give me the time of day, though and help out! lol
We need a new kernel, because Linux is compromised.
How so?
@@jimh472 - If you have to ask, then you're not in the know.
@@JohnCrawford1979 Correct. I'm not. But I earnestly want to understand. Can you elaborate?
Firstly wee need not an OS but rather a standard , and this standards should be rock-solid and easy to comprehend.
This documentation should be writen (on purpose) so that anyone could easy comprehend it for few days fully.
And documentation should cover only fundamentals , things that are really meaningfull like OpenGL or Input/Output,
That what it means to be fundamental , be essential.
Bare minimum to do pretty much anything.
Upon this fundamentals there could be built an OS (still as documentation, as a standard)
And this standard should be a bare minimum as well , fully described reguarding its inner working,
Nothing should be left unnatended.
So you have nice system that have very structured small and consise description that have only an essential staff there.
You could lern it for very short period of time, use to create pretty much anything
Everything there could be created in the sam fasion.
Everything to make it easy to memorize, easy to comprehend, easy to imagine and there should be no doubt how this thing
would work(from top to bottom) , just rock-solid thing that you could learn in few days and use the rest of your life.
This staff should be done on purpose so that is easy to understand , easy to use
And you could know everything you need about this system to use and develope pretty much enything and everything in a few days.
Such a description is a must. (from the people for the people)
And this system should be hardware free (not tied to any particular hardware , something like java but with functionality of
C++ but with ease of use of assembler - you get the idea)
There should be only standard how this thing should work exactly on the hardware and software level(documentetion)
because ties with hardware is a bad thing , and there might be a proprietary stuff but nevertheless this thin should have
strictly determined behavior. (like OpenGL but better because there could be some loose things here and there, people should have confidence that this stuff would behave exactly this way)
For such a system hardware could be built at any time, so noone could restrict it
And it could be implemented in existing one if needed, x86 , arm , ... whatever , ... bring it on!
And if there is no hardware (everything propriatary) we could built it ourselves , or use existing non propriatary stuff.
===
To implement the system we should have a bunch of Basic functions(essentials) , list of this staff and its exact behavior
should be in documantation. (standard of behavior - like OpenGL)
And this functions should be enough to built system upon them.
Then you should implement only them and you essentially done porting such a system to a new platform.
===
For such a system could be writen its own compiler with strictly determined behavior
There shouldnt be any question or ambiguity how it works.
Comprehensible description no less important then the Thing itself. , it should be written on purpose to leave no blank spaces,
ans so that you could learn it in a few days.(and use for life)
Keep it simple to use , to learn , to develop easy to implement/port to any platform and simultaneously effective
Simple principles ,and all of them should be explicitly stated , nothing should be left in assumption that people should know it,
or should anderstand that,
Description is like a number each statement could make picture more crisp(more digits after a comma)
if you anderstand something it doesnt mean that everyone else do , and system it is for everyone,
nothing should be left unattended.(thats the Basis of such a system , from the people for the people with care)
Good description so important that without it make no sence ,
So you should team up with the people who could make such a description that people who prior dont even know
what computer is could weild this stuff skillfuly in a few days.
Be easy to use and comprehend is a main thing.
===
Thats my take on this matter
One idea I've had for a while is to have a distributed ledger to handle software publishing and dependencies, and have that be linked to a package manager. Of course in not actually hosting the software, that's a whole other issue, but at least in terms of bookkeeping it in git style. If there's one massive advantage a new OS should have over others is to never go down the route of dependency hell.
I think BSD is a good alternative for an operating system.
because it is more centralized and coordinated than Linux.
It's not. The same core issues exist in BSD.
@@JodyBruchon
no perfection...
as for me, a good alternative than Linux...
I'd like to hear more of your thoughts about Rust.
@@user-yx3wk7tc2t From what I've gleaned, Rust's only notable benefit is that it's supposed to be safe against buffer overflows and such. That isn't worth throwing out C to pull off. The "community" around Rust is slam full of SJW types. If someone recognizes your name and something you said that they don't like then your participation will yield cancel culture targeted at you. It's cancer and just not worth it.
at least no one spoke yet of using temple os lol
ActiveDirectory/Cerebus is such a godsend on Windows. Neither Mac nor Linux have achieved anything close.
Kerberos?
@@JodyBruchon Yeah that's what I meant, been a while since I cracked open a Powershell book lol.
Have you considered contacting Brendan Eich?
I probably need to have something to show him first.
Everything on the internet takes up more space. higher quality images and video. More advertising. the internet site itself is of a higher resolution and there is more going on. More security running in the background.
All this takes up more space in memory and requires more CPU. power.
But win 11 itself probably runs a lot of unnecessary junk in the background
Hay, it takes a lot of resources to maintain a viable surveillance state AND continuously scrape other peoples hard work off the internet to plagiarize using "A.I.".
I would love to see that! Wish I knew enough of programming to get on/start a project like that.
I totally agree on what you said about functionality vs user interface ... damn, I'd rather have an OS that looks like Windows 95/98, but has way better functionality and offer me programs that could maybe, potentially replace mega corp shit like Adobe (but for real, not like KDE/Gnome bullshit).
I really like the idea/point about parallel processes, opposed to what we currently have.
I think the most important thing (after fully functional system is established potentially) would be making sure it would have completely proprietary and non-opensource kernel + components like WM or desktop environment, so we don't end up with same mess that Linux became now.
Wouldn't it be better to just take Linux and make an actually polished distro? Then you could at least partially use work done before. You could make the main focus for the distro usability for the average user. Also Rollin rambles are my favorite videos on the channel. Great for driving to work.
No way. The point is to dump the technical debt of the existing systems. I explicitly don't want to make something that's just another Linux distro.
@@JodyBruchon 10-4. Just wondering would it be easier to completely customize the Linux kernel, and maybe some software, or create a new kernel from scratch? This is coming from someone with just a little coding experience.
We need to simplify the computing infrastructure so that complex operating systems are not needed. We should be able to do basic computing stuff using very primitive operating systems that don't take years for one person to write, or even completely without an operating system at all - just a bootable program should suffice for most use cases. We need to develop simple permacomputing-friendly common protocols and file formats, and not normalize bloated operating systems.
@@computermaster360 I'm already on it.
@@tikkasen_urakointiTempleOS?
maybe i'm not fully on topic here, but what do you think about Qubes OS?
That's a Linux distribution.
@@JodyBruchon yeah sorry i was only part way through when i posted that. you had mentioned some things about every system data harvesting and such, so i wondered if you had considered which options may be best to avoid that. i see you were also getting into more low level stuff. starting from the ground up makes sense and surely would be a major task. i wish i had some good insights or advice to help, but for now i will just wish you the best in whatever you do. :)
"We need a new OS"
That's like wanting people to use a new currency, it's possible, but it'll take time for people to get used to. Even Linux (or how you say 😆) people are just giving it a chance now because of how 💩 Windows 11 is.
Linux is hot garbage, with a toxic community. That's why they can't even give it away.
Steam OS on the Deck is PHENOMENAL. I WAY prefer it to Windows.
It's still Linux though.
@@Shock_Treatmentwhat's wrong with Linux?
>we need a whole new OS and programs
yeah sure but we won't get any, we might get a new OS but the problem would remain the same as it is with Linux ATM because the reality of the matter is that software development will never favor tiny market-shares.
And any project that grows to any relevant size will be INEVITABLY coopted, maybe not in the lifetime of the owner, but after it for certain.
You can not fix computing without fixing the underlying political issues.
well, have a look at reactos, it is not Linux or the others but is under GNU license and is still under heavy development yes but someone did what you want but the work is complex and very long . get a look at the source code or just try a nightly build to see what it is .
Maybe start with a standard "POSIX2"/"POSIXII"
Edit: Name needs workshopping.
EPOSIS
Don't worry. I'm on it.