I was browsing my media and decided to rewatch this, as I hadn't looked at it in fifteen years or so.
I was mainly struck by the unalloyed optimism of pretty much everyone who contributed, including Linus, Richard Stallman, Eric Raymond, Alan Cox, Ted T'so, Eric Allman and many other original neckbeards (I use that appellation affectionately, and in a bunch of cases, literally).
In the 20-plus years since the film was released, much has changed.
I think much of the optimism embodied by RMS and the FSF has waned a good deal (and more's the pity), and the complete reversal of Microsoft from Ballmer's "Free Software is communism" to Nadella's embrace of GNU/Linux in both Azure and WSL, to the co-opting of Linux for Google/Android, as well as aging and slow drift towards retirement/death/irrelevance of those who championed Free Software for nearly four decades have really hurt the movement, while boosting Open Source.
I think that refocusing on "free as in beer" instead of "free as in freedom" across the development community may have been inevitable as GNU/Linux (although I guess it could have been GNU/Hurd or one of the BSDs) became mainstream a couple decades after the commoditization of IBM PC-like hardware.
That got me thinking, where does that leave us and "who are the new neckbeards tht can carry the vision of Free Software into the middle of the century?" Are there really any such folks with the passion and drive to champion Free Software moving forward?
Or is Free Software (as originally defined and advocated for by RMS and the FSF) dying a slow death in favor of "Open Source" and more permissive licenses like MIT and Apache?
What will Open Source look like in 2050, 52 years after Bruce Perens and the OSI's Open Source definition?
(Score: 1, Interesting) by Anonymous Coward on Saturday June 11 2022, @12:30PM (11 children)
s/permissive/non-reciprocal/g
(Score: 1, Interesting) by Anonymous Coward on Saturday June 11 2022, @12:39PM (10 children)
Except that's never been the case, it's expensive to maintain a separate code base and to maintain comparability for any custom patches. Sure it is speed allowed and does happen, but it's never been the kind of issue that justified limiting the rights of users to use the code as they saw fit.
(Score: 1, Insightful) by Anonymous Coward on Saturday June 11 2022, @12:49PM (9 children)
You're talking about the developers not the end users. The copyleft licenses are the ones that retain the rights of the end users. Non-reciprocal licenses might give an illusion of being more enticing for developers but, yes, eventually businesses realize that maintaining a separate code base is impractical and that they eventually learn to start contributing upstream -- except when they go under first. Then all that work is lost.
(Score: 1, Insightful) by Anonymous Coward on Saturday June 11 2022, @12:56PM (7 children)
How often is that even an issue though? If a business isn't successful enough to remain in business, or be bought by somebody else, it's a pretty good bet that they aren't doing anything that's irreplaceable.
You're also downplaying the significance of being able to dip a toe in and not be forced to figure out how to keep your code from going through a forced relicensing to meet the requirements.
As I stated, there's little reason to be restricting the rights of people to use the code as they see fit, the nightmare scenarios that were envisioned when the first versions of the gpl were being drafted never happened.
(Score: 1, Insightful) by Anonymous Coward on Saturday June 11 2022, @01:00PM (3 children)
(Score: 1, Insightful) by Anonymous Coward on Saturday June 11 2022, @01:29PM (2 children)
That's a dumb argument to make. There is precisely 0 benefit to typical users unless developers get involved. In the early days, it may have been reasonable to assume enough knowledge for a random user to modify and fix bugs, but that never happened. Developers capable of adding features, fixing bugs and auditing code bases is a fraction of the total user base for most projects of note.
What's more, the whole bit about whether or not code has to share the same license is 100% about developers. It's asinine to suggest that code that gets contributed under a different license needs to be excluded purely because the license of the software that it patches has a viral clause in it. More open licenses allow for that code to be segregated into a different bit of the source tree and linked from there. It has worked just fine for decades.
At the end of the day, your whole point is pretty dumb. In practice, notable software tends to get sent back upstream for expediency's sake. And what software doesn't, tends to not be very good and not really any different from closed source software, except that only a fraction of it would need to be reverse engineered if need be. For all the inconveniences, there's simply insufficient justification for it. When the first licenses were written to include the viral clause, it may have made more sense, there were a greater percentage of computer users that could program, and there wasn't a good sense of where things would go. At this point, it's clearly not necessary and just a licensing pain. Especially when it rarely gets litigated.
(Score: 3, Insightful) by Anonymous Coward on Saturday June 11 2022, @01:53PM
Your assertion has already been proven wrong. Take a look at the documentary again, especially at the parts with Linus:
https://youtu.be/zPt_e9Cdk08 [youtu.be]
He goes on a lot about community building. However, if you need a stronger, more in-your-face statement to that effect, then take a look at the Q&A session he had ten years ago:
https://youtu.be/MShbP3OpASA?t=1875s [youtu.be]
He reiterates that the copyleft helped build a global community.
> Especially when it rarely gets litigated.
And there, amid the yammering, you hit the nail on the head. Perhaps it has something to do with both the OSI and the LF having been subverted from the inside against their original goals. Litigation is not the only approach but with M$ at the helm of both organizations no action of any kind is going to happen.
(Score: 0) by Anonymous Coward on Sunday June 12 2022, @08:07AM
No, it isn't, and you are stupid for suggesting it is. And, you are dumb, and callous, and a reprobate. Superscilious and a buffon, a doo-doo head and a poltroon! Your mother was a hamster, and your father smelt of elderberries. Phattah! Your hole counter-argument (if there is one besides the name calling), is pretty dumb, stupid, even, and not too bright. Do you use a Mac, or code for Microsoft? (sorry, did not mean to sink so low, right runaway.)
(Score: 5, Insightful) by Thexalon on Saturday June 11 2022, @03:00PM (2 children)
Except this happening [apple.com] and this happening [android.com], and for those old enough to remember this happening [wikipedia.org]. Those were all against the philosophy that was trying to be enacted via the GPL, even if Android kinda sorta gets away with it.
The more restrictive licenses exist for a reason: They're basically a demand that the works in question and any innovations with regards to those works are and always will be a public resource rather than something that can be artificially restricted by a private entity for profit. And I for one am not interested in hearing any whining about how it's so unfair that you can't take code you didn't write and didn't pay anybody to write and claim legal control over it in an effort to make money, because you have exactly zero right to somebody else's volunteer time to line your own pockets.
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 1, Interesting) by Anonymous Coward on Saturday June 11 2022, @05:22PM (1 child)
That's rather ideological of you, I don't know how you got fairness from any of what I've posted. It's got nothing to do with fairness, it's got to do with how ridiculous it is to not being to use open source code from a different license even without jumping through a bunch of hoops or getting the licensing changed. So ZFS required extra work because it couldn't simply be placed in a folder of contributed code and simply linked in.
The list you've got is hardly a doomsday scenario, it sucks, but you make it sound like nobody ever takes GPL code puts it into a closed binary and doesn't distribute the code. I'm pretty sure that does happen, and licensing won't solve it, lawsuits when found would.
I'm simply making the point that the fears that gpl was intended to address haven't ever gotten to the point that the viral clause is really necessary and that it does have costs in terms of adoption and ability to use code with other licenses.
(Score: 2, Touché) by Anonymous Coward on Saturday June 11 2022, @06:22PM
There needs to be a "-1, point missed" mod option.
(Score: 1) by khallow on Sunday June 12 2022, @01:39PM
Show this has any relevance to the end user. For example, my employer treat most software as a black box. We push the buttons, things we want happen. If they don't happen, the vendor fixes the problem.
How it's licensed and how available the code base is have no relevance to that limited process.
What's impractical about it? Proprietary code wouldn't exist in the first place, if this were always true.
(Score: 5, Interesting) by Anonymous Coward on Saturday June 11 2022, @01:04PM (4 children)
What kind of knowledge generation is characterized by discovery and experiment, and is freely shared among the community engaged in creating it?
It's called science.
Open source software is really just science: That's all.
The only difference is of course it is 'published' within its own technical means. That's the point.
Apart from that, a project on github is really no different from a scientific paper published in a scientific journal: Better in some cases, because the barrier to discovery, reproduction and extension is so much lower.
If you're for keeping knowledge 'private' and proprietary, you are basically arguing against science: The very thing that has done so much globally for the human condition since the renaissance.
Companies that have normalized proprietary software as a product (really, copyright *licenses* as a product) have basically pulled such a massive scam, that they've normalized it to the point that it has the aura of legitimacy: Even among engineers, who should be ethical enough to know better. It's still common practice to think nothing of simply supporting whatever big-name software house by buying into proprietary software.
But the whole closed-source business movement, being a scam, has been wildly successful: Just as you'd imaging some people basically cheating the world would be. It's considered legitimate, and continues to try to subvert the very means of computation production out from under us all.
In opposing open sourced software: by the very fact of the existence of the legitimacy of 'closed source' software business, they are eroding the very fundamental basis of our whole technologically based free society.
Our vulnerability in the cyber world has already begun causing social damage.
Already, you buy a computer, but you are not its owner. It has a 'TPM', but many misunderstand just who is being trusted: The chip, the 'platform', is the thing being trusted: by the manufacturer, against whoever has the computer.
Whatever feature or benefit it offers you, as the owner, is quite beside the point: It's entirely so the rent-seekers, the owners of IP who can just outright buy law like the extensions of copyright and the DMCA to force their control into your personal property, can take full control despite anything you may try to do.
Work around it? Criminal.
Now, it may be that MSFT is no longer as evil as it once was: But forcing the TPM down everyone's throats, probably at the behest of the likes of Disney, et al; Argues strongly against it.
It is not in your favour to own a devices literally designed to hold its secrets FROM you, and to go to massive lengths to do so.
What ever side benefits it offers you for use as a keystore for your own convenience is besides the point: just a cover story to make you think it works for you.
I mean, this is a subsystem able to communicate in parallel to the whole system with the internet: Entirely *around* the OS. It's there as Disney's 'inside guy' to ensure you aren't making a copy of the data they'll let you see, for your fee. It can scan and checksum arbitrary memory in your system, so it can be confirmed with utter confidence that you haven't arbitrarily changed a thing.
They'll try to tell you it 'to protect you'. It is not. It is to protect their interests: In owning your platform. It is literally a spy chip. A unified and soon to be all-encompassing way to do an end-run about your freedom to process as you like on your hardware.
And a direct threat to science itself.
Not just 'computer science'.
Not just OSS.
But the very concept of science as a social mechanism for change for the better.
The danger isn't really now - although the first signs of the corruption have set in:
OSS should be taught in computer science class. And 'IT' such as taught in 'schools of IT' should just be considered 'computer magic'. But it is presented as 'technology' when it bears much more resemblance to magic.
You can either learn computer science, or IT magic. Which is fast becoming literal handwaving, and has always been all about the 'in the know' knowledge such as what registry key to put where, what to do for what pitfall. (Fix the pitfall? Never! Except in OSS: Some would be kept, but at least really well documented, and there'd be a fork, with a similar substitute that 'fixes' such a thing.).
Smart devices have been both good and bad for OSS: They've given a reason for some big players to get on board, whilst at the same time there are what, two companies selling an actual Linux phone?
They're appliances, but have become more personal than a PC to very many people. Who are locked out of using them to their full potentials: Which is sad.
But at the same time, the phenomenon of the internet, particularly the web, and web 2.0 at that, has become a mechanism for openness. You can do anything in a browser.
It seems really unlikely at this point that they'll ever be able to lock that down, at least.
The danger is far down the track yet. But the loss of clarity, of truth, has already begun and is well underway.
(Score: 4, Insightful) by Anonymous Coward on Saturday June 11 2022, @05:28PM (3 children)
> It's called science.
Well, speaking as a scientist here, science is full of MBAs with PhDs looking for lock-in opportunities, gate-keeping and rent extraction. There used to be a kind of selflessness that it was for the good of all but in my time it has shifted to gaming the funding bodies. Those who secure grants by fair means or foul are treated like Great Men by universities enamored with Elon Musk-style Great Men who have the mysterious vision which is nothing more than empty puffery by people unencumbered with having to make anything work themselves. You know, that shitty annoying thing that motivates the rest of us to spend days/weeks/months fixing our shit.
(Score: 1, Insightful) by Anonymous Coward on Sunday June 12 2022, @04:50AM
Universities are basically corporate structures nowadays that sell educational products on Disneyland campuses. Any historical link with free-thinking and creativity is gone.
(Score: 1, Insightful) by Anonymous Coward on Sunday June 12 2022, @06:48AM (1 child)
It's worse even. Thoreau said in the 19th Century: there are no longer philosophers, only professors of philosophy. A hundred and 50 years later there are no longer scientists, only professors of science ready to scold you for inappropriate use of departmental resources, and failure to meet leadership priorities.
(Score: 1, Interesting) by Anonymous Coward on Sunday June 12 2022, @08:02AM
Oh, dear! Someone was caught doing a bit in binking on the side, without giving credit where credit was due? And Conservatives complain about not being welcome in academia!
(Score: 5, Insightful) by Rosco P. Coltrane on Saturday June 11 2022, @01:53PM (2 children)
Yeah open source has won. But here's what happened: instead of fighting open source, like the software monopolies of yore like Microsoft, the big data monopolies of today have co-opted it. Open source software is now nothing more than a free tool for them to exploit where the real money is: data. It's just a way for them to do what they need to do on the cheap.
Case in point: when open source advocates cite the successes of free software, they'll tell you: "Most of the internet run Linux on servers, and Android run on Linux". Well guess what. both use case are terrible examples of what the open-source advocates of yesteryear wanted to avoid.
Servers run Linux? Fuck if I care. Servers process my data behind the scene, is what I care about. As for Android, it's busy STEALING my data. All I see here, both in the case of servers and in the case of Android, is giant big data corporations saving money thanks to idealists creating software for them for free.
Open source is yesterday's fight. Open source won the battle, and lost the proverbial war.
What we need is new idealists hell-bent on freeing up data. There isn't any, and I doubt there will be, because the big data monopolies are well and truly entrenched by now, and won't be displaced easily. Microsoft was an easy adversary back when they were a software company. Taking them on now they're a big data player, or Google, or Amazon, or Akamai, or CloudFlare... good luck with that. The future is bleak for those who value liberty and privacy.
(Score: 3, Insightful) by Anonymous Coward on Saturday June 11 2022, @02:46PM
It's not just the big companies who insist on data collection. (You forgot the telecom companies too.) The most powerful government in the world, the US federal government, deputizes those "private" companies to collect data for it (when it's not taking the data on its own.) That's quite a list of adversaries to take on if you want to defend privacy.
realistically, your data will leak out. The only way to prevent that is to do all your work disconnected from the internet.
(Score: 2, Touché) by Anonymous Coward on Saturday June 11 2022, @08:45PM
Go read the GPL and come back with a citation showing it promised to keep your data private or to restrain the users of the code from making money in ways you consider unethical.
(Score: 3, Interesting) by meustrus on Saturday June 11 2022, @04:16PM (2 children)
What's holding back free software now is Unix. Unix is designed for time sharing the computer, which was a fine solution for sharing resources at the time but which has been obsolete since Gmail and the normalization of high-speed JavaScript engines in the browser. JavaScript has a lot of flaws, but it is a better ecosystem for two reasons: it decentralizes the actual computations running the UI and such, and the computing model makes it so service providers are not as affected by poor user security practices. The actual connection to the server is secure by default, and like it or not, JavaScript is a more secure computing environment than a local application with user mode kernel access.
The free POSIX model is a snapshot of the state of the art of networked computing 30 years ago. Where is the free equivalent of Web 1.0? Where is my private Altavista? My private Netflix? Where is the successor to OpenLDAP that can actually address everyone on a GNU social network?
...thing is, these things do exist. Getting them up and running is just as batteries-not-included now as doing anything non-trivial on a desktop Linux in the 2000s. There are scale problems. And if you're the kind of person that can do it yourself, you can get a job that'll pay so much you could just buy every service from someone else.
That's the real difference. Being free software instead of open source means leaving a LOT of money on the table. So we open source Elasticsearch and Terraform and libvirt and the whole Apache suite and we say that anyone can technically build their own Google. And we would do it ourselves, if we had the time and the inclination.
The old neckbeards though? Let's not mince words, they were practically communists. GNU isn't about the quality of software available to you, it's about the lifestyle that gives rise to software in the first place. It's not supposed to be used to extract capital first and foremost, it's supposed to be a mechanism to solve big problems WITHOUT resorting to something so dangerous as regulated greed.
GNU lost the war because POSIX as a time sharing system lost the war, and that happened a long time ago. Ultimately, we only got a glimpse at true computing freedom because, for a time, everyone involved was already free. University types that had spare time to mess around with electronics. That's not the only people in the garden anymore, so it looks like computer users have lost a lot of freedom when maybe all that happened was that computing opened up to people who already lacked that kind of economic freedom. If that's true, then we're just as stuck as the rest of society on needing to solve for democracy again.
If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
(Score: 2) by turgid on Saturday June 11 2022, @06:42PM (1 child)
It did not.
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 2, Disagree) by meustrus on Monday June 13 2022, @03:46PM
GNU winning the war looks like AGPL everywhere. Apache/MIT style open source won. I thought that point was in TFA, but I may have conflated the comment section with the article.
If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
(Score: 2, Interesting) by Anonymous Coward on Saturday June 11 2022, @04:35PM (8 children)
(Score: 2, Interesting) by Anonymous Coward on Saturday June 11 2022, @04:50PM (7 children)
Adding: I've given up contributing to NetBSD because its coreteam & management is dominated by people who're chomping at the bit to introduce sociopolitical gatekeeping, and know of three other 20+ year BSD coders who've done the same.
Goes off to check if the Hitler quotes are still in fortune(6)
(Score: 0) by Anonymous Coward on Saturday June 11 2022, @06:30PM (2 children)
So you want to be mean to trannies and they won't let you?
(Score: 3, Touché) by Anonymous Coward on Saturday June 11 2022, @11:30PM (1 child)
I don't want the trannies to be mean and exclude me from playing with everyone when I don't pay attention to whatever cause du jour the NetBSD Twitter Clique wants to get in with.
(Score: 1, Funny) by Anonymous Coward on Sunday June 12 2022, @04:04AM
I'm reporting you to the central committee of the tranny conspiracy. There will be consequences for this.
(Score: 0) by Anonymous Coward on Saturday June 11 2022, @07:28PM (1 child)
They've always been like that. Theo was kicked out due to that same tendency towards political infighting.
(Score: 1, Interesting) by Anonymous Coward on Saturday June 11 2022, @11:07PM
Disagree. NetBSD used to be the 'cancer-free' FOSS OS. It didn't have FreeBSD's hatred of GNU tools (the FSF even hosted the first server sunlamp on its network; NBSD Core never believed that BSD Purity Was The Path To Success), and it's bullshit what you're saying about Theo's expulsion (NBSD was encumbered with 4.4BSD code at the time and couldn't go full-open; guess why it's "OpenBSD" and not "SecureBSD"; OBSD needed to be hosted in Canada for a reason; also Theo's the one with the personality disorder..)
(Score: 0) by Anonymous Coward on Sunday June 12 2022, @07:29AM (1 child)
That is depressing to hear. NetBSD was always what I had in mind if anything ever happened to Slackware.
(Score: 2) by turgid on Sunday June 12 2022, @09:35AM
Me too. I put NetBSD on an old SPARC box last year.
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 0) by Anonymous Coward on Saturday June 11 2022, @07:48PM
--EOL--
(Score: 1, Insightful) by Anonymous Coward on Saturday June 11 2022, @08:48PM (2 children)
Not sure when it happened, but it should be somewhere around early '10s. I noticed that the FOSS communities changed around that time. Before that time software projects had a website, mailing list, forum board, etc. and a community formed around that software project. Contact lines were short. These days, although some software projects seem to have resisted this a bit more than others, software projects are not much more than a github page, with a bug tracker. There are hardly any communities formed around it. At least that's how I feel about it.
(Score: 2, Informative) by Anonymous Coward on Saturday June 11 2022, @11:37PM (1 child)
The Internet as a whole has forgotten how to Community... Thank the Billionaries.
(Recent epiphany: it's been said that Usenet was the 'first social media'... But imho it was more a 'collaboration media', where we got in contact with others doing similar things through newsgroups, and then developed on that through email..)
(Score: 0) by Anonymous Coward on Sunday June 12 2022, @07:18AM
what a bunch of bullshit trying to retroactively claim ownership over something.
Social media was nothing more than noob friendly geocities personal website hosting platform, that offered some interconnectivity, ranking, polling, and real time chat with other people's personal websites...
(Score: 1) by piss_drinker on Sunday June 12 2022, @03:33AM
The "big players" that have contributed to Linux etc. all come with business or academic goals. The "Movement" was always more like an argument, over how things are done. When the newly introduced idea gained enough traction, the established side compromised. What was then lost, is the contributors that had vision being the majority. This has changed vision and structuring.
Linux was originally not as I have described. It was a pet project, for building an operating system; on a machine that the vendor no longer supported. Once its union with GNU took off, it was academic and soon business applicable. But many early contributors carried the same spirit as Linus originally had.
But eventually the "Larger" group of contributors ignored the interests of the original spirit. And those there from the early days accepted the new direction. However, there have been many that complained. Some even walked away. But eventually the spirit was heavily displaced. Soon major distributions dropped support for 32bit machines. Older 3D capable video cards became essentially vesa devices. 386 support was dropped. It makes sense that these drops happen. But feeding the modern churn of progress was not the initial intent. Code inclusions, that specifically benefit large companies, became normal. This code needed to be mainline, as to establish clearly that something else added could cause breakage.
It is my opinion that not much is going to change here. I'd look to groups like Aros OS, KolibiOS, and Freedos; if you want to see that spirit again. Hardware is the problem. Moving target or obsolescence. That used to be a Linux problem. I guess its only a problem if you want to support people everywhere, not just a smaller group of determined devs. But even more problematic is adhering to modern security and technology. Its not just having a modern enough device. You need secure connections and updated libraries, or you can only play in your own little world.
(Score: 0) by Anonymous Coward on Sunday June 12 2022, @07:20AM
Open Source misses the point!
(Score: 1, Interesting) by Anonymous Coward on Sunday June 12 2022, @07:53AM (5 children)
I think the Free Software Movement, it's GNU project, and the Linux Kernel were a marriage of convenience. They both had something the other needed, desperately; but, neither side really agreed with each other on much of anything.
One side was a rag tag crew of renegades, fighting against corporate horse shit, that was slowly, and continues to, basically destroy anything good and sacred in this world. And the other side, was a bunch of very technically competent people, who, 'liked the idea,' of a 'free (as in freedom) operating system; but, not so much they were willing to take up arms and join the fight.
Marriages of convenience can work out very well for both involved. But, when the prime reason for the marriage, such as money, or family ties, becomes a moot point, the reason for sticking together begins to get in the way.
GNU and the FSF are about Software Freedom. For them, it's a political, philosophical, sociological, and culture thing. It's everything.
Open Source, I can't really speak to. I think Open Source wants everything Free Software has to offer; but, isn't willing to, or doesn't want it to be a battle. It's more of a sales pitch for free software. The problem with that is, Free Software aint fuckin' pretty. It's a patchwork quilt of all kinds of stuff, coming from all kinds of people, and many of those things and people, wouldn't want to be in the same room as eachother, much less work on something together.
So, nothing lasts forever, and marriages of convenience, in our case, ain't so bad some times; but, nothing lasts forever.
We are stuck with Linux as Kernel now. Probably, no amount of work, could ever get a new GPL3 kernel to be competitive against Linux,BSD,Binbows,or MacOS... Unless we start seeing a move away from x86x64 that is friendly towards freedom, and allows us to move in a direction where we can begin again at the kernel...
(Score: 3, Interesting) by turgid on Sunday June 12 2022, @10:05AM (4 children)
Do we need a new kernel? Linux and the BSDs follow the unix philosophy. At the time Linux was coming out initially, it was criticised for being old-fashioned, a monolith when microkernels were the thing in research.
At the time, hardware was very slow compared with what we have now. People were still using the i386 at 16 to 25MHz with 4MB RAM and maybe 20 to 40MB of disk. Microkernels had quite an overhead, so for practical reasons the monolithic kernel won.
The unix kernel design is very simple. It contains some early examples of Object Orientation, but very primitive. "Everything is a file" is such an example. Except when it isn't (ioctl()). The Bell Labs people were already on to Plan 9 [wikipedia.org] and then moved on to Inferno [wikipedia.org], both of which have some much more modern innovations.
From the Inferno wikipedia page:
The mighty Sun Microsystems, once at the forefront of Unix System V (Solaris) development dropped the ball on innovation and declared that the unix design was "good enough" and would not be moving on. Java was the future.
I've dabbled a bit in kernels. I've done some Linux driver development and some embedded work. I've also played with a couple of embedded real time OSs (including FreeRTOS). I've always wanted to write my own, but when I was younger I never had the chops, and now I don't have the time. I was inspired in the 8-bit days when I had a multitiasking system on a Z80 that ran out of an 8k ROM with 16k of RAM. An RTOS can be tiny, and probably should be. However, a kernel for a PC, workstation or server has to deal with a lot of hardware, peripherals, all sorts of fancy things including NUMA these days.
I think that if a new kernel or OS comes along, it will probably start small, out of an RTOS, and grow organically due to a particular need. RTOS kernels tend to be small and simple enough for one person to learn reasonably thoroughly. I think that it will be some sort of microkernel and it will have a Linux layer on top as a "personality" so that it can instantly use the last 30-40 years of Linux and unix software. 20 years ago, there were a number of Linux microkernel ports, for example to PowerPC (remember the Power Macs?). They were probably ahead of their time.
Hypervisors have come along too. Everyone runs VMs these days. There's no reason you can't have as many different kernels as you like running on a single machine, hardware permitting, these days.
My final rant/observation is that in the Linux world, what people see is the GUI. For most people, that's all they care about. And because reasons, the popular GUIs, the ones most people see, try to follow Windows and/or mobile phone UIs. There's precisely zero innovation, which is a shame. I don't use any of these modern GUIs except when I am forced to use Ubuntu for work and Windows. (Don't get me started on WSL. People aren't even bothering to set up a Linux system any more. This is The End if we're not careful). To get work done, which I do in my spare time because work time is hampered by Windows and Ubuntu and corporate craziness, I use Slackware, Window Maker and vim.
By the way, when I last had to write a Linux driver, the O'Reiley book "Linux Device Drivers, 3rd Edition" was about a decade behind the current state of the kernel. I had to spend about a week learning the deltas and there was no 4th Edition of the book. There's a barrier to entry. You've got to have been doing the job continuously for the last decade, learning from the forums, or you will have a mountain to climb.
I am an old man shouting at clouds.
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: -1, Redundant) by Anonymous Coward on Sunday June 12 2022, @06:11PM
" People aren't even bothering to set up a Linux system any more."
We don't need those dumb Windows-using whores ruining Ganuus+Linpox anyways!
(Score: 3, Insightful) by hendrikboom on Sunday June 12 2022, @10:01PM (2 children)
Yes, capability-based kernels were slow. Perhaps that can be mitigates somewhat by careful design. Perhaps it's also possible to use hardware to assist in this.
Nut what we're getting instead is virtual machines and bytecode interpreters.
It's entirely plausible that these mechanisms to protect programs from each other and to protect the users' own interests from malicious software are more expensive than having the OS handle capabilities at a lower level. I suspect no one has the numbers to determine this.
What would be needed to make a revolution in this direction is a capability-based system that is capable of emulating a Linux userspace for legacy use. (and maybe a Windows userspace?) New code could instead be written to use the new, underlying capability-enforcing kernel efficiently.
Much like the Linux VM's now built on top of Android and on top of Windows at significant overhead as a sop to those who still want to use Linux-based software, and wish to pretend they are safe against predation by the underlying system.
Except there's still no capability kernel.
I'm not looking forward to the overhead of running each individual app in its own complete Linux VM.
I suspect the overhead would be lower if the underlying systems had been designed from the start to handle kind of capability restrictions instead of having it kludged on top of something fundamentally different.
-- hendrik
(Score: 2) by turgid on Monday June 13 2022, @09:02PM (1 child)
A VM for every application? It makes you wonder about the design of the OS...
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 3, Insightful) by hendrikboom on Monday June 13 2022, @09:41PM
Yes, exactly. That's kind of what you get if you try to put a capability architecture on top of an existing OS.