Unix is dead. Long live Unix! It's the end of an era. As The Reg covered last week, IBM has transferred development of AIX to India. Why should IBM pay for an expensive US-based team to maintain its own proprietary flavor of official Unix when it paid 34 billion bucks for its own FOSS flavor in Red Hat? Here at The Reg FOSS desk, we've felt this was coming …
Maybe you have mistaken it for Lunix that was released for C64 in 1993?
LUnix ... It supports TCP/IP networking
From a time when many people thought that the significant defining feature of unix was unix networking.
"Open Source" meant that you could try to build and compile software to run on a different unix, but that was normally a major effort, done by somebody else. There wasn't even a standard c library, let alone a standard unix. But if you could connect your LUnix to a unix network, and talk to the other unix machines -- that was another win for unix.
From a time when many people thought that the significant defining feature of unix was unix networking.
There were lots of options for MS-DOS networking. Unix was pretty unique in support TCP/IP early on, though DOS got TCP/IP as well.
https://en.wikipedia.org/wiki/PC/TCP_Packet_Driver
And at least one imitation Unix system never got networking:
https://en.wikipedia.org/wiki/Coherent_%28operating_system%29
Unix was also pretty unique in that networking was included in the OS distribution for free and workstations were shipping with Ethernet (MAU) ports by 1985...
For practically everything else (*) networking was an extra and only available in proprietary flavours. The PC was open and hence had a wide variety of third-party networking solutions.
I forget when PC's started shipping with pre-installed Ethernet adaptors - my first laptop (a 386) in circa 1995 required a 3Com PCMCIA card, my 2000 laptop had an RJ45 Ethernet port and I think also desktops were shipping with motherboard RJ45 Ethernet ports.
Unix was also pretty unique in that networking was included in the OS distribution for free
That was certainly NOT true for Xenix and OpenServer at least:
"You have Openserver Host installed, not Enterprise. Host only contains support for serial attached terminals. You have to upgrade to Enterprise [for TCP/IP]"
https://www.unix.com/unix-for-dummies-questions-and-answers/38558-sco-unix-tcp-ip-not-licensed-error.html
https://www.scosales.com/ta/kb/125299.html
"I've found a source for the "TCP/IP 1.2.0 Supplement" for SCO Xenix 386. However, I don't have the corresponding activation keys. "
https://groups.google.com/g/comp.unix.sco.misc/c/2Sfyp5bKyrY/m/3hs6oXiYph8J
I'd say you're pretty close on PC ethernet. Dial-up modems were the go-to through the `90s, only in the very late 90s did ethernet NICs start getting included/integrated. The iMac G3 in August 15, 1998 had one. Plenty of systems sold in the early 2000s still did not, requiring an add-in NIC card.
>Unix was also pretty unique in that networking was included in the OS distribution for free
Had to do a little research on the early 80's to refresh the memory, as I remember having to use UUCP over RS-232 to move source code between different Unix/Xenix boxes in 1985, because as you noted many boxes didn't have networking and/or used different and incompatible removable media and formats. Yet was happily using Ethernet/802.3LAN on Sun.
It seems TCP/IP networking was a key differentiator between AT&T Unix and BSD, with it only being formally included in AT&T's SVR4 in 1988. However, this is only part of the story. We forget that Ethernet, was only made commercially available in 1980 and was Standardised as IEEE 802.3 LAN in 1983. Thus TCP/IP was effectively a WAN protocol suite.
Hence Xenix, for example, was based on AT&T's 1978 code and would have predated the BSD introduced TCP/IP LAN capabilities.
The Unix new boys (eg. Sun etc.), who were riding the 80's microcomputer wave, primarily focused on workstations supported by file/print/email servers and thus included LAN capabilities out-of-the-box, which was included in the BSD distro. I suggest it is this free LAN which really sealed it for TCP/IP LAN, everything else (including OSI when it arrived) was an additional cost.
I used TCP/IP on AT&T R&D Unix 5.2.6 on an Amdahl mainframe around 1988, using a Channel based Ethernet adapter.
The TCP/IP stack was a STREAMS based implementation that came out of AT&T's Indian Hill location in Chicago, although I believe that there was also a Wollongong port on the platform as well.
I used it to integrate some SunOS systems with the 5EES exchange development, just before they moved the exchange emulation from mainframe on to Sun based environments (initially on Motorola Sun 3/280 and then Sun4 SPARC based systems).
On the same network, we also had SVR2 UNIX running on a VAX (apparently supplied by DEC, but it was not Ultrix or Digital UNIX which were BSD derrived), and even 3B2 and 3B15 computers, as well as some PC's running PC/NFS. We had a mix of 10base5 and 10base2, and even Starlan10 early twisted pair Ethernet.
But we did not just use NFS. We also had AT&T RFS, a full Unix semantics remote fileystem, which at the time I thought was superior to NFS.
> From a time when many people thought that the significant defining feature of unix was unix networking.
Isn't it?
I look after an entire networked environment. To tease out the OS from the dependencies which live on the network is quite the challenge.
This isn't a bad thing: the fact of this integration into a connected environment means it's far more efficient, more available, and less surprising to end users, than the "other" OS. Unless the network goes down.
The analogy with cellphones isn't lost on me. Oh what's that? They either run a GNU OS or a BSD derivative?
Android does contain the Linux kernel, but it contains very little of the GNU operating system, at least according to this
https://www.gnu.org/philosophy/android-and-users-freedom.en.html
You can install Busybox on it, which is a stripped-down version of GNU, but that is an optional extra that most people don't use.
Vanilla AT&T UNIX didn't include TCP/IP networking until late in the System V series after it merged in Sun's BSD-based code. It was a common add-on by third parties as a selling point.
If you look at source from the day, you'll see IfDefine statements for a shedload of *IX variants because there were that many versions with differing options based on what that build had/hadn't optioned in.
Wonder why Xinuos dropped their BSD-based update to SCO-the-undead's OpenServer; licensing problems after they resurrected their parent's lawsuit?
As an enlightened, modern parent, I try to be as involved as possible in the lives of my six children. I encourage them to join team sports. I attend their teen parties with them to ensure no drinking or alcohol is on the premises. I keep a fatherly eye on the CDs they listen to and the shows they watch, the company they keep and the books they read.
You could say I’m a model parent. My children have never failed to make me proud, and I can say without the slightest embellishment that I have the finest family in the USA.Two years ago, my wife Carol and I decided that our children’s education would not be complete without some grounding in modern computers. To this end, we bought our children a brand new Compaq to learn with. The kids had a lot of fun using the handful of application programs we’d bought, such as Adobe’s Photoshop and Microsoft’s Word, and my wife and I were pleased that our gift was received so well. Our son Peter was most entranced by the device, and became quite a pro at surfing the net. When Peter began to spend whole days on the machine, I became concerned, but Carol advised me to calm down, and that it was only a passing phase. I was content to bow to her experience as a mother, until our youngest daughter, Cindy, charged into the living room one night to blurt out: “Peter is a computer hacker!”
As you can imagine, I was amazed. A computer hacker in my own house! I began to monitor my son’s habits, to make certain that Cindy wasn’t just telling stories, as she is prone to doing at times. After a few days of investigation, and some research into computer hacking, I confronted Peter with the evidence. I’m afraid to say, this was the only time I have ever been truly disappointed in one of my children. We raised them to be honest and to have integrity, and Peter betrayed the principles we tried to encourage in him, when he refused point blank to admit to his activities. His denials continued for hours, and in the end, I was left with no choice but to ban him from using the computer until he is old enough to be responsible for his actions. After going through this ordeal with my own family, I was left pondering how I could best help others in similar situations. I’d gained a lot of knowledge over those few days regarding hackers. It’s only right that I provide that information to other parents, in the hope that they will be able to tell if their children are being drawn into the world of hacking. Perhaps other parents will be able to steer their sons back onto the straight and narrow before extreme measures need to be employed.
To this end, I have decided to publish the top ten signs that your son is a hacker. I advise any parents to read this list carefully and if their son matches the profile, they should take action. A smart parent will first try to reason with their son, before resorting to groundings, or even spanking. I pride myself that I have never had to spank a child, and I hope this guide will help other parents to put a halt to their son’s misbehaviour before a spanking becomes necessary.
Has your son asked you to change ISPs? Most American families use trusted and responsible Internet Service Providers, such as AOL. These providers have a strict “No Hacking” policy, and take careful measures to ensure that your internet experience is enjoyable, educational and above all legal. If your child is becoming a hacker, one of his first steps will be to request a change to a more hacker friendly provider. I would advise all parents to refuse this request. One of the reasons your son is interested in switching providers is to get away from AOL’s child safety filter. This filter is vital to any parent who wants his son to enjoy the internet without the endangering him through exposure to “adult” content. It is best to stick with the protection AOL provides, rather than using a home-based solution. If your son is becoming a hacker, he will be able to circumvent any home-based measures with surprising ease, using information gleaned from various hacker sites.
Are you finding programs on your computer that you don’t remember installing? Your son will probably try to install some hacker software. He may attempt to conceal the presence of the software in some way, but you can usually find any new programs by reading through the programs listed under “Install/Remove Programs” in your control panel. Popular hacker software includes “Comet Cursor”, “Bonzi Buddy” and “Flash”. The best option is to confront your son with the evidence, and force him to remove the offending programs. He will probably try to install the software again, but you will be able to tell that this is happening, if your machine offers to “download” one of the hacker applications. If this happens, it is time to give your son a stern talking to, and possibly consider punishing him with a grounding.
Has your child asked for new hardware? Computer hackers are often limited by conventional computer hardware. They may request “faster” video cards, and larger hard drives, or even more memory. If your son starts requesting these devices, it is possible that he has a legitimate need. You can best ensure that you are buying legal, trustworthy hardware by only buying replacement parts from your computer’s manufacturer. If your son has requested a new “processor” from a company called “AMD”, this is genuine cause for alarm. AMD is a third-world based company who make inferior, “knock-off” copies of American processor chips. They use child labor extensively in their third world sweatshops, and they deliberately disable the security features that American processor makers, such as Intel, use to prevent hacking. AMD chips are never sold in stores, and you will most likely be told that you have to order them from internet sites. Do not buy this chip! This is one request that you must refuse your son, if you are to have any hope of raising him well.
Does your child read hacking manuals? If you pay close attention to your son’s reading habits, as I do, you will be able to determine a great deal about his opinions and hobbies. Children are at their most impressionable in the teenage years. Any father who has had a seventeen year old daughter attempt to sneak out on a date wearing make up and perfume is well aware of the effect that improper influences can have on inexperienced minds. There are, unfortunately, many hacking manuals available in bookshops today. A few titles to be on the lookout for are: “Snow Crash” and “Cryptonomicon” by Neal Stephenson; “Neuromancer” by William Gibson; “Programming with Perl” by Timothy O’Reilly; “Geeks” by Jon Katz; “The Hacker Crackdown” by Bruce Sterling; “Microserfs” by Douglas Coupland; “Hackers” by Steven Levy; and “The Cathedral and the Bazaar” by Eric S. Raymond. If you find any of these hacking manuals in your child’s possession, confiscate them immediately. You should also petition local booksellers to remove these titles from their shelves. You may meet with some resistance at first, but even booksellers have to bow to community pressure.
Actually that's an advantage, if you know how to hack things that the chances that you will be hacked are much lower than anyone who just thinks that they are smart enough to notice when they are hacked. Teach your kids how to HAK5 your home network and you will be so safe!
Well, after your first paragraph, I don't think you could be referred to as a "model parent" but rather what is commonly known as a "helicopter parent"
Also, after reading further, I call bullshit on your story overall, as "two years ago", you would hardly been able to buy a "brand new Compaq", as that brand was discontinued by HP back in 2013 and is only licensed for trademark use by 3rd parties overseas...
And skimming over the rest of the rather longish post of yours, I am not sure if this whole post was an attempt at sarcastic parody. If that was your intention, I think you failed. Miserably...
How much time does your child spend using the computer each day? If your son spends more than thirty minutes each day on the computer, he may be using it to DOS other peoples sites. DOSing involves gaining access to the “command prompt” on other people’s machines, and using it to tie up vital internet services. This can take up to eight hours. If your son is doing this, he is breaking the law, and you should stop him immediately. The safest policy is to limit your children’s access to the computer to a maximum of forty-five minutes each day.
Does your son use Quake? Quake is an online virtual reality used by hackers. It is a popular meeting place and training ground, where they discuss hacking and train in the use of various firearms. Many hackers develop anti-social tendencies due to the use of this virtual world, and it may cause erratic behaviour at home and at school. If your son is using Quake, you should make hime understand that this is not acceptable to you. You should ensure all the firearms in your house are carefully locked away, and have trigger locks installed. You should also bring your concerns to the attention of his school.
Is your son becoming argumentative and surly in his social behaviour? As a child enters the electronic world of hacking, he may become disaffected with the real world. He may lose the ability to control his actions, or judge the rightness or wrongness of a course of behaviour. This will manifest itself soonest in the way he treats others. Those whom he disagrees with will be met with scorn, bitterness, and even foul language. He may utter threats of violence of a real or electronic nature. Even when confronted, your son will probably find it difficult to talk about this problem to you. He will probably claim that there is no problem, and that you are imagining things. He may tell you that it is you who has the problem, and you should “back off” and “stop smothering him.” Do not allow yourself to be deceived. You are the only chance your son has, even if he doesn’t understand the situation he is in. Keep trying to get through to him, no matter how much he retreats into himself.
Is your son obsessed with “Lunix”? BSD, Lunix, Debian and Mandrake are all versions of an illegal hacker operation system, invented by a Soviet computer hacker named Linyos Torovoltos, before the Russians lost the Cold War. It is based on a program called ” xenix”, which was written by Microsoft for the US government. These programs are used by hackers to break into other people’s computer systems to steal credit card numbers. They may also be used to break into people’s stereos to steal their music, using the “mp3” program. Torovoltos is a notorious hacker, responsible for writing many hacker programs, such as “telnet”, which is used by hackers to connect to machines on the internet without using a telephone. Your son may try to install ” lunix” on your hard drive. If he is careful, you may not notice its presence, however, lunix is a capricious beast, and if handled incorrectly, your son may damage your computer, and even break it completely by deleting Windows, at which point you will have to have your computer repaired by a professional. If you see the word “LILO” during your windows startup (just after you turn the machine on), your son has installed lunix. In order to get rid of it, you will have to send your computer back to the manufacturer, and have them fit a new hard drive. Lunix is extremely dangerous software, and cannot be removed without destroying part of your hard disk surface.
Has your son radically changed his appearance? If your son has undergone a sudden change in his style of dress, you may have a hacker on your hands. Hackers tend to dress in bright, day-glo colors. They may wear baggy pants, bright colored shirts and spiky hair dyed in bright colors to match their clothes. They may take to carrying ” glow-sticks” and some wear pacifiers around their necks. (I have no idea why they do this) There are many such hackers in schools today, and your son may have started to associate with them. If you notice that your son’s group of friends includes people dressed like this, it is time to think about a severe curfew, to protect him from dangerous influences.
Is your son struggling academically? If your son is failing courses in school, or performing poorly on sports teams, he may be involved in a hacking group, such as the infamous ” Otaku” hacker association. Excessive time spent on the computer, communicating with his fellow hackers may cause temporary damage to the eyes and brain, from the electromagnetic radiation. This will cause his marks to slip dramatically, particularly in difficult subjects such as Math, and Chemistry. In extreme cases, over-exposure to computer radiation can cause schizophrenia, meningitis and other psychological diseases. Also, the reduction in exercise may cause him to lose muscle mass, and even to start gaining weight. For the sake of your child’s mental and physical health, you must put a stop to his hacking, and limit his computer time drastically. I encourage all parents to read through this guide carefully. Your child’s future may depend upon it. Hacking is an illegal and dangerous activity, that may land your child in prison, and tear your family apart. It cannot be taken too seriously."
You seem to forget that most good security people started by getting themselves into trouble and were mercifully helped early enough to convert that natural curiosity into something positive. That's the job - believe in your child's better nature and coach it to emerge. Assuming malintent when a child displays a talent that fits their natural interest and learning process, THAT is the way to drive a kid down the wrong road.
Having worked with Unix in the 80's and 90's, both Posix and the Open Groups licencing of the UNIX™ branding through the passing of a suite of conformance tests was very necessary to bring the various Unix source-code variants into line and at least maintain a common feature set.
Looking at the demise of proprietary Unix, I suspect Andriod in it's current form consisting of open-source and closed vendor additions from both Google and the phone manufacturer has a limited life.
I also see Linux at some stage having to develop a LINUX™ conformance testing suite.
[Author here]
> I think that's a backronym.
The name "Linux" was invented by Ari Lemmke, who hosted the FTP site it was initially posted on. He named it after Torvalds.
Torvalds' chosen name was "Freax".
https://en.wikipedia.org/wiki/History_of_Linux#:~:text=Ari%20Lemmke%20at%20Helsinki%20University,the%20server%20without%20consulting%20Torvalds.
I’m old, and probably dim, but other than Unixware 7 I can’t find anything that looks like Linux here: UNIX® Certified Products. I thought I remembered (when I was deploying this stuff), that UnixWare came out of System V?
Think it refers to this part of the article "There used to be two Chinese Linux distros which had passed the Open Group's testing and could use the Unix trademark: Inspur K/UX and Huawei EulerOS. Both companies have let the rather expensive trademark lapse, though. But the important detail here is that Linux passed and was certified as a UNIX™."
So, 2 Linux distros were certified, but they lapsed. So, if they passed, then realistically nearly all modern Linux distros would pass.
From memory I think a company did put their Linux implementation through the X/Open (aka Open Group's) POSIX testing - a necessary requirement for inclusion in Government contracts.
If POSIX has been replaced by UNIX™ as a requirement for Government contracts then there will be distributions out there that will have been tested.
[Author here]
I'm sorry but you're way out of date. This has been incorrect for 30 years. It was true from 1991 to 1993, not since.
Unix means "passes Open Group testing". That is the strict legal definition, since 1993, when Novell bought UNIX from AT&T and gave the trademark to the Open Group.
Linux passed. Linux is a UNIX™.
I learned from the original article that an organisation that publishes a specific operating system that uses the Linux kernel and an appropriate set of applications in the userland could apply for testing and then, if successful, pay the trademark fee for UNIX.
So for example the OneBornEveryMinute Inc publishes SilverLining Linux and applies to the Open Group for testing and is successful, and then pays their fee from their huge reserves of VC funded capital. We can say that SilverLining Linux is a UNIX.
Does that actually say *anything* about Joe's YOLO Linux (maintained by one geezer and run from a garden shed somewhere in Washwood Heath)? Even if Joe uses the same userland and even the same build scripts as SilverLining Linux?
Icon: Have I misunderstood? A common experience these days...
No,. I don't think you have misunderstood.
"pay a fee to use the trademark" says it all to me, especially if, as the author states, Windows could probably pass the tests too, ie if you pays your money, you win the Unix badge.
On the other hand, he does raise an interesting point. Apart from Windows, the OSs most of us use or come in contact with are FOSS these days, and even Windows has FOSS elements in it, even apart from WSL.
Just having the same source and build scripts doesn't necessarily mean Joe ends up with the same binaries as OneBornEveryMinute or that they're correct. They probably will be, but no one knows for sure until tested. Joe might unknowingly be using a borked version of GCC (though admittedly, that is a remote possibility), or something else is wrong in his set up, or someone else has got to his set up and corrupted it.
That's basically what the compliance testing is about; it cares not one hoot for the source code, just the runtime behaviour of the compiled result meets test expectations derived from the specification of what a Unix is.
You see the impact of rigorous compliance testing in the medical instruments business. A device is specified, designed, built and tested against the specification. If it involves software, the compliance passes to serially manufactured examples of that instrument if, and only if, the software load is exactly the same as used in compliance testing.
The problems set in when the OS the manufacturer chose to use gets patched; you cannot apply the patch to the instrument because you've now changed the software load. That's why a load of networked hospital gear gets so easily hacked; all the latent flaws in it cannot be fixed without a lot of money spent on re-testing. Or at least, this was the official regulatory position until quite recently.
An approach that can be followed is to use an OS that's fairly obscure and not likely to fall victim to "casual" exploitation. So where Linux, Windows are pretty common, and readily susceptible even only a short time after certification, something like INTEGRITY, QNX, VxWorks or SEL4 are less likely to fall unintended prey to an attack (they're either very good, obscure (i.e. not a common target), or both).
@bazza: even if Joe's binaries are by some miracle correct he still can't use the UNIX trademark as he has not submitted YOLO linux for testing and has not paid the fees. Correctness is not the issue.
So therefore 'Linux' is not, generally, a Unix. Some specific operating systems that use the Linux kernel and GNU-like userlands may well be Unixen.
That is my understanding. Corrections welcome.
Icon: mostly retired teacher. I had to deal with accreditation processes decades ago and had to gently inform course teams that, no, they could not 'update' or 'streamline' the course submissions without re-accrediting the courses even though in many cases the proposed changes had merit. Same idea as bazzas medical instruments (although probably less dangerous).
No corrections from this direction, that is also my understanding, other than correctness is a supposedly guaranteed side effect of paying the fees and sporting the trademark. In that sense, the "Unix" trademark and its reliance on laws around trademarks isn't so very different to GPL's reliance on copyright law to get a desired outcome never envisaged by the legislation.
Basically, it's down to what one is looking for. If one is looking for some certainty about the specification of an OS genuinely being "Unix" in all respects, one might want to choose an OS that bears the trademark which, so long as all it legit, means that Unix is what you've most definitely got.
However, if one is content to take someone's word for it, and/or rely on the experiences of others, and generally be content that it is probably close enough for one's purpose, then Ubuntu-up one's machine and off one can go. Of course, most people these days fall into this category, though it's interesting observing that the spread of systemD is making a lot of people not very happy about it.
The Importance of Accrediting Education
Re: education vs medical instruments; I'm not sure I agree. Education can set the minds of an entire civilisation for a long time, a dodgy medical instrument is more about a fewer number of individuals. Just look at what's being taught in schools in, say, China, North Korea, Russia today... Education is the most potent and powerful tool available to malignant regimes.
So I'd say that a strong course accreditation process, even if it's more commonly having to deal with "trivial" matters like good ideas for course content updates, etc. is actually a key plank in defending liberal democratic civilisation. Whilst we are fortunate that people such as yourselves have not often had to stand up strong and firm protecting education in the interests of Our Way of Life (tm), it's not unknown here in the UK. There has been some trouble with schools set up by some people who are not well meaning looking to use them to impose their own ideas that are contradictory to the values of the UK on their pupils. Such educational establishments / curricula failing accreditation (badly, in such cases) is a key part of a Judge being able to order, "this is illegal, and it shall stop."
Unix means "passes Open Group testing".
Let me stop you right there. Passing the UNIX standard does not make an OS a UNIX OS. It merely makes it "UNIX-like" or "UNIX compliant". You know, when UNIX and UNIX-like OS-es used to compete for POSIX standard compliance, nobody said "Hey, this UNIX is compliant, it is now officially called POSIX."
People don't know that what IBM forgot you still cannot do with Linux or VMware , but that's IBM's fault.
To think that Linux can do what you can do with IBM AIX s quite sad. (I do both for 25y). I could explain for an hour what sh... we have from Linux in the past years, this is no longer OS rather than Distro's (kids) who do what ever they want (Arch excluded).
Ha ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha,aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, ha ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha, aha ha ha, haaaa ha ha, ha ha ha....
If operating systems were airlines. Unix Express:
Passengers bring a piece of the airplane and a box of tools with them to the airport. They gather on the tarmac, arguing about what kind of plane they want to build. The passengers split into groups and build several different aircraft but give them all the same name. Only some passengers reach their destinations, but all of them believe they arrived.
... I boldly predicted that Linux would take over and become the dominant O/S. Most of my AIX colleagues scoffed at me at the time (you know who you are!!). I could already see that HP-UX was on the way out, and Solaris was looking shaky.
What I didn't predict is that Linux would completely take over and devour the whole world.
I'm not entirely sure this is such a good thing now, as a bit of competition seems like a good thing. Is the Linux community in danger of stagnating or becoming complacent, due to having no competition? Or does that not really matter, as someone will always have an itch that they want to scratch?
I'm not as certain about things now as I used to be when I was younger and knew less.
It's not quite the _whole_ world.
At the bottom end you still have RTOS's and bare metal for smaller/hard real time applications on CPU's that can't take the bloat of modern (or any) linux.
There are also a few schisms (e.g. with or without the dreaded systemd?), and therefore forks, although the programming API's remain fairly unchanged.
But yes, I agree it will continue eating Micro$haft's lunch, especially with the pressure of exploits, ransomware, viruses, and the company itself borking everyone's machines.
Linux isn't fully secure.
Windows isn't fully insecure.
Not that relevant. You can get most things reasonably secure. What matters is how much effort you have to put in to get there, and how long it will take for an update to nuke your hard won protection.
The latter is easy to answer in Microsoft's case: one week. Or less, if you didn't start counting from patch Tuesday.
> a bit of competition seems like a good thing
Linux's driving force isn't competition, it's urges: If enough (1 or more) developers decide they need [feature], [feature] will be implemented, maintained and improved. It's as simple as that.
On one side it's a good, solid way to make sure things are worked on, safe from the poisoned hand of marketing. On the other it's highly susceptible to fads, and you will be dragged, kicking and screaming, through all the silly IT trends which come and go.
Probably an unpopular comment here. But the IT world as a whole seems at least no less susceptible to latest fads and shiny ideas as anywhere other field. And since it's what underlies all of the modern world this affects everything from potato farming to poetry writing.
Add in FOSS elements that rely on the enthusiasm of one or more volunteer contributors who may have a certain feature preference, or need to do more paid work and spend less time fixing bugs and changing superficial features, or just don't have the enthusiasm to redraft old ideas (Or just get older).
So there is a randomness of development with stuff being abandoned because it's not profitable or not interesting enough, or suddenly changed,, or contain quirky features that a developer just liked (or perhaps because of someone in the marketing dept.if it's commercial software) and refusing to respond to what ordinary users might want. And then there is this thing we're all aware of- of a huge monolithic bit of software that's in everything, but is relying on one part-time FOSS developer who's never been paid for all the work he's put into it and is now approaching retirement age and would rather be out fishing.
Linux is the dominant enterprise OS precisely BECAUSE its license allows it to be driven by whatever urge turns you on; crowdsourcing the best enterprise OS ever being amongst them. And if I were you, I wouldn't tell the Emperor Penguin he doesn't have whims any more, or he might bite your fingers.
Really, a huge amount of the commits from individual dev's are whatever their passion projects are. The big money lets companies push their own whims, but they are still usually that, the interests of a single entity based on it's interests, budget, and needs.
There isn't anyone handing out homework assignments here, and I'd call it fair to say that only a fractional percentage of the kernel commits or userland are really collaboratively planned between the big players or prioritized solely on the interests of the greater good.
Seems like the classic open source patch comes from someone that has the right skills and is annoyed that no one else has submitted a pull request for it yet.
Stagnating? No, I don't think that's the major risk. One only has to look at the vast array of CPU architectures that Linux covers to see that, if anything, its development is more dynamic than any other OS.
To me, the bigger worry is that Linux tries to become all things to all people and ends up satisfying no one - or that enough people with big enough voices get away with making questionable changes that end up getting rolled into all distributions eventually (systemd, I'm looking at you).
yeah, systemd is what got me thinking about what could possibly happen. It just feels a bit like putting all your eggs in one basket, and then every year becoming more and more reliant on that particular basket until there's no one around who knows how to do anything with any other baskets (I think I've strangled that analogy to death). I'm probably worrying about nothing.
Linux IS "all things to all people". It currently runs on everything from the top 500 supercomputers, the majority of phones and tablets, to the tiniest of IOT devices.
The fact that one can port it and adapt it to suit one's dreams, whims, or marketing fads, (and others can embrace. adapt, or reject them) is arguably its greatest strength.
Well, the billions of installs of OSes other than Linux say that Linux is not all things to all people...
And that's before enumerating the various applications for which Linux is not suitable (eg hard real time, high assurance safety critical, microcontrollers, etc)
15 years ago I was working for a very large outsourcing company in the UK, they were hiring old timers who were last ones standing who knew a particular O/S very well indeed. These old boys would do about 5 mins work a day, rest of the time they'd just sit and read the paper. It was cheaper to simply pay these old boys a full wage to do almost nothing for peace of mind as the systems they watched were so ultra critical the owners were paying anything the outsourcer demanded.
So if you know something old, hang on to your notes as your time will come again!
It's already happened, albeit only for a few days. I've had HP-UX work paying ~£1,000 / day. I can't see it happening too often though, and it was a pain-in-the-arse to be honest as you had to discover everything on site where no one seemed to know anything about the systems.
32-bit time_t lives on in a lot of file formats.
Sure, but I suspect most of those file formats can be saved for another 68 years by presuming the fields are 32 bit unsigned and turning them into 64 bit time_t on input. Most will be file archives of one sort or another or calendars and no Unix file was ever created before 1970.
"I suspect most of those file formats can be saved for another 68 years by presuming the fields are 32 bit unsigned and turning them into 64 bit time_t on input."
Making that a reality will mean lucrative consulting contracts for the Old Fart UNIX people like me that the OP presumably had in mind. With a bit of luck, upcoming time_t projects will be my pension plan.
hm 16 year agos SGI discontinued Irix, the best unix Ever,
still have a green 2xR10K 175MHz/2GBmem Octane/MXI in the attic,
It doesnt power up, So I'm not sure whats wrong It was last powered-up around 2010.
https://en.wikipedia.org/wiki/IRIX
- xfs now the default filesystem on redhat
- CXFS - clustered filesystem long before Lustre or ZFS
- Geometry Engine and Open GL
- openMP - shared mem MPI
- 4DWM - Still miss the minimize into icon, you could easily have 100 windows arranged on a 1024x1280 screen.
- 4DWM also had backing store so, for the first time you could drag full video windows around.
- smake - sgis parallel version of make, when you had 4+ processors it was very nice
As an SGI (& Sun and SCO and..) admin back in the '90s, I also loved working with Irix. However... its TCP/IP implementation was ... full of holes IIRC - I distinctly remember having issues trying to get tcp_wrappers working correctly back on our Indigos that were running Irix 5.0 or 5.0.1 (my memory is hazy on the specifics).
Irix 5.3 for the win - things were much more stable by this point!
Irix was odd - shipped probably the most insecure configuration of any workstation Unix but had posix acls and posix capabilities at least by Irix 6.5. Based on comments in config files a version, Trusted Irix?, supported MAC.
Gcc couldn't generate the right function calling register usage for networking code - had to use the licensed mipspro compiler to build tcpwrappers amongst other stuff. Unless you could afford a license you had to be cheeky and wrap the compiler in a script that filtered out the licensing warnings so configure and make would work.
Near the end you really had to work overtime to secure these machines on the internet - I can recall host firewall ipfilter, tcpwrappers, replace rpcbind, replace naming service with resolver libraries built from isc bind, replace sendmail with postfix. And 3 ABIs (2 32bit and 1 64bit.)
Pretty ghastly really.
Ended up running netbsd or openbsd instead on an Octane that just ran a squid proxy.
The big Sony? crt monitors were eventually purloined by windows box owners. Never did decide whether Northern/Southern hemisphere monitor thing was real or some sgi techy having a lend.
One think I was grateful for was the libc support for zoneinfo files (unofficially) presumably snuck in by Olsen.
Because zonefiles are architecture independent (like terminfo) you could grab the binary zone files from another Unix and set TZ=:Australia/Sydney and done. No such luck with hpux-10.x
I think the key thing there are the words "proprietary Unix" in the original article.
It's not so much the proprietary bit but the way it was enacted with extortionate pricing to customers so when the open source and Unix-like Linux came on the scene, there was a massive and irreversible shift over to Linux.
In one sense though, Unix does still live on in the form of Linux, the BSDs and macOS.
The pricing was high, but I can remember the very high quality of Sun's documentation being a real bonus for developing on Solaris.
With Linux, and especially the mire that is the ever changing Gnome, SystemD, GTK, etc you're being actively hindered by the developers of those things. Change for change's sake in pursuit of some undefined uncertain goal with no certainty that it'll stop there is not encouraging for application developers.
Which is why Windows in particular and MacOS (despite its frequent-ish trimmings of old code) are still the platforms of choice for application development.
But Uinx™ was always open source: it was essentially unusable otherwise. This, together with generous ARPA grants, effectively kickstarted both the internet and the whole idea of software, ie. code that could be distributed independent of hardware. Who knows, historians may at some point come to view proprietary software as an anomaly.
"But Uinx™ was always open source: it was essentially unusable otherwise. "
Both these statements are false.
V7 and System III and System V ran just fine for the organisations who *bought* binary licences from whatever name AT&T chose for its software sales operation that month. And for those who had a source code licence, that licence meant they could only share code with others who also had one - academia mostly.
BSD (which had/has a "do whatever you want with this" licence) only became freely available open source after the lawsuit with AT&T in the 1990s determined BSD didn't have any AT&T code. Before that ruling, UC Berkeley could only supply BSD UNIX to those who already had an AT&T licence.
[Author here]
> Are FreeBSD and all the other FOSS-supported BSDs dead?
You miss the point of the article.
Linux passed Open Group testing. That means that, legally, Linux™ is a UNIX™.
None of the BSDs have. They are not, legally, Unixes.
Linux is. Apart from Apple macOS -- also a UNIX™ -- that means that Linux is now the only UNIX still in active development.
This is one of those "you'll get more answers than there are people answering" situations. IMHO Unix is pretty well synonymous with System V and Linux will always remain a "Unix-like", or somewhat less like given the invasive nature of Poetterware in most distros; whereas BSD was always The Other True Unix and feels more like the genuine article. Maybe I have a narrow view as I started with Ultrix-32 in 1986/87 and in my first real job I was constrained by a herd of System V boxes (R3 and sourceless: what fun!) after which it was SVR4 of various flavours and the odd encounter with OSF/1. I encountered Linux at the start of my SVR4 phase and actually preferred it in a lot of ways but it's remained a Unix-like for me. I think various Posix and OpenGroup certifications just end up with madness like MVS being counted as a Unix and nobody wants to go there.
You miss the point of the article.
It's you who's missing the point. You're using a definition which may well be correct from a legal PoV but is wildly at odds wiith the opinion of most folk who use UNIX
UNIX is an OS. It comes in many flavours and has many names. Getting a pass mark from Open Group testing is just a worthless marketing term. It gives a vendor the right to use the string "UNIX" in their sales blurbs. Whether or not that vendor's OS resembles a reasonable definition of UNIX doesn't matter.
Your chosen definition is little different to the "product of the year" marketing bollocks doled out by these chancers: https://www.productoftheyear.co.uk.
No version of Linux has passed the certification testing. Ever. Nor could it, Linux is a kernel and the specification describes an environment. Even then an incomplete one - admin tools are largely beyond its scope as are many other aspects you take for granted. Even at the most basic level - for example, there is no notion of what is or isn't a system call.
What has been certified in the past is a couple of specific versions of specific distros. Distros that specifically aimed at conformance and not your vanilla Linux distro. Remember that POSIX compliance is explicitly not an aim of the GNU project.
When Windows NT was released Microsoft claimed (or implied) that its POSIX accreditation meant that it was Unix.
It wasn't.
BSD *is* Unix in any meaningful sense.
Linux is, by design, Unix-like. Very Unix-like, to the extent that its software will compile for other Unix or Unix-like systems.
I have no idea what LinuxTM is supposed to be be.
-A.
So BSD is not Unix because nobody's paid for the copyrighted name?
What a weird hill to die on.
Some greybeards here remember the original attempts to trademark "Unix" and sneered at it back then. From the looks of the comments, a few of them are still here.
You don't seem to be convincing them.
*wishing there was a popcorn icon*
'twas always thus, etc. Vax Unix had to be called Ultrix; BSD had to be called BSD; even Motorola-licensed System V had to call its kernel /sysV68 instead of /unix because of some trademark wrangling that was never explained. It was a world where MVS could call itself Unix (or so someone claimed: I never bothered to verify if that was really the case) and BSD couldn't, but nobody called MVS Unix and everyone called BSD Unix.
Unix has been trademarked since the 1990s and nowadays operates purely in that space. Literally nobody cares about Unix certification except marketing people.
People who work in tech understand that the BSDs are "real Unix" (and have been for decades), to the extent that if I saw "Unix" on a CV, I would assume they used BSD.
This article may have had legs if it appeared on a marketing blog and was aimed at non-technical people. Trying to persuade a bunch of tech people that Linux is Unix and BSD isn't Unix just looks dumb.
Like most of the cases filed by POTUS No 45, they lack merit or even standing. How else would several of his lawyers be facing disbarment then?
This is true for many of the cases filed in the USA today. Just because anyone can sue anyone over anything does not mean that they should do so.
This post has been deleted by its author
Linux is styled after a System V Unix and the article talks about legality and trademarks, which is only half a story. As I recall in the early 1990's there was one or more court battles between SysV Unix from Bell Labs and the BSD Unix derivates from Berkeley concerning trademarks that swung partly in BSD's favour (one would need to dig up the court result (again)). I don't recall the entire story (there was a 25th anniversary book about Unix history that outlined it all at the time). I do remember that 4.4BSD came about and was then open sourced around 1994/95 and from that came NetBSD followed by OpenBSD and FreeBSD.
My point being that, while Linux is a SystemV clone that passes some certification and trademark licensing, the BSD family are still technically Unix too. Now the use of the name `Unix` vs `UNIX(tm)` is probably where lots of confusion lies to this day.
Being old I've been aware of and tinkered with Linux on and off for many years. Not deeply. ( Currently I have a little Lenovo MiiX convertible that I've truly converted- to PeppermintOS). And I had some experience of messing with a UNIX based system donkeys years ago.
But right from the start LINUX was described to me and in the stuff I read while learning about it as being a "UNIX-like OS".
So from a purely historic angle, that's what it is; "UNIX like".
I don't really care about the similarities or differences to be honest. As far as I'm concerned an OS is something you beat into submission and then leave tucked out of sight while you get on with doing stuff on a computer.
I use macOS and Linux. If macOS is officially UNIX then to call Linux UNIX-like is pretty close to laughable. At work, I also occasionally use Solaris. The differences are a little more obvious but that’s more down to a lack of recent evolution where Solaris is concerned.
MacOS had a lot of FreeBSD in it, so there is that. I still view BSD as more of a Unix than Linux is, though to say "therefore MacOS is Unix" is a bit of a stretch, whatever the trademarks say. The whole time I've worked in IT it's been the techies' experience of stuff vs. the marketroids' pedantry...
They haven't been separately maintained since the late 1990's. After POSIX.1-1996 and The Single UNIX Specification version 2 (SUSv2, 1997) were published, IEEE, The Open Group and ISO got together and formed The Austin Group, a technical working group which created a joint standard that was published as POSIX.1-2001, SUSv3 and ISO 9945-1:2002 (likewise for subsequent revisions). They are three names for one and the same document. (Actually that should be "the base volumes of SUSv3", as SUSv3 also has a separate XCurses volume.)
The distinction between POSIX(TM) and UNIX(R) in the joint standard is that it has a lot of functionality that is optional for POSIX conformance but is mandatory for UNIX conformance (in particular the "XSI option", but other optional things too).
Yes, we know that Linux doesn't use the original kernel, and isn't part of the Open Group or whatever, but it is the spiritual successor to AT&T's original offering. Linux won. Linux won everything. AIX is dead. Solaris is dead. HPUX (pronounced "H Pukes") is, thankfully, dead. SCO is dead but they'll zombie up every couple of years to sue someone.
Unix won. Thanks to Linux. Do you remember who had the early lead during the Unix Wars? It was Microsoft, with Windows NT. Go back to the late 1990's and hearing about "moving to NT" was as common as hearing about "moving to the cloud" is today. It was Linux that saved us from the dystopia of a Windows monoculture. Not IBM, not Sun, not HP.
Linux has become what Brian Valentine told us Windows would become: "The fabric of standard computing."
The fate that has befallen AIX, Solaris, and HPUX will also strike Windows Server in due time. The writing is on the wall. No one runs Windows Server anymore except to run Microsoft's own server software, and most of that they'd rather have you run on their cloud. Azure networking runs on Linux. Azure Cloud Shell runs on Linux! There is no future for Windows Server, any more than there is a future for AIX. It's merely a matter of time.
God bless the Linux Operating System, the new holder of the Unix crown.
Money talks. If the money says you can have something that does everything Unix can do, without licensing (if not support) costs, then most will listen to the money regardless. After the OS wars of the 80s and 90s, I find it ironic that one man’s idea now dominates so completely. It’s going to take a paradigm shift to usurp a free thing that copies a thing invented in the 60s (not long before I was born). I’d bet on fusion and General AI happening first (and expect both sides of the bet to outlive me).
One day in the late 1980s / early 1990s, I was debugging some C code that I had ported over to our RS/6000, and noticed something interesting after a malloc() call.
...instead of the expected value of all zeroed out memory (0x00.....), the new memory block was initialized to repeating 0xdeadbeef.
Clever people those IBMers. It's easy to spot in any hex output, and if you ever see "0xdeadbeef" in your pointers then you know you've got something wrong.
(Someone has it that it's a play on being 'dead meat' if you end up with a pointer into unallocated memory. But I hadn't heard that before.)
None of the various incarnations of BSD or Linux have recently passed the certification required to be UNIX(TM). Nor do they have any source code which came from any AT&T Unix release which wasn't made freely available at some point. As I remember it, the only reason that the *BSDs are even around is because AT&T lost in their legal efforts to claim there was any protectable Unix source code in 386BSD which would have allowed them to stop the initial release. Prior to 386BSD, to make a functional BSD system; you had to include some AT&T owned Unix source code and have a bonafide Unix license. Any modern BSDish system which anybody uses was derived from the original non-infringing 386BSD. Amusingly, BSD no longer being Unix-based happened at the same time as Linux appeared (1991-1993 time frame). I was at a Usenix conference around that time when BSD/386 (a proprietary fork of 386BSD) was announced and saw the BSD/386 guys blasted by Bill Jolitz for taking his efforts to make a freely distributable BSD and immediately make a proprietary system out of it. That pretty much cemented my switch to Linux from BSD-based systems.
AT&T vs UCB lawsuit: I seem to recall that there were some very minimal amounts of AT&T derived code in 386BSD; but at the same time AT&T had grabbed a lot of BSD's code for networking utilities, etc. without following the required copyright license. UCB, as a result, countersued for copyright infringement against AT&T. When the dust settled, BSD remained freely distributable. So BSD was once Unix, but I think you would have decide the "Ship of Theseus" question in the positive; to say modern releases are. While the Linux kernel to the best of my knowledge never had any Unix code it, many early network programs in Linux distributions were ports of the BSD code. So at that time, you could say that Linux distributions were derived from BSD which was derived from Unix, so Linux distributions were Unix as well. :-) Unless, you decide the "Ship of Theseus" question so that modern BSDs are no longer Unix. How about we just agree that they all do a good job of filling the same niches that "real" Unices did and more.
We use a mix of MacOS, Linux and FreeBSD, which means that everything we need to work does so, without fail, without uncontrolled TBs of weekly updates that break other things and without sudden UI changes that would require retraining. It's also much, MUCH faster and safer.
The killer feature, however, is that it is also cheaper. And by cheaper I mean substantially so. As far as we can tell, running Windows is bad for users, for shareholder value and for information safety so why the heck people persist in suffering breach after breach after ransomware attack is beyond me. Unless, of course, they've only ever worked with an OS which is only really good for gaming and/or their management also buys from salespeople selling extensions and double glazing..
The facts are there.
Well, I have seen a huge evolution in Unix over wide range of versions from Unix V7 with BSD enhancement which I ran on a Fortune 32:16 in 1984, SCO, Slackware Linux (Kernel .99beta), Red Hat and CentOs even Mitel Unix on a VOIP switch ending with Raspian OS on a variety of PI's in my retirement. Then there are embedded OS's on switches and routers mobile phones etc. So what is Unix? It's a tricky one.
I suppose the only sensible answer is does it pass the Duck Test, in this case the Duck being the environment described in K&R The C Programming Language.
I suppose a few essentials are:
/dev
/etc
/bin
and mtab, fstab, a kernel and a login prompt. Not required the Mount command with a syntax you can never quite guess on a new machine
Other desirable features a boot screen that scrolls to fast to see, and gives you processor speed estimated in bogomips.
Does the VMWare hypervisor count, it certainly looks quite a bit Unixly, but then so did Netware4 under the hood.
I have never seen it on an Arduino/Atmega328 but maybe that's stretching versatility too far.
While the various BSD flavors are very much UNIX-like, most of them fail to meet all of the requirements needed to obtain Single UNIX Specification certification. If you want to use the UNIX name for your OS, you have to pass SUS certification. The article you posted says the same thing.
FWIW, MacOS 10.5 is UNIX 03 compliant.
I'm old enough to remember installing 5735-HAL IBM TCP/IP FOR MVS Version 2.2 waaay back in 1992.
Otherwise known in IBM speak as the Transmission Control Protocol/Internet Protocol Programme Offering, or in plain English an unsupported experimental product. I still have the scars and nightmares about SMP/E.
And all that pain so the MVS image could have a single IP address! It was the first time I used 'ping.
Then within 3 years, along came the fully POSIX compliant OS/390.
Some might be interested in the timelines of UNIX™ and Unix-like operating system development. It's complicated.
(?over) simplified - Jason Eckhert: Ultimate UNIX timeline
Rather complicated - Éric Lévénez: Unix History. Éric's homepage has links to other histories.
Groklaw did have a project to document the history and ownership of UNIX™, but I don't think it completed before Groklaw ceased being updated.
> OS with its differences for no appreciable reason (NIH it seems)
Around early-90's? IBM had a goal to replace all AT&T code in AIX, so they could control their version of the O/S (and avoid royalties), stated by marketing droids and announced publicly in reseller industry mags. This was after being bitten by Microsoft controlling DOS & Windows, and when Sun Microsystems had done a deal with AT&T for SVR4 which IIRC a perpetual unlimited-user reseller license. Which gave Sun a pricing advantage on the O/S vs IBM, HP, etc.
POSIX/SUS didn't specify e.g. admin file formats, so IBM could do what they liked there in replacing AT&T code. Hence stanza files and other practical incompatibilities.
AFAIK IBM never finished disentangling AIX from AT&T before switching that focus to Linux.
Eh, Unix - you must mean that super-set of C code that deviated from the POSIX standards by being not 100% POSIX compliant, as to is it still around, yes very much so - as Open Indiana - Hipster
RedHat got chewed for trying to add bit's to the original Korn Shell in it's 2020 release, an had to roll it back to 93u+ again.. Not the first time some /dev has ended up in hot water or gone to jail for illegally to modify the telephone companies property and then getting caught doing it.
The QT - Quicktime hack in X is still making it's rounds whilst everyone else is eyeballing Serenity OS on a minimal hip install. Just need to delete the Webkit and QT based browser and you end up with a half decent workstation.
Ah the differences between Bourne Shell (Sanity), Bourne Again Shell (Insanity) and the phear-som Korn (Kiddie Porn Shell).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
