OSnews Decries 'The Mass Extinction of Unix Workstations' (osnews.com) 284
Anyone remember the high-end commercial UNIX workstations from a few decades ago — like from companies like IBM, DEC, SGI, and Sun Microsystems?
Today OSnews looked back — but also explored what happens when you try to buy one today> : As x86 became ever more powerful and versatile, and with the rise of Linux as a capable UNIX replacement and the adoption of the NT-based versions of Windows, the days of the UNIX workstations were numbered. A few years into the new millennium, virtually all traditional UNIX vendors had ended production of their workstations and in some cases even their associated architectures, with a lacklustre collective effort to move over to Intel's Itanium — which didn't exactly go anywhere and is now nothing more than a sour footnote in computing history.
Approaching roughly 2010, all the UNIX workstations had disappeared.... and by now, they're all pretty much dead (save for Solaris). Users and industries moved on to x86 on the hardware side, and Linux, Windows, and in some cases, Mac OS X on the software side.... Over the past few years, I have come to learn that If you want to get into buying, using, and learning from UNIX workstations today, you'll run into various problems which can roughly be filed into three main categories: hardware availability, operating system availability, and third party software availability.
Their article details their own attempts to buy one over the years, ultimately concluding the experience "left me bitter and frustrated that so much knowledge — in the form of documentation, software, tutorials, drivers, and so on — is disappearing before our very eyes." Shortsightedness and disinterest in their own heritage by corporations, big and small, is destroying entire swaths of software, and as more years pass by, it will get ever harder to get any of these things back up and running.... As for all the third-party software — well, I'm afraid it's too late for that already. Chasing down the rightsholders is already an incredibly difficult task, and even if you do find them, they are probably not interested in helping you, and even if by some miracle they are, they most likely no longer even have the ability to generate the required licenses or release versions with the licensing ripped out. Stuff like Pro/ENGINEER and SoftWindows for UNIX are most likely gone forever....
Software is dying off at an alarming rate, and I fear there's no turning the tide of this mass extinction.
The article also wonders why companies like HPE don't just "dump some ISO files" onto an FTP server, along with patch depots and documentation. "This stuff has no commercial value, they're not losing any sales, and it will barely affect their bottom line.
Today OSnews looked back — but also explored what happens when you try to buy one today> : As x86 became ever more powerful and versatile, and with the rise of Linux as a capable UNIX replacement and the adoption of the NT-based versions of Windows, the days of the UNIX workstations were numbered. A few years into the new millennium, virtually all traditional UNIX vendors had ended production of their workstations and in some cases even their associated architectures, with a lacklustre collective effort to move over to Intel's Itanium — which didn't exactly go anywhere and is now nothing more than a sour footnote in computing history.
Approaching roughly 2010, all the UNIX workstations had disappeared.... and by now, they're all pretty much dead (save for Solaris). Users and industries moved on to x86 on the hardware side, and Linux, Windows, and in some cases, Mac OS X on the software side.... Over the past few years, I have come to learn that If you want to get into buying, using, and learning from UNIX workstations today, you'll run into various problems which can roughly be filed into three main categories: hardware availability, operating system availability, and third party software availability.
Their article details their own attempts to buy one over the years, ultimately concluding the experience "left me bitter and frustrated that so much knowledge — in the form of documentation, software, tutorials, drivers, and so on — is disappearing before our very eyes." Shortsightedness and disinterest in their own heritage by corporations, big and small, is destroying entire swaths of software, and as more years pass by, it will get ever harder to get any of these things back up and running.... As for all the third-party software — well, I'm afraid it's too late for that already. Chasing down the rightsholders is already an incredibly difficult task, and even if you do find them, they are probably not interested in helping you, and even if by some miracle they are, they most likely no longer even have the ability to generate the required licenses or release versions with the licensing ripped out. Stuff like Pro/ENGINEER and SoftWindows for UNIX are most likely gone forever....
Software is dying off at an alarming rate, and I fear there's no turning the tide of this mass extinction.
The article also wonders why companies like HPE don't just "dump some ISO files" onto an FTP server, along with patch depots and documentation. "This stuff has no commercial value, they're not losing any sales, and it will barely affect their bottom line.
Obligatory car analogy (Score:5, Interesting)
Re: (Score:2)
I wonder if the same thing will happen when the current generation of cars become antiques. How are people in the 2070s with nostalgia for cars from the 2020s going to find each of the 500+ chips and necessary firmware to get them running?
It depends on the requirement. Just because the current car has 50 modules with 500 chips doesn't mean a retrofit couldn't replace many of them with some general purpose hardware and new software. As an example, I drive a 1999 rx300 which has an integrated entertainment and climate control system with a common display. You can now replace the lot with something built around an android tablet, like this: https://www.ebay.com/itm/40351... [ebay.com]
Re: (Score:2)
In 1999, you didn't have 50 computers talking to themselves over proprietary encrypted channels just to turn the dome light on.
My 1999 Lexus is entirely computer controlled. The door module talk to the body control module which talks to the ecu which determines when to turn the interior lights on and off. That's how they slowly dim out a certain amount of time after the doors are closed, and how they are controlled by the remote key fob. You can't even raise/lower a window without the computer making it happen. There is a network bus everything communicates on.
Re: (Score:3)
In 1999, you didn't have 50 computers talking to themselves over proprietary encrypted channels just to turn the dome light on.
My 1999 Lexus is entirely computer controlled. The door module talk to the body control module which talks to the ecu which determines when to turn the interior lights on and off. That's how they slowly dim out a certain amount of time after the doors are closed, and how they are controlled by the remote key fob. You can't even raise/lower a window without the computer making it happen. There is a network bus everything communicates on.
Your 1999 Lexus is not going to be of interest to collectors in 2050. Any of them that are still on the road will be driven by people who can't afford anything modern but can hack together something functional to get from place to place. As long as the mechanical parts are sound though, people will kludge together ways to make it work. Everything other than the ECU and TCU are mostly not really necessary. The nav system will no longer work with 8G. Digital gauges may be replaced with analog ones. Ligh
Re: (Score:3)
There are 2 kinds of restoration cars.
The kind where non-original / modern replacement parts are used and the kind built entirely from original parts.
Obviously, the fully original parts version is much more desirable in the car restoration world.
It's just how it is and makes sense.
No disagreement here. Restoration cars and keep it running for a long as possible beaters are entirely different things. Some cars will be worth restoring. The rest will be worth simply getting every possible mile out of, with or without all the factory electronics.
Re: Obligatory car analogy (Score:3)
I think you might be slightly overestimating the collector appeal of your 1999 Lexus if you think it will fetch more than its worth in scrap metal by 2050. By the time it would be worth any more than that, the car would be too far gone to restore unless someone who really hates money and loves bad investments decides to return it to life.
Re: (Score:3)
someone who really hates money and loves bad investments
Arguably that describes most car collectors.
Re:Obligatory car analogy (Score:5, Interesting)
They actually did in 1999. I owned a Mercedes and Volvo from those years and everything talked on the can bus. The Volvo was straight garbage. When the auto dimming mirror failed you couldn’t just grab one from a junkyard or ebay because it had to be programmed to the car. You had to open it up change the mirror itself. I bought a Chinese knockoff reader that ran Volvo’s own software and talked to every system on the car.
The Mercedes had a module under the rear seat that was guaranteed to get wet somehow and corrode the connectors right off. This module ran the pneumatic locks and played hell with the can bus when it started going bad. I documented the wiring and soldered new connectors on for a fix.
Re: (Score:2)
I doubt the issues would be technological, but more legal, as the article also encounters. Unless we get right to repair, then obsoleted chips would still be held to ransom by rights holders.
Re: (Score:2)
I wonder if the same thing will happen when the current generation of cars become antiques. How are people in the 2070s with nostalgia for cars from the 2020s going to find each of the 500+ chips and necessary firmware to get them running?
If they are cars people still care about in 50 years (which is not most of them), there will be specialist suppliers, just as there are for 50 year old cars today. Chips and modules are not that special that you can't find some factory in China to make them cheaply. Probably won't cost any more than new manufacture engine, body and suspension parts do for antiques today.
Re: (Score:2)
I bet in 2070, 12v and a ground will still turn a dome light on. But I'm sure there will be other solutions by then. A few years ago timing belts were scary...and then people learned how to work with them. A few years ago, EFI was scary...then people learned how to work with it. A few years ago CAN was scary...then people learned how to work with it.
Solution? code embedded microcontrollers (Score:4, Interesting)
2009 Cadillac XLR. Made less than 1000 of them. "Smart" LED taillights started failing right away. An OEM replacement is sometimes available for one side, for over $5000, if you can find an owner with an unused working spare. Which will also fail shortly. The other side, AFAIK, has no remaining spares at all.
Answer? Crack the housing open and install an aftermarket circuit board using a Rpi, opensource code, and commercially available LED's.
Open source programmable drivetrain modules to replace complex Tesla proprietary crap is a thing among people using the Tesla drivelines in hot rods and custom cars.
Embedded systems coding will solve all these problems.
Re: (Score:2)
Retro computer enthusiasts have this issue already. Lots of custom chips that are not made anymore. Dwindling supply of spares.
Sometimes they can be replaced by a newer part that does the same thing. Otherwise you are basically screwed.
Still bitter about the NT takeover (Score:2)
I'm still bitter that the death of the UNIX Workstation really started in earnest before Linux became a viable alternative, leading to a mass migration of that market to NT that we'll probably never fully recover from.
Re: Still bitter about the NT takeover (Score:2)
"This stuff has no commercial value, they're not losing any sales, and it will barely affect their bottom line.
Really? We make a boatload of money supporting customers who use these "no commercial value" systems. A good chunk of my day job has become keeping our junk runniing on each new release of these "no commercial value" systems that are allegedly not being maintained or sold any longer.
Linux took over commercial Linux, not WinNT (Score:4, Informative)
I'm still bitter that the death of the UNIX Workstation really started in earnest before Linux became a viable alternative, leading to a mass migration of that market to NT that we'll probably never fully recover from.
What are you talking about? Commercial UNIX users migrated to PC hardware running Linux. They did not migrate to WinNT in large numbers.
The WinNT users you see were were largely migrating from DOS, OS/2 and Win9x. People putting Windows servers on the internet were coming from the PC world not the UNIX world.
Re: (Score:2)
Re: (Score:2)
A niche market was UNIX and they largely went Linux. Sure some small segment of this niche that had legit HW needs stayed with it until the end, but most of this niche only needed *NIX. Linux was a good enough *NIX.
Linux converted few PC users. WinNT converted few *NIX users. People largely stuck with what they knew.
Mac made major inroads when they switched to Intel CPUs, they doubled their market share. Intel CPUs allowed M
Re: (Score:3)
Linux converted few PC users.
Linux converted hell of a lot of PC users, and it continues to.
Re: (Score:2)
I'm still bitter that the death of the UNIX Workstation really started in earnest before Linux became a viable alternative, leading to a mass migration of that market to NT that we'll probably never fully recover from.
What are you talking about? Commercial UNIX users migrated to PC hardware running Linux. They did not migrate to WinNT in large numbers.
The WinNT users you see were were largely migrating from DOS, OS/2 and Win9x. People putting Windows servers on the internet were coming from the PC world not the UNIX world.
NT was railing Unix back when Linux was in its diapers. Maybe by the time Windows 2000 had taken over Unix whipping, Linux was just barely being taken seriously, on servers, definitely not workstations.
Re: (Score:2)
NT was railing Unix back when Linux was in its diapers ...
That was marketing. UNIX users didn't care. They could evaluate Linux easy enough on any convenient PC.
Maybe by the time Windows 2000 had taken over Unix whipping, Linux was just barely being taken seriously, on servers, definitely not workstations.
I think you overestimate what most workstation users needed. They did not get UNIX workstations because they needed the performance, they needed to run *NIX software and there was no other option. This simple fact made for an easy transition to Linux. It was a good enough *NIX. The handful of users who did need the performance of workstation were able to stay with them. At my University the PC labs were co
Re:Still bitter about the NT takeover (Score:4, Interesting)
Yeah, the article seems a couple decades out of touch.
Frankly, with the creation of OS X, Unix workstations enjoyed an unprecedented resurgence, and are probably still near historic highs in absolute numbers. (depending on your definition of workstation I suppose) And BSD Unix never went anywhere, and doesn't look poised to do so any time soon.
The Unixes that have died seem to be the niche corporate models running on proprietary hardware that was typically massively overpriced (at least from a $/MIPS), and which frankly were already becoming very niche products by the time my interests crossed their path in the mid 90s. I never really got into Unix simply because the writing was already on the wall by the time I first laid eyes on it. That was also the era that GUIs were both finally intruding into the back office, and beginning to evolve rapidly - except for in the proprietary Unix world, where AmigOS looking GUIs were still not exactly rare.
I suspect the whole fragmented proprietary design of Unix was a huge part of its downfall. Lacking both the massive development budget of Microsoft, or the cooperpetitive code sharing of the even more fragmented Linux community, Unix just couldn't evolve fast enough to keep up. I kind of suspect BSD survived mostly because of the volunteer appeal amongst the ideological purists. It's a real shame a more modern kernel was a running vaporware joke all through those formative years - it would have been really interesting having BSD competing on more even footing with Linux. Then again there were a lot of marginal years for Linux in there, when it's survival beyond the hard-core enthusiasts crowd didn't look at all assured. A more even splitting of the volunteer community might have doomed Linux to the same fringe-of-the-fringe status, without giving BSD the edge to get out either. But BSD might also have emerged the clear winner, and spawned many more proprietary spinnoffs. Though I do suspect that the fact that the GPL made Linux was mostly immune to such spinoffs is a large part of why so many corporations poured so many resources into the evolution of a non-proprietary code base.
Re: (Score:2)
But BSD might also have emerged the clear winner, and spawned many more proprietary spinnoffs. Though I do suspect that the fact that the GPL made Linux was mostly immune to such spinoffs is a large part of why so many corporations poured so many resources into the evolution of a non-proprietary code base.
I would argue this is why everything is cloud/api based. It's the GPL loophole. If you distribute a modified Linux, you have to make it open source. But if you modify things and then just sell access to your API or rent out time on your server, now you can charge money for GPL code and modify that code without publishing your modifications.
Would the cloud/api trend taken over anyway? Probably. But I don't think it would be as absolute as it is today, where software vendors will do anything possible to make
BSD is alive and well (Score:5, Insightful)
Much of the Unix Workstation world and developers use MacOS, which is based on BSD, not System/V. That means ever iPhone and IPad runs BSD. BSD is arguably more widespread than Windows.
Android is basically Linux, and there are more Android Phones than IPhones, the combined market for Linux and BSD outclasses Windows by an order of magnitude.
Unix Workstations just moved into your jacket pocket. An IPad is more powerful than a Cray X/MP was back in the day, never mind a Unix Workstation.
Re: (Score:2)
leading to a mass migration of that market to NT that we'll probably never fully recover
Why is this a problem? NT was and is a good replacement for Unix now that Windows has matured. For things that you need a Unix like environment for you have Linux is various flavors. Unix had a good run but time have changed.
Re: (Score:2)
The Navigator's Guild? (Score:5, Insightful)
The premise of the blurb brings up a good question. What happens when your hardware is obsolete? I saw this with a client with CAD/CAM in the 90's. Hardware and OS were no longer supported but they had all the data. End of the line.
I think we're headed for Technology Guild kind of future. Look around, nobody understands what's going on, and the data genie is out of the bottle. Apple and Google will become sentient and people will forget how to make the tech. We'll control it for a while, some people will write the policies, but eventually CHOMP, the last need to think will disappear. People are just like that. Lazy. but as Amazon (also sentient) lowers the prices so damn low you'd be crazy NOT to sign in blood....
Technologists will be caught in their own trap. Like a genie, everyone thinks they are the one who will be able to control it. Elon? Marky? The future looks closed.
Re: (Score:3)
You would be surprised how serviceable a Tesla is. Tesla wants you to think that it's all sealed units and only they can work on it, but there are plenty of people fixing them out of warranty or because Tesla abandoned them after a crash.
Re: (Score:3)
Teslas are a bad example - frankly most of the big name EVs are. There's currently a big "be fancy and high tech" trend in big-name EVs, possibly because the batteries still carry a premium that needs to be emotionally justified among buyers.
Keep in mind though that EVs predate the invention of the internal combustion engine, and through most of their market history a huge part of their appeal has been the fact that they're far simpler, cheaper, more serviceable, and less prone to needing service in the fir
Re: (Score:3)
You can get a surprisingly capable and comfortable short-range eCar for under $1000, or something far closer to a "normal" car for under $5000, and if anything does go wrong with the drive train (or most other components) you can order replacement parts from dozens of different factories, few of which even have any special connection to cars.
If they start making credible pickups and SUVs they might even catch on.
Re: (Score:3)
Most of the world isn't fixated on fake utility vehicles.
Buy a XEON box and run your favorite linux. (Score:3)
Oh please. Just buy a high end XEON CPU based windows box, ditch Windows, and run your favorite version of Linux on it.
It will be more powerful than any of those old dedicated UNIX workstations.
Re: (Score:2)
Re: (Score:2)
Oh its more than geneology.
BSD is a straight up bonafide SYSV descended UNIX.
(I'd argue Linux is a genuine UNIX too. Its posix compliant , a first class citizen of the workstation and mainframe worlds, and has most of the assumptions necessary for most of the old UNIX software to be compiled with little to no modification.
Comment removed (Score:5, Informative)
Re: (Score:2)
- Mitch Hedberg, probably.
Re: (Score:2)
>> (I'd argue Linux is a genuine UNIX too. Its posix compliant , a first class citizen of the workstation and mainframe worlds, and has most of the assumptions necessary for most of the old UNIX software to be compiled with little to no modification.
i realize it's unix-like, but it's not unix. if you don't believe me ask stallman, man. linuxs torvalds didn't write a unix-like os, he just wrote a kernel suitable for one. he wrote the fact that it is not unix write into the name of the damn userland
Re: Buy a XEON box and run your favorite linux. (Score:2)
Java was originally designed to run on cable boxes.
And does anyone remember WebTV for browsing to inter webs?
Re: (Score:2, Troll)
Oh its more than geneology.
BSD is a straight up bonafide SYSV descended UNIX.
(I'd argue Linux is a genuine UNIX too. Its posix compliant , a first class citizen of the workstation and mainframe worlds, and has most of the assumptions necessary for most of the old UNIX software to be compiled with little to no modification.
I'd argue Linux used to be before Systemd. It is much less UNIX like now.
Re: (Score:2)
So by default, that flag was off to prevent that forced-write-after-read behaviour.
I don't know if the situation's changed now.
Re: (Score:2)
Oh please. Just buy a high end XEON CPU based windows box, ditch Windows, and run your favorite version of Linux on it. It will be more powerful than any of those old dedicated UNIX workstations.
Power isn't the only reason people run the old UNIXes. They often had a consistent desktop design philosophy that was pushed everywhere. IRIX had a nice web/html styled UI for controlling system functions. CDE has it's way of doing things. It's like Gnome vs KDE vs MATE vs XFCE but with Unixes having their own DEs to compete against those Linux ones.
Re: (Score:2)
Buy a Xeon box and run your favorite Linux as host and favorite Unix in a virtual machine.
Re: (Score:2)
You dont even have to spends lots of money. Lots of 3 year old Xeons out there from discarded servers. And slightly older RAM is cheap too. I have a 16 core Xeon with 128 G RAM which cost less than $1000. It is a very capable WS, completely outclasses my laptop and gives me 90 of the performance for 20% of the price.
Nowadays a modern Ryzen is going to kick that machines backside for similar amounts of money and less Wattage too.
Not part of the current revenue vision! (Score:2)
AIX (Score:4, Informative)
Re: (Score:2)
As I understand it, AIX is still being sold by IBM. IBM also takes backwards compatibility very seriously. So buying a modern AIX workstation would not be that much different from buying an older AIX workstation from IBM.
All of this is 100% true. I started in the mid-90s on an RS6000 and decades later find I'm back working with AIX. Hell, IBMi (the rebranding of AS/400) is also still alive and kicking. So yes, "very seriously."
Re: (Score:2)
There’s something to be said for the AS/400 hardware and software. Those boxes ran forever. Sit down at the latest OS release and the menus haven’t changed in decades.
Re: (Score:2)
I learned to program on AS400s
Yes, RPG (ii, iii, 400).
Then I started to explore programming on linux and encountered the concept of memory allocation and management.
"Malloc? What's that? Doesn't the operating system handle memory management?"
That's why OS400 (IBM i) costs a packet.
Re: (Score:2)
Macs (Score:5, Informative)
Re:Macs (Score:5, Informative)
Exactly. Macs (ever since NeXT/OSX) have been easily the best Unix workstations you could buy. Since then they have always been hands-down a better price/performance proposition than the wares of Sun/Oracle, DEC and SGI. Back in the early days Apple deliberately courted that cohort (even though one assumes it was comparatively small) by advertising that they had UNIX(TM) certification from the Open Group, something not really achievable by Linux or the open-source BSDs.
These days you can buy a 20-core 64-bit RISC-powered pizza box(ish) with absurdly good (by "Unix Workstation" standards) graphics performance and memory bandwidth for "reasonable" money. And it'll still run all of your professional Unix commercial CAD tools and all of your old-as-dirt Unix closed and open-source software. What's not to like?
Re: (Score:3, Insightful)
"What's not to like?"
The proprietary hardware, the lack of repairability/upgradability, and Apple trying to lock everyone into their walled garden by nudging developers to use their App Store, refusing to run unsigned code, etc. There are many things that are rubbish about Apple.
Re:Macs (Score:4, Insightful)
The proprietary hardware ...
LOL, and commercial UNIX ran on?
Re: (Score:2)
Re: (Score:3)
Generally the hardware could be opened and things like storage and RAM upgraded ... unlike most modern iMacs.
Yes, however that is not proprietary. Notice I did not challenge the not user serviceable/upgrade complaint, that one complaint is valid.
Re: (Score:2)
Oh man- a lot of that stuff was absolutely proprietary back in the day. And even if you could find something compatible- it certainly wasn't supported!
Re: (Score:2)
LOL, and commercial UNIX ran on?
Proprietary hardware that was beaten into the ground because Linux had all the necessary Unix features.
Re:Macs (Score:5, Interesting)
Also the UI.
I'm forced to use a Macbook at work and I hate it. I've asked for a Linux machine and I've been told "Mac IS Linux" which is absurd and insulting. Not only is it a heavily modified variant of [I *think] NetBSD (and only the kernel at that), the main gripe that I have with it is that I can't use MATE or some other DE that I'm more productive with. I'm so accustomed to the "legacy" of X11 that things like alt+click to drag and resize windows, and highlighting text plus middle-click to paste, are basic behaviours that are ingrained in my brain and not being able to do them on an OS X machine makes me feel like I have a disability. The window manager is beyond atrocious. Being able to do basic personalization like customize fonts has been stripped from me and OMG does this thing ever look horrendous on an external monitor.
Because Apple wants everyone to do things "The Apple Way."
I'm a Linux user precisely because I'm that "power user" who likes to do things MY way.
Someone who has used Mac products their entire lives will not see "what is not to like." And absolutely there is the aspect of familiarity that is at play here. The problem is when you take someone who is, at heart, "a hacker"; someone that likes to customize absolutely everything, strip things down to the bare essentials to remove everything superfluous and have things set up in a way that maximizes their productivity according to their own personal tastes and preferences ... the Apple ecosystem is downright hostile.
Even tools like iTerm2 feel like a poor substitute for what I take for granted on every single Linux machine that I run (which is all of my personal devices). As an x86 user my entire life, the keyboard bindings are infuriating. You can (and I have) done *some* customization to swap CTRL/CMD but it's not a magic solution since there are edge cases that don't play nicely.
Re:Macs are not part of the walled garden (Score:4, Insightful)
Re: (Score:2)
You couldn’t be more wrong.
Re:Macs are not part of the walled garden (Score:4, Interesting)
To wit:
bash-3.2$ spctl -v --assess erl
erl: rejected
source=no usable signature
bash-3.2$ erl -version
Erlang (SMP,ASYNC_THREADS) (BEAM) emulator version 13.1.1
But hey, we shouldn't let the truth get in the way of a good story, amirite GP?
Re: (Score:2)
There's WSL too. You can run GUI UNIX apps now.
Windows hardware compatibility and Outlook for your corporate network, along side a real UNIX system.
Re: (Score:2)
Re: (Score:2)
The problem with OSX PCs is that other than the CPUs there's not much in them that's interesting/unique anymore. There used to be high end workstations running weird data buses and SCSI cards/drives. Now you buy a Mac and it's basically a midrange PC. Even on the high end you're not getting fancy storage kit. There's nothing that differentiates a Mac from a high end HP other than the CPU. Apple have gone very boring. They even got rid of optical audio on their laptops.
Ok, since the topic is Unix... just want to say all of that applies to Linux servers too. There's nothing exciting to talk about, and Linux always sucked at doing anything interesting/unique anyway because it lacks vertical integration with fancy new hardware. Linux can't even do SCSI well, never mind when it's really plugged into a fibre channel fabric, the closest thing it has to a "weird data bus" that was kind of cool.
Intel PC architecture won the race to the bottom, that's the common thread in these s
macOS is really NeXTSTEP with a different GUI (Score:2)
Aren't modern Macs basically high-end Unix workstations running a somewhat gimped version of BSD ...
If by "gimped" you mean a Mach kernel.
... with a pretty skin over it? (On a proprietary CPU of late).
It not a skin, its functionality goes way beyond a Linux display manager. The Cocoa frameworks are pretty much a complete OS API. Its much like Android. While UNIX is available, Android and macOS developers can largely get things done without it.
It would be more accurate to say macOS is NextOS with a different GUI. Cocoa is largely built upon NeXTSTEP.
Re: (Score:2)
The Cocoa APIs still have NS prefixes last time I looked. On MacOS and iOS.
I remember being hell of a lot impressed with NextStep when I saw it back in 92 or so. It was some exotic stuff
Re: (Score:2)
Macs have never been Unix Workstations. When you read "workstation" think "full ATX x 2" sized cases with expansion slots for more than just a GPU. $6000 CPUs. Ugly as Windows 3.1 UIs. The most creative thing back in those workstation days was an SGI case, and maybe some specific high-style Cray supercomputer installs.
"Gimped version of BSD" - it's a pretty complete BSD Unix with a Mach microkernel, so unlike monolithic kernels of Free / Open BSD and Linux, it had an interesting concept for a kernel. Apple
Re: Macs (Score:2)
When you read "workstation" think "full ATX x 2" sized cases with expansion slots for more than just a GPU. $6000 CPUs. Ugly as Windows 3.1 UIs.
I remember working on SPARC stations (IPC and IPX) at university and they do not correspond to your description in any way. But they still got replaced by Linux PCs eventually, with larger cases.
This is evolution in action (Score:2)
Unix firms didnâ(TM)t invest, charged exorbitant prices for hardware and software. Linux did it just as well but for less money. Why do we need to keep this software running? Itâ(TM)s being replaced and soon will be like BeOS, NextStep, or OS/2
I'll tell you what killed them (Score:5, Informative)
The price.
If there's one thing I remember most about Unix workstations (or Unix-anything), it's the eye-watering cost of the hardware, and even more eye-watering cost of the software that ran on it.
The second thing I remember about Unix workstations is the complete lack of standardization on the UI side. In short, if you bought Solaris, you pretty much had to use graphical programs that ran on Solaris. And there wasn't many, because - surprise surprise - software vendors didn't want to support a million splintered different variants of Unix and whatever GUI they happened to run, and much prefered the unified Windows environment.
As a result, you pretty much chose the workstation and Unix OS that ran on it according to the software you wanted to use - or rather, needed to use, because nobody wanted to spend that much moola unless they had to.
Good riddance commercial Unix.
Re: (Score:2)
I liked that every vendor had a different window manager. The 4dwm under IRIX was particularly nice. They eventually all standardized on CDE which was butt ugly and too late to matter.
Linux or FreeBSD was close enough (Score:3)
Re:Linux or FreeBSD was close enough (Score:4, Informative)
Linux or FreeBSD was close enough / compatible enough.
It was more than that. I was at a small company in the early 2000s and we needed new hardware. Our choice was to buy a Sun server or workstation, or, for half the price, buy an x86 machine running Linux that was twice as fast. The only downside of the x86 was the 3.5GB per-process limit (or thereabout -- there were kernel patches to get the limit closer to 4GB).
still under multiple copyrights and licenses... (Score:4, Insightful)
Why don't they just dump the software out there?
Because they don't own it all, and they don't have distribution rights to give out other people's stuff.
Re: (Score:3)
Re: (Score:2)
Pretty much everything is out there if you look. SunOS, Solaris, IRIX, OpenVMS, Ultrix, HP-UX, OS/400, OS/390, zOS, AIX just to name some. What are you looking for?
Re: (Score:2)
The operating systems can be found, but most of the application software is unobtainable.
This is a shame, because it severely limits the possibilities of the retro-workstation-computer world.
Meanwhile, in the retro-home-computer world, we have pretty much all the software that ever existed, and a vibrant hobbyist community still able to get their hands on it and even develop new things.
Re: (Score:2)
Oh yeah I agree software for these platforms has evaporated. Literally no traces left. I’d love to find some unix cad or pcb layout software. The OpenVMS software source book is like 800 pages long.
CAD software still alive, remembering these O/Ss (Score:2)
The CAD software is still out there, and support for all of these old O/Ss os still in the source control tree of the various vendors.
I worked at Chronologic, Viewlogic, Synopsys, Verisity & Cadence and can assure you that the is all still there; carefully carried forward because no one working there now knows what any of it does, nor is brave enough to do more than #ifdef 0 around code that generated compiler errors with newer C compilers; but left in the code base.
The top level patch of each tool to R
So what? (Score:2)
I'm not going to waste time on the article - the signal-to-noise ratio on most contemporary Slashdot links is basically at the background noise level anymore. But, if someone braver (or more foolhardy) than I am reads it, perhaps they can answer the following question:
Does this article explain why the extinction of these old workstations should be considered a bad thing? Unless you're a fan of HP-UX, AIX, or the like I suppose. But personally, I think the way Linux (and to a lesser degree, the free BSDs) di
Re: (Score:2)
Well, I wasted my time on TFA. The reason is: "Because I'm totally nostalgic about that old hardware and software, and I wanted to play around with it a bit, so I could write an article about my experience." Oh, also "I'm totally mad at companies for not supporting a product that they abandoned a decade or two ago. How dare they not indulge my whim to stroll down memory lane?"
I mean, I get the nostalgia factor, but of course there's simply no practical reason to concern ourselves with the demise of these
I know why companies won't do it (Score:5, Interesting)
The article also wonders why companies like HPE don't just "dump some ISO files" onto an FTP server, along with patch depots and documentation.
I know why companies won't do it and I can give you an example from personal experience that sort of related. Companies won't do it because they enjoy being dicks.
I worked recently for more than a decade for a company in the bottom part of the Fortune 500. I'm not going to name them because I'm not supposed to say anything bad about them for a while longer, so I won't mention them at all. I worked on a SAAS program we sold that was pretty darn successful. In fact, it was so good that we had problems moving its customers to its replacement, which was not as good and much more expensive for them to pay for. This program I supported had a reporting function that we bought from a third party. They promised us it could do custom reports and after we started using it, it wasn't as well developed as we hoped, but we were kind of stuck with it, so we beat on the vendor and they finally got it working correctly. This product was about 3 years away from an end of life shutdown, so we needed to keep using the reporting program. We had been licensing it year by year. The vendor got a new CEO and he basically told us to go to hell, that he would never allow us to license it again. We were never given a reason for this. We told him that we would sign an agreement and promise to never under any circumstances ask for modifications or help with it after he expressed concern that his company would still have to do work on it if they licensed it to us. It took some real pleading by our upper management and probably an increased license fee, but in the end we were able to license it until our program was shutdown. We offered to buy the code outright to keep using it and they said "Nope. Not interested." In fact, they stopped selling it. All we wanted was them to take our money to let us keep using their program with the understanding that we couldn't get them fix anything if we ever found a problem. Free money for them, right? They still didn't want it. In the last year we used it, they warned us that at the end of the calendar year they would stop the license and nothing we said and no amount of money we offered them would allow us to keep using it. We shut down the program before the end of the year. They just seemed to really enjoy having the power to be dicks and say no more than they wanted to get free money for licensing a program that they no longer had to develop or devote resources to.
Itanium a sour footnote in computing history (Score:2)
Win NT 'vs' Unix (Score:3)
I still remember the study that said "A Windows NT Server costs 1/4 what a Unix Server costs." The only problem was that, to match the performance of that single Unix server, you needed at least 5 WinNT boxes, because NT was so poor at running multiple concurrent (disparate) services.
Re: (Score:2)
That and making the slightest change to NT required a reboot.
SoftWindows lived on for a while (Score:4, Interesting)
SoftWindows' core technology eventually led to a product called Win4Lin which was the fastest of any vm technologies for running windows 98 on Linux at the time. I got quite a bit of good use out of it. It basically ran the windows kernel as a process under Linux similar to user-model-Linux running a Linux kernel as a process.
It's over man, let it go.... (Score:5, Interesting)
The Dec Alpha, Ultra Sparc, SGU Onyx's with the Jurassic Park logo.
Those were the days of big UNIX running CAD and Simulation software like NASTRAN.
But then x86 got better, I mean a lot better and quickly. Intel moved from the Pentium 4 to the Xeon for performance. And then Linux arrived killing off SCO Unix and Slowaris (Solaris).
In no time we were building Intel Beowulf Linux clusters to run FEA simulations in a fraction of the time instead of using Dec Alpha's and Ultrasparcs.
Fast forward to today. We have the Amazon and Azure cloud that seem to defy the laws of hardware physics.
I remember fondly lugging the giant CRT monitors through the design center, making space in the raised floor datacenter for the big Sun E4500s, but sadly, just like the days of the mainframe, it's over, time to let it go.
Re:It's over man, let it go.... (Score:4, Insightful)
Realistically, when Mac OS X came out it became the user-friendly UNIX workstation alternative. The bash shell is there if you need it, and it has much better commercial software support than AIX or HP-UX ever had.
Re: (Score:2)
Mainframes aren’t going anywhere. If you use a credit card then you’re using a mainframe.
Better VT emulators (Score:2)
The world tried thin-computing (eg. X-windows) and no-one cared. At least we don't need smart hardware, we need software that can handle large copy/paste buffers, command-based I/O redirection (shell/pipe function), the DEC VT241 / DEC VT340 / Sixel command-sets.
Re: Better VT emulators (Score:2)
NsCDE brings that CDE look to modern setups (Score:2)
Copyright needs obligations (Score:3)
Ideally the requirement to provide a description of the hardware to run the software would force companies to release the software to the public domain before the hardware becomes extinct to avoid the fines.
High School Grant (Score:2)
We'd make simple fortran programs and some cool equation visualizer 3d plane thing. Money well spent.
Mostly dead... But not completely (Score:3)
https://raptorcs.com/content/b... [raptorcs.com]
I have 2 POWER9 based boxes with the motherboards, CPU's, and RAM purchased from Raptor. There were some growing pains at first, but with some people actually using Linux (and FreeBSD!) on POWER outside of IBM... the experience has become better.
For example, I found that some software didn't like the default 64k page size. I compiled a kernel with 4k page support, and while it fixed some issues... I found other software that had weirdness on POWER with 4k page sizes. Now everything works form e under 64k pages.
Too bad Polaris (the project to port OpenSolaris/Illumos to PowerPC) never went very far.
My only real issue is that we're still stuck with POWER9 has POWER10 has some non-open stuff in iit that can't be vetted... But there's hope POWER11 will work out.
End of an era (Score:4, Interesting)
When I started with my current employers we were a Sun shop, big servers in the back room, most people (including me) had an Ultra 5 on their desk. High tech for the time.
With changing markets and technology the company was no longer viable and as part of the implosion/downsizing we got rid of a lot of hardware. We literally couldn't give away several pallets of SparcStations. We did better with a pile of Ultra 5s. I took one home, put more memory in it (Sun had weird ideas about what constituted standard memory), installed Debian Linux and played with it. It used a lot of PC peripherals, which made it easier to live with than the SparcStations. I even put a PCI USB card in it, long before Sun started shipping servers with USB.
As time went on the headcount continued downwards, as did the Sun hardware count. Eventually we were down to one Sun box and my boss asked if I could port what we did with it to Linux. I did. The era was over...
...laura
Anybody want to buy an old SGI workstation? (Score:3)
Anybody want to buy an old SGI workstation? No, seriously, there's on in my basement...
It is a shame there's no incentive for code dumps with at least zero liability and there's no mechanism for using things abandoned by the copyright holder and, more generally, that our copyright is so onerously long that the prior point matters. But... I'm not at all sure old workstations are in my top 10 important cases of software where that matters.
I still lament the outliner called "More" from around 1991 MacOS that I have still never found the equal of.
Netcraft confirms it (Score:2)
What a PITA! (Score:2)
Then 32-bit Intel/AMD and 64-bit AMD/Intel processors running Linux killed off the pricey workstations.
Supporting software on all of those evolving combinations of processor architectures (SPARC, MIPS, Cray, Alpha, Itanium, PA-RISC, PowerPC) and OSes (Solaris, Ultrix, IRIX, AIX, HP-UX, Unicos) was a nightmare and pain-in-the-ass! Other than a little nostalgia, why in the world would one want to bring that mess back to life?
Most of them died because apps didn't need Unix (Score:2)
...the apps just needed more than 640K of RAM.
My workplace had about 40 Unix workstations, at peak, that were the CAD drafting machines. The draftsmen learned as much Unix as they had to, to move files around, do their drafting work. Maybe a head draftsman learned shell scripting to manage the backup copying.
As soon as WinNT could run the CAD program, there was NO reason to keep using Unix; it's sole advantage was to be able to run the CAD program at all.
Unix shines at running servers. But workstations