8 ms·

It’s absolutely mind-boggling to me that everybody is ignoring the relevant difference between Unix™ and Plan 9: Unix source code was given away (or cheaply lic

by teddyh 4y ago

It’s absolutely mind-boggling to me that everybody is ignoring the relevant difference between Unix™ and Plan 9: Unix source code was given away (or cheaply licensed) to hardware companies which all wrote their own operating systems on top (SunOS, Ultrix, HP-UX, etc. etc.). This made Unix the common factor of very many commercial workstation environments. Plan 9? It was sold directly as a commercial product, for no hardware platform in particular. Nobody wanted to buy it. People liked Unix because it was free – either really free, via BSD, or as a Unix derivative provided at no cost when people bought their workstations. A new revolutionary operating system had absolutely no reason for anybody to buy it: No commercial developers wanted to develop to a platform without users, and no users wanted a platform without software.

teddyh 4y ago
I’m reminded of a similar blunder made by IBM with their PS/2 series of computers. The initial IBM PC, PC XT and PC AT models were very popular, because they could be cloned by third-party manufacturers, which in turn created an enormous market for third-party software and hardware accessories and peripherals. But IBM wanted a bigger piece of that pie than just the few sales of genuine (and very expensive) IBM hardware, so they created the IBM PS/2 line: Heavily locked down, with its own new proprietary bus (MCA), connectors and peripherals; all heavily patented, so only IBM could produce them. It was a disaster. The only thing which endured was the PS/2 style connectors for mouse and keyboards, and those were made obsolete with USB. Nothing beside remains.
- lproven 4y ago
  > Nothing beside remains. This is wrong. The influence is far wider than you realise. The PS/2 standard for keyboards and mice is perhaps the only one that bears the name PS/2 but it is very far from the only standard that came out of the PS/2 range. VGA is a PS/2 standard. The VGA port and the still-ubiquitous SVGA plug and the SVGA modes is a PS/2 standard. My Core i7 nVME-only native-UEFI laptop still has both these ports on its vendor-supplied docking station. 3.5" HD 1.4MB floppy drives are a PS/2 standard. TTBOMK the 3.5" hard disk bay is also a PS/2 range innovation: while tower PS/2 machines came with 5.25" hard disks in the early models, the smaller desktop machines came with 3.5" hard disks. That was certainly the first time I ever saw a 3.5" hard disk and I vividly remember being boggled by how small it was. PS/2s predated the rise of CD drives and optical media. The desktop ones didn't have 5.25" bays. That drove the success of 3.5" hard disks. But the floppy drive was more influential. If you have an old PC with a floppy drive, then it's a PS/2 format floppy drive. More to the point, the PS/2 range is the direct reason that Microsoft Windows was a success. The widespread success of Windows is what created the conditions for Linux to be written. The reasons that the PC industry exists and thrives today is the IBM PS/2 range. I had to try to explain this here on HN recently, as in just this autumn, and while I can't find that comment, here's the blog post I made from it: https://liam-on-linux.dreamwidth.org/87108.html https://liam-on-linux.dreamwidth.org/87108.html And that itself caused considerable discussion here: https://news.ycombinator.com/item?id=33019019 https://news.ycombinator.com/item?id=33019019
  - teddyh 4y ago
    All right, yes, PS/2 spearheaded many more widely-used technologies than the keyboard and mouse port. Plan 9 also pioneered many things now widely used, including UTF-8. Now, does this in any way invalidate the analogy? If not, why go to such lengths in disproving a minor detail? The topic was Plan 9, and I made a comparison to IBM PS/2. Why write a minor essay about some detail about PS/2 which I got wrong, when that wasn’t the topic?
    - lproven 4y ago
      See the quote of yours that I started with. You are saying nothing else remains of the PS/2 than its mouse/keyboard port. That is not true. Lots of things from the PS/2 remain apart from that. That's it. That is the point. Not metaphors, not concepts, not influence. You said it only left 1 thing behind. It didn't. It shaped the whole PC industry from then on. All PCs after the PS/2 used PS/2 ports and PS/2 displays and PS/2 format storage media. OK, I concede, the stuff about the PS/2 driving Windows is a different issue: that's the inverse, the negative, the shape it left. But the ports and the media still stand.
      - teddyh 4y ago
        Yes, but how does this relate to Plan 9?
- jsiepkes 4y ago
  Certainly not the last time a company failed because they thought they could make big bucks by creating a vendor lock-in. Just look at Intel with Optane.
- simne 4y ago
  PS/2 series closeness, was not a problem. Real problem where other vendors, which offered new hardware (with for example, new CPUs), before IBM. First case was Compaq-386, which considered by wide public as brand, and appear on market, before 386 machine from IBM. As I know, Compaq-386, was not much commercial success, but it shown people, they could buy excellent hardware from other brands, not only from IBM. And after this happen, it made PS/2 series not so interest, as before.
  - indymike 4y ago
    The issue with PS/2 was the new MicroChannel bus that was incompatible with existing expansion cards (except for the 30/286 which had a couple of old-style AT bus connectors). This was a real problem for people that had devices that required an expansion cards (tape drives, network adapters, memory expansion, DTP display adapters, multi serial port cards) - you couldn't just move up to IBM. As a result the clone manufacturers came out with EISA - a card format that had all the benefits of MicroChannel, but was compatible (with a really over-engineered connector) with older AT bus cards. I believe ALR and Compaq were among the first with EISA. Incidentally, Compaq really put it to IBM with their lunch box style portable 386 and 486 models.
    - rstupek 4y ago
      I worked at Compaq just after the 486 model and those things were nice for their time
      - simne 4y ago
        Yeah, I seen Compaqs of near times. Have some quirks, but overall impression good.
    - simne 4y ago
      Unfortunately, practical EISA was too expensive, to compete with MCA. MCA was completely defeated only when happen PCI. Because of this, some time exists Vesa, and some other exotics.
simne 4y ago
"People liked Unix because it was free" Good point, but have You hear abot, https://en.wikipedia.org/wiki/Perfect_is_the_enemy_of_good https://en.wikipedia.org/wiki/Perfect_is_the_enemy_of_good That is the point, that better things must be hugely better, than ordinary, to win. For example, my friend worked in startup, based on Erlang. Once at public talk one guy asked "have you hear about Erjang, a virtual machine for Erlang, which runs on JVM, it is up to 5 times faster than Beam?". Answer was "5 times faster is not enough, to consider switch of platform, in my case, need 10 times faster".
nix23 4y ago
>Unix source code was given away (or cheaply licensed) Yeah that's not true, at the beginning it was ~10k per processor. But you can thank Berkeley for rewriting Unix and replace all the AT&T parts. And yes SunOS was from the BSD lineage. https://www.quora.com/How-much-did-Unix-cost-during-the-70s-or-80s https://www.quora.com/How-much-did-Unix-cost-during-the-70s-...
- monocasa 4y ago
  It was free up through and including V7. Because of the specifics of AT&T's monopoly on the phone system, they were banned from commercializing other technologies. It was only starting with System III and 32V that they were able to charge for it since their monopoly was taken away and they were broken up into the 'baby bells'.
  - nix23 4y ago
    >It was free up through and including V7 They could not sell a product but the licenses.
    - monocasa 4y ago
      They could not sell licenses. Hell, it wasn't even clear that software had copyright protections at all anyway, with computer programs being added to us code as explicitly copyrightable in the Computer Software Copyright Act of 1980. They charged their tape copying fee that was a couple hundred dollars or so, but that's the kind of thing even the GPL explicitly says is OK for distribution. They would later charge a license fee for even the older versions, but they didn't originally. > In the early 1970s AT&T distributed early versions of Unix at no cost to government and academic researchers, but these versions did not come with permission to redistribute or to distribute modified versions, and were thus not free software in the modern meaning of the phrase. After Unix became more widespread in the early 1980s, AT&T stopped the free distribution and charged for system patches. As it is quite difficult to switch to another architecture, most researchers paid for a commercial license. https://en.wikipedia.org/wiki/History_of_free_and_open-source_software https://en.wikipedia.org/wiki/History_of_free_and_open-sourc...
nine_k 4y ago
What works: make something good for building infrastructure and give it away. It will grow enormously. You will only take a small piece off this giant pie, but it will be large enough in absolute terms. Examples: Unix, IBM PC, Netscape, Android, Chrome, a huge lot of others. What also works: make something completely unique which nobody is able to competently reproduce or replace. Lock it down heavily, reap the benefits. Examples: Apple, maybe Nintendo. Hardly anyone else: the order is really tall. What does not work: make something fine but not uniquely great, lock it down. Failures are by the thousand.
- empthought 4y ago
  Isn't "make something fine but not uniquely great" just what Oracle does, as a rule?
  - nine_k 4y ago
    They are sufficiently uniquely great at getting government and big business contracts though. Their leverage is not necessarily in their technology (which can be of all kinds, form excellent to barely usable), but in the skill of selling it to particular markets, and access to that markets earned by decades of effort.
throwawaylinux 4y ago
It's mind boggling that there are still some people who don't think Plan 9 was a text book case of second system syndrome.
- coldtea 4y ago
  It's mind boggling that someone will consider it a case of second system syndrome. Second system syndrome is not about "trying to make improvements to a first version of something and failing in the market". It's about trying to make improvements and ending up with a convoluted mess. The Wiki definition is literally: "The second-system effect or second-system syndrome is the tendency of small, elegant, and successful systems to be succeeded by over-engineered, bloated systems, due to inflated expectations and overconfidence.". That's nothing like Plan-9, whose developers insisted on a simple elegant small design, with the original ethos of Unix, kept it KISS at all times, and if anything made it even simpler and more coherent than the original (which is why Linux copied all the major Plan-9 features over the years). Failing in the market is orthogonal to this being (or, in this case, not being) a "second system syndrome". The reason for that was that it was a proprietary platform, the existing unices were "good enough" for it to gain traction, and it came at a time that companies turned first Windows NT and soon after Linux. Besides ALL traditional UNIX vendors failed and most folded at the same period.
  - throwawaylinux 4y ago
    It is second system syndrome, and it absolutely is about the market. No wonder so many academic types fall to second-system syndrome, they don't think about their users only themselves. Second system syndrome is absolutely related to market success when you're talking about a commercial system like this -- how else are you going to gauge success? Transparent distributed single system, for example. The strange and unexplainable obsession by so many computer systems developers. The reality is almost nobody wants it, yet Plan9 was almost entirely designed around this concept. A complete boondoggle. Even the nerdy people who get a kick out of this kind of thing will set up a single system distributed cluster and run a few commands and say "hey that's neat" and then go back to using Windows or Linux, complete with tools for remote administration of systems, and applications that can do clustering and distributed systems in the cases where it's necessary or helpful.

by teddyh 4y ago

teddyh 4y ago

I’m reminded of a similar blunder made by IBM with their PS/2 series of computers. The initial IBM PC, PC XT and PC AT models were very popular, because they could be cloned by third-party manufacturers, which in turn created an enormous market for third-party software and hardware accessories and peripherals. But IBM wanted a bigger piece of that pie than just the few sales of genuine (and very expensive) IBM hardware, so they created the IBM PS/2 line: Heavily locked down, with its own new proprietary bus (MCA), connectors and peripherals; all heavily patented, so only IBM could produce them. It was a disaster. The only thing which endured was the PS/2 style connectors for mouse and keyboards, and those were made obsolete with USB. Nothing beside remains.

lproven 4y ago
> Nothing beside remains. This is wrong. The influence is far wider than you realise. The PS/2 standard for keyboards and mice is perhaps the only one that bears the name PS/2 but it is very far from the only standard that came out of the PS/2 range. VGA is a PS/2 standard. The VGA port and the still-ubiquitous SVGA plug and the SVGA modes is a PS/2 standard. My Core i7 nVME-only native-UEFI laptop still has both these ports on its vendor-supplied docking station. 3.5" HD 1.4MB floppy drives are a PS/2 standard. TTBOMK the 3.5" hard disk bay is also a PS/2 range innovation: while tower PS/2 machines came with 5.25" hard disks in the early models, the smaller desktop machines came with 3.5" hard disks. That was certainly the first time I ever saw a 3.5" hard disk and I vividly remember being boggled by how small it was. PS/2s predated the rise of CD drives and optical media. The desktop ones didn't have 5.25" bays. That drove the success of 3.5" hard disks. But the floppy drive was more influential. If you have an old PC with a floppy drive, then it's a PS/2 format floppy drive. More to the point, the PS/2 range is the direct reason that Microsoft Windows was a success. The widespread success of Windows is what created the conditions for Linux to be written. The reasons that the PC industry exists and thrives today is the IBM PS/2 range. I had to try to explain this here on HN recently, as in just this autumn, and while I can't find that comment, here's the blog post I made from it: https://liam-on-linux.dreamwidth.org/87108.html https://liam-on-linux.dreamwidth.org/87108.html And that itself caused considerable discussion here: https://news.ycombinator.com/item?id=33019019 https://news.ycombinator.com/item?id=33019019
- teddyh 4y ago
  All right, yes, PS/2 spearheaded many more widely-used technologies than the keyboard and mouse port. Plan 9 also pioneered many things now widely used, including UTF-8. Now, does this in any way invalidate the analogy? If not, why go to such lengths in disproving a minor detail? The topic was Plan 9, and I made a comparison to IBM PS/2. Why write a minor essay about some detail about PS/2 which I got wrong, when that wasn’t the topic?
  - lproven 4y ago
    See the quote of yours that I started with. You are saying nothing else remains of the PS/2 than its mouse/keyboard port. That is not true. Lots of things from the PS/2 remain apart from that. That's it. That is the point. Not metaphors, not concepts, not influence. You said it only left 1 thing behind. It didn't. It shaped the whole PC industry from then on. All PCs after the PS/2 used PS/2 ports and PS/2 displays and PS/2 format storage media. OK, I concede, the stuff about the PS/2 driving Windows is a different issue: that's the inverse, the negative, the shape it left. But the ports and the media still stand.
    - teddyh 4y ago
      Yes, but how does this relate to Plan 9?
jsiepkes 4y ago
Certainly not the last time a company failed because they thought they could make big bucks by creating a vendor lock-in. Just look at Intel with Optane.
simne 4y ago
PS/2 series closeness, was not a problem. Real problem where other vendors, which offered new hardware (with for example, new CPUs), before IBM. First case was Compaq-386, which considered by wide public as brand, and appear on market, before 386 machine from IBM. As I know, Compaq-386, was not much commercial success, but it shown people, they could buy excellent hardware from other brands, not only from IBM. And after this happen, it made PS/2 series not so interest, as before.
- indymike 4y ago
  The issue with PS/2 was the new MicroChannel bus that was incompatible with existing expansion cards (except for the 30/286 which had a couple of old-style AT bus connectors). This was a real problem for people that had devices that required an expansion cards (tape drives, network adapters, memory expansion, DTP display adapters, multi serial port cards) - you couldn't just move up to IBM. As a result the clone manufacturers came out with EISA - a card format that had all the benefits of MicroChannel, but was compatible (with a really over-engineered connector) with older AT bus cards. I believe ALR and Compaq were among the first with EISA. Incidentally, Compaq really put it to IBM with their lunch box style portable 386 and 486 models.
  - rstupek 4y ago
    I worked at Compaq just after the 486 model and those things were nice for their time
    - simne 4y ago
      Yeah, I seen Compaqs of near times. Have some quirks, but overall impression good.
  - simne 4y ago
    Unfortunately, practical EISA was too expensive, to compete with MCA. MCA was completely defeated only when happen PCI. Because of this, some time exists Vesa, and some other exotics.

simne 4y ago

"People liked Unix because it was free" Good point, but have You hear abot, https://en.wikipedia.org/wiki/Perfect_is_the_enemy_of_good https://en.wikipedia.org/wiki/Perfect_is_the_enemy_of_good That is the point, that better things must be hugely better, than ordinary, to win. For example, my friend worked in startup, based on Erlang. Once at public talk one guy asked "have you hear about Erjang, a virtual machine for Erlang, which runs on JVM, it is up to 5 times faster than Beam?". Answer was "5 times faster is not enough, to consider switch of platform, in my case, need 10 times faster".

nix23 4y ago

>Unix source code was given away (or cheaply licensed) Yeah that's not true, at the beginning it was ~10k per processor. But you can thank Berkeley for rewriting Unix and replace all the AT&T parts. And yes SunOS was from the BSD lineage. https://www.quora.com/How-much-did-Unix-cost-during-the-70s-or-80s https://www.quora.com/How-much-did-Unix-cost-during-the-70s-...

monocasa 4y ago
It was free up through and including V7. Because of the specifics of AT&T's monopoly on the phone system, they were banned from commercializing other technologies. It was only starting with System III and 32V that they were able to charge for it since their monopoly was taken away and they were broken up into the 'baby bells'.
- nix23 4y ago
  >It was free up through and including V7 They could not sell a product but the licenses.
  - monocasa 4y ago
    They could not sell licenses. Hell, it wasn't even clear that software had copyright protections at all anyway, with computer programs being added to us code as explicitly copyrightable in the Computer Software Copyright Act of 1980. They charged their tape copying fee that was a couple hundred dollars or so, but that's the kind of thing even the GPL explicitly says is OK for distribution. They would later charge a license fee for even the older versions, but they didn't originally. > In the early 1970s AT&T distributed early versions of Unix at no cost to government and academic researchers, but these versions did not come with permission to redistribute or to distribute modified versions, and were thus not free software in the modern meaning of the phrase. After Unix became more widespread in the early 1980s, AT&T stopped the free distribution and charged for system patches. As it is quite difficult to switch to another architecture, most researchers paid for a commercial license. https://en.wikipedia.org/wiki/History_of_free_and_open-source_software https://en.wikipedia.org/wiki/History_of_free_and_open-sourc...

nine_k 4y ago

What works: make something good for building infrastructure and give it away. It will grow enormously. You will only take a small piece off this giant pie, but it will be large enough in absolute terms. Examples: Unix, IBM PC, Netscape, Android, Chrome, a huge lot of others. What also works: make something completely unique which nobody is able to competently reproduce or replace. Lock it down heavily, reap the benefits. Examples: Apple, maybe Nintendo. Hardly anyone else: the order is really tall. What does not work: make something fine but not uniquely great, lock it down. Failures are by the thousand.

empthought 4y ago
Isn't "make something fine but not uniquely great" just what Oracle does, as a rule?
- nine_k 4y ago
  They are sufficiently uniquely great at getting government and big business contracts though. Their leverage is not necessarily in their technology (which can be of all kinds, form excellent to barely usable), but in the skill of selling it to particular markets, and access to that markets earned by decades of effort.

throwawaylinux 4y ago

It's mind boggling that there are still some people who don't think Plan 9 was a text book case of second system syndrome.

coldtea 4y ago
It's mind boggling that someone will consider it a case of second system syndrome. Second system syndrome is not about "trying to make improvements to a first version of something and failing in the market". It's about trying to make improvements and ending up with a convoluted mess. The Wiki definition is literally: "The second-system effect or second-system syndrome is the tendency of small, elegant, and successful systems to be succeeded by over-engineered, bloated systems, due to inflated expectations and overconfidence.". That's nothing like Plan-9, whose developers insisted on a simple elegant small design, with the original ethos of Unix, kept it KISS at all times, and if anything made it even simpler and more coherent than the original (which is why Linux copied all the major Plan-9 features over the years). Failing in the market is orthogonal to this being (or, in this case, not being) a "second system syndrome". The reason for that was that it was a proprietary platform, the existing unices were "good enough" for it to gain traction, and it came at a time that companies turned first Windows NT and soon after Linux. Besides ALL traditional UNIX vendors failed and most folded at the same period.
- throwawaylinux 4y ago
  It is second system syndrome, and it absolutely is about the market. No wonder so many academic types fall to second-system syndrome, they don't think about their users only themselves. Second system syndrome is absolutely related to market success when you're talking about a commercial system like this -- how else are you going to gauge success? Transparent distributed single system, for example. The strange and unexplainable obsession by so many computer systems developers. The reality is almost nobody wants it, yet Plan9 was almost entirely designed around this concept. A complete boondoggle. Even the nerdy people who get a kick out of this kind of thing will set up a single system distributed cluster and run a few commands and say "hey that's neat" and then go back to using Windows or Linux, complete with tools for remote administration of systems, and applications that can do clustering and distributed systems in the cases where it's necessary or helpful.