6 ms·

The post is very good but has a huge glaring hole in it. They say Sun died because of AMD64 but that’s incorrect. It was Linux. I know this because I was in en

The post is very good but has a huge glaring hole in it. They say Sun died because of AMD64 but that’s incorrect. It was Linux. I know this because I was in enterprise software during this era and the demise of Sun due to Linux was stunning. Intel and Windows killed SGI, which was a huge coup. However Sun was killed because during that time, Linux went from a toy OS into a fully supported enterprise OS that was “good enough”. I remember how at one point in time, Oracle on Solaris was just about as rock solid as you could get. But slowly and surely more enterprise software moved over to Linux, especially Oracle. And within 10 years Sun went from one of the “4 horsemen of the Internet” to being sold to Oracle. It was sad because I loved Solaris as a technology, I even had a Sun workstation at home. That said, how quickly Linux killed Solaris/Sun is a great reminder of how quickly things change in tech.

BirAdam 4y ago
From the article: > SPARC didn’t keep its edge. The x86 platform became extremely competitive with the arrival of AMD64. The 64 bit x86 machines were cheaper than SPARC machines, and when coupled with Linux they could run all of the UNIX software any web company needed. This fact was proven by the likes of Google and Facebook. The core of Sun’s business was destroyed.
- remote_phone 4y ago
  64 bit came after Linux. And amd64 didn’t become popular until Intel moved to it unwillingly. Enterprise didn’t mess Around with amd at that point.
  - BirAdam 4y ago
    Yes. Linux came before AMD64, but that’s irrelevant. Corporate UNIX was more popular than Linux in the 1990s. AMD64 launched with the Opteron in 2003. In 2004, it was making AMD hundreds of millions per quarter and it was beating both Sun and Intel in server sales. Intel and HP were trying to push Itanium at this point, and Itanium was performing quite poorly. By 2004, AMD Opteron was also doing better than IBM Power. AMD was was crushing it. EDIT: To give an idea of AMD at this point, a few customers of AMD buying Opteron were The Weather Channel, Siemens, Boeing, Lockheed, Ferrari, Racksaver, Volkswagon, Rackspace, University of Kentucky (these are the ones to have publicly stated they adopted Opteron whose news articles to the effect I could find). Revenue of Q3 2003 was $1.2 billion (about $2 billion today), and that was the first quarter to include AMD64.
NoZebra120vClip 4y ago
It was not merely Linux, but it was also PC commodity hardware that rapidly caught up to workstation-quality. Toward the end, you had Suns with PCI buses and stuff. The UEFI frameworks, and high-quality peripherals that could make it to the mass market, were poised to beat the pants off all the bespoke stuff coming out of workstation manufacturers like Sun, HP, Digital, SGI, and IBM. Sun was very much a hardware VAR, and they made great software because people would purchase premium hardware to run that great software on. The dual prongs of Linux plus commodity PC hardware damaged both of those selling points simultaneously, as it were, and no Unix workstation vendor could survive. The more adaptable ones embraced Linux and they embraced that commodity hardware (IBM, RHEL...) but even with their Java initiative, Sun could not survive the onslaught.
samstave 4y ago
Sun has the best logo ever... But Intel didnt kill SGI, or MIPS for that matter. SGI killed themselves for a too-expensive-walled garden. I dont know who the industrial design team was at SGI for all their cases, but their cases are still legendary beautiful artifacts of computing (like CRAY) It sucks that FB now operates out of SUNs HQ... Fun fact ; when we designed the Lucas' Presidio Complex and moved ILM and other Lucas entities in there, they had a boatload of old SGI full rack machines (SGI machines that were the same size as a 42u cabinet) and several were turned into keggerators. Sadly I could have taken some of these amazing SGIs, but I was in an apartment in SF at the time and didnt have space for them.... it would have been worth it for me to rent a storage space and grab those machines... sadly many ~$500,000 SGI cabinets just went to the dump. (2004)
- hulitu 4y ago
  > But Intel didnt kill SGI, or MIPS for that matter. Itanium ?
  - samstave 4y ago
    Pricing killed them, not Intel in specific....
  - jabl 4y ago
    SGI managed to hang on for quite many years after their 3D graphics workstation market tanked by pivoting from Irix/MIPS to Linux/Itanium (later x86-64)(saving a lot of SW and HW development cost) and selling those big shared memory supercomputers. Sure, those were expensive boutique products, but they sat in their own niche that commodity HW couldn't easily touch. I think what ultimately doomed this was that with increasing core count CPU's the market that really needed much bigger shared memory systems became smaller and smaller. They were finally bought out by HPE, and while HPE still sells their Altix brand I hear the development of the shared memory supercomputer line has ended and now the Altix memory fabric has been utilized in the enterprise Superdome line.
- DonHopkins 4y ago
  Excerpt from my post about FORTH and Homer and Associates from the discussion of Forth vs Lisp: https://news.ycombinator.com/item?id=29261868 https://news.ycombinator.com/item?id=29261868 Coco Conn and Paul Rother wrote this up about what they did with FORTH at HOMER & Assoc, who made some really classic music videos including Atomic Dog, and hired Charles Moore himself! Here's what Coco Conn posted about it, and some discussion and links about it that I'm including with her permission: [...] Flying Logos for 1989 Siggraph Electronic Theater: https://www.youtube.com/watch?v=9hIOfEiy4lc https://www.youtube.com/watch?v=9hIOfEiy4lc >First shown at the 1989 Siggraph Electronic Theater to a rave response, this 3 minute humourous film went on to win several top computer graphic awards that same year including Niccograph of Japan. >Coco: This was a show favorite at the SIGGRAPH film show that year. The year before the conference committee decided that showing demos wasn't the way to go anymore. Peter wrote Flying Logos as a way to sneak our demo reel into the show by turning it into a story. It worked and we made it into the film show. >Don: I truly believe that in some other alternate dimension, there is a Flying Logo Heaven where the souls of dead flying logos go, where they dramatically promenade and swoop and spin around each other in pomp and pageantry to bombastic theme music. It would make a great screen saver, at least! Somewhere the Sun Logo and the SGI Logo are still dancing together. [...] From the discussion of The Dawn and Dusk of Sun Microsystems: https://news.ycombinator.com/item?id=34134114 https://news.ycombinator.com/item?id=34134114 That's right: Vaughan Pratt designed the original square Sun logo. https://en.wikipedia.org/wiki/Vaughan_Pratt https://en.wikipedia.org/wiki/Vaughan_Pratt And Sun Science Officer and Nixon Enemy John Gage is the genius who rotated it 45 degrees. https://en.wikipedia.org/wiki/John_Gage https://en.wikipedia.org/wiki/John_Gage Nixon Enemies List entry for John B. Gage: https://www.enemieslist.info/enemy.php?ID=463 https://www.enemieslist.info/enemy.php?ID=463 Another one of the greatest logos of all time is the SGI logo, designed by none other than Scott Kim! https://en.wikipedia.org/wiki/Scott_Kim https://en.wikipedia.org/wiki/Scott_Kim http://xahlee.info/UnixResource_dir/sgi_logo.html http://xahlee.info/UnixResource_dir/sgi_logo.html https://www.amazon.com/Inversions-Scott-Kim/dp/1559532807 https://www.amazon.com/Inversions-Scott-Kim/dp/1559532807 https://scottkim.com/ https://scottkim.com/
samstave 4y ago
Heh. I was Head of IT for a company that manufactured the physical media for all SUN OS products. (meaning if you bought any SUN software, and it came on a CD and in a box- we manufactured all of that, and shipped it as if we were SUN) We would receive updates to the releases via FTP. We had several FTP servers which SUN would PUT their files on and we had a nightly batch job grab the files and transfer to the machines that burned the CDs... I implemented Linux on those machines with the help of hiriing a few Linux consultants I knew... Dave Sifry, Chris DiBona and a few others. I called Dave in and told him "If I were you, I'd start a Linux support company/consulting company" A few weeks later he came to me and stated that they had founded a company "LINUXCARE" At one point he was worth $100 million off that... Chris was/is huge at google... but yeah, another annecdote was that this is also when XML became a thing, and we were one of the first companies to adopt XML out of SUN because they wanted to PUT XML docs on our FTP server for their releases... and LINUXCARE was tasked with creating our CRON job scripts for watching the directories for updates and batch jobs....
pjmlp 4y ago
While I use Linux distributions since 1995, starting with Slackware 2.0, I only used it in production after 2006. Until then it was Xenix, DG/UX, Solaris, HP-UX, Aix. Ironically, when we at Nokia Networks were considering HP-UX to Linux migration for some of our software, a parallel effort for Solaris was in motion as plan B.
sn1de 4y ago
The x86 hardware and the Linux OS both had to make the jump to 64 bit. Sun was still able to make hay in the period where Linux was well established because 32 bit memory restrictions kept if from being a viable option for a lot of higher end computing needs. It was inevitable that Linux would eventually make the migration to 64 bit once the hardware was there, but Sun got there first on both the hardware and software fronts. Why Sun didn't react more radically when 64 bit Lintel was able to go head to head with Sun's precious 'enterprise' solutions is unknown to me. Maybe they were just in mass denial? Maybe they thought the technical hight priests within their organization would manifest new technical advantages? Maybe they just knew that if they dropped their pricing to address the new price/performance reality that their company was no longer economically viable? There was a lot of denial about Linux at the time, and there were plenty of legitimate reasons to question whether or not Linux could really make the leap to compete head to head with the leading commercial Unix offerings. Meanwhile, major efforts were being made by the likes of IBM and, ironically, Oracle, to contribute to efforts to make Linux a 'no compromise' commercial offering. Remember that companies like IBM had watched their own Unix offerings suffer from Solaris on the high end and Linux on the low end. The smart ones realized that Linux could be the answer to the hard sell they were running into with their own Unix and put engineering efforts towards making that happen. I'm not saying Linux wouldn't have gotten there without them, but there was a strategic shift that happened to accelerate the technical ascension of Linux, and that also caught Sun off guard. I suspect they had looked at the historical trajectory of Linux and over estimated how long it would take Linux to 'catch up' to Solaris. Linux went 'hockey stick' on Sun.
- remote_phone 4y ago
  No. AMD64 wasn’t as popular especially in enterprise environments until just before Intel started to support it. Intel had the failed Itanium but that wasn’t the reason for Linux’s success. Linux on 32bit was still dominant at that point and the cause. I was supporting a cross-platform enterprise environment and was deep into this at the time. I saw it from the front line’s and I even implemented my company’s port to Linux. Everyone was skeptical about Itanium. Enterprises companies in most part didn’t have the luxury to experiment.
lizknope 4y ago
Linux was most of it but it needed hardware to run on. I wrote in another comment in this thread that our semiconductor design software was all on Solaris. We started shifting to Linux around 2002. We still needed some Suns for the large jobs that required over 4GB RAM. When AMD64 came out that was the end of Sun for us. The chip design software vendors quickly ported to it.
TristanBall 4y ago
I was a sysadmin in a windows desktops + sun/windows server env for about 10 years from 1999 I loved our sun boxes, hated that we could only ever afford their "midrange" models, and then only from grey market resellers, or outright second hand. Everything being said here is pretty much true.. but the additional factors I'd call out were vmware and to a lesser extent "componentised" infrastructure systems.. I'm thinking blade servers and sans. If I bought a sun box, it was only a sun box. If I bought an intel/amd, then it could be linux or windows or both. Yes we could, and did, hook the sun boxes to SAN's, but I don't remember sun ever having a particularly strong blade offering.. or maybe it was just priced and like sun hw, either way.. Dell Blades let me do incremental upgrades at price points within individual projects sign off budgets, and dell delivery times + San boot + esx meant I could usually have new systems up and available in a couple of weeks. That sounds bad in the cloud era, but before that I'd probably still be trying to get quotes or negotiate discounts, so it was fantastic. Linux and the other bits above let me say "yes" to user requests more often, with lower friction and shorter deliver times. I can't tell you just how nice that was. Then I started work at an aix/linux/windows shop and the whole cycle repeated...

BirAdam 4y ago

From the article: > SPARC didn’t keep its edge. The x86 platform became extremely competitive with the arrival of AMD64. The 64 bit x86 machines were cheaper than SPARC machines, and when coupled with Linux they could run all of the UNIX software any web company needed. This fact was proven by the likes of Google and Facebook. The core of Sun’s business was destroyed.

remote_phone 4y ago
64 bit came after Linux. And amd64 didn’t become popular until Intel moved to it unwillingly. Enterprise didn’t mess Around with amd at that point.
- BirAdam 4y ago
  Yes. Linux came before AMD64, but that’s irrelevant. Corporate UNIX was more popular than Linux in the 1990s. AMD64 launched with the Opteron in 2003. In 2004, it was making AMD hundreds of millions per quarter and it was beating both Sun and Intel in server sales. Intel and HP were trying to push Itanium at this point, and Itanium was performing quite poorly. By 2004, AMD Opteron was also doing better than IBM Power. AMD was was crushing it. EDIT: To give an idea of AMD at this point, a few customers of AMD buying Opteron were The Weather Channel, Siemens, Boeing, Lockheed, Ferrari, Racksaver, Volkswagon, Rackspace, University of Kentucky (these are the ones to have publicly stated they adopted Opteron whose news articles to the effect I could find). Revenue of Q3 2003 was $1.2 billion (about $2 billion today), and that was the first quarter to include AMD64.

NoZebra120vClip 4y ago

It was not merely Linux, but it was also PC commodity hardware that rapidly caught up to workstation-quality. Toward the end, you had Suns with PCI buses and stuff. The UEFI frameworks, and high-quality peripherals that could make it to the mass market, were poised to beat the pants off all the bespoke stuff coming out of workstation manufacturers like Sun, HP, Digital, SGI, and IBM. Sun was very much a hardware VAR, and they made great software because people would purchase premium hardware to run that great software on. The dual prongs of Linux plus commodity PC hardware damaged both of those selling points simultaneously, as it were, and no Unix workstation vendor could survive. The more adaptable ones embraced Linux and they embraced that commodity hardware (IBM, RHEL...) but even with their Java initiative, Sun could not survive the onslaught.

samstave 4y ago

Sun has the best logo ever... But Intel didnt kill SGI, or MIPS for that matter. SGI killed themselves for a too-expensive-walled garden. I dont know who the industrial design team was at SGI for all their cases, but their cases are still legendary beautiful artifacts of computing (like CRAY) It sucks that FB now operates out of SUNs HQ... Fun fact ; when we designed the Lucas' Presidio Complex and moved ILM and other Lucas entities in there, they had a boatload of old SGI full rack machines (SGI machines that were the same size as a 42u cabinet) and several were turned into keggerators. Sadly I could have taken some of these amazing SGIs, but I was in an apartment in SF at the time and didnt have space for them.... it would have been worth it for me to rent a storage space and grab those machines... sadly many ~$500,000 SGI cabinets just went to the dump. (2004)

hulitu 4y ago
> But Intel didnt kill SGI, or MIPS for that matter. Itanium ?
- samstave 4y ago
  Pricing killed them, not Intel in specific....
- jabl 4y ago
  SGI managed to hang on for quite many years after their 3D graphics workstation market tanked by pivoting from Irix/MIPS to Linux/Itanium (later x86-64)(saving a lot of SW and HW development cost) and selling those big shared memory supercomputers. Sure, those were expensive boutique products, but they sat in their own niche that commodity HW couldn't easily touch. I think what ultimately doomed this was that with increasing core count CPU's the market that really needed much bigger shared memory systems became smaller and smaller. They were finally bought out by HPE, and while HPE still sells their Altix brand I hear the development of the shared memory supercomputer line has ended and now the Altix memory fabric has been utilized in the enterprise Superdome line.
DonHopkins 4y ago
Excerpt from my post about FORTH and Homer and Associates from the discussion of Forth vs Lisp: https://news.ycombinator.com/item?id=29261868 https://news.ycombinator.com/item?id=29261868 Coco Conn and Paul Rother wrote this up about what they did with FORTH at HOMER & Assoc, who made some really classic music videos including Atomic Dog, and hired Charles Moore himself! Here's what Coco Conn posted about it, and some discussion and links about it that I'm including with her permission: [...] Flying Logos for 1989 Siggraph Electronic Theater: https://www.youtube.com/watch?v=9hIOfEiy4lc https://www.youtube.com/watch?v=9hIOfEiy4lc >First shown at the 1989 Siggraph Electronic Theater to a rave response, this 3 minute humourous film went on to win several top computer graphic awards that same year including Niccograph of Japan. >Coco: This was a show favorite at the SIGGRAPH film show that year. The year before the conference committee decided that showing demos wasn't the way to go anymore. Peter wrote Flying Logos as a way to sneak our demo reel into the show by turning it into a story. It worked and we made it into the film show. >Don: I truly believe that in some other alternate dimension, there is a Flying Logo Heaven where the souls of dead flying logos go, where they dramatically promenade and swoop and spin around each other in pomp and pageantry to bombastic theme music. It would make a great screen saver, at least! Somewhere the Sun Logo and the SGI Logo are still dancing together. [...] From the discussion of The Dawn and Dusk of Sun Microsystems: https://news.ycombinator.com/item?id=34134114 https://news.ycombinator.com/item?id=34134114 That's right: Vaughan Pratt designed the original square Sun logo. https://en.wikipedia.org/wiki/Vaughan_Pratt https://en.wikipedia.org/wiki/Vaughan_Pratt And Sun Science Officer and Nixon Enemy John Gage is the genius who rotated it 45 degrees. https://en.wikipedia.org/wiki/John_Gage https://en.wikipedia.org/wiki/John_Gage Nixon Enemies List entry for John B. Gage: https://www.enemieslist.info/enemy.php?ID=463 https://www.enemieslist.info/enemy.php?ID=463 Another one of the greatest logos of all time is the SGI logo, designed by none other than Scott Kim! https://en.wikipedia.org/wiki/Scott_Kim https://en.wikipedia.org/wiki/Scott_Kim http://xahlee.info/UnixResource_dir/sgi_logo.html http://xahlee.info/UnixResource_dir/sgi_logo.html https://www.amazon.com/Inversions-Scott-Kim/dp/1559532807 https://www.amazon.com/Inversions-Scott-Kim/dp/1559532807 https://scottkim.com/ https://scottkim.com/

samstave 4y ago

Heh. I was Head of IT for a company that manufactured the physical media for all SUN OS products. (meaning if you bought any SUN software, and it came on a CD and in a box- we manufactured all of that, and shipped it as if we were SUN) We would receive updates to the releases via FTP. We had several FTP servers which SUN would PUT their files on and we had a nightly batch job grab the files and transfer to the machines that burned the CDs... I implemented Linux on those machines with the help of hiriing a few Linux consultants I knew... Dave Sifry, Chris DiBona and a few others. I called Dave in and told him "If I were you, I'd start a Linux support company/consulting company" A few weeks later he came to me and stated that they had founded a company "LINUXCARE" At one point he was worth $100 million off that... Chris was/is huge at google... but yeah, another annecdote was that this is also when XML became a thing, and we were one of the first companies to adopt XML out of SUN because they wanted to PUT XML docs on our FTP server for their releases... and LINUXCARE was tasked with creating our CRON job scripts for watching the directories for updates and batch jobs....

pjmlp 4y ago

While I use Linux distributions since 1995, starting with Slackware 2.0, I only used it in production after 2006. Until then it was Xenix, DG/UX, Solaris, HP-UX, Aix. Ironically, when we at Nokia Networks were considering HP-UX to Linux migration for some of our software, a parallel effort for Solaris was in motion as plan B.

sn1de 4y ago

The x86 hardware and the Linux OS both had to make the jump to 64 bit. Sun was still able to make hay in the period where Linux was well established because 32 bit memory restrictions kept if from being a viable option for a lot of higher end computing needs. It was inevitable that Linux would eventually make the migration to 64 bit once the hardware was there, but Sun got there first on both the hardware and software fronts. Why Sun didn't react more radically when 64 bit Lintel was able to go head to head with Sun's precious 'enterprise' solutions is unknown to me. Maybe they were just in mass denial? Maybe they thought the technical hight priests within their organization would manifest new technical advantages? Maybe they just knew that if they dropped their pricing to address the new price/performance reality that their company was no longer economically viable? There was a lot of denial about Linux at the time, and there were plenty of legitimate reasons to question whether or not Linux could really make the leap to compete head to head with the leading commercial Unix offerings. Meanwhile, major efforts were being made by the likes of IBM and, ironically, Oracle, to contribute to efforts to make Linux a 'no compromise' commercial offering. Remember that companies like IBM had watched their own Unix offerings suffer from Solaris on the high end and Linux on the low end. The smart ones realized that Linux could be the answer to the hard sell they were running into with their own Unix and put engineering efforts towards making that happen. I'm not saying Linux wouldn't have gotten there without them, but there was a strategic shift that happened to accelerate the technical ascension of Linux, and that also caught Sun off guard. I suspect they had looked at the historical trajectory of Linux and over estimated how long it would take Linux to 'catch up' to Solaris. Linux went 'hockey stick' on Sun.

remote_phone 4y ago
No. AMD64 wasn’t as popular especially in enterprise environments until just before Intel started to support it. Intel had the failed Itanium but that wasn’t the reason for Linux’s success. Linux on 32bit was still dominant at that point and the cause. I was supporting a cross-platform enterprise environment and was deep into this at the time. I saw it from the front line’s and I even implemented my company’s port to Linux. Everyone was skeptical about Itanium. Enterprises companies in most part didn’t have the luxury to experiment.

lizknope 4y ago

Linux was most of it but it needed hardware to run on. I wrote in another comment in this thread that our semiconductor design software was all on Solaris. We started shifting to Linux around 2002. We still needed some Suns for the large jobs that required over 4GB RAM. When AMD64 came out that was the end of Sun for us. The chip design software vendors quickly ported to it.

TristanBall 4y ago

I was a sysadmin in a windows desktops + sun/windows server env for about 10 years from 1999 I loved our sun boxes, hated that we could only ever afford their "midrange" models, and then only from grey market resellers, or outright second hand. Everything being said here is pretty much true.. but the additional factors I'd call out were vmware and to a lesser extent "componentised" infrastructure systems.. I'm thinking blade servers and sans. If I bought a sun box, it was only a sun box. If I bought an intel/amd, then it could be linux or windows or both. Yes we could, and did, hook the sun boxes to SAN's, but I don't remember sun ever having a particularly strong blade offering.. or maybe it was just priced and like sun hw, either way.. Dell Blades let me do incremental upgrades at price points within individual projects sign off budgets, and dell delivery times + San boot + esx meant I could usually have new systems up and available in a couple of weeks. That sounds bad in the cloud era, but before that I'd probably still be trying to get quotes or negotiate discounts, so it was fantastic. Linux and the other bits above let me say "yes" to user requests more often, with lower friction and shorter deliver times. I can't tell you just how nice that was. Then I started work at an aix/linux/windows shop and the whole cycle repeated...