> Linux would have made a much better choice as a transitional platform from IRIX.
It would have, but this was 1998, so very few of us knew this at the time. I don't know if you remember the tech zeitgeist in 1998, but Linux was still a pretty hardcore geek alternative at the time, maybe like Vim today. NT had things like journaled filesystems.
You can delve into history to see what the Linux community was talking about at the time:
"Intel and Netscape are now officially investors in Red Hat Software. A year ago such a thing would have been almost unthinkable. But times have changed."
"With this release, the GNOME desktop is reaching a usable state."
"Expect that, with the introduction of pre-installed Linux systems from vendors like Gateway, IBM, Dell or others, the number of people searching out a Linux VAR in order to get a pre-installed Linux system will rapidly dwindle."
"140,000 Mexican school labs to be outfitted with Linux and GNOME." (That was the "Red Escolar" or "Scholar Net" project. Its failure is described in http://www.wired.com/science/discoveries/news/2001/08/45737: "RedEscolar, or "Scholar Net," has put computers in 4,500 schools, but fewer than 20 are equipped with GNU/Linux machines. The rest are running Windows 95 or 98.".)
So, from the standpoint of 2011, it's easy to see that committing head-on to Linux could have saved SGI. But, at the time, it was a risky decision.
They did actually start to hitch their wagon to Linux a year later: Scott Bekker "SGI Backs Away from Windows NT - Silicon Graphics - Company Business and Marketing". ENT. FindArticles.com. 11 Feb, 2011. http://findarticles.com/p/articles/mi_m0FOX/is_15_4/ai_55675...
Rick Belluzzo was the CEO who made the NT decision. WP says had been working at HP for the last 23 years, not Microsoft, but he did take a job at Microsoft after being deposed at SGI. WP also says his only degree is a BS in accounting from Golden Gate University, a vocational school, so he's probably not very smart.
Here's what he said about working with Bill Gates: "I was always surprised how much Bill knew. I wonder how you find time in the day to be on top of how many things Bill seemed to be on top of. Bill is so passionate about technology, and he really is in the right role to focus the vast majority of his time."
To me, that sounds like typical non-hacker cluelessness. So maybe it's not that surprising that Rick was blind to the finer points of technical risk, compatibility, portability, and how the free-software community worked. (He was also largely responsible for committing HP to Itanic, effectively killing PA-RISC.)
The the historic state of Linux at the time makes all the difference. I should have been more general and said "Open Platform." Because at the time BSD was a much less risky choice -- it was more mature & didn't have the GPL license stigma being pushed so hard.
The outcome of the SGI WinNT decision is even more interesting when juxtaposed with Apple's decisions during the same time period.
The adoption & use of FreeBSD as the base of OS X's Posix layer shows how SGI could have gained all the benefit of a set of core tools that they needed to stop spending money on maintaining while still retaining control of the best parts of IRIX.
In 1998 both Apple and SGI were on the ropes. Apple embraced and integrated open source, SGI turned to a closed platform. Apple immediately got geeks like me on-board (still have my OS X 10.0.x disk around here somewhere). The difference between the two companies & where they stand now is the ultimate example of how open source can help a corporation succeed.
Apple was pushing very hard at the time to get geeks on. They actually flew their evangelist and lead dev for OS 7 out to my undergrad CompSci program in '99. They really wanted to get a foothold with a dev community & integrating FreeBSD not only allowed them to save on the cost of developing a whole new posix layer, but also was a great way to get a new community of developers interested in the platform.
The Apple story is interesting. I switched from Windows to Mac in 2000 to get on the OS X Public Beta, and it really was all about the BSD underpinnings for me.
But at that point Apple was already on an upward trajectory thanks to the iMac and iBook, with the iPod to come the following year. Granted OS 9 was a dog that has long since seen its day, and the technology in OS X was necessary to create a modern operating system that would enable the kind of UX that Jobs sought. But I question how much of it can really be credited to open source or the geek early adopters per se; certainly it helped, but was it pivotal?
Yeah, I guess it does sound like I'm crediting Open Source and the geeks with singlehandedly saving Apple.
I don't think it was open source itself that saved Apple, but the decision -- as revolutionary as it was at the time -- proved to be an excellent business decision. They added value to their product (hardware sales) by integrating Open Source. And as a bonus, they appealed to a specific audience without having to invest significant resources in selling to that audience.
It was a brilliant business decision because it not only opened a new channel of potential customers, it opened up a whole new set of possibilities for integrating Open Source software; they could leverage the movement. And that's really the significance of the decision in this context.
Apple, Nokia and SGI all focused on their core business -- making hardware -- but only Apple saw the value of leveraging a movement to add value to their hardware/brand. All three were in desperate positions, but only Apple saw the merit of leveraging the Open Source movement for the good of the business.
Where Nokia could have built on the goodwill of their brand by adopting an open and vibrant platform -- one where they could have as much control as they desired -- they have chosen to chain their fortune to a third party that has relatively little vested in their success.
The biggest irony here (other than Apple's role in Nokia's demise) is that choosing Android comes with very little risk compared to the risk that Apple took: there's ongoing development underwritten by a single organization; there's no licensing or quality stigma about the platform; there's a worldwide consortium of companies already using the platform; and there's already a significant and growing market share.
They have missed the opportunity to leverage the movement.
It would have, but this was 1998, so very few of us knew this at the time. I don't know if you remember the tech zeitgeist in 1998, but Linux was still a pretty hardcore geek alternative at the time, maybe like Vim today. NT had things like journaled filesystems.
You can delve into history to see what the Linux community was talking about at the time:
http://old.lwn.net/archives/ http://old.lwn.net/1998/1001/
"Intel and Netscape are now officially investors in Red Hat Software. A year ago such a thing would have been almost unthinkable. But times have changed."
"With this release, the GNOME desktop is reaching a usable state."
http://old.lwn.net/1998/1008/
"Expect that, with the introduction of pre-installed Linux systems from vendors like Gateway, IBM, Dell or others, the number of people searching out a Linux VAR in order to get a pre-installed Linux system will rapidly dwindle."
http://old.lwn.net/1998/1015/
"Oracle has announced that it will ship and support a Linux distribution."
http://old.lwn.net/1998/1022/press.php3
"Like sex in high school, everyone's talking about Linux, but is anyone doing it?"
http://old.lwn.net/1998/1029/
"140,000 Mexican school labs to be outfitted with Linux and GNOME." (That was the "Red Escolar" or "Scholar Net" project. Its failure is described in http://www.wired.com/science/discoveries/news/2001/08/45737: "RedEscolar, or "Scholar Net," has put computers in 4,500 schools, but fewer than 20 are equipped with GNU/Linux machines. The rest are running Windows 95 or 98.".)
So, from the standpoint of 2011, it's easy to see that committing head-on to Linux could have saved SGI. But, at the time, it was a risky decision.
They did actually start to hitch their wagon to Linux a year later: Scott Bekker "SGI Backs Away from Windows NT - Silicon Graphics - Company Business and Marketing". ENT. FindArticles.com. 11 Feb, 2011. http://findarticles.com/p/articles/mi_m0FOX/is_15_4/ai_55675...
Rick Belluzzo was the CEO who made the NT decision. WP says had been working at HP for the last 23 years, not Microsoft, but he did take a job at Microsoft after being deposed at SGI. WP also says his only degree is a BS in accounting from Golden Gate University, a vocational school, so he's probably not very smart.
Here's what he said about working with Bill Gates: "I was always surprised how much Bill knew. I wonder how you find time in the day to be on top of how many things Bill seemed to be on top of. Bill is so passionate about technology, and he really is in the right role to focus the vast majority of his time."
To me, that sounds like typical non-hacker cluelessness. So maybe it's not that surprising that Rick was blind to the finer points of technical risk, compatibility, portability, and how the free-software community worked. (He was also largely responsible for committing HP to Itanic, effectively killing PA-RISC.)