What about the Actimates animatronic dolls? There were dolls in the shape of Barney the Dinosaur, Arthur, and the Teletubbies. When placed in front of a specially encoded computer game of VCR tape, the dolls would respond to parts of the on screen action.
Fred was editor-in-chief, but I think I have the time period wrong. According to this web page Fred was editor-in-chief from 1988 to 1991. This was after the change from Robert Tinney paintings on the cover to photographs, but still while Steve Ciarcia had his Circuit Cellar column there.
I always remember Fred Langa as Byte's Editor-in-chief for the last four years of the magazine's existence (1994-1998). That time that was essentially the magazine's death march into irrelevancy.
I'm not saying that he was solely responsible for what happened to Byte, but it was on his watch.
On the other hand, that might imply that his experience does extend beyond those used for Windows Magazine.
When you buy a RTOS, you usually aren't getting compiled executable code. You usually get source code that you need to port to the hardware you are building.
Data sheets like this implies that Green Hills adheres to this common practice. So all the open source is more trustworthy than a black box arguments don't apply. Anyone who wishes to deploy a system based on Green Hills' RTOS can audit the code, it isn't hidden from them. Also, this PDF linked says:
INTEGRITY178B has been audited and approved by the FAA for DO178B Level A use.
Which to me implies that it has had a more thorough external audit than most open source packages.
One final argument is that an RTOS is usually very small. Their Velocity RTOS can run in 3KB of RAM. When the OS is stripped down to something that small, a full audit seems like a much less daunting task.
This implies that he isn't arguing security through obscurity. He is arguing for the cathedral approach vs. the bazaar. Don't get me wrong, he still is spreading FUD. Its just a different FUD than you think. He is ignoring the role that Linus Torvalds and some of his trusted lieutenants like Alan Cox play in planning a direction, vetting ideas, and protecting the stability of the code base. Patches don't just come out of the blue from anonymous sources and applied without any examination, no matter what Dan O'Dowd may think.
Still miss Hypercard as a development environment
on
HyperCard Gone for Good
·
· Score: 5, Interesting
The Hypercard environment suited a very iterative development style, perhaps more so than anything else that I have worked on since then. Data was automatically persistent. Switching from running a program to editing a method handler was just clicking on a graphics palette. You could be using a program, see something you don't like, click on a selection tool, click on something, and fix it.
It very much had the feeling of being able to tinker with the engine while the car is running. I suspect that working with Lisp Machines and Smalltalk environments was similar, but unfortunately I missed those boats. (except for being able to play around with Squeak now.)
My first professional software development job was writing a series Hypercard stacks. I remember one time realizing that I had hit an architectural dead end, and needed to refactor a bunch of methods (although I didn't learn the term refactor until much later.) I was lamenting having to make those changes all across all the code base until it suddenly hit me, I could write a hypercard script to make the changes. I put something home stack that said "for each backgroud... for each card in... for item in.... set the script of it to...." and it was all done.
The license with Microsoft specifically covered non-Macintosh applications, because Microsoft said that they didn't want wildly differing interfaces for their applications for different operating systems. The license also gave a time limit between when the signing of the license and when Microsoft could release mouse and windows based products. Unfortunately, since the time period was given relative to the signing of the license, and not the release of the Macintosh, and the Macintosh's launch date slipped. Apple didn't get the large time period of exclusivity that they wanted.
Andy Hertzfeld's folklore.org has some interesting stories on the subject like A Rich Neighbor Named Xerox (I just love Gate's line: "Well, Steve, I think there's more than one way of looking at it. I think it's more like we both had this rich neighbor named Xerox and I broke into his house to steal the TV set and found out that you had already stolen it." I love it even if Andy isn't sure that he has the episode in which it was uttered correct.) Andy also implies in MacBasicthat Microsoft had them over a barrel in renewing the Applesoft BASIC contract and wouldn't renew unless signed both the look-and-feel contract and insisted on dropped the MacBasic product. The Applesoft Basic contract was set to expire in September 1985, and Apple couldn't afford to remove that feature from the Apple II and let it's sales drop at that point in time.
I think you are confusing anonymous code blocks with closures.
The map and sort functions can take an anonymous code block, but they aren't creating closures.
It is a closure if, when it executes, it refers to my variables that were in a visible scope at the time that it was defined. You might want to re-read Function Template from the perlref man page to become a little clearer on the subject.
The sort function doesn't use a closure, it created two new global called $a and $b. The map function doesn't use globals either, it redefines $_.
With this kind of error of understanding, it takes away from the rest of your posting.
Well, there are a lot of people who try to emulate an early 21st century version of the lifestyle they read in those books, and then people who couldn't be bothered to read but found the style and attitude cool. Eventually it spreads out to people who see The Matrix and decide to dress in black leather trenchcoats and sunglasses.
esr's jargon file entry for cyberpunk includes the cultural extension to the original literary definition.
The term cyberpunk was invented to describe a scifi subgenre that started to bloom in the late seventies to early eighties (around the same time that punk did) that intended to shake scifi out of the tame complacent mode that scifi had gotten itself into. (similar to the rock around the time period punk came along.) The cyperpunk authors were new, and their work had some obvious weaknesses, but if you put aside the analytical criticisms and read it for pure enjoyment, you would get a better story than you would from the authors with decades of work under their belt, or reading their decades old, but really good work yet again(Do I have to keep putting these cyberpunk <-> punk rock comparisons in parenthesis? When I'm talking about the old and out of touch, think Asimov, Baen, Pournelle, Jagger, Townshend, etc. When I'm talking about new and explosive for the day, think of Gibson, Sterling, the Sex Pistols, the clash and the Ramones ) If you think that the term cyberpunk has nothing to do with punk rock, you are the one that is confused.
Of course, you could co-opt the word for your own meaning (to some extent, the cyberpunk movement that grew out of the cyberpunk scifi movement already has) then why don't you just steal "hacker". Its already been misused by a large part of the general public for just this purpose.
Unfortunately, if google changed their links now to recorded redirects, there would be a large outcry about privacy concerns.
When google added a user cookie (the first popular search engine to do so.) people were concerned about someone's search behavior being tracked. When they started selling AdWords, people wondered if their google cookie could then expose their browsing habits. Right now, people are concerned about their Orkut project.
There are only so many times they can feign innocence on the matter.
I've often wondered if Google would do well as taking the
clicking of the "Next" link as an implicit lack of confidence in the current pages search and having an abandoned search be an implicit vote of confidence. After all, if you have stopped searching, the current page of search results likely contains a good answer. If you click next, then the current page is likely not to have one.
What if a system's search could draw information from all the applications within the system. For example, if your electronic datebook had a day long entry for "wedding" and the photo manager has photos taken on that date, then a search for "wedding photos" would be able to find out when the wedding was, match it up with the date the photos were taken, and come up with the appropriate set.
To some extent Apple tried this with "Newton Intelligence" on the MessagePad. If you wrote "Thursday, Lunch with Bob at Redbones" It would (after you fixed all of its handwriting recognition mistakes) look up Bob and Redbones in the address book, look in the calendar for the next occurrence of a Thursday, and schedule a noon time appointment.
Newton Intelligence really only amounted to a small set of interapplication tricks, but it was assumed that as the popularity of the units increased, the functionality would be extended. (which pretty much tells you what happened to it.)
There isn't a syntactic similarity between Tcl and Applescript, but there are some deeper similarities.
In Tcl and the Tk Toolkit John Ousterhout describes Hypercard as an inspiration for Tk for an easy to use to to create graphical applications.
The design of Tcl command extensions is also very similar to the classes and events that an application exposes to Applescript. Tcl statements are in the form "command arg...". When you extend Tcl, you take libraries, maybe even existing libraries, and add them as new commands to the language. So while Applescript exposes your applications component functions to scripting, Tcl suggests that you decompose your application to components and tie them together with scripting. Tcl's scripting language subroutines also look like additional commands, and base language commands, a library extension's command, and user written subroutines are essentially interchangeable.
That last point about the commands being interchangeable, and Tk's "send" command combine to create something very similar to Applescript. You can send arbitrary Tcl expressions to a Tk application, have it evaluated based on the context of the remote interpreter, and have the result returned to you.
set result [ send Editor.text get 1.end ]
and
tell application "Editor" get the text of window 1 end tell
may look a bit different, but they seem very similar to me functionally.
Clascal and Object Pascal are actually two different languages. Clascal was a language for the Apple
Lisa. Apple then scrapped Clascal, invited Wirth to Cupertino and produced Object Pascal for the Mac.
Re:USB toothbrush
on
USB Menorah
·
· Score: 2, Informative
According to Darl, the GPL is designed to take profit incentive out of copyright ownership. Copyright is defined in the Constitution. In the case of Eldrid v. Ashcroft, the Supreme Court decided that the constitution's description of copyright existing to an advancement of arts and sciences includes the possibility of profit as a motivation. So, according to Darl, The GPL is unconstitutional.
I've read other odd things that he has said, but this seems to be a new one.
One particular example that might be similar to your case is American Geophysical Union v. Texaco Inc., where a Texeco scientist was photocopying articles from Journals we subscribed to. Since the research was done to to strengthen Texaco's balance sheet, the purpose wasn't absolutely non-profit. Since an entire article was photocopied, it fell against the substantiality factor. The end result is that the court didn't consider it fair use.
No. The Sony vs. Universal case which came up with the space shifting argument didn't say that home taping was fair use. It just said that Sony wasn't to blame if and when it wasn't. The court saw that there were ranges of uses, between taping a soap opera while one was at work, to large libraries of archived shows. They also saw that there were some content producers who didn't mind taping (the producers of sporting events, for example) and some that did.
If Universal wanted to make a politically unpopular and financially disastrous move, they could have started suing individual VCR users in order to better determine where the fair use line is drawn.
If that's not fair use, then I don't know what is.
I'm pretty sure that you don't know what fair use is.
Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright.
Jon Johansen can consider what he is doing research. Once he publishes QTFairUse and you use it, which category of fair use are you claiming? Are you going to try to say that it is reporting? How about comment?
The legal concept of fair use is more than the consumer of a work saying "I'm using it, and I think I'm being fair."
I thought the theme was "important computers that have some connection to Texas". Dell, Tandy, and Compaq came out of Texas. Apple used to have large manufacturing facilities there. That's six out of ten right there.
You have a point that hardware advances have played a significant role. The idea of free redistribution of an operating system wasn't very useful until the user could
afford a machine that ran a useful operating system. When a home user's machine only had a system monitor and program loader, the idea of an OS is unfathomable.
I don't think I buy the argument that no one outside of a university environment would have any interest in operating
systems. Having a fascination to learn and explore is one thing. Having the kind of nature to flourish in a university environment is another. (Its great when the fall into place in the same person at the same time, real life can occasionally seem a bit messier.) I know people whose interest in computers started by getting the opportunity to use machines from their fathers work. I also know people who dated or avoided breaking up with their girlfriends just so they could get or maintain access to a University's or business' machines.
Until the the Net1 release, the Berkeley code was intermingled with Bell Labs code, considered a derived work, and needed the purchase of an AT&T license. Your "BSD yesterday" corresponds to about BSD 4.4-lite, from 1994.
In a way OzPixel's post got it wrong too. People in the academic environment got the freedoms of liberal distribution, but people outside of the university environment who were interested in learning about or using these technologies were out of luck. Linux expanded to a wider audience than BSD was capable of reaching.
Sun has two designations for ceasing hardware support. There is "End-of-life", where they cease producing the production, and there is "end-of-service-life", where they cease providing replacement parts.
I'm assuming they do some sort of calculation for the likelihood of failure, the cost of storage, the income from support contracts, and figure out how many extra units to store for parts, (and the price for support contract renewal.)
Of course, they produce product for long after you would think they would. The Sun 220R had an EOL date of 05/2002, so you can order replacement parts until 2007.
I guess there are occational exceptions. If you look at the Sun-4c Page you'll notice that the EOSL date for models like the SPARCStation 1 and 2 were extended. I'm guessing that enough people kept signing support contracts for them, which made it worth it for Sun to keep the hardware around.
This EOSL document has an interesting list of Sun's products and their EOSL date.
All this, of course, is just buying parts from Sun. There is a big 3rd party market for older mainframe and minicomputer components. I know of a large Boston area company whose publishing company is finally being moved off of the PDP-11 system they have been using for the last 20 years. From what I understand, as companies decommission systems built on old DEC hardware, people will buy it up for recondition and resale.
OK, I take it back. Everyone should forget entirely about offsite, redundant storage because it might contain some information that is distantly related to some information the Anonymous Coward above disagrees with.
Slashdot Mirror
What about the Actimates animatronic dolls? There were dolls in the shape of Barney the Dinosaur, Arthur, and the Teletubbies. When placed in front of a specially encoded computer game of VCR tape, the dolls would respond to parts of the on screen action.
Fred was editor-in-chief, but I think I have the time period wrong. According to this web page Fred was editor-in-chief from 1988 to 1991. This was after the change from Robert Tinney paintings on the cover to photographs, but still while Steve Ciarcia had his Circuit Cellar column there.
I always remember Fred Langa as Byte's Editor-in-chief for the last four years of the magazine's existence (1994-1998). That time that was essentially the magazine's death march into irrelevancy.
I'm not saying that he was solely responsible for what happened to Byte, but it was on his watch.
On the other hand, that might imply that his experience does extend beyond those used for Windows Magazine.
When you buy a RTOS, you usually aren't getting compiled executable code. You usually get source code that you need to port to the hardware you are building.
Data sheets like this implies that Green Hills adheres to this common practice. So all the open source is more trustworthy than a black box arguments don't apply. Anyone who wishes to deploy a system based on Green Hills' RTOS can audit the code, it isn't hidden from them. Also, this PDF linked says:
Which to me implies that it has had a more thorough external audit than most open source packages.One final argument is that an RTOS is usually very small. Their Velocity RTOS can run in 3KB of RAM. When the OS is stripped down to something that small, a full audit seems like a much less daunting task.
This implies that he isn't arguing security through obscurity. He is arguing for the cathedral approach vs. the bazaar. Don't get me wrong, he still is spreading FUD. Its just a different FUD than you think. He is ignoring the role that Linus Torvalds and some of his trusted lieutenants like Alan Cox play in planning a direction, vetting ideas, and protecting the stability of the code base. Patches don't just come out of the blue from anonymous sources and applied without any examination, no matter what Dan O'Dowd may think.
The Hypercard environment suited a very iterative development style, perhaps more so than anything else that I have worked on since then. Data was automatically persistent. Switching from running a program to editing a method handler was just clicking on a graphics palette. You could be using a program, see something you don't like, click on a selection tool, click on something, and fix it.
... for each card in ... for item in .... set the script of it to ...." and it was all done.
It very much had the feeling of being able to tinker with the engine while the car is running. I suspect that working with Lisp Machines and Smalltalk environments was similar, but unfortunately I missed those boats. (except for being able to play around with Squeak now.)
My first professional software development job was writing a series Hypercard stacks. I remember one time realizing that I had hit an architectural dead end, and needed to refactor a bunch of methods (although I didn't learn the term refactor until much later.) I was lamenting having to make those changes all across all the code base until it suddenly hit me, I could write a hypercard script to make the changes. I put something home stack that said "for each backgroud
The license with Microsoft specifically covered non-Macintosh applications, because Microsoft said that they didn't want wildly differing interfaces for their applications for different operating systems. The license also gave a time limit between when the signing of the license and when Microsoft could release mouse and windows based products. Unfortunately, since the time period was given relative to the signing of the license, and not the release of the Macintosh, and the Macintosh's launch date slipped. Apple didn't get the large time period of exclusivity that they wanted.
Andy Hertzfeld's folklore.org has some interesting stories on the subject like A Rich Neighbor Named Xerox (I just love Gate's line: "Well, Steve, I think there's more than one way of looking at it. I think it's more like we both had this rich neighbor named Xerox and I broke into his house to steal the TV set and found out that you had already stolen it." I love it even if Andy isn't sure that he has the episode in which it was uttered correct.) Andy also implies in MacBasicthat Microsoft had them over a barrel in renewing the Applesoft BASIC contract and wouldn't renew unless signed both the look-and-feel contract and insisted on dropped the MacBasic product. The Applesoft Basic contract was set to expire in September 1985, and Apple couldn't afford to remove that feature from the Apple II and let it's sales drop at that point in time.
Of course, elements of the automotive UI were patented, the most famous be the intermittent wipers patent. Many aspects of early telephone and telegraph dials were patented.
I think you are confusing anonymous code blocks with closures. The map and sort functions can take an anonymous code block, but they aren't creating closures.
It is a closure if, when it executes, it refers to my variables that were in a visible scope at the time that it was defined. You might want to re-read Function Template from the perlref man page to become a little clearer on the subject.
The sort function doesn't use a closure, it created two new global called $a and $b. The map function doesn't use globals either, it redefines $_.
With this kind of error of understanding, it takes away from the rest of your posting.
Well, there are a lot of people who try to emulate an early 21st century version of the lifestyle they read in those books, and then people who couldn't be bothered to read but found the style and attitude cool. Eventually it spreads out to people who see The Matrix and decide to dress in black leather trenchcoats and sunglasses.
esr's jargon file entry for cyberpunk includes the cultural extension to the original literary definition.
The term cyberpunk was invented to describe a scifi subgenre that started to bloom in the late seventies to early eighties (around the same time that punk did) that intended to shake scifi out of the tame complacent mode that scifi had gotten itself into. (similar to the rock around the time period punk came along.) The cyperpunk authors were new, and their work had some obvious weaknesses, but if you put aside the analytical criticisms and read it for pure enjoyment, you would get a better story than you would from the authors with decades of work under their belt, or reading their decades old, but really good work yet again(Do I have to keep putting these cyberpunk <-> punk rock comparisons in parenthesis? When I'm talking about the old and out of touch, think Asimov, Baen, Pournelle, Jagger, Townshend, etc. When I'm talking about new and explosive for the day, think of Gibson, Sterling, the Sex Pistols, the clash and the Ramones ) If you think that the term cyberpunk has nothing to do with punk rock, you are the one that is confused.
Both punk movements did their job of shaking up the status quo, and relegated themselves to part of the establishment. What we have for big budget scifi movies are things like The Matrix, not Capt. Kirk Saves Whales From Extinction With Transparent Aluminum. Punk music is now being used for television commercials and television theme songs.
Of course, you could co-opt the word for your own meaning (to some extent, the cyberpunk movement that grew out of the cyberpunk scifi movement already has) then why don't you just steal "hacker". Its already been misused by a large part of the general public for just this purpose.
Unfortunately, if google changed their links now to recorded redirects, there would be a large outcry about privacy concerns. When google added a user cookie (the first popular search engine to do so.) people were concerned about someone's search behavior being tracked. When they started selling AdWords, people wondered if their google cookie could then expose their browsing habits. Right now, people are concerned about their Orkut project.
There are only so many times they can feign innocence on the matter.
I've often wondered if Google would do well as taking the clicking of the "Next" link as an implicit lack of confidence in the current pages search and having an abandoned search be an implicit vote of confidence. After all, if you have stopped searching, the current page of search results likely contains a good answer. If you click next, then the current page is likely not to have one.
What if a system's search could draw information from all the applications within the system. For example, if your electronic datebook had a day long entry for "wedding" and the photo manager has photos taken on that date, then a search for "wedding photos" would be able to find out when the wedding was, match it up with the date the photos were taken, and come up with the appropriate set.
To some extent Apple tried this with "Newton Intelligence" on the MessagePad. If you wrote "Thursday, Lunch with Bob at Redbones" It would (after you fixed all of its handwriting recognition mistakes) look up Bob and Redbones in the address book, look in the calendar for the next occurrence of a Thursday, and schedule a noon time appointment.
Newton Intelligence really only amounted to a small set of interapplication tricks, but it was assumed that as the popularity of the units increased, the functionality would be extended. (which pretty much tells you what happened to it.)
There isn't a syntactic similarity between Tcl and Applescript, but there are some deeper similarities. In Tcl and the Tk Toolkit John Ousterhout describes Hypercard as an inspiration for Tk for an easy to use to to create graphical applications.
The design of Tcl command extensions is also very similar to the classes and events that an application exposes to Applescript. Tcl statements are in the form "command arg ...". When you extend Tcl, you take libraries, maybe even existing libraries, and add them as new commands to the language. So while Applescript exposes your applications component functions to scripting, Tcl suggests that you decompose your application to components and tie them together with scripting. Tcl's scripting language subroutines also look like additional commands, and base language commands, a library extension's command, and user written subroutines are essentially interchangeable.
That last point about the commands being interchangeable, and Tk's "send" command combine to create something very similar to Applescript. You can send arbitrary Tcl expressions to a Tk application, have it evaluated based on the context of the remote interpreter, and have the result returned to you.
- and
-
may look a bit different, but they seem very similar to me functionally.Clascal and Object Pascal are actually two different languages. Clascal was a language for the Apple Lisa. Apple then scrapped Clascal, invited Wirth to Cupertino and produced Object Pascal for the Mac.
anything YOU cant power with usb
Yes, anything needing more than 500mA of power.According to Darl, the GPL is designed to take profit incentive out of copyright ownership. Copyright is defined in the Constitution. In the case of Eldrid v. Ashcroft, the Supreme Court decided that the constitution's description of copyright existing to an advancement of arts and sciences includes the possibility of profit as a motivation. So, according to Darl, The GPL is unconstitutional.
I've read other odd things that he has said, but this seems to be a new one.
Before you did that, they should probably take a look at the four factors that cpt kangarooski took me to task for omitting.
Fair Use: Overview and Meaning for Higher Education has an interesting analysis of court cases where fair use was argued, and whether it prevailed or not.
One particular example that might be similar to your case is American Geophysical Union v. Texaco Inc., where a Texeco scientist was photocopying articles from Journals we subscribed to. Since the research was done to to strengthen Texaco's balance sheet, the purpose wasn't absolutely non-profit. Since an entire article was photocopied, it fell against the substantiality factor. The end result is that the court didn't consider it fair use.
No. The Sony vs. Universal case which came up with the space shifting argument didn't say that home taping was fair use. It just said that Sony wasn't to blame if and when it wasn't. The court saw that there were ranges of uses, between taping a soap opera while one was at work, to large libraries of archived shows. They also saw that there were some content producers who didn't mind taping (the producers of sporting events, for example) and some that did.
If Universal wanted to make a politically unpopular and financially disastrous move, they could have started suing individual VCR users in order to better determine where the fair use line is drawn.
If that's not fair use, then I don't know what is.
I'm pretty sure that you don't know what fair use is.
Jon Johansen can consider what he is doing research. Once he publishes QTFairUse and you use it, which category of fair use are you claiming? Are you going to try to say that it is reporting? How about comment?
The legal concept of fair use is more than the consumer of a work saying "I'm using it, and I think I'm being fair."
I thought the theme was "important computers that have some connection to Texas". Dell, Tandy, and Compaq came out of Texas. Apple used to have large manufacturing facilities there. That's six out of ten right there.
You have a point that hardware advances have played a significant role. The idea of free redistribution of an operating system wasn't very useful until the user could afford a machine that ran a useful operating system. When a home user's machine only had a system monitor and program loader, the idea of an OS is unfathomable.
I don't think I buy the argument that no one outside of a university environment would have any interest in operating systems. Having a fascination to learn and explore is one thing. Having the kind of nature to flourish in a university environment is another. (Its great when the fall into place in the same person at the same time, real life can occasionally seem a bit messier.) I know people whose interest in computers started by getting the opportunity to use machines from their fathers work. I also know people who dated or avoided breaking up with their girlfriends just so they could get or maintain access to a University's or business' machines.
Until the the Net1 release, the Berkeley code was intermingled with Bell Labs code, considered a derived work, and needed the purchase of an AT&T license. Your "BSD yesterday" corresponds to about BSD 4.4-lite, from 1994.
See Twenty Years of Berkeley Unix From AT&T-Owned to Freely Redistributable for details.
In a way OzPixel's post got it wrong too. People in the academic environment got the freedoms of liberal distribution, but people outside of the university environment who were interested in learning about or using these technologies were out of luck. Linux expanded to a wider audience than BSD was capable of reaching.
Sun has two designations for ceasing hardware support. There is "End-of-life", where they cease producing the production, and there is "end-of-service-life", where they cease providing replacement parts.
I'm assuming they do some sort of calculation for the likelihood of failure, the cost of storage, the income from support contracts, and figure out how many extra units to store for parts, (and the price for support contract renewal.)
Of course, they produce product for long after you would think they would. The Sun 220R had an EOL date of 05/2002, so you can order replacement parts until 2007.I guess there are occational exceptions. If you look at the Sun-4c Page you'll notice that the EOSL date for models like the SPARCStation 1 and 2 were extended. I'm guessing that enough people kept signing support contracts for them, which made it worth it for Sun to keep the hardware around.
This EOSL document has an interesting list of Sun's products and their EOSL date.All this, of course, is just buying parts from Sun. There is a big 3rd party market for older mainframe and minicomputer components. I know of a large Boston area company whose publishing company is finally being moved off of the PDP-11 system they have been using for the last 20 years. From what I understand, as companies decommission systems built on old DEC hardware, people will buy it up for recondition and resale.
OK, I take it back. Everyone should forget entirely about offsite, redundant storage because it might contain some information that is distantly related to some information the Anonymous Coward above disagrees with.