People seem to forget that OS/2 was the Linux of the early/mid 1990's -- it was the only real alternative for x86 users outside of the Windows family, everyone who was anyone was trying it in 1992-1992 when OS/2 2.0 was a US$50 upgrade from any version of Windows, and for folks who wanted to run DOS programs it was the only 32-bit multitasking game in town until the fall of 1995 (when Win95 was finally released). Windows NT 3.x simply didn't cut it for DOS users.
The only DOS program of significance I can think of which wouldn't work under OS/2 was Doom 1.2 and later, but that was because iD Software changed the sound code after the v1.1 release in such a way that it happened to brake under OS/2. No hostile intent -- they simply didn't care (as evidenced by various USENET discussions about the subject in 1995).
Right. Older DOS programs which would fail under Windows 95 would also fail under OS/2 and vice versa, mainly things such as low-level disk utilities, programs which used VCPI memory management, etc. Such things would not run alongside native 32-bit programs in either OS.
Of course, both OSes could handle such cases by booting to DOS (Windows 9x called it "DOS Mode" and included it in the box, while OS/2 supported a special Dual Boot configuration with a real DOS that was installed separately and which worked in almost the same way).
Also keep in mind that the number of users of OS/2 was very, very, very, small.
Not so -- educated estimates in 1995/1996 had OS/2 sitting between 15% and 20% of the total desktop market, much of it business users, so that meant several million users at least, and Sun estimated a few years ago that there might have been as many as 20 million OS/2 users.
So the degree of compatibility is still really unknown since the majority of DOS and 16-bit Windows applications have never been tested on OS/2.
IBM had the actual source code to Windows 3.1 by agreement with Microsoft, rewrote it as a DPMI client, and included it with OS/2 as its "WinOS2" subsystem.
For all practical purposes it *was* Windows 3.1. Until Microsoft started releasing WIN32S.DLL modules in 1993 and later which added new APIs every few months, OS/2's WinOS2 subsystem was almost 100% compatible with Windows 3.1. Its high level of compatibility was never an issue.
Operating systems like OS/2 were able to retain almost 100% compatibility with DOS and 16-bit Windows applications without making the kinds of architectural sacrifices that Microsoft has made over the past 10+ years.
Why? They simply created a Virtual DOS Machine that was sophisticated enough to handle things properly, including running multiple isolated copies of a rewritten Windows 3.1 concurrently to protect 16-bit processes from each other.
Win32 compatibility doesn't require any of that.
The bloat we're seeing is simply poor technical design on Microsoft's part, and the "backwards compatibility" card is just something they played to explain some of the stupid stop-gap decisions made with their Windows 9x line.
It depends on what you're working on. In the type of work that I do (working on vertical in-house mainframe apps for airlines), two years isn't enough time to learn more than the surface basics about a given application area, much less learn deeply about the app as a whole or the idiosyncrasies of the technology in use. In a year or so one can become useful in terms of making large code changes, but it takes more time than that before one knows enough to make larger decisions.
Just learning the specialized end-user vocabulary can be quite an undertaking.
I spent ten years (between contract time and employee time) at Northwest Airlines in flight ops on their WorldFlight system, and I only learned details of perhaps 20% of the system during that time. There was always room for growth. One of the guys I worked with started when the application did (in 1966), and retired from the airline after over 30 years in that environment, and *he* still didn't know it all (though he did know a LOT ).
When all is said and done it's what you're comfortable with. I view folks who bounce between jobs every 3-5 years as people who can't handle sitting in one place. It might show an ability to shift gears quickly, but it might also show a lack of interest in long-term application development. Some businesses design software to last for years or decades, not months, and that requires a somewhat different mindset than more volatile types of projects like web development or writing shrinkwrapped software for public consumption.
That's exactly what my wife and I do -- the books are categorized by general type (fiction, cookbooks, reference, etc.), and within each category we sort by author. More or less.
It seems to work quite well.:-)
I use the same method for my CD collection. Three groups (Rock, Rock Collections, and Classical), each sorted by performer or composer.
I've only had to send facsimile documents to others a few times, and at least two of them referred to it in all caps so I picked up on that. If it's not correct, I'll adjust the practice, but frankly I don't use them except when I'm buying or selling a house...
Both IBM (PROFS) and UNISYS (OfisLink) had mainframe e-mail systems which combined both mail and meeting/calendar functionality a number of years before MS did it.
It's a functional expectation of old-school corporate e-mail, not an MS "innovation"...
Apparently signatures sent via FAX are legally binding, but signatures sent by e-mail might not be? I know my real estate agent insisted that certain things be sent via FAX for legal reasons...
Okay. Thanks. I agree that the "IBM is a competitor in the PC market" thought may have been a factor among hardware vendors, particularly earlier on (during the v1.x days) when IBM was a stronger competitor in the PC market, but I think it was less of an influence than other factors later on.
I work for SITA in Atlanta -- we certainly use Unisys Clearpath Dorado servers here (which are modern variants of the Sperry UNIVAC 1100-series mainframe), but those are completely unrelated to the ES7000 line (and unrelated to UNIX, Windows, or z/OS).
I'm not aware of ES7000's in use here. That doesn't mean we don't use them, but I've not seen them (and the UNIX hardware I've seen here is all Sun and IBM).
OS/2 failed primarily because IBM weren't trusted to give the same OS/2 to Dell, Compaq et al that they were giving to their own PC division.
There was no tying (either rumored or otherwise) of the OS/2 operating system to IBM hardware, unless you're talking about something which happened in the v1.x days...?
The real competition between OS/2 and Windows occurred after 1992, when IBM's OS/2 (the 32-bit version, or versions from v2.0 onwards) was brought to market as a direct competitor.
It failed for a plethora of reasons, amongst them:
* Poor marketing and focus on the product by IBM as a whole.
* Various tactics on the part of Microsoft (some brilliant, many found to be illegal by the US Federal Court during the anti-trust case) to discourage ISV's from developing for OS/2, to prevent PC vendors from bundling it on new computers, and to spread FUD about its own platforms (remember "Chicago"?) and paint them as superior solutions to OS/2.
* The IBM PCCO's hostility towards OS/2. IBM's PC Company was very much pro-Windows.
* The trade press (in general) was hostile to anything not Microsoft. Some exceptions did exist, and OS/2 won dozens of awards, but head-to-head reviews (when they were done) usually seemed to hedge in favor of Windows. This in spite of some benchmarks and other tests done by folks like David Barnes and pubs like InfoWorld which showed OS/2 absolutely destroying Windows NT in various client and server performance tests.
They didn't improve COMMAND.COM in MS-DOS in over a decade...so why are you surprised that the pattern continues with Windows? Only potential profit centers receive attention.
How else would you suggest the relative success or failure of each department be described?
Colors?
Some arbitrary scale?
At least "grades" are almost universally understood in the US. A department which received an "F" is obviously not a success story when it comes to computer security...
The Fossil and Fossil/Abacus watch models failed because of timing (Fossil took too long to get them to market), not because of a lack of utility. You can get them on eBay for around US$50, or at places like Tiger Direct, and they can be extremely useful devices if all you need is a 66MHz DragonBall SuperVZ-powered B&W Palm running PalmOS 4.1.
Both models are effectively a super-fast Palm m105 (3-4 times faster) with 4MB of Flash ROM sitting on your wrist.
A watch running DateBK3, Parens, BigClock, Pilot-DB, and Yotta is actually pretty useful. No color, Bluetooth, no MP3 playing or other fancy stuff, but it can beam stuff back and forth to other palms via its IR port, and it can play almost all the old Palm games that I've fed it.
Learn more about the Abacus and Fossil watches here.:-)
The Ultimate Boot CD is a nice collection of memory, CPU, partition, filesystem, benchmarking, and BIOS utilities, and the "full" version of the UBCD contains INSERT as well as all of the other stuff. Quite a nice collection of utilities and diagnostic software on one CD.
At a previous job, they used IBM's OfficeVision (formerly known as PROFS), which did a fairly nice job of handling calendars, and before that (when I worked at Unisys) we used a mainframe application called OfisLink.
There are lots of mail systems which handle shared calendars, scheduling, etc., but most folks simply aren't aware of the alternatives because "everyone uses Outlook/Exchange"...
Even normal business users need to remember two or three passwords, and programmers might need to remember several dozen, often on systems with completely imcompatible password requirements.
That makes it harder to use a single mnemonic device to remember complex passwords.
Who knows what the future will bring? Some organizations will *always* need veteran software developers (who generally do a mix of analysis, design and coding, not just coding), but it's hard to say if those places will become the norm or the exception.
Yes, in the world of computer programming, 36 is by far way over the hill.
That wasn't the case in any of the four software shops I've worked in.:-) Heck, I was laid off in 2002 because at age 39 and 13+ years of experience I was tied for the bottom spot experience-wise on a team of 14 people.
That's one of the big differences between working in a larger mainframe shop with complex vertical in-house applications and writing shrinkwrapped software, though. The former tends to have a lot of people who have a decade or more of core in certain areas, while the latter doesn't care so much about such things because the software being produced is of a very different nature.
Slashdot Mirror
A piece of software bundled with the defacto monopoly operating system can (and has) destroy the potential market for that type of software.
If you want to learn more about the legal basis for various countries' actions agaist Microsoft, do some Google searches related to anti-trust laws.
Brake? Give me a break! :-) And 1992-1992? Nice date range, Rich. Try 1992-1993. Sheesh!
:-)
If typos were nickels, I'd be a hundredaire, at least!
People seem to forget that OS/2 was the Linux of the early/mid 1990's -- it was the only real alternative for x86 users outside of the Windows family, everyone who was anyone was trying it in 1992-1992 when OS/2 2.0 was a US$50 upgrade from any version of Windows, and for folks who wanted to run DOS programs it was the only 32-bit multitasking game in town until the fall of 1995 (when Win95 was finally released). Windows NT 3.x simply didn't cut it for DOS users.
The only DOS program of significance I can think of which wouldn't work under OS/2 was Doom 1.2 and later, but that was because iD Software changed the sound code after the v1.1 release in such a way that it happened to brake under OS/2. No hostile intent -- they simply didn't care (as evidenced by various USENET discussions about the subject in 1995).
Right. Older DOS programs which would fail under Windows 95 would also fail under OS/2 and vice versa, mainly things such as low-level disk utilities, programs which used VCPI memory management, etc. Such things would not run alongside native 32-bit programs in either OS.
Of course, both OSes could handle such cases by booting to DOS (Windows 9x called it "DOS Mode" and included it in the box, while OS/2 supported a special Dual Boot configuration with a real DOS that was installed separately and which worked in almost the same way).
Also keep in mind that the number of users of OS/2 was very, very, very, small.
Not so -- educated estimates in 1995/1996 had OS/2 sitting between 15% and 20% of the total desktop market, much of it business users, so that meant several million users at least, and Sun estimated a few years ago that there might have been as many as 20 million OS/2 users.
So the degree of compatibility is still really unknown since the majority of DOS and 16-bit Windows applications have never been tested on OS/2.
IBM had the actual source code to Windows 3.1 by agreement with Microsoft, rewrote it as a DPMI client, and included it with OS/2 as its "WinOS2" subsystem.
For all practical purposes it *was* Windows 3.1. Until Microsoft started releasing WIN32S.DLL modules in 1993 and later which added new APIs every few months, OS/2's WinOS2 subsystem was almost 100% compatible with Windows 3.1. Its high level of compatibility was never an issue.
Operating systems like OS/2 were able to retain almost 100% compatibility with DOS and 16-bit Windows applications without making the kinds of architectural sacrifices that Microsoft has made over the past 10+ years.
Why? They simply created a Virtual DOS Machine that was sophisticated enough to handle things properly, including running multiple isolated copies of a rewritten Windows 3.1 concurrently to protect 16-bit processes from each other.
Win32 compatibility doesn't require any of that.
The bloat we're seeing is simply poor technical design on Microsoft's part, and the "backwards compatibility" card is just something they played to explain some of the stupid stop-gap decisions made with their Windows 9x line.
One of their best ones (Wish You Were Here) was a Warner Brothers release.
Besides, the Apple logo has been used on a number of soemwhat recent CD releases from (I think) Capitol Records.
I found a copy for US$2 at a user book store. :-)
It depends on what you're working on. In the type of work that I do (working on vertical in-house mainframe apps for airlines), two years isn't enough time to learn more than the surface basics about a given application area, much less learn deeply about the app as a whole or the idiosyncrasies of the technology in use. In a year or so one can become useful in terms of making large code changes, but it takes more time than that before one knows enough to make larger decisions.
Just learning the specialized end-user vocabulary can be quite an undertaking.
I spent ten years (between contract time and employee time) at Northwest Airlines in flight ops on their WorldFlight system, and I only learned details of perhaps 20% of the system during that time. There was always room for growth. One of the guys I worked with started when the application did (in 1966), and retired from the airline after over 30 years in that environment, and *he* still didn't know it all (though he did know a LOT ).
When all is said and done it's what you're comfortable with. I view folks who bounce between jobs every 3-5 years as people who can't handle sitting in one place. It might show an ability to shift gears quickly, but it might also show a lack of interest in long-term application development. Some businesses design software to last for years or decades, not months, and that requires a somewhat different mindset than more volatile types of projects like web development or writing shrinkwrapped software for public consumption.
That's exactly what my wife and I do -- the books are categorized by general type (fiction, cookbooks, reference, etc.), and within each category we sort by author. More or less.
:-)
It seems to work quite well.
I use the same method for my CD collection. Three groups (Rock, Rock Collections, and Classical), each sorted by performer or composer.
I've only had to send facsimile documents to others a few times, and at least two of them referred to it in all caps so I picked up on that. If it's not correct, I'll adjust the practice, but frankly I don't use them except when I'm buying or selling a house...
Both IBM (PROFS) and UNISYS (OfisLink) had mainframe e-mail systems which combined both mail and meeting/calendar functionality a number of years before MS did it.
It's a functional expectation of old-school corporate e-mail, not an MS "innovation"...
Apparently signatures sent via FAX are legally binding, but signatures sent by e-mail might not be? I know my real estate agent insisted that certain things be sent via FAX for legal reasons...
Okay. Thanks. I agree that the "IBM is a competitor in the PC market" thought may have been a factor among hardware vendors, particularly earlier on (during the v1.x days) when IBM was a stronger competitor in the PC market, but I think it was less of an influence than other factors later on.
I work for SITA in Atlanta -- we certainly use Unisys Clearpath Dorado servers here (which are modern variants of the Sperry UNIVAC 1100-series mainframe), but those are completely unrelated to the ES7000 line (and unrelated to UNIX, Windows, or z/OS).
I'm not aware of ES7000's in use here. That doesn't mean we don't use them, but I've not seen them (and the UNIX hardware I've seen here is all Sun and IBM).
There was no tying (either rumored or otherwise) of the OS/2 operating system to IBM hardware, unless you're talking about something which happened in the v1.x days...?
The real competition between OS/2 and Windows occurred after 1992, when IBM's OS/2 (the 32-bit version, or versions from v2.0 onwards) was brought to market as a direct competitor.
It failed for a plethora of reasons, amongst them:
* Poor marketing and focus on the product by IBM as a whole.
* Various tactics on the part of Microsoft (some brilliant, many found to be illegal by the US Federal Court during the anti-trust case) to discourage ISV's from developing for OS/2, to prevent PC vendors from bundling it on new computers, and to spread FUD about its own platforms (remember "Chicago"?) and paint them as superior solutions to OS/2.
* The IBM PCCO's hostility towards OS/2. IBM's PC Company was very much pro-Windows.
* The trade press (in general) was hostile to anything not Microsoft. Some exceptions did exist, and OS/2 won dozens of awards, but head-to-head reviews (when they were done) usually seemed to hedge in favor of Windows. This in spite of some benchmarks and other tests done by folks like David Barnes and pubs like InfoWorld which showed OS/2 absolutely destroying Windows NT in various client and server performance tests.
They didn't improve COMMAND.COM in MS-DOS in over a decade...so why are you surprised that the pattern continues with Windows? Only potential profit centers receive attention.
How else would you suggest the relative success or failure of each department be described?
Colors?
Some arbitrary scale?
At least "grades" are almost universally understood in the US. A department which received an "F" is obviously not a success story when it comes to computer security...
What national lottery?
I'm not sure that the feds profit all that much from the dozens of casinos on Native American reservations, either. Or maybe they do...?
The Fossil and Fossil/Abacus watch models failed because of timing (Fossil took too long to get them to market), not because of a lack of utility. You can get them on eBay for around US$50, or at places like Tiger Direct, and they can be extremely useful devices if all you need is a 66MHz DragonBall SuperVZ-powered B&W Palm running PalmOS 4.1.
:-)
Both models are effectively a super-fast Palm m105 (3-4 times faster) with 4MB of Flash ROM sitting on your wrist.
A watch running DateBK3, Parens, BigClock, Pilot-DB, and Yotta is actually pretty useful. No color, Bluetooth, no MP3 playing or other fancy stuff, but it can beam stuff back and forth to other palms via its IR port, and it can play almost all the old Palm games that I've fed it.
Learn more about the Abacus and Fossil watches here.
Cube? The Abacus version isn't all that big. Grow some wrists. :-)
The Ultimate Boot CD is a nice collection of memory, CPU, partition, filesystem, benchmarking, and BIOS utilities, and the "full" version of the UBCD contains INSERT as well as all of the other stuff. Quite a nice collection of utilities and diagnostic software on one CD.
At a previous job, they used IBM's OfficeVision (formerly known as PROFS), which did a fairly nice job of handling calendars, and before that (when I worked at Unisys) we used a mainframe application called OfisLink.
There are lots of mail systems which handle shared calendars, scheduling, etc., but most folks simply aren't aware of the alternatives because "everyone uses Outlook/Exchange"...
Even normal business users need to remember two or three passwords, and programmers might need to remember several dozen, often on systems with completely imcompatible password requirements.
That makes it harder to use a single mnemonic device to remember complex passwords.
Who knows what the future will bring? Some organizations will *always* need veteran software developers (who generally do a mix of analysis, design and coding, not just coding), but it's hard to say if those places will become the norm or the exception.
That wasn't the case in any of the four software shops I've worked in. :-) Heck, I was laid off in 2002 because at age 39 and 13+ years of experience I was tied for the bottom spot experience-wise on a team of 14 people.
That's one of the big differences between working in a larger mainframe shop with complex vertical in-house applications and writing shrinkwrapped software, though. The former tends to have a lot of people who have a decade or more of core in certain areas, while the latter doesn't care so much about such things because the software being produced is of a very different nature.