It might not be possible to answer this, since the SW-CMM stops at zero on the low end. Well, to be fair, I guess Microsoft is solidly at zero, since they never were evaluated in the first place.
Anyway, even if they were evaluated and claimed to be at 3 or 4, it is likely only one project was evaluated, and, then, they go to claim it company-wide...at least, that's how many other companies do it.
The CMMs are another one of those things that a person can learn a lot from, but implementing them is no small feat. Truly implementing the CMMs really takes a whole-organization commitment...perhaps just like Extreme Programming? That's probably why CMM-oriented companies are still rare, because the human inertia already in place can be prohibitive. Alternatively, the technologies the PHBs choose to implement the CMMs can be prohibitively expensive and complicated ($150,000 for a requirements management system...sure! We need only three system administrators and two slots in the bureaucarcy for it...that brings the total to...).
Software engineering, even after a half a century, is still in its infancy. Companies are still hiring half-assed fresh meat as Software Engineers and hopping onto every new high-risk technology invented. Just look at all of the available sites on the WWW--that's a pretty good indication of software quality these days. Simply, it sucks. I know that there are a few very serious software houses out there (military comes to mind), but for the most part, Software Engineering will continue to be a joke for quite some time.
Anyway, if you browse the full component lists at SunSolve, you can see exactly what model numbers Sun uses in their systems. For example, the IDE-based Blade workstations list various Seagate drives, and the FibreChannel Blades list a mix of Seagate and Fujitsu drives. Again, I'd bet these drives have been selected for specific MTBF criteria (e.g., 24x7xMany Years).
Please don't classify Sun in the same boat as your local Best Buy, Circuit City, or other "white box" pushers.
This is not a good turn of events for anyone who buys hard drives from Sun.
Actually, this article is largely irrelevant regarding Sun computers. Sun sells tested stable configurations...I'd bet that their drives are just fine. Also, 10K RPM SCSI drives are a different breed from the IDE trash discussed in the article.
Basically, Sun could purchase drives from a different statistical pool than the ones known to fail. Whether they actually do, I don't know, but their reputation is at stake in matters like these.
Firewire is hot-swappable. Try that with a external SCSI Drive. (not a hot swappable disk, the entire drive)
Well, technically an external SCSI drive can be hot swapped (I've done it), but having the bus idle while swapping is important. It's just not a good idea to yank out a drive when the bus is active. I haven't worked with true hot-swap drives; my hope is that they leave behind an intact bus when they are removed (otherwise, I guess hot-swap would be pretty moot).
I don't doubt there is a similar pro-Solaris, pro-SPARC, anti-Linux, anti-Intel memo within Sun's sales organization.
This is probably true, but Sun, at least, can dispense with a little pride by actually marketing Linux-based products. I'm sure things like the LX50 server and Sun Linux pain the SPARC-advocates, but the public wants something other than SPARC at the really low-end. At least be thankful that Sun is listening to the market rather than trying to bludgeon it like Microsoft does.
...and generally also requires hiring expert technicians.
You'd rather have non-experts running systems that your business depends on? All but the smallest networks require experts from initial network architecture to end-user support.
Microsoft products just do what they're supposed to do, and do them reasonably well.
Other options, such as OS/2, Beos, Wordperfect, etc. have all come and gone at Microsoft's whims yet they were all arguable better than MS' offerings. Microsoft's products do approximately what their marketing department says they do, but not nearly as well as advertised. It's pretty rare that I'm actually impressed by one of Microsoft's products after I get to use it.
This is why Microsoft has the edge, it's EASY.
Not really. Microsoft's current edge really is founded on their downright predatory business and marketing tactics over the past two decades.
The overhead to pay MS for software and support is less than hiring Senior UNIX Admins, and that's basically what it all comes down to.
It really depends on the size of the network. UNIX scales better in cost as the size of the network increases. Unless, of course, the company is locked into Exchange, for example.
Can you imagine sitting in that thing as it tracked a frisky storm with shifting winds? In the right conditions you'd be doing 360's, perhaps with significant centrifugal accelerations at the ends.
I guess if they were smart, they would design some form of damping into the pivot. A sleeve bearing with highly viscous grease would probably be sufficient or, perhaps, an impeller immersed in heavy oil. Of course, the natural rotational inertia of the airframe may be sufficient, too, but I really don't know. A sophisticated homeowner might engage the autopilot to keep the airframe stable in a storm, but that would require too much maintaince of the hydraulics (I'd rather just watch TV and drink beer (yes, I'm an American), but I guess someone else might be more gung-ho about it).
...with Spark's [sic] RISC instruction set and Sun's insistance on keeping both hardware and software closed...
What is closed, exactly, about Sun's hardware and software? Solaris is extremely well documented (if not fully documented; regardless, it makes Windows envious), CDE and X Windows are standardized, SPARC is an IEEE standard, and other important hardware components, such as SBus, PCI, SCSI, and IDE, are also standardized. If their hardware is so closed, why do Linux, NetBSD, and OpenBSD run well on Sun-branded machines with comendable peripheral support?
I think the only thing that makes Solaris "closed" is that Sun's compiler (~$1000, now-a-days) is required to build their Solaris source distribution. For a business, a few thousand dollars isn't that big of a deal, especially given that the Sun compiler provides optimizations targeted for each type of SPARC cpu.
In fact, I look forward to SPARC-based systems being one of the safe-havens from Intel and AMD if/when Palladium and "user untrusted" computing tries to take over. PowerPC and MIPS will also be important in the Palladium age. It's important to keep our options open, especially when our current popular "open" systems turn against us.
There are well-educated and thoughtful members of any religion. That fact doesn't make those religions more credible; it only establishes that even well-educated and thoughtful people are not immune to the siren song of religious ideology.
The Solaris vi man page refers to both ex and ed, so I guess we're both right. The Solaris ex man page says, "ex is a superset of ed...", so, perhaps, reading the ex man page makes vi even more powerful. That's impressive.
It doesn't even have a line number indicator on the status bar.
vi is configurable. Try ":set all" to see what you can play around with.
...the position that appends to the end of the line is not available.
Use 'a' instead of 'i'. There is an append mode in vi.
It's bad enough having two modes.
For a pure console environment (no GUI, no mouse), vi's modes really aren't a bad thing. Command mode allows cut'n'paste, navigation, global search and replace, etc. You only enter new data in insert mode. Another thing about vi: its power isn't unleashed until you've read the 'ed' man page.
...cruft is eliminated only when something new comes along that's ten times better. Only ten times. And vi still exists. Amazing.
This is true. There has not been a programming environment invented that is ten times better than vi (or Emacs, to be fair). I have used Java IDEs, used Visual C++, etc., and all of them added only complexity to the development process. Complexity is a large project's worst enemy. And I'm talking about true complexity, here, not the apparent simplicity of an enormous GUI application like Visual C++. For example, when Visual C++ breaks...how do you know what went wrong? What if the binary build system goes awry? What if the VC plugin doesn't work right? What is the best way to keep binary project files under good change management? And so forth. Why add things to worry about, when vi, make, sccs, etc., do everything in a predictable managable way?
One option in UNIX is to use SCCS or RCS as the versioning system. User applications could invoke the appropriate command rather than overwrite a file on the filesystem. The end user would never need to know, either. The applicaition could also provide an intuitive version browser that allows any prior version to be reloaded (again, SCCS and RCS would be invisible).
A lot of the technology to reduce cruft is already spread around out there. Mostly, we just need thoughful uses of that technology to make the low-cruft world possible.
If you *would* bother to read the article you'd notice that the premier cause of cruft is trying to keep everything "backwards-compatible".
Actually, Windows long file names is cruft to the extreme. Long filenames in one UI, 8.3 in another UI, spaces accepted in some places not in others. Where are the long file names even stored? When looking right at the filesystem, say from a UNIX system, only 8.3 is displayed. This means MS kludged the long filenames into some other "hidden" place.
Also, spaces in any file or directory name is simply terrible. For example, how does one handle such spaces in a shell 'for' loop, when the spaces are assumed to be delimiters in the file list? The underlying problem is that the space character is being overloaded in ways that are incompatible across uses of those spaces.
I really think any other solution would have been better than the one MS chose. MS really should be using emulation environments to break themselves of trying to implement everything they every tried to do in the last 20 years with each new release of their OS.
...because who doesn't know graffiti by this point?
I enjoy using grafitti, but I wonder if it is slightly biased towards right-handed people. I tend to have trouble with 'o' coming out as 's' or 'y' pretty regularly, and there are other occasional quirks. Otherwise, I prefer graffiti to the on-screen keyboard.
So, this leads to my question: would this 'ink' stuff work as well for left handers (i.e., people who write "backwards" or "slanty")?
The decision-making capabilities of the American people can be questionable at times.
Everytime I think about these things, I'm reminded of the basic theme to Star Wars. When will the citizens of the US vote for GWB to be our supreme and all-powerful tyrant?
While I agree with basic sentiment, the problem is that action is required now.
What sort of action? Bars on our windows, armed militiamen at every street corner, and a federal database of everything each person says from birth to death??? Anything less than this is just shades of the same situation and would still fail to address the real issues.
The fundamental problem with the fear about "terrorism" is based in our own society. It is basically an extension to the trend to fear our own damned neighbors; wierdo nuts from the middle east are just a special case. Why is it that we are so untrusting that simply walking down a city street at night can be nerve-wrecking. I really think "terrorism" is the least of our troubles.
I'm glad I humored you. However, I am serious, because the underlying issue to all of this is long-term risk.
Fully commented source code counts a documentation.
I count uncommented source code as better than nothing. Source code is a real instance of algorithms and data structures. Granted, it can be understood more quickly if it is commented, but even uncommented source code can be studied to learn what it does. Simply having the ability to act indepently of a particular vendor regarding critical company data is much more significant than whether it can be done conveniently.
Or of Abiword? Or any other proprietry file format used by an Open Source application?
If Abiword is Open Source, how can its file format be proprietary? Even if there is currently only one piece of software implementing Abiword's file format, anyone else, such as an Open Office developer, can study the Abiword source code and associated documentation to make a new implementation.
We are moving off the subject of Open Source and onto Open Standards.
I agree, because proprietary and free implementations can co-exist with respect to truly open standards. This is really the ideal situation. Sometimes, however, the standards themselves can be so complex that no one can achieve a truly compliant implementation. The challenge, here, is to invent families of genuinely comprehensible standards rather than all-in-one mega standards.
I have worked with mega standards in the past, where there was always something else to implement no matter how hard we worked. Also, validating such implementations approaches impossible without gargantuan test suites and man power. I hope that efforts, such as the Open Office common file format project, create useful digestible chunks of functionality that make implementations widely feasible.
I suspect that in fact Office will be using open file formats, and that I am in fact right. Time will tell.
Yes, time will tell. I would be speachless if Microsoft actually did open up the Office formats. If they did, however, my instinct tells me to look elsewhere for their lock-in strategies (i.e., they may use Office 11 to bait more customers into a different mode of lock-in).
I have heard somewhere that when a company start cutting engineers, then the company REALLY is not doing so well.
It depends on the proportions. When you see the core talent leaving...that is the sign of trouble. Who wants to be left behind in a company that lost the people who knew everything leaving everyone else to feel pretty damn clueless about what to do next? Only the people remaining at Corel will truly know where the company stands.
Regardless of cost, Open Source software is inherently safe from volatility among commercial vendors like Microsoft. Open Source software is, by definition, fully documented and non-proprietary. Yes, source code does count as documentation, because it can be used to understand things like binary data formats when printed manuals are not available. The source code can save your ass, given that you'd be completely out of luck trying to interpret proprietary data. Yes, it may be inconvienient, but, at least, you aren't bound by the testicles to Microsoft's whims about forward and backward compatibility, licensing, planned obselesence, etc.
Documentation created today should be readable tomorrow and ten years from now. Is that true of Microsoft Word or Powerpoint? Now, how about text, LaTeX, and Open Office? I do believe that Word is the most dangerous file format invented...do you know how many companies have all their documentation in Office formats? Wouldn't they feel safer knowing that their documentation isn't fundamentally bound to one company's products? Unfortunately, they don't think about such things. Perhaps that is darwinism on a large scale.
And considering that Office 11 is apparently openly based on XML file formats, this is a sticking point in your theory.
It is likely that Microsoft's XML-format files will, in fact, be proprietary in nature. Remember, XML does not imply open, but, instead, it implies structured. Microsoft can use a proprietary DTD along with binary encoded data in between tags to make the Office 11 format no better than any current or past Office file format.
Yes it is. Microsoft is frustrating to no end. They are like a playground bully that kicks kids down every time they try to get up. They are the class clown that twists and repeats other people's words to nullify any chance at an intelligent argument. Microsoft's behavior invokes an irrational emotion that in a natural setting results in such a bully being slaughtered in revenge.
Microsoft to many software engineers is the conceptual equivalent of industrial-age barons, factory housing, and company stores. They are corrupt, everyone suspects or knows that, but so many people are financially enslaven to Microsoft's business model that a revolt is unthinkable.
I predict that the next decades of the "information age" will mirror the aftermath of the 19th and 20th century industrial monopolies, where only intense victimization of tens of thousands of workers stirs a largely ignorant and sluggish Congress into action.
As long as the masses of people implementing IT "solutions" are so damn ignorant of how thoroughly they are being so stabbed in the back by Microsoft's breadth of lock-in and leeching contracts, we will not see freedom of information, press, academia, nor social equality in the software and internet industries.
To fill the same car with petrol in the USA costs £15.07 or $24.11
It costs me only US$15 for a good two weeks of short-haul commuting (One of the nice things about a compact car, BTW). If I had one of those VW diesel Golfs (more oodles of MPG), oh boy would that be nice!
Slashdot Mirror
By the way, what SEI CMM level _is_ Microsoft at?
It might not be possible to answer this, since the SW-CMM stops at zero on the low end. Well, to be fair, I guess Microsoft is solidly at zero, since they never were evaluated in the first place.
Anyway, even if they were evaluated and claimed to be at 3 or 4, it is likely only one project was evaluated, and, then, they go to claim it company-wide...at least, that's how many other companies do it.
The CMMs are another one of those things that a person can learn a lot from, but implementing them is no small feat. Truly implementing the CMMs really takes a whole-organization commitment...perhaps just like Extreme Programming? That's probably why CMM-oriented companies are still rare, because the human inertia already in place can be prohibitive. Alternatively, the technologies the PHBs choose to implement the CMMs can be prohibitively expensive and complicated ($150,000 for a requirements management system...sure! We need only three system administrators and two slots in the bureaucarcy for it...that brings the total to...).
Software engineering, even after a half a century, is still in its infancy. Companies are still hiring half-assed fresh meat as Software Engineers and hopping onto every new high-risk technology invented. Just look at all of the available sites on the WWW--that's a pretty good indication of software quality these days. Simply, it sucks. I know that there are a few very serious software houses out there (military comes to mind), but for the most part, Software Engineering will continue to be a joke for quite some time.
Anyway, if you browse the full component lists at SunSolve, you can see exactly what model numbers Sun uses in their systems. For example, the IDE-based Blade workstations list various Seagate drives, and the FibreChannel Blades list a mix of Seagate and Fujitsu drives. Again, I'd bet these drives have been selected for specific MTBF criteria (e.g., 24x7xMany Years).
Please don't classify Sun in the same boat as your local Best Buy, Circuit City, or other "white box" pushers.
This is not a good turn of events for anyone who buys hard drives from Sun.
Actually, this article is largely irrelevant regarding Sun computers. Sun sells tested stable configurations...I'd bet that their drives are just fine. Also, 10K RPM SCSI drives are a different breed from the IDE trash discussed in the article.
Basically, Sun could purchase drives from a different statistical pool than the ones known to fail. Whether they actually do, I don't know, but their reputation is at stake in matters like these.
Firewire is hot-swappable. Try that with a external SCSI Drive. (not a hot swappable disk, the entire drive)
Well, technically an external SCSI drive can be hot swapped (I've done it), but having the bus idle while swapping is important. It's just not a good idea to yank out a drive when the bus is active. I haven't worked with true hot-swap drives; my hope is that they leave behind an intact bus when they are removed (otherwise, I guess hot-swap would be pretty moot).
I don't doubt there is a similar pro-Solaris, pro-SPARC, anti-Linux, anti-Intel memo within Sun's sales organization.
This is probably true, but Sun, at least, can dispense with a little pride by actually marketing Linux-based products. I'm sure things like the LX50 server and Sun Linux pain the SPARC-advocates, but the public wants something other than SPARC at the really low-end. At least be thankful that Sun is listening to the market rather than trying to bludgeon it like Microsoft does.
Sun does essentially the same thing...
...but is substantially more expensive...
...has less application support...
...and generally also requires hiring expert technicians.
Okay.
How so?
It depends on your point-of-view.
You'd rather have non-experts running systems that your business depends on? All but the smallest networks require experts from initial network architecture to end-user support.
Microsoft products just do what they're supposed to do, and do them reasonably well.
Other options, such as OS/2, Beos, Wordperfect, etc. have all come and gone at Microsoft's whims yet they were all arguable better than MS' offerings. Microsoft's products do approximately what their marketing department says they do, but not nearly as well as advertised. It's pretty rare that I'm actually impressed by one of Microsoft's products after I get to use it.
This is why Microsoft has the edge, it's EASY.
Not really. Microsoft's current edge really is founded on their downright predatory business and marketing tactics over the past two decades.
The overhead to pay MS for software and support is less than hiring Senior UNIX Admins, and that's basically what it all comes down to.
It really depends on the size of the network. UNIX scales better in cost as the size of the network increases. Unless, of course, the company is locked into Exchange, for example.
Can you imagine sitting in that thing as it tracked a frisky storm with shifting winds? In the right conditions you'd be doing 360's, perhaps with significant centrifugal accelerations at the ends.
I guess if they were smart, they would design some form of damping into the pivot. A sleeve bearing with highly viscous grease would probably be sufficient or, perhaps, an impeller immersed in heavy oil. Of course, the natural rotational inertia of the airframe may be sufficient, too, but I really don't know. A sophisticated homeowner might engage the autopilot to keep the airframe stable in a storm, but that would require too much maintaince of the hydraulics (I'd rather just watch TV and drink beer (yes, I'm an American), but I guess someone else might be more gung-ho about it).
...with Spark's [sic] RISC instruction set and Sun's insistance on keeping both hardware and software closed...
What is closed, exactly, about Sun's hardware and software? Solaris is extremely well documented (if not fully documented; regardless, it makes Windows envious), CDE and X Windows are standardized, SPARC is an IEEE standard, and other important hardware components, such as SBus, PCI, SCSI, and IDE, are also standardized. If their hardware is so closed, why do Linux, NetBSD, and OpenBSD run well on Sun-branded machines with comendable peripheral support?
I think the only thing that makes Solaris "closed" is that Sun's compiler (~$1000, now-a-days) is required to build their Solaris source distribution. For a business, a few thousand dollars isn't that big of a deal, especially given that the Sun compiler provides optimizations targeted for each type of SPARC cpu.
In fact, I look forward to SPARC-based systems being one of the safe-havens from Intel and AMD if/when Palladium and "user untrusted" computing tries to take over. PowerPC and MIPS will also be important in the Palladium age. It's important to keep our options open, especially when our current popular "open" systems turn against us.
There are well-educated and thoughtful members of any religion. That fact doesn't make those religions more credible; it only establishes that even well-educated and thoughtful people are not immune to the siren song of religious ideology.
I think you mean read the 'ex' man page.
The Solaris vi man page refers to both ex and ed, so I guess we're both right. The Solaris ex man page says, "ex is a superset of ed...", so, perhaps, reading the ex man page makes vi even more powerful. That's impressive.
It doesn't even have a line number indicator on the status bar.
...the position that appends to the end of the line is not available.
...cruft is eliminated only when something new comes along that's ten times better. Only ten times. And vi still exists. Amazing.
vi is configurable. Try ":set all" to see what you can play around with.
Use 'a' instead of 'i'. There is an append mode in vi.
It's bad enough having two modes.
For a pure console environment (no GUI, no mouse), vi's modes really aren't a bad thing. Command mode allows cut'n'paste, navigation, global search and replace, etc. You only enter new data in insert mode. Another thing about vi: its power isn't unleashed until you've read the 'ed' man page.
This is true. There has not been a programming environment invented that is ten times better than vi (or Emacs, to be fair). I have used Java IDEs, used Visual C++, etc., and all of them added only complexity to the development process. Complexity is a large project's worst enemy. And I'm talking about true complexity, here, not the apparent simplicity of an enormous GUI application like Visual C++. For example, when Visual C++ breaks...how do you know what went wrong? What if the binary build system goes awry? What if the VC plugin doesn't work right? What is the best way to keep binary project files under good change management? And so forth. Why add things to worry about, when vi, make, sccs, etc., do everything in a predictable managable way?
This project is about correcting OCR errors, not spelling / grammar.
Yes. I remember reading that Tolkien had some trouble with editors who thought they could spell better than he did. IIRC, it was a real mess at first.
So, to reiterate, don't second-guess the authors' intents.
One option in UNIX is to use SCCS or RCS as the versioning system. User applications could invoke the appropriate command rather than overwrite a file on the filesystem. The end user would never need to know, either. The applicaition could also provide an intuitive version browser that allows any prior version to be reloaded (again, SCCS and RCS would be invisible).
A lot of the technology to reduce cruft is already spread around out there. Mostly, we just need thoughful uses of that technology to make the low-cruft world possible.
If you *would* bother to read the article you'd notice that the premier cause of cruft is trying to keep everything "backwards-compatible".
Actually, Windows long file names is cruft to the extreme. Long filenames in one UI, 8.3 in another UI, spaces accepted in some places not in others. Where are the long file names even stored? When looking right at the filesystem, say from a UNIX system, only 8.3 is displayed. This means MS kludged the long filenames into some other "hidden" place.
Also, spaces in any file or directory name is simply terrible. For example, how does one handle such spaces in a shell 'for' loop, when the spaces are assumed to be delimiters in the file list? The underlying problem is that the space character is being overloaded in ways that are incompatible across uses of those spaces.
I really think any other solution would have been better than the one MS chose. MS really should be using emulation environments to break themselves of trying to implement everything they every tried to do in the last 20 years with each new release of their OS.
...because who doesn't know graffiti by this point?
I enjoy using grafitti, but I wonder if it is slightly biased towards right-handed people. I tend to have trouble with 'o' coming out as 's' or 'y' pretty regularly, and there are other occasional quirks. Otherwise, I prefer graffiti to the on-screen keyboard.
So, this leads to my question: would this 'ink' stuff work as well for left handers (i.e., people who write "backwards" or "slanty")?
The decision-making capabilities of the American people can be questionable at times.
Everytime I think about these things, I'm reminded of the basic theme to Star Wars. When will the citizens of the US vote for GWB to be our supreme and all-powerful tyrant?
While I agree with basic sentiment, the problem is that action is required now.
What sort of action? Bars on our windows, armed militiamen at every street corner, and a federal database of everything each person says from birth to death??? Anything less than this is just shades of the same situation and would still fail to address the real issues.
The fundamental problem with the fear about "terrorism" is based in our own society. It is basically an extension to the trend to fear our own damned neighbors; wierdo nuts from the middle east are just a special case. Why is it that we are so untrusting that simply walking down a city street at night can be nerve-wrecking. I really think "terrorism" is the least of our troubles.
Where did VRML go? It was really hyped up several years ago.
LMAO
I'm glad I humored you. However, I am serious, because the underlying issue to all of this is long-term risk.
Fully commented source code counts a documentation.
I count uncommented source code as better than nothing. Source code is a real instance of algorithms and data structures. Granted, it can be understood more quickly if it is commented, but even uncommented source code can be studied to learn what it does. Simply having the ability to act indepently of a particular vendor regarding critical company data is much more significant than whether it can be done conveniently.
Or of Abiword? Or any other proprietry file format used by an Open Source application?
If Abiword is Open Source, how can its file format be proprietary? Even if there is currently only one piece of software implementing Abiword's file format, anyone else, such as an Open Office developer, can study the Abiword source code and associated documentation to make a new implementation.
We are moving off the subject of Open Source and onto Open Standards.
I agree, because proprietary and free implementations can co-exist with respect to truly open standards. This is really the ideal situation. Sometimes, however, the standards themselves can be so complex that no one can achieve a truly compliant implementation. The challenge, here, is to invent families of genuinely comprehensible standards rather than all-in-one mega standards.
I have worked with mega standards in the past, where there was always something else to implement no matter how hard we worked. Also, validating such implementations approaches impossible without gargantuan test suites and man power. I hope that efforts, such as the Open Office common file format project, create useful digestible chunks of functionality that make implementations widely feasible.
I suspect that in fact Office will be using open file formats, and that I am in fact right. Time will tell.
Yes, time will tell. I would be speachless if Microsoft actually did open up the Office formats. If they did, however, my instinct tells me to look elsewhere for their lock-in strategies (i.e., they may use Office 11 to bait more customers into a different mode of lock-in).
I have heard somewhere that when a company start cutting engineers, then the company REALLY is not doing so well.
It depends on the proportions. When you see the core talent leaving...that is the sign of trouble. Who wants to be left behind in a company that lost the people who knew everything leaving everyone else to feel pretty damn clueless about what to do next? Only the people remaining at Corel will truly know where the company stands.
Regardless of cost, Open Source software is inherently safe from volatility among commercial vendors like Microsoft. Open Source software is, by definition, fully documented and non-proprietary. Yes, source code does count as documentation, because it can be used to understand things like binary data formats when printed manuals are not available. The source code can save your ass, given that you'd be completely out of luck trying to interpret proprietary data. Yes, it may be inconvienient, but, at least, you aren't bound by the testicles to Microsoft's whims about forward and backward compatibility, licensing, planned obselesence, etc.
Documentation created today should be readable tomorrow and ten years from now. Is that true of Microsoft Word or Powerpoint? Now, how about text, LaTeX, and Open Office? I do believe that Word is the most dangerous file format invented...do you know how many companies have all their documentation in Office formats? Wouldn't they feel safer knowing that their documentation isn't fundamentally bound to one company's products? Unfortunately, they don't think about such things. Perhaps that is darwinism on a large scale.
And considering that Office 11 is apparently openly based on XML file formats, this is a sticking point in your theory.
It is likely that Microsoft's XML-format files will, in fact, be proprietary in nature. Remember, XML does not imply open, but, instead, it implies structured. Microsoft can use a proprietary DTD along with binary encoded data in between tags to make the Office 11 format no better than any current or past Office file format.
It's about loving Freedom
Yes it is. Microsoft is frustrating to no end. They are like a playground bully that kicks kids down every time they try to get up. They are the class clown that twists and repeats other people's words to nullify any chance at an intelligent argument. Microsoft's behavior invokes an irrational emotion that in a natural setting results in such a bully being slaughtered in revenge.
Microsoft to many software engineers is the conceptual equivalent of industrial-age barons, factory housing, and company stores. They are corrupt, everyone suspects or knows that, but so many people are financially enslaven to Microsoft's business model that a revolt is unthinkable.
I predict that the next decades of the "information age" will mirror the aftermath of the 19th and 20th century industrial monopolies, where only intense victimization of tens of thousands of workers stirs a largely ignorant and sluggish Congress into action.
As long as the masses of people implementing IT "solutions" are so damn ignorant of how thoroughly they are being so stabbed in the back by Microsoft's breadth of lock-in and leeching contracts, we will not see freedom of information, press, academia, nor social equality in the software and internet industries.
To fill the same car with petrol in the USA costs £15.07 or $24.11
It costs me only US$15 for a good two weeks of short-haul commuting (One of the nice things about a compact car, BTW). If I had one of those VW diesel Golfs (more oodles of MPG), oh boy would that be nice!