DDOS is vandalism, simply. We'd cover it if a mob was throwing rocks at Microsoft, RedHat og IBMs head quarters, would we not? This is just as much vandalism, and there *is* no justification for vandalism.
This is actually why I am no fan of groups like Anonymous, and do not cheer when they for example hit ISIS targets. There's bound to be a lot of collateral damage, and a lot of innocent people will be targeted.
As for Linode? God knows what they or some of their customers have done. But this time - like a lot of other innocent people - it actually hit *me* too. I got tired of running a physical machine at home just so I could run all my services, so I got myself a server at Linode. It has a lot better WAF, too.
So am I satisfied with Linode? You bet. This is not their fault. I am sure they are bound to learn from it, to quicker stop this in the future. But it is not easy, to stop mobs. Because DDOS groups are just that: Mobs! No matter how much you sympatize with a cause, a mob is never justified.
Address space is large enough unless we do something seriously fucked up. The IPv6 adress space has enough Ip-adresses that every atom of the surface of the earth can have 40.000 adresses.
Or, to divide it up a bit:
A "local network" will probably get a/64. This is *enough*, trust me, it's so much addresses that it can comtain the entire ipv4 address space - SQUARED. Noone will ever need more adresses than that in a local network. A typical "end site" (a company, or even maybe a home user) would probably get a/48, or 65536 local networks. Again, *enough*. An ISP would very often have one or perhaps several/32s. That means it can have 2^16 = 65536 "customers" who each have enough ip-adresses.
However, there are recommendations to limit the assignments for "home" users to/56. This makes for only 256 local networks in your home.
So if an ISP has a/32, we can imagine the following example: Half of it, that is a/49, is allocated in/48 networks, allowing for 32768 corporate customers. The other half of it is allocated in/56s, allowing for 32768*256 = 196608 home users.
Currently, one/3 is allocated to global unicast adress space. This gives space for 2^(32-3)=2^29 = roughly 534 million ISP allocations. Or, in another word, approximately one ISP per 10th of todays inhabitants in the world.
I'm in IT myself, and I know how difficult it is to come up with good test-data for your testing...so what's better than production data?
I'm not saying it is so, but it could very well be that the testers have loaded into it this years candidates, made up some likely result, and run the software to see that it works...
1) Discover that the turkey that was supposed to be cooked at home by one of us is still raw upon arriving at the cabin. Remember, if it's frozen, this needs to happen a day or two before the turkey dinner! In hour case, that was the key to success. 2) On the morning of the dinner, stuff the sauna oven full of wood. Apply match, and heat. 3) Repeat 2 until you are beyond the useful range of the thermometer - for example 150 deg C/300 deg F (testet values) 4) To keep turkey from getting dry, apply water as normally. Dress in tin-foil (no idea if it helps...but it didn't hurt!) 5) Cover every bit of your skin. Take a deep breath, enter the sauna with the turkey, put it on the topmost seat of the sauna. Remember - breathing too deaply can burn your lungs! 6) About every halp hour, again cover every bit of your skin, enter your large oven (eh, the sauna), unwrap tin-foil, add water, wrap again in tin-foil. Warning: Spilling water on the sauna oven will create dangerous water vapour that could again burn your lungs! 6) Refill wood whenever the thermometer drops towards the useful range again. 7) Wait an appropriate amount of time. Example: 8-9 hours with a 9 kg/20 lb turkey. 8) Bon appetit!
Yah. The real heros bringing us the PC revolution was the guys reverse engineering the hardware/BIOS, and made cheap clones. The OS was just what became the de facto standard.
As we all know, DOS won over CP/M. CP/M was technically superior at the time, but lost for political and/or contract reasons, whatever.
Digital Research then went on to create a better DOS to compete. MS fought it with all means it could, and it went into oblivition.
At early stages, MS Windows was just a graphical shell on top of DOS. It wasn't particulary good either. There were competing graphical shells, for example Digital Research' GEM. Digital Research lost the patent lawsuit that MS essentially won, and GEM was limited to have only two windows simultaneously...who knows what it could have been.
MS has not had the technical best/superior solutions at any time. It was just better at legal and marketing stuff than anyone else.
The PC revolution would have come with or without MS. We'll never know how much innovation MS have killed on its way where it is, so to hail it as a savior is just plain stupid.
There is a simple answer: The days that you can live on selling OSes and Office suites and finance the rest of your business through that *will* end. Sooner or later. Microsoft understands that, of course, so they're getting desperate to generate significant income through other markets than retail software.
They have never been good at it, though, the only game they are good at, is milking their monopoly.
But what happens when the monopoly eventually disappears?
We've seen a lot of strange moves from Microsoft., and this one will not be the last one.
It's not about removing functionality. It's about letting the integrators (that means Dell, HP etc) decide a little bit about the functionality. If they could decide to leave out IE and install Firefox for their users instead, then there'd be not much to gripe about.
If they could decide to leave out WMP and install vlc instead?
That would mean IE would have to compete on functionality, as would WMP, to compete with the alternatives.
For IE vs. Firefox, it's starting to become a real competition again, though. But IE still has an edge in being pre-installed, so only by IE being exceptionally bad, Firefox can actually compete. And let's face it, it has been exceptionally bad. I fear what would happen if IE should become "good enough" again.
Me, I think the 42S was the best thing that ever came out. Still have it, although since I'm more or less always by a computer now, and don't have that much advanced mathematic needs either, it's not been much used the last 10 years.
One of the best thing about them, was their durability. They were nearly unbreakable.
Before my 42S, I had a 10C (which I later gave to my brother), and a 15C, which I still have.
In addition to the other comments: If it's their own code, and only theirs, they are free to license it under any license they will, even if it's already licensed under GPL. It's called dual-licensing, and is a well-known practise.
I remember when the "google equivalent" was a web page with all the worlds web servers was listed. It rapidly got divided into one page per country....
Windows XP SP2 was a great stride forward in Windows security. They did much of what they should have done ages ago, and the security people even was allowed to break functionality for a lot of applications. I respect those people who lead that work tremendously, even though I personally hate using Windows, and avoid it as much as possible.
I believe this to be a good move. I don't think all people at Microsoft is evil. Even though some people probably are;-)
I can testify that it's damned difficult to get the hang of Windows after having learnt Unix or Linux well.
Not only that, I don't really see what's in it for me. I mean, for me Windows is:
- Harder - Less useful for Work (granted, I could of course change job or department, but..) - More expensive - Illogical (nothing ever makes sense) - Less stable.
All of these arguments can probably be reversed. The point of this post isn't to show that Linux is any better, only that ANY switch is difficult.
This story is about the place I used to work. We had Unix on the desktop, this was at the time that universities and academic institutions had good deals with Unix vendors on their workstations. Now, we grew, and got more administrative/economy people, and at one point management said that it was time to standardize platform on Windows. Those economists just love their excel spreadsheets. None of the engineers were particulary happy with it. But it wasn't until one of the secretaries uttered a tiny little "I think I would rather prefer staying with Unix" that the plan was reversed.
Now, the situation is a dual platform, mostly, Windows or Linux. And when I last year met this secretary, 10 years after I left the place, she was still a Linux user, still rather non-technical, but *use* it, she could!
So no - Linux isn't harder to use - only different. That is my firm belief.
Myself, I did once try to use Windows seriously on the desktop. That was when I first entered commercial world, in a consultant company. It even lasted around 2 years. I was, however, always a Unix guy, professionally. It simply made no sense for me having Windows on the desktop. After 2 years, I finally gave up and installed Linux. That was in 98/99. It felt like coming home after a long and tiring journey. Since then, I've been a happy Linux user, and I see more and more of my colleages in IT using Linux at the deskop.
I didn't mean this. I'm dead sure that a significant amount of important customers have to switch to open formats before they themselves turn to open formats. Their history proves it.
There is also another inherent danger in choosing the Microsoft standard - who's to say that Microsoft, once getting it approved by a standard, will actually *follow* their own standard? What's to prevent them, once having an ISO-standard that's widely accepted, to implement small but important, non-open extensions in their next version?
I'm not saying they will, I say that I don't think it's beyond them, if they think that it'll benefit them against their competitors, and that they can get away with it. And what good will their once so open standard be then? It will be worth *nothing*, once again.
I might be paranoid. But in some cases, I think you just have to be.
There is of course also a danger that the same thing happens to ODT. The difference is that the reference application here is an already quite known and wide-spread Open Source-application, and it's already implemented in several products by several vendors. The opportunity for any of those vendors to get away with it without losing customers is just so much less. History has shown that Microsoft can get away with a lot, and nothing, not even lawsuits, can really prevent them.
For me, it's not so much about getting everyone to use Open Source as to ensure that *I* have the freedom to use it, and still be a part of society (i.e be able to communicate with government etc).
It's definitely true that it's open STANDARDS that matter. There is, however, a large pitfall: Don't let vendors like Microsoft redefine what an open standard actually is. They tried a little while ago, with their previous office XML standard...
Could you give an example that I could verify?
I found parts of the gphoto2 code clear enough that I was able to help make my Nikon D70 more supported. gtkam, btw, and I guess kamera for kde, are two applications with a simple user interface that normally just works. This even makes the state of digital camera support at least on par with Windows, where the norm is more that each camera vendor writes his own camera support/control application, and not joining forces with the others.
"Does open source software live up to the promises of helping people to learn from the code and allowing numerous people to contribute to it?"
And I think, to the second (relevant) question, there answer is no, it is not living up to that promise.
Afraid I disagree with you somewhat. Granted, you *will* need to be interested in learning to code, and have some limited coding experience. And many programs *are* complex by nature. It's not realistic to be able to grasp the full structure of a program from the start, but rather to dive into specific bits and pieces, and gradually work from there to understand more of the totality.
Programming is not easy, it's that simple. You can not blame it all on lack of educational value of the code in question. Open source code is, after all, not primarily written to teach people how to code.
Given time and effort, most people with a small talent for coding *can* participate in open source projects. Not all projects are equally suited for everyone, but I'm dead sure that if you're a coder, there is at least *one* program out there in your area of interet that you'll be able to participate in, which will give you a valuable programming experience.
Yes, some Open Source software is definitely too complex and doesn't have readable code to outsiders. But there is also software who does. I'm no great coder, but I do contribute here and there, when I have the knowledge.
Thinking about all the "Aunt Jane"s is not entirely correct either. Just because *you* are never going to be able to modify some software, doesn't mean that it isn't valuable to you that someone else (than the original author) can. This last point is often overlooked by those who have not yet seen the point in open source.
But some code is poorly structured and not well documented. That's bound to be true both with open source and proprietary code. It's no more of a problem with open source than with proprietary code, it's just that you happen to be able to see it. At the worst, the open source will degenerate to something equivalent with free-as-in-beer closed source, because noone except the author can change it. It's still redistributable, though. And if everything else fails, you can always rewrite the code, and look at the code to hopefully understand *some* nits and bits here and there.
We can only hope it's easily crackable. Fat chance they'll support anything else than Windows, and Mac to "show that they're multiplatform". Do you expect to use it on anything else? Wait for the crack...
The pitfalls could be summarized into these three points, as I see it:
1) Patents/licenses.
Do Microsoft have any patents to any methods/techniques in the XML schemas? Patents seem to be granted on pretty much anything, nowadays (that's another discussion), but even if it's non-valid, open source-developers can rarely afford to either contest or license use of a patent.
If Microsoft makes a blanket license to use any patented method they might have claim on relating to the format, no questions asked, and with a right to sublicense, kudos to them. If not, it's not an open format.
There was also some technicalities regarding "a conforming implementation". Does this mean that you're not allowed to implement support for any extensions that are non-conforming to the specification? Are Microsoft the only ones allowed to do that? (Microsoft doesn't actually have a good track-record for following specifications - not even their own ones).
That leads us into point 2:
2) Is Microsoft itself going to conform to the specification, or are they going to embrace and extend their own formats? If they are, this means that the situation won't be much better than today, as we're forever stuck with reverse-engineering "the newest Microsoft Office formats". Making an XML specification itself changes nothing. The value in this XML specification coming from Microsoft, is that it promises interoperability with and long-term-archivability of documents written in Microsoft office, something that's been problematic up to now.
If this is just a "snapshot", however, something that some version of Microsoft office once used, but you can't be sure that *any* Microsoft Office-document can be opened with just implementing the specification, we gain nothing. Nothing at all. Then, it's just a fake bone, a PR-stunt, to keep off ODF competition.
ODF of course have the same problems, but at least that format comes from the open source world, which means that at least the open source implementations (that are likely to become the "reference implementations") can be studied to see what the hell they have changed and why they're not conforming.
A Unix admin with some experience has had the opportunity to become more and more effective. It has to do with tweaking the routine, making shell scripts which makes your job much easier, and generally working with the command line. As time goes and knowledge comes, one can have a remarkable arsenal of scripts and tools at hand. Since most stuff is quite portable (you can compile bash or any other shell of your preference for every Unix there is, I think), and the *basic* unix things can be expected to be there always, one tends to rely on it in ones day-to-day tasks, and reuse whatever can be reused as new Unix-machines comes in.
Windows, however, isn't like that. At least not initially. Good Windows-admins know their way in the GUIs, know exactly where to click, and can navigate quickly to get stuff done.
I know you can script, you can do *some* stuff from the command line, but it quickly becomes a challenge, and of the wrong kind. You can get a bit of the way with Cygwin and such, but you'll end up constantly trying to make Windows into Unix.
Never mind that all monitoring-tools, scripts, things set up to run through cron, and all that stuff, has to be changed. No, a Unix admin truly does *not* want to migrate to windows. I know, I am one.
Slashdot Mirror
DDOS is vandalism, simply. We'd cover it if a mob was throwing rocks at Microsoft, RedHat og IBMs head quarters, would we not? This is just as much vandalism, and there *is* no justification for vandalism.
This is actually why I am no fan of groups like Anonymous, and do not cheer when they for example hit ISIS targets. There's bound to be a lot of collateral damage, and a lot of innocent people will be targeted.
As for Linode? God knows what they or some of their customers have done. But this time - like a lot of other innocent people - it actually hit *me* too. I got tired of running a physical machine at home just so I could run all my services, so I got myself a server at Linode. It has a lot better WAF, too.
So am I satisfied with Linode? You bet. This is not their fault. I am sure they are bound to learn from it, to quicker stop this in the future. But it is not easy, to stop mobs. Because DDOS groups are just that: Mobs! No matter how much you sympatize with a cause, a mob is never justified.
Period.
Address space is large enough unless we do something seriously fucked up. The IPv6 adress space has enough Ip-adresses that every atom of the surface of the earth can have 40.000 adresses.
Or, to divide it up a bit:
A "local network" will probably get a /64. This is *enough*, trust me, it's so much addresses that it can comtain the entire ipv4 address space - SQUARED. Noone will ever need more adresses than that in a local network. /48, or 65536 local networks. Again, *enough*. /32s. That means it can have 2^16 = 65536 "customers" who each have enough ip-adresses.
A typical "end site" (a company, or even maybe a home user) would probably get a
An ISP would very often have one or perhaps several
However, there are recommendations to limit the assignments for "home" users to /56. This makes for only 256 local networks in your home.
So if an ISP has a /32, we can imagine the following example: /49, is allocated in /48 networks, allowing for 32768 corporate customers. /56s, allowing for 32768*256 = 196608 home users.
Half of it, that is a
The other half of it is allocated in
Currently, one /3 is allocated to global unicast adress space. This gives space for 2^(32-3)=2^29 = roughly 534 million ISP allocations. Or, in another word, approximately one ISP per 10th of todays inhabitants in the world.
There are several /3s not yet allocated.
I guess there is enough.
Or, if you live in a country which still haven't protected DRM by law, you should always remove DRM as part of your buying process.
Well. That's not *exaclty* what I meant.
But the actual particiants were known, so to test the post-election part of the system, all you need to do is to mock together some results.
Again: That might be a plausible, though slightly more boring explanation :)
Then again, it could also be a test of the procedures for reporting faked results...we'lll probably never know :)
I'm in IT myself, and I know how difficult it is to come up with good test-data for your testing...so what's better than production data?
I'm not saying it is so, but it could very well be that the testers have loaded into it this years candidates, made up some likely result, and run the software to see that it works...
And apparently it did! ;)
1) Discover that the turkey that was supposed to be cooked at home by one of us is still raw upon arriving at the cabin. Remember, if it's frozen, this needs to happen a day or two before the turkey dinner! In hour case, that was the key to success.
2) On the morning of the dinner, stuff the sauna oven full of wood. Apply match, and heat.
3) Repeat 2 until you are beyond the useful range of the thermometer - for example 150 deg C/300 deg F (testet values)
4) To keep turkey from getting dry, apply water as normally. Dress in tin-foil (no idea if it helps...but it didn't hurt!)
5) Cover every bit of your skin. Take a deep breath, enter the sauna with the turkey, put it on the topmost seat of the sauna. Remember - breathing too deaply can burn your lungs!
6) About every halp hour, again cover every bit of your skin, enter your large oven (eh, the sauna), unwrap tin-foil, add water, wrap again in tin-foil. Warning: Spilling water on the sauna oven will create dangerous water vapour that could again burn your lungs!
6) Refill wood whenever the thermometer drops towards the useful range again.
7) Wait an appropriate amount of time. Example: 8-9 hours with a 9 kg/20 lb turkey.
8) Bon appetit!
I have seen the same happen with SVN. Trust me, people who don't understand version control systems will *always* find a way to fuck up! :)
0-click-shopping: Buy something *unless* you press a button :)
might not be able to *write* the entire collection of Shakespeare, but with this setup, I'm quite sure that they would be able to digitize it!
Yah. The real heros bringing us the PC revolution was the guys reverse engineering the hardware/BIOS, and made cheap clones. The OS was just what became the de facto standard.
As we all know, DOS won over CP/M. CP/M was technically superior at the time, but lost for political and/or contract reasons, whatever.
Digital Research then went on to create a better DOS to compete. MS fought it with all means it could, and it went into oblivition.
At early stages, MS Windows was just a graphical shell on top of DOS. It wasn't particulary good either. There were competing graphical shells, for example Digital Research' GEM. Digital Research lost the patent lawsuit that MS essentially won, and GEM was limited to have only two windows simultaneously...who knows what it could have been.
MS has not had the technical best/superior solutions at any time. It was just better at legal and marketing stuff than anyone else.
The PC revolution would have come with or without MS. We'll never know how much innovation MS have killed on its way where it is, so to hail it as a savior is just plain stupid.
There is a simple answer: The days that you can live on selling OSes and Office suites and finance the rest of your business through that *will* end. Sooner or later. Microsoft understands that, of course, so they're getting desperate to generate significant income through other markets than retail software.
They have never been good at it, though, the only game they are good at, is milking their monopoly.
But what happens when the monopoly eventually disappears?
We've seen a lot of strange moves from Microsoft., and this one will not be the last one.
It's not about removing functionality. It's about letting the integrators (that means Dell, HP etc) decide a little bit about the functionality. If they could decide to leave out IE and install Firefox for their users instead, then there'd be not much to gripe about.
If they could decide to leave out WMP and install vlc instead?
That would mean IE would have to compete on functionality, as would WMP, to compete with the alternatives.
For IE vs. Firefox, it's starting to become a real competition again, though. But IE still has an edge in being pre-installed, so only by IE being exceptionally bad, Firefox can actually compete. And let's face it, it has been exceptionally bad. I fear what would happen if IE should become "good enough" again.
- Vegard
Nope.
I never (enter) hang (enter) the RPN (enter) of got
Btw, anyone that like RPN should *really* take a look at the postscript language. It will all feel natural.
- Vegard
Me, I think the 42S was the best thing that ever came out. Still have it, although since I'm more or less always by a computer now, and don't have that much advanced mathematic needs either, it's not been much used the last 10 years.
One of the best thing about them, was their durability. They were nearly unbreakable.
Before my 42S, I had a 10C (which I later gave to my brother), and a 15C, which I still have.
(yes, these were all scientific versions)
- Vegard
In addition to the other comments: If it's their own code, and only theirs, they are free to license it under any license they will, even if it's already licensed under GPL. It's called dual-licensing, and is a well-known practise.
- Vegard
I remember when the "google equivalent" was a web page with all the worlds web servers was listed. It rapidly got divided into one page per country....
Windows XP SP2 was a great stride forward in Windows security. They did much of what they should have done ages ago, and the security people even was allowed to break functionality for a lot of applications. I respect those people who lead that work tremendously, even though I personally hate using Windows, and avoid it as much as possible.
;-)
I believe this to be a good move. I don't think all people at Microsoft is evil. Even though some people probably are
- Vegard
I can testify that it's damned difficult to get the hang of Windows after having learnt Unix or Linux well.
Not only that, I don't really see what's in it for me. I mean, for me Windows is:
- Harder
- Less useful for Work (granted, I could of course change job or department, but..)
- More expensive
- Illogical (nothing ever makes sense)
- Less stable.
All of these arguments can probably be reversed. The point of this post isn't to show that Linux is any better, only that ANY switch is difficult.
This story is about the place I used to work. We had Unix on the desktop, this was at the time that universities and academic institutions had good deals with Unix vendors on their workstations. Now, we grew, and got more administrative/economy people, and at one point management said that it was time to standardize platform on Windows. Those economists just love their excel spreadsheets. None of the engineers were particulary happy with it. But it wasn't until one of the secretaries uttered a tiny little "I think I would rather prefer staying with Unix" that the plan was reversed.
Now, the situation is a dual platform, mostly, Windows or Linux. And when I last year met this secretary, 10 years after I left the place, she was still a Linux user, still rather non-technical, but *use* it, she could!
So no - Linux isn't harder to use - only different. That is my firm belief.
Myself, I did once try to use Windows seriously on the desktop. That was when I first entered commercial world, in a consultant company. It even lasted around 2 years. I was, however, always a Unix guy, professionally. It simply made no sense for me having Windows on the desktop. After 2 years, I finally gave up and installed Linux. That was in 98/99. It felt like coming home after a long and tiring journey. Since then, I've been a happy Linux user, and I see more and more of my colleages in IT using Linux at the deskop.
I didn't mean this. I'm dead sure that a significant amount of important customers have to switch to open formats before they themselves turn to open formats. Their history proves it.
There is also another inherent danger in choosing the Microsoft standard - who's to say that Microsoft, once getting it approved by a standard, will actually *follow* their own standard? What's to prevent them, once having an ISO-standard that's widely accepted, to implement small but important, non-open extensions in their next version?
I'm not saying they will, I say that I don't think it's beyond them, if they think that it'll benefit them against their competitors, and that they can get away with it. And what good will their once so open standard be then? It will be worth *nothing*, once again.
I might be paranoid. But in some cases, I think you just have to be.
There is of course also a danger that the same thing happens to ODT. The difference is that the reference application here is an already quite known and wide-spread Open Source-application, and it's already implemented in several products by several vendors. The opportunity for any of those vendors to get away with it without losing customers is just so much less. History has shown that Microsoft can get away with a lot, and nothing, not even lawsuits, can really prevent them.
For me, it's not so much about getting everyone to use Open Source as to ensure that *I* have the freedom to use it, and still be a part of society (i.e be able to communicate with government etc).
It's definitely true that it's open STANDARDS that matter. There is, however, a large pitfall: Don't let vendors like Microsoft redefine what an open standard actually is. They tried a little while ago, with their previous office XML standard...
- Vegard
Could you give an example that I could verify?
I found parts of the gphoto2 code clear enough that I was able to help make my Nikon D70 more supported. gtkam, btw, and I guess kamera for kde, are two applications with a simple user interface that normally just works. This even makes the state of digital camera support at least on par with Windows, where the norm is more that each camera vendor writes his own camera support/control application, and not joining forces with the others.
"Does open source software live up to the promises of helping people to learn from the code and allowing numerous people to contribute to it?"
And I think, to the second (relevant) question, there answer is no, it is not living up to that promise.
Afraid I disagree with you somewhat. Granted, you *will* need to be interested in learning to code, and have some limited coding experience. And many programs *are* complex by nature. It's not realistic to be able to grasp the full structure of a program from the start, but rather to dive into specific bits and pieces, and gradually work from there to understand more of the totality.
Programming is not easy, it's that simple. You can not blame it all on lack of educational value of the code in question. Open source code is, after all, not primarily written to teach people how to code.
Given time and effort, most people with a small talent for coding *can* participate in open source projects. Not all projects are equally suited for everyone, but I'm dead sure that if you're a coder, there is at least *one* program out there in your area of interet that you'll be able to participate in, which will give you a valuable programming experience.
Only true to some extent.
Yes, some Open Source software is definitely too complex and doesn't have readable code to outsiders. But there is also software who does. I'm no great coder, but I do contribute here and there, when I have the knowledge.
Thinking about all the "Aunt Jane"s is not entirely correct either. Just because *you* are never going to be able to modify some software, doesn't mean that it isn't valuable to you that someone else (than the original author) can. This last point is often overlooked by those who have not yet seen the point in open source.
But some code is poorly structured and not well documented. That's bound to be true both with open source and proprietary code. It's no more of a problem with open source than with proprietary code, it's just that you happen to be able to see it. At the worst, the open source will degenerate to something equivalent with free-as-in-beer closed source, because noone except the author can change it. It's still redistributable, though. And if everything else fails, you can always rewrite the code, and look at the code to hopefully understand *some* nits and bits here and there.
We can only hope it's easily crackable. Fat chance they'll support anything else than Windows, and Mac to "show that they're multiplatform". Do you expect to use it on anything else? Wait for the crack...
The pitfalls could be summarized into these three points, as I see it:
1) Patents/licenses.
Do Microsoft have any patents to any methods/techniques in the XML schemas? Patents seem to be granted on pretty much anything, nowadays (that's another discussion), but even if it's non-valid, open source-developers can rarely afford to either contest or license use of a patent.
If Microsoft makes a blanket license to use any patented method they might have claim on relating to the format, no questions asked, and with a right to sublicense, kudos to them. If not, it's not an open format.
There was also some technicalities regarding "a conforming implementation". Does this mean that you're not allowed to implement support for any extensions that are non-conforming to the specification? Are Microsoft the only ones allowed to do that? (Microsoft doesn't actually have a good track-record for following specifications - not even their own ones).
That leads us into point 2:
2) Is Microsoft itself going to conform to the specification, or are they going to embrace and extend their own formats? If they are, this means that the situation won't be much better than today, as we're forever stuck with reverse-engineering "the newest Microsoft Office formats". Making an XML specification itself changes nothing. The value in this XML specification coming from Microsoft, is that it promises interoperability with and long-term-archivability of documents written in Microsoft office, something that's been problematic up to now.
If this is just a "snapshot", however, something that some version of Microsoft office once used, but you can't be sure that *any* Microsoft Office-document can be opened with just implementing the specification, we gain nothing. Nothing at all. Then, it's just a fake bone, a PR-stunt, to keep off ODF competition.
ODF of course have the same problems, but at least that format comes from the open source world, which means that at least the open source implementations (that are likely to become the "reference implementations") can be studied to see what the hell they have changed and why they're not conforming.
- Vegard
For me, it has nothing to do with humiliation.
A Unix admin with some experience has had the opportunity to become more and more effective. It has to do with tweaking the routine, making shell scripts which makes your job much easier, and generally working with the command line. As time goes and knowledge comes, one can have a remarkable arsenal of scripts and tools at hand. Since most stuff is quite portable (you can compile bash or any other shell of your preference for every Unix there is, I think), and the *basic* unix things can be expected to be there always, one tends to rely on it in ones day-to-day tasks, and reuse whatever can be reused as new Unix-machines comes in.
Windows, however, isn't like that. At least not initially. Good Windows-admins know their way in the GUIs, know exactly where to click, and can navigate quickly to get stuff done.
I know you can script, you can do *some* stuff from the command line, but it quickly becomes a challenge, and of the wrong kind. You can get a bit of the way with Cygwin and such, but you'll end up constantly trying to make Windows into Unix.
Never mind that all monitoring-tools, scripts, things set up to run through cron, and all that stuff, has to be changed. No, a Unix admin truly does *not* want to migrate to windows. I know, I am one.