you probably pay somewhere between $20 and $60 per month per machine in electricity no matter how fast or slow; a bunch of those pay for a new low-end machine quickly
you pay for that sort of thing in space (=bigger house) and power; the new systems are often smaller and more energy efficient. upgrading really might make sense...
Microsoft has been retreating constantly over the last few years. Well, another way of looking at it is that they only have given as much as they absolutely had to. Either way, they have moved towards more open systems. The problem is that this opens them up to competition: as long as they had an undocumented mess of software running on underpowered hardware, nobody could come in and compete. These days, you can emulate much of the Win32 APIs without much trouble.
So, Microsoft needs a security blanket in this new world, so they seem to be kicking around several ideas: we could charge for this, we could charge for that. Will it work? I kind of doubt it. If their infrastructure is truly open and well documented, people will offer competing services. If it is closed, it's just the status quo, and users are less and less fooled by it.
As for SOAP and XML that could go both ways. They help a little with interoperability, but Microsoft still has a wide range of options of making things proprietary and non-interoperable while nominally complying with the standards.
DM6000 is what, $2500? I wouldn't be surprised if the primary purpose of this phone isn't simply to flush out people who have what some people might consider, rightly or wrongly, a "suspicious concern for privacy". It will definitely flush out people who have too much money and aren't very smart.
Hardware encryption itself is also both flawed and unnecessary. With hardware, you can't tell what bugs or backdoors may be in there, and if you discover anything, you can't fix it.
There are options that are cheaper, more secure, and more standard around. Current laptops can do real time speech compression and encryption just fine, with software that uses known strong algorithms and is demonstrably without backdoors. You can plug in any of the wireless PCMCIA cards in there and have secure phone conversations over the Internet, not just another ccompatible ell phone user. If you need something smaller, a WinCE or LinuxCE handheld with a cellular phone/modem CF card will probably be a realistic option pretty soon.
Most web site compatibility problems don't come from the servers being used (very few ever do).
Indeed. Most web site compatibility problems actually come from using proprietary software and gimmicky features. You know, things like ASP, VBscript, frames, Flash, and that sort of stuff. The same kind of misfeatures are also responsible for delays.
They are most concerned with getting the site working with the majority of the web browsers and users out there. (Which just so happens to be IE 4/5 on Windows 95/98/NT/etc...) Given their timetables, it would not be feasable to start out supporting everyone and everything...
Sure it would be. If they stuck to plain HTML, not only would they be compatible with all browsers, they'd also be able to deliver faster, lower costs, and keep their site accessible to people with disabilities. Anything else is a waste of money for a government site.
Read on. You get a choice between Microsoft IE on Windows or Microsoft IE on MacOS. You may be able to use Netscape 4 on Windows or MacOS with some add-on (like most users are going to know how to install that), and Netscape 6 won't really work. Users of other operating systems can't enter the site at all (try it from Linux).
That pretty much comes down to "use Microsoft Internet Explorer" on Windows or MacOS, and that's exactly what people are complaining about here.
And it's not the web site, it's the digital certificates that cause problems
The web site seems to just refuse to talk to people if they don't come from a Windows or MacOS machine. They could simply check for the certificates when they are needed, but they actually seem to check for the browser id string. What if Opera or Konqueror get the necessary certificates tomorrow? What if they already have them?
As for saying you can't do your tax return on-line in Netscape? Bollocks. It uses straight HTTPS. I've just completed mine. Did any of you think to check for yourselves? I doubt it, you just saw "it was written by Microsoft" and your knees starting jerking.
Maybe you just have a knee-jerk reaction when people criticize your favorite monopolist? You seem so far in the Windows world that you don't even notice the inconveniences Microsoft causes to competitors.
There are very few cases in which institutions must have immediate access to my records when I'm not in a position to explicitly approve it. Medical records is pretty much the only instance I can think of. However, it appears that most people can be treated just fine for emergencies without having access to their medical records. And for people for whom this is either a real issue or a real concern, they can carry a medical bracelet. With current storage technology, bracelets could contain complete medical records, encrypted with a key accessible to physicians only (the latter isn't a strong protection, but it protects against casual reading).
In all other cases, I should be able to furnish personal information on a case-by-case basis, and I should have the expectation that the company does not store the information beyond the business transaction we are engaged in unless I give my explicit permission.
I can't quite figure out why McNealy is so stupid about this. Giving people tools to protect private information is actually good for Sun's business: it benefits from lots of little smart devices that negotiate with other smart devices.
Perhaps McNealy isn't very sensitive to any of this because he is pretty well off. If you have sufficient amounts of money, you don't have to worry that much about privacy. For example, mere mortals have to file insurance claims, which distributes their records far and wide throughout the insurance industry. If you have enough money, you just pay out of pocket and only your doctor will know.
In any case, the Europeans will not fall for this "we'll just regulate ourselves" stuff. US companies better get used to conforming to some simple privacy rules or they won't continue to do business in Europe.
Got to love those clueless CIO's and phb's who insist on using NT over Unix. Well, if they want to pay 3x fold for an NT solution over a unix one, I guess thats their problem.:-)
Unfortunately, it's everybody's problem. The money attracts more marketing and more product development. NT's problems get fixed by even higher-priced add-ons and alternatives disappear. The market is supposed to punish stupid companies like that, but if everybody's doing it, that doesn't help.
Linux and BSD are somewhat less susceptible to this because Microsoft can't just displace them: they get developed whether or not the developers make money with it. Consulting services for them, on the other hand, make money just like for NT.
I'll be called a "Microsoft shill" or an "astroturfer", but truth is truth: Microsoft's latest server offerings are extremely secure, scalable, and reliable.
Windows 2000 finally delivers the kind of kernel that is necessary to build Internet services. But Microsoft still falls way short on many fronts:
Windows 2000 is two years old, with lots of new APIs, administrative procedures, and protocols. Nobody has yet figured out all the security holes that that creates. It took people a few years to discover how poor VMS security was, even though initially VMS was touted as such a great system. In contrast, the Linux security and service model has two decades of practical experience behind it; people know how to make Linux systems secure.
You cannot strip down a Windows 2000 system to its essentials like you can with Linux: a high performance Linux Internet server needs nothing more than the kernel, the server, and a few POSIX utilities.
You cannot review the Windows 2000 system architecture or inspect its source code; how can you trust it? In particular given Microsoft's lousy track record?
Microsoft markets Windows 2000 to novices, pretending that anybody can run a secure Internet server; they need to take the responsibility for their chosen target market. Linux is unapologetically unfriendly to novices when it comes to server installations.
With Windows 2000, Microsoft has finally addressed one necessary requirement for even playing in this market. But there are lots of other necessities where they still fall short. And as long as Microsoft has notions of world domination and one system fits all on their mind, I don't think they can succeed. In fact, even if Microsoft were to focus, I don't think it makes much sense for the industry to commit to their system: why pay a lot of money for Microsoft's proprietary system if you can get something cheaper and vendor neutral for less already?
The Gecko layout engine seems mostly functional. What seems to be holding them up is all the other stuff.
I still think that the Mozilla team would do wonders for their PR and their user community if they released platform specific browsers for UNIX, Windows, and MacOS based on just Gecko and a minimal UI around that. For Gtk+, that could be Galeon, and for Windows, there is already something like that in their tree. People would be awed by the speed and flexibility of Gecko, and most remaining serious bugs in Gecko would get fixed quickly.
Gecko is also what matters most to conmpanies who want to embed a browser; they are not going to use all of Mozilla anyway.
If someone gives me source code, and says that I have license to "use" it - to me, that means that I have the right to modify it, because that's one of the ways to use source code.
That is another valid interpretation. However, Darren's interpretation also seems reasonable. Since he wrote the code and since he added the explanation to clarify, I think one should take his intent and interpretation at face value and live with it. Saying to BSD contributors "hey, we want this code, and by some legal means we can use it whether you like it or not" sets a bad precedent, I think.
That's why a license should be explicit about such things and list the long litany of things you can do: "use, modify, redistribute, create derivative works,..." and why it is essential to spell out things in excrutiating detail before contributions to open source projects are accepted.
As a "regular company", the expectation is that Mandrake will turn a profit or go out of business. It doesn't make that much sense to donate money to them: if they turn a profit, that money will go to its stockholders or owners, and if they don't, they'll disappear.
There is also a practical problem with donating: it doesn't show up as sales or installed base, and it's hard for people to "buy" Mandrake that way for corporate use. At the very least, they should have a pro-forma bundle of Mandrake with some non-free software and sell a downloadable license.
Still, I think a community-supported organization is probably best done as a "not-for-profit". They are financially more transparent. It inspires a bit more trust that donations aren't just going to go to finance someone's sports car. It may even make donations tax deductible. In particular if Mandrake is in trouble, I think they should seriously consider this. Of course, it doesn't hold the promise of instant riches that an IPO does.
With gigabit interconnects (optical Firewire, Ethernet), there is less and less reason for the traditional "big box with slots". At those speeds, you can even run graphics cards and video capture. So, I think we can have it all: expandability, cheap off-the-shelf parts, convenience, looks, and compact designs.
I think Darren is right: there is no change. As I read it, his license only ever allowed "redistribution" and "use", not "modification". If the BSD project wanted to have the code fall under the more liberal BSD license, they should have made that a requirement before accepting it.
This seems like a fairly benign case: the code is self contained and can be removed if Darren doesn't change the license (and he is under no obligation). If this code were more integral to the system, it would potentially be a big problem. In fact, I wonder now whether the BSD project has really made sure in other, more critical cases that all contributions to the project really are made under the BSD copyright.
The FSF is picky about copyright assignments and licenses for a reason. Open source projects really need to pay attention to this or they put their whole user community at risk.
I think if you actually gave the Linux PDAs a try, you might think somewhat differently. The applications aren't quite as streamlined yet, but the platform and UI work like a charm. They could catch up with the Palm in a fraction of the time it took the Palm to get where it is now; all it takes is attracting a few core developers.
A standard unix startup includes device initialization, filesystem integrity checks, etc. and then login. In other words, it's designed to set up everything, then allow the user to work. In a handheld, I want to be able to work NOW.
Both the Palm and a Linux handheld "boot" after a hard reset. After that, they suspend when you turn off the power and come on instantly when you turn it back on again. Linux can and does initialize devices on the fly.
My Palm is terrific because it's small, quick, and streamlined. Data entry is simple, synchronization is simple, and it works well with my iBook, Windows PC, and Linux PC.
The AgendaVR is smaller than the Palm, it is as quick and responsive, it has multiple choices for data entry, and sit ynchronizes very easily. You don't even need any proprietary apps or protocols--it uses rsync.
the only benefit I see is (theoretically) easier development because a Unix programmer should be able to easily write code for a Linux-based PDA like the Yopy. But writing Palm code is already pretty simple,
I've done some porting to the AgendaVR. It took me 10 minutes to get a frame and SSL capable web browser with mouse support cross-compiled and running on the AgendaVR from its original UNIX sources. Other applications ported as easily. Some will benefit from slight modifications to their UI, but that's it: all the libraries and environment are there. Porting to Linux handhelds is orders of magnitude simpler than PalmOS or WinCE in my opinion.
Well if you had bothered to check the FAQ, you would have read the following statement.
No need to get all huffy about it. My point is simply that they shouldn't be bothering with "W" at all: they end up spending a lot of time trying to solve a problem that just doesn't exist. If eventually they offer XFree86, they'll have two window systems on the same platform. How is that for memory overhead, usability problems, and developer confusion?
And all that effort is invested because people mistakenly believe X11 is big and slow. A 200MHz RISC chip, as found in the Yopy, was the top of the line graphics workstation just a few years ago; X11 was developed for, and ran fine on, machines with less than 1/10th the power.
People frequently say: "Why would I want Linux on a handheld?" "Why bother with X11?" "Why bother with bash?" I think it's pretty clear that if you designed a handheld OS from scratch, it wouldn't quite look like Linux with X11 and a bunch of command line apps sitting around.
However, even something like the AgendaVR (66MHz MIPS) is powerful enough to run Linux 2.4 with XFree86 4.0 quite comfortably. And the benefits of having standard Linux running on it are significant. You may not want to actually use normal desktop linux applications or command line programs while tapping away on your machine. However, having a standard environment makes a lot of libraries and tools port very easily. And because you can log into your little handheld and even run X11 apps in both directions, debugging gets really eas.
For the AgendaVR, it took me maybe an hour to set up the cross-compilation environment, get PPP up and running, and read the documentation. After that, porting applications and tools to the AgendaVR was a snap: it took maybe 10 minutes to get a web browser and a scripting engine cross-compiled for the AgendaVR.
In that regard, I'm wondering whether the Yopy is doing the right thing by going with a different window system. If the AgendaVR can run XFree86 4.0 without problems, it should fly on the Yopy with its faster processor and bigger memory. I think the Yopy would be better off going with XFree86 4.0 as well. (And calling its window system the "W Window System" is nothing short of confusing, since X11's predecessor was called "W".)
Newsgroup charters are ignored for just about everything else, but all of a sudden they are supposed to be a legally binding document when it comes to copyrights? That doesn't wash, in particular since most people have never seen the newsgroup charter before posting.
In fact, the NYT will often explicitly verify for E-mail feedback whether it is OK to publish them.
Slashdot Mirror
you probably pay somewhere between $20 and $60 per month per machine in electricity no matter how fast or slow; a bunch of those pay for a new low-end machine quickly
you pay for that sort of thing in space (=bigger house) and power; the new systems are often smaller and more energy efficient. upgrading really might make sense...
So, Microsoft needs a security blanket in this new world, so they seem to be kicking around several ideas: we could charge for this, we could charge for that. Will it work? I kind of doubt it. If their infrastructure is truly open and well documented, people will offer competing services. If it is closed, it's just the status quo, and users are less and less fooled by it.
As for SOAP and XML that could go both ways. They help a little with interoperability, but Microsoft still has a wide range of options of making things proprietary and non-interoperable while nominally complying with the standards.
Hardware encryption itself is also both flawed and unnecessary. With hardware, you can't tell what bugs or backdoors may be in there, and if you discover anything, you can't fix it.
There are options that are cheaper, more secure, and more standard around. Current laptops can do real time speech compression and encryption just fine, with software that uses known strong algorithms and is demonstrably without backdoors. You can plug in any of the wireless PCMCIA cards in there and have secure phone conversations over the Internet, not just another ccompatible ell phone user. If you need something smaller, a WinCE or LinuxCE handheld with a cellular phone/modem CF card will probably be a realistic option pretty soon.
Indeed. Most web site compatibility problems actually come from using proprietary software and gimmicky features. You know, things like ASP, VBscript, frames, Flash, and that sort of stuff. The same kind of misfeatures are also responsible for delays.
They are most concerned with getting the site working with the majority of the web browsers and users out there. (Which just so happens to be IE 4/5 on Windows 95/98/NT/etc...) Given their timetables, it would not be feasable to start out supporting everyone and everything...
Sure it would be. If they stuck to plain HTML, not only would they be compatible with all browsers, they'd also be able to deliver faster, lower costs, and keep their site accessible to people with disabilities. Anything else is a waste of money for a government site.
That pretty much comes down to "use Microsoft Internet Explorer" on Windows or MacOS, and that's exactly what people are complaining about here.
The web site seems to just refuse to talk to people if they don't come from a Windows or MacOS machine. They could simply check for the certificates when they are needed, but they actually seem to check for the browser id string. What if Opera or Konqueror get the necessary certificates tomorrow? What if they already have them?
As for saying you can't do your tax return on-line in Netscape? Bollocks. It uses straight HTTPS. I've just completed mine. Did any of you think to check for yourselves? I doubt it, you just saw "it was written by Microsoft" and your knees starting jerking.
Maybe you just have a knee-jerk reaction when people criticize your favorite monopolist? You seem so far in the Windows world that you don't even notice the inconveniences Microsoft causes to competitors.
In all other cases, I should be able to furnish personal information on a case-by-case basis, and I should have the expectation that the company does not store the information beyond the business transaction we are engaged in unless I give my explicit permission.
I can't quite figure out why McNealy is so stupid about this. Giving people tools to protect private information is actually good for Sun's business: it benefits from lots of little smart devices that negotiate with other smart devices.
Perhaps McNealy isn't very sensitive to any of this because he is pretty well off. If you have sufficient amounts of money, you don't have to worry that much about privacy. For example, mere mortals have to file insurance claims, which distributes their records far and wide throughout the insurance industry. If you have enough money, you just pay out of pocket and only your doctor will know.
In any case, the Europeans will not fall for this "we'll just regulate ourselves" stuff. US companies better get used to conforming to some simple privacy rules or they won't continue to do business in Europe.
Unfortunately, it's everybody's problem. The money attracts more marketing and more product development. NT's problems get fixed by even higher-priced add-ons and alternatives disappear. The market is supposed to punish stupid companies like that, but if everybody's doing it, that doesn't help.
Linux and BSD are somewhat less susceptible to this because Microsoft can't just displace them: they get developed whether or not the developers make money with it. Consulting services for them, on the other hand, make money just like for NT.
Windows 2000 finally delivers the kind of kernel that is necessary to build Internet services. But Microsoft still falls way short on many fronts:
With Windows 2000, Microsoft has finally addressed one necessary requirement for even playing in this market. But there are lots of other necessities where they still fall short. And as long as Microsoft has notions of world domination and one system fits all on their mind, I don't think they can succeed. In fact, even if Microsoft were to focus, I don't think it makes much sense for the industry to commit to their system: why pay a lot of money for Microsoft's proprietary system if you can get something cheaper and vendor neutral for less already?
I still think that the Mozilla team would do wonders for their PR and their user community if they released platform specific browsers for UNIX, Windows, and MacOS based on just Gecko and a minimal UI around that. For Gtk+, that could be Galeon, and for Windows, there is already something like that in their tree. People would be awed by the speed and flexibility of Gecko, and most remaining serious bugs in Gecko would get fixed quickly.
Gecko is also what matters most to conmpanies who want to embed a browser; they are not going to use all of Mozilla anyway.
That is another valid interpretation. However, Darren's interpretation also seems reasonable. Since he wrote the code and since he added the explanation to clarify, I think one should take his intent and interpretation at face value and live with it. Saying to BSD contributors "hey, we want this code, and by some legal means we can use it whether you like it or not" sets a bad precedent, I think.
That's why a license should be explicit about such things and list the long litany of things you can do: "use, modify, redistribute, create derivative works, ..." and why it is essential to spell out things in excrutiating detail before contributions to open source projects are accepted.
There is also a practical problem with donating: it doesn't show up as sales or installed base, and it's hard for people to "buy" Mandrake that way for corporate use. At the very least, they should have a pro-forma bundle of Mandrake with some non-free software and sell a downloadable license.
Still, I think a community-supported organization is probably best done as a "not-for-profit". They are financially more transparent. It inspires a bit more trust that donations aren't just going to go to finance someone's sports car. It may even make donations tax deductible. In particular if Mandrake is in trouble, I think they should seriously consider this. Of course, it doesn't hold the promise of instant riches that an IPO does.
Dan Gillmore made a similar point in his article "Microsoft's dominance may just be beginning".
With gigabit interconnects (optical Firewire, Ethernet), there is less and less reason for the traditional "big box with slots". At those speeds, you can even run graphics cards and video capture. So, I think we can have it all: expandability, cheap off-the-shelf parts, convenience, looks, and compact designs.
This seems like a fairly benign case: the code is self contained and can be removed if Darren doesn't change the license (and he is under no obligation). If this code were more integral to the system, it would potentially be a big problem. In fact, I wonder now whether the BSD project has really made sure in other, more critical cases that all contributions to the project really are made under the BSD copyright.
The FSF is picky about copyright assignments and licenses for a reason. Open source projects really need to pay attention to this or they put their whole user community at risk.
A standard unix startup includes device initialization, filesystem integrity checks, etc. and then login. In other words, it's designed to set up everything, then allow the user to work. In a handheld, I want to be able to work NOW.
Both the Palm and a Linux handheld "boot" after a hard reset. After that, they suspend when you turn off the power and come on instantly when you turn it back on again. Linux can and does initialize devices on the fly.
My Palm is terrific because it's small, quick, and streamlined. Data entry is simple, synchronization is simple, and it works well with my iBook, Windows PC, and Linux PC.
The AgendaVR is smaller than the Palm, it is as quick and responsive, it has multiple choices for data entry, and sit ynchronizes very easily. You don't even need any proprietary apps or protocols--it uses rsync.
the only benefit I see is (theoretically) easier development because a Unix programmer should be able to easily write code for a Linux-based PDA like the Yopy. But writing Palm code is already pretty simple,
I've done some porting to the AgendaVR. It took me 10 minutes to get a frame and SSL capable web browser with mouse support cross-compiled and running on the AgendaVR from its original UNIX sources. Other applications ported as easily. Some will benefit from slight modifications to their UI, but that's it: all the libraries and environment are there. Porting to Linux handhelds is orders of magnitude simpler than PalmOS or WinCE in my opinion.
No need to get all huffy about it. My point is simply that they shouldn't be bothering with "W" at all: they end up spending a lot of time trying to solve a problem that just doesn't exist. If eventually they offer XFree86, they'll have two window systems on the same platform. How is that for memory overhead, usability problems, and developer confusion?
And all that effort is invested because people mistakenly believe X11 is big and slow. A 200MHz RISC chip, as found in the Yopy, was the top of the line graphics workstation just a few years ago; X11 was developed for, and ran fine on, machines with less than 1/10th the power.
However, even something like the AgendaVR (66MHz MIPS) is powerful enough to run Linux 2.4 with XFree86 4.0 quite comfortably. And the benefits of having standard Linux running on it are significant. You may not want to actually use normal desktop linux applications or command line programs while tapping away on your machine. However, having a standard environment makes a lot of libraries and tools port very easily. And because you can log into your little handheld and even run X11 apps in both directions, debugging gets really eas.
For the AgendaVR, it took me maybe an hour to set up the cross-compilation environment, get PPP up and running, and read the documentation. After that, porting applications and tools to the AgendaVR was a snap: it took maybe 10 minutes to get a web browser and a scripting engine cross-compiled for the AgendaVR.
In that regard, I'm wondering whether the Yopy is doing the right thing by going with a different window system. If the AgendaVR can run XFree86 4.0 without problems, it should fly on the Yopy with its faster processor and bigger memory. I think the Yopy would be better off going with XFree86 4.0 as well. (And calling its window system the "W Window System" is nothing short of confusing, since X11's predecessor was called "W".)
In fact, the NYT will often explicitly verify for E-mail feedback whether it is OK to publish them.
Since most USENET posts still aren't posted through Google, it seems like that's not much of a cover.