Statistics has been used for decades in handwriting input, OCR, speech recognition, systems like T9, and other input modalities. Dasher seems pretty cumbersome in comparison to most of those.
And the fact that it only generates "correct" input can be a real problem: names, foreign words, etc. just don't come out right.
Also, Raskin's complaints about Windows and OS X being similar could come down to other explanations: 1) convergent evolution or 2) Microsoft blatantly ripped off Apple in look and feel and continues to do so
There's a third explanation: neither of them is original and they are both slightly different commercial implementations of systems that have been around since the 1970's. If you go back and read the original ideas by people like Kay, you'll see that there is very little new on either the Mac or Windows.
I'm sure you could load all the important data (say 5kb - you don't need any presentation once you have the XUL loaded) in less than a second, which is certainly faster than desktop apps on a slower computer, plus no-startup time or install time.
My comment wasn't about the data, it was about where you put GUI code. Even if your ping times were typical delays for entire transactions, 170ms is too much of a delay for interactive response (and actual HTTP transactions are often much longer). If you want interactive response, you must put code on the client, and the only option for that right now is JavaScript.
The whole point is that you use both.
The point of my posting was to say that most application programmers probably aren't going to take any platform seriously that only has JavaScript as its client-side programming language, so if Firefox wants to go in that direction, it needs to do something. Allowing people to use Java or C# instead of Javascript for writing the client-side logic XUL apps might be a solution.
In fact, Java almost got it right: a simple, efficient, sandboxed client-side langage. What it was lacking originally was a good toolkit, but XUL fills that niche. And what it was lacking later was a commitment by Sun to keep it a client-side language, but Firefox could fill that niche.
Javascript is suprisingly powerful, but I agree not ideal for doing business logic in.
A good GUI tends to be a whole lot more complicated than the business logic behind it.
I'm not sure what that "No" is in response to. Are you saying that plain XUL without client-side programming is enough? I don't think it is: if you use XUL, you end up using JavaScript, extensively. With HTTP round-trip times of several seconds over real-world Internet connection, you just can't make a good UI by programming everything in PHP on the server.
This sounds a lot like late 90's,.com era speak to me.
Yes, and 5 years is hardly enough to realize a vision fully. It took hypertext, OOP, and Java-like languages nearly 40 years to make it into the mainstream.
I am using firefox to type up this comment, and yes it is a great browser, but it's not going to change the way the world does business.
So, the.com era speak has largely come true. Howver, current browser-based UIs are still pretty bad (like this fixed size text entry box). There is a lot of room for improvement there. An open, backwards-compatible platform giving people the ability to write much better web-based UIs could make all the difference. That might be Firefox or it might be something else.
What an apt metaphor: an intelligent, young, adventurous member of the species "homo sapiens" (Netscape) gets gored by a bunch of dumb, overweight beasts with sharp horns (Microsoft).
A lot more applications should have moved to the web over the last decade. Microsoft prevented that because they were not ready for it yet, even though the industry was. Instead, we got nearly another decade of poorly written VB, Office, and Access applications.
Firefox is a great browser, and there are a number of useful plug-ins available for it. It's also supported on many platforms.
But I have my doubts whether it's a good applications development platform as it is. Out of the box, you get, what, XUL and JavaScript? I'm sorry, but that doesn't strike me as a good platform for application development. In particular, JavaScript is just far too flaky to develop anything significant or complicated in it, and a lot of libraries just don't exist for JavaScript at all. And, like it or not, even if you put part of the application on the server, things still get complicated if you want a high quality GUI.
Maybe if Firefox shipped with a small, efficient JVM or CLR runtime and JIT that tie into the DOM, XUL, HTML, SVG, and event handlers (but without most of the bloated class libraries that Sun or Microsoft want to force on you), it could become a full platform. It would be even better if it included a small IDE out of the box.
As it is, I think it will remain limited to simple web apps created by rather dedicated Firefox hackers (and thank you for it, it is a great browser).
And how many people who live in the city of San Francisco cannot afford a computer? The median income is $74,000 per year.
That's probably because the median income only counts those who are actually employed. But San Francisco has a large population of unemployed, illegals, and/or homeless. Those people could be helped quite a bit by widespread and cheap Internet access.
Note that in those statistics median household income is only slightly above the national median, while the median income (i.e., the income of those employed) is considerably higher.
First, Sony sells both content and hardware, and their major business decisions seem to take into account the needs of their content business (look at their line of digital audio players). Second, the hardware manufacturers don't like the standard to be "as open as possible" either; they like to control who can enter the market and who cannot as well.
Are there any open standards (Dirac maybe?) that are at an advanced enough stage of development to be used as an alternative?
You've got to be kidding--the last thing the SMTPE and the motion picture industry wants is an open standard. They want something that is heavily patented because that gives them control. They just want the patent holders to be companies that can be pushed around by the content providers.
The ideal standard for the SMTPE would be something that is heavily patented, where the patents are held by labs and companies too small to make a business out of their own inventions, small companies that are happy with scraps and handouts from the motion picture industry.
Microsoft used shark-style tactics using his monopolistic power to get what it wanted and crush opposition... film at 11. Is this even news?
No, that's not news. What's news is that an important industry standards body noticed in time and is trying to prevent it.
Microsoft overpromised it seems, at least on the feature set. But cheated and lied?
I think if you "overpromise" in order to gain business advantages worth billions of dollars, that counts as "cheated and lied". In fact, it might count as "fraud".
Maybe we have gotten a little too jaded in this industry, but this kind of behavior should not be acceptable.
One-time passwords don't completely protect you from that either, since a compromised machine can intercept and alter what you type. So, among other things, it can install some kind of back door during your session.
That's the reason why there has to be patents on medications... without that IP-based protection, nobody would pay to do the research that creates new drugs.
The US public already pays a large chunk of US drug development, but it then also has to pay inflated prices for the patented drugs that come out of it. Overall, an economic analysis suggests that it would be cheaper to fund drug development entirely through taxes and make all new drugs generic.
It's an interesting dilema... if we pull out of funding the world's research, that research just isn't going to get done.
The main reason this research is being done preferentially in the US is because the US has chosen to pay a premium for the privilege, and that attracts companies and scientists. If the US changed its policy, the research wouldn't stop, it would simply move to other countries and end up being cheaper overall for everybody, including Americans.
In different words, the US "pulling out of funding the world's research" would be the best thing that could happen to the world's research; currently, the US has just been monopolizing the market through unfair indirect subsidies--"dumping" of research, if you will. It's no different from when other nations try to corner the milk or steel market by creating an unfair advantage there.
If everybody in the US stops paying for the research (by going to Mexico or Canada or whereever), there simply won't be much in the way of new drug development.
There may not be much in the way of commercial drug development, but that's not big loss because the market incentives for drug development aren't working: the drugs that companies have the highest incentive to develop are drugs that don't cure but ameliorate chronic conditions.
Drug development can be handled just fine through public funding. It already is to a significant degree anyway.
That's no argument. Classic and Carbon apps don't adhere to current Apple GUI guidelines either, yet Apple supports them.
The other reason is that, as someone before me has posted, there's absolutely no benefit to Apple for releasing their GUI elements to the public domain.
Of course there would be no benefit to Apple's commercial interests in shipping X11: their business model is obviously still to create a proprietary GUI with a captive developer community. That's the point.
When I actually want to get things done, Apple has a product that works out of the box and doesn't require me to poke around at arcane details.
That's not a question of software quality, it's a question of how you choose to buy your computers. If you buy a supported PC with Linux preinstalled, then Linux works "out of the box" as well.
I've used computers for twenty years at this point, and Linux is still something I have to wrestle with. I keep a partition to fool with, but it's becoming less and less attractive as I fiddle with the BSD substrate in OS X
It's a free country--you can keep on badmouthing and lying about Apple competitors, including Linux, if you like. But you are a fool if you think that kind of behavior is going to result in any kind of allegiance between Apple and FOSS.
Also, consider that Apple developed almost all of the GUI features that we now take for granted, including overlapping windows, pull-down menus, and drag-and-drop. They also pioneered playing movies and audio on PC's, shipping the first CD-ROM drive,
Apple developed none of those tings: they came variously from PARC, SRI, and AT&T. As for the "pioneering" audio, video, and the CD-ROM bit, I don't consider being first-to-ship products that other companies spent hundreds of millions of dollars to create a badge of honor.
in fact. I mean, Microsoft copies Apple, Linux/Gnome/KDE copy Microsoft, and then a Linux guy has the temerity to write an article about how Apple has never done anything for Linux?
That history is also completely wrong. Apple, Microsoft, and X11 all copied extensively from PARC GUIs. And X11 and X11 toolkits actually predate MS Windows and chunks of Windows were based on it, rather than the other way around.
Some people can't wait the required time for a doctor to become available, so they end up doing flying over to India to get it done. Universal health care isn't perfect either.
So, in the UK, everybody gets health care, but the wealthy can choose when and where to get it. In the US, a large chunk of the population doesn't get health care at all, and the wealthy still can choose when and where to get it. Seems like the UK is better in pretty much every way.
Here, you can find a complete list of Apple's ties to Open Souce.
Go through that list: of the things that Apple releases in open source form, none are a regular part of BSD or Linux installations. The open source Apple has released is open source mainly of interest to Apple developers. That's rather self-serving.
I'm sorry, but what? You're saying that OS X is a "relatively proprietary operating system?"
Big parts of OS X are proprietary, including the GUI. Furthermore, Apple is working hard to keep it that way. For example, while Apple could easily and smoothly integrate X11 into the system, in the same way they integrated Classic and Carbon, enabling fully integrated X11 apps, they steadfastly refuse to. When asked why they tell people to port their apps to their proprietary GUI, Cocoa.
The place that OS X is now is where Linux needs to be fast, stable, pretty, and usable. So far, the Linux community can manage three out of the four, but there are serious problems with the usability and appearance aspects. [...] I see Apple and the FOSS community as allies and not enemies,
I see. So, you think that for people like you to go around stating that Linux usability sucks makes Apple and the FOSS community "allies"? I don't think so. You're making enemies, not friends and allies, that way.
Actually, the FOSS community doesn't care about Apple much either way; they only care when Apple marketing or Apple fans try to discredit or disparage the quality of open source sofware, like you just did again.
You can fault Microsoft for many things, but not for not redesigning their software. They threw out their entire kernel and moved over to a new kernel, designed from scratch. That could be either good or it could be bad, depending on how well they designed their new system. I think it's pretty clear they didn't do such a good job.
OS X, for better or for worse, hasn't been redesigned that much. Under the covers, it's still pretty much the same NeXTStep system from the late 1980's.
Yes, and that's a logical conclusion: people here often have hands-on experience. When Microsoft comes out on top in such an evaluation, it contradicts their experience, and they usually quickly find the flaws in the analysis.
This report, on the other hand, is right on: it has lots of data, even using Microsoft's own metrics, and its analysis of the technical issues is thorough and accurate.
Slashdot Mirror
Current x86 chips (like the AMD) also have 64bit pointers.
Statistics has been used for decades in handwriting input, OCR, speech recognition, systems like T9, and other input modalities. Dasher seems pretty cumbersome in comparison to most of those.
And the fact that it only generates "correct" input can be a real problem: names, foreign words, etc. just don't come out right.
Also, Raskin's complaints about Windows and OS X being similar could come down to other explanations: 1) convergent evolution or 2) Microsoft blatantly ripped off Apple in look and feel and continues to do so
There's a third explanation: neither of them is original and they are both slightly different commercial implementations of systems that have been around since the 1970's. If you go back and read the original ideas by people like Kay, you'll see that there is very little new on either the Mac or Windows.
I'm sure you could load all the important data (say 5kb - you don't need any presentation once you have the XUL loaded) in less than a second, which is certainly faster than desktop apps on a slower computer, plus no-startup time or install time.
My comment wasn't about the data, it was about where you put GUI code. Even if your ping times were typical delays for entire transactions, 170ms is too much of a delay for interactive response (and actual HTTP transactions are often much longer). If you want interactive response, you must put code on the client, and the only option for that right now is JavaScript.
The whole point is that you use both.
The point of my posting was to say that most application programmers probably aren't going to take any platform seriously that only has JavaScript as its client-side programming language, so if Firefox wants to go in that direction, it needs to do something. Allowing people to use Java or C# instead of Javascript for writing the client-side logic XUL apps might be a solution.
In fact, Java almost got it right: a simple, efficient, sandboxed client-side langage. What it was lacking originally was a good toolkit, but XUL fills that niche. And what it was lacking later was a commitment by Sun to keep it a client-side language, but Firefox could fill that niche.
Javascript is suprisingly powerful, but I agree not ideal for doing business logic in.
A good GUI tends to be a whole lot more complicated than the business logic behind it.
I'm not sure what that "No" is in response to. Are you saying that plain XUL without client-side programming is enough? I don't think it is: if you use XUL, you end up using JavaScript, extensively. With HTTP round-trip times of several seconds over real-world Internet connection, you just can't make a good UI by programming everything in PHP on the server.
This sounds a lot like late 90's, .com era speak to me.
.com era speak has largely come true. Howver, current browser-based UIs are still pretty bad (like this fixed size text entry box). There is a lot of room for improvement there. An open, backwards-compatible platform giving people the ability to write much better web-based UIs could make all the difference. That might be Firefox or it might be something else.
Yes, and 5 years is hardly enough to realize a vision fully. It took hypertext, OOP, and Java-like languages nearly 40 years to make it into the mainstream.
I am using firefox to type up this comment, and yes it is a great browser, but it's not going to change the way the world does business.
So, the
What an apt metaphor: an intelligent, young, adventurous member of the species "homo sapiens" (Netscape) gets gored by a bunch of dumb, overweight beasts with sharp horns (Microsoft).
A lot more applications should have moved to the web over the last decade. Microsoft prevented that because they were not ready for it yet, even though the industry was. Instead, we got nearly another decade of poorly written VB, Office, and Access applications.
Firefox is a great browser, and there are a number of useful plug-ins available for it. It's also supported on many platforms.
But I have my doubts whether it's a good applications development platform as it is. Out of the box, you get, what, XUL and JavaScript? I'm sorry, but that doesn't strike me as a good platform for application development. In particular, JavaScript is just far too flaky to develop anything significant or complicated in it, and a lot of libraries just don't exist for JavaScript at all. And, like it or not, even if you put part of the application on the server, things still get complicated if you want a high quality GUI.
Maybe if Firefox shipped with a small, efficient JVM or CLR runtime and JIT that tie into the DOM, XUL, HTML, SVG, and event handlers (but without most of the bloated class libraries that Sun or Microsoft want to force on you), it could become a full platform. It would be even better if it included a small IDE out of the box.
As it is, I think it will remain limited to simple web apps created by rather dedicated Firefox hackers (and thank you for it, it is a great browser).
And how many people who live in the city of San Francisco cannot afford a computer? The median income is $74,000 per year.
That's probably because the median income only counts those who are actually employed. But San Francisco has a large population of unemployed, illegals, and/or homeless. Those people could be helped quite a bit by widespread and cheap Internet access.
Note that in those statistics median household income is only slightly above the national median, while the median income (i.e., the income of those employed) is considerably higher.
The median income is so high because there are so many people here with so much money.
If the median is $74000, it makes no difference whether the people above the median all make $75000 or $7.5m, the median will be unaffected.
First, Sony sells both content and hardware, and their major business decisions seem to take into account the needs of their content business (look at their line of digital audio players). Second, the hardware manufacturers don't like the standard to be "as open as possible" either; they like to control who can enter the market and who cannot as well.
There are lots of them, depending on what you mean by "starter":
Are there any open standards (Dirac maybe?) that are at an advanced enough stage of development to be used as an alternative?
You've got to be kidding--the last thing the SMTPE and the motion picture industry wants is an open standard. They want something that is heavily patented because that gives them control. They just want the patent holders to be companies that can be pushed around by the content providers.
The ideal standard for the SMTPE would be something that is heavily patented, where the patents are held by labs and companies too small to make a business out of their own inventions, small companies that are happy with scraps and handouts from the motion picture industry.
Microsoft used shark-style tactics using his monopolistic power to get what it wanted and crush opposition... film at 11. Is this even news?
No, that's not news. What's news is that an important industry standards body noticed in time and is trying to prevent it.
Microsoft overpromised it seems, at least on the feature set. But cheated and lied?
I think if you "overpromise" in order to gain business advantages worth billions of dollars, that counts as "cheated and lied". In fact, it might count as "fraud".
Maybe we have gotten a little too jaded in this industry, but this kind of behavior should not be acceptable.
One-time passwords don't completely protect you from that either, since a compromised machine can intercept and alter what you type. So, among other things, it can install some kind of back door during your session.
That's the reason why there has to be patents on medications... without that IP-based protection, nobody would pay to do the research that creates new drugs.
The US public already pays a large chunk of US drug development, but it then also has to pay inflated prices for the patented drugs that come out of it. Overall, an economic analysis suggests that it would be cheaper to fund drug development entirely through taxes and make all new drugs generic.
It's an interesting dilema... if we pull out of funding the world's research, that research just isn't going to get done.
The main reason this research is being done preferentially in the US is because the US has chosen to pay a premium for the privilege, and that attracts companies and scientists. If the US changed its policy, the research wouldn't stop, it would simply move to other countries and end up being cheaper overall for everybody, including Americans.
In different words, the US "pulling out of funding the world's research" would be the best thing that could happen to the world's research; currently, the US has just been monopolizing the market through unfair indirect subsidies--"dumping" of research, if you will. It's no different from when other nations try to corner the milk or steel market by creating an unfair advantage there.
If everybody in the US stops paying for the research (by going to Mexico or Canada or whereever), there simply won't be much in the way of new drug development.
There may not be much in the way of commercial drug development, but that's not big loss because the market incentives for drug development aren't working: the drugs that companies have the highest incentive to develop are drugs that don't cure but ameliorate chronic conditions.
Drug development can be handled just fine through public funding. It already is to a significant degree anyway.
I've used X11 and it doesn't exactly adhere.
That's no argument. Classic and Carbon apps don't adhere to current Apple GUI guidelines either, yet Apple supports them.
The other reason is that, as someone before me has posted, there's absolutely no benefit to Apple for releasing their GUI elements to the public domain.
Of course there would be no benefit to Apple's commercial interests in shipping X11: their business model is obviously still to create a proprietary GUI with a captive developer community. That's the point.
When I actually want to get things done, Apple has a product that works out of the box and doesn't require me to poke around at arcane details.
That's not a question of software quality, it's a question of how you choose to buy your computers. If you buy a supported PC with Linux preinstalled, then Linux works "out of the box" as well.
I've used computers for twenty years at this point, and Linux is still something I have to wrestle with. I keep a partition to fool with, but it's becoming less and less attractive as I fiddle with the BSD substrate in OS X
It's a free country--you can keep on badmouthing and lying about Apple competitors, including Linux, if you like. But you are a fool if you think that kind of behavior is going to result in any kind of allegiance between Apple and FOSS.
Also, consider that Apple developed almost all of the GUI features that we now take for granted, including overlapping windows, pull-down menus, and drag-and-drop. They also pioneered playing movies and audio on PC's, shipping the first CD-ROM drive,
Apple developed none of those tings: they came variously from PARC, SRI, and AT&T. As for the "pioneering" audio, video, and the CD-ROM bit, I don't consider being first-to-ship products that other companies spent hundreds of millions of dollars to create a badge of honor.
in fact. I mean, Microsoft copies Apple, Linux/Gnome/KDE copy Microsoft, and then a Linux guy has the temerity to write an article about how Apple has never done anything for Linux?
That history is also completely wrong. Apple, Microsoft, and X11 all copied extensively from PARC GUIs. And X11 and X11 toolkits actually predate MS Windows and chunks of Windows were based on it, rather than the other way around.
Some people can't wait the required time for a doctor to become available, so they end up doing flying over to India to get it done. Universal health care isn't perfect either.
So, in the UK, everybody gets health care, but the wealthy can choose when and where to get it. In the US, a large chunk of the population doesn't get health care at all, and the wealthy still can choose when and where to get it. Seems like the UK is better in pretty much every way.
Here, you can find a complete list of Apple's ties to Open Souce.
Go through that list: of the things that Apple releases in open source form, none are a regular part of BSD or Linux installations. The open source Apple has released is open source mainly of interest to Apple developers. That's rather self-serving.
I'm sorry, but what? You're saying that OS X is a "relatively proprietary operating system?"
Big parts of OS X are proprietary, including the GUI. Furthermore, Apple is working hard to keep it that way. For example, while Apple could easily and smoothly integrate X11 into the system, in the same way they integrated Classic and Carbon, enabling fully integrated X11 apps, they steadfastly refuse to. When asked why they tell people to port their apps to their proprietary GUI, Cocoa.
The place that OS X is now is where Linux needs to be fast, stable, pretty, and usable. So far, the Linux community can manage three out of the four, but there are serious problems with the usability and appearance aspects. [...] I see Apple and the FOSS community as allies and not enemies,
I see. So, you think that for people like you to go around stating that Linux usability sucks makes Apple and the FOSS community "allies"? I don't think so. You're making enemies, not friends and allies, that way.
Actually, the FOSS community doesn't care about Apple much either way; they only care when Apple marketing or Apple fans try to discredit or disparage the quality of open source sofware, like you just did again.
You can fault Microsoft for many things, but not for not redesigning their software. They threw out their entire kernel and moved over to a new kernel, designed from scratch. That could be either good or it could be bad, depending on how well they designed their new system. I think it's pretty clear they didn't do such a good job.
OS X, for better or for worse, hasn't been redesigned that much. Under the covers, it's still pretty much the same NeXTStep system from the late 1980's.
Yes, and that's a logical conclusion: people here often have hands-on experience. When Microsoft comes out on top in such an evaluation, it contradicts their experience, and they usually quickly find the flaws in the analysis.
This report, on the other hand, is right on: it has lots of data, even using Microsoft's own metrics, and its analysis of the technical issues is thorough and accurate.
The last time a US government did that was many decades ago, and even then, it wasn't exactly for selfless reasons.