One of the features that's been discussed as a long-term objective is for multiple user support in BeOS. The foundation for it exists (for instance, BeOS does possess RWX and file ownership concepts from the standard UNIX worlds, although given that it is currently a single user OS, most applications freely ignore it and the operating system isn't too heavy in enforcing them).
Whether the multiuser support will be more similar to Linux, in that multiple users can use multiple shells, or NT, where only a single user can be on the machine at a given time, but the OS supports multiple user contexts, is unknown (and the level of multiple user support that exists doesn't really give much of an indication).
Ages ago, Be had a pre-release version of a JDK running on their operating system using native threads. Apparently, Java ran *extremely* well using the BeOS native threading.
Assuming they aren't bound by NDAs, and that BeOS still possessed the code, documentation, know-how, or all three, there isn't any reason they simply couldn't adapt the techniques they used from the old JDK to the new one (with appropriate modifications, yes).
To upgrade from a dot-oh release (4.0) to a point release (4.x) = $0 To upgrade from a point release (4.x) to a point release (4.y) = $0 To upgrade from a dot-oh release (4.0) or a point release (4.x) to the next doh-oh release (5.0) = $25
I think you may have read his statement incorrectly - I know I did the first time through.
He says, "For example, BeOS could have little messages showing up saying that their OS won't work if Windows is installed on another partition (even though it certainly does, as win3.1 did with DR-DOS)..."
What he means is, "For example, BeOS could have little messages showing up saying that their OS won't work if Windows is installed on another partition (even though it certainly does work, as win3.1 did with DR-DOS)..."
*IIRC*, and I may not, I believe it was that legal briefs have to be a certain size or less, including footnotes. Wordperfect's word counter does include footnotes in the total count, but Microsoft Word does not.
In layman's terms, lawyers could abuse MS Word to file briefings that are actually longer than the law allows, putting others of their profession at a disadvantage. Rather than up the word count, the legal community has taken a rather dim view of those who try and circumvent (or perhaps just push a little) the limit.
From the article I read, apparently MS has been contacted about this issue. Whether they've since changed things, I don't know (this was before Office 2K was released, so it's possible this may be a moot point by now).
Well, I've always gotten the part about why it was a crummy (and not very intelligent) thing to do; I was just wondering why it's illegal, or at least civilly vulnerable.
Disclaimer: IANAL I believe the reason it's illegal is because it was artificial. Windows didn't require anything MS-DOS-specific in order to run; Microsoft merely implemented some checks and utilized some poorly-/un-documented tricks in order to prevent Windows 95 from operating on DR-DOS. Similarly, IE was hacked into the existing codebase of Windows for the "inseperable" defense they used in the DOJ trial.
If a product legitimately requires some underlying product to run, then I don't believe it's tying.
The whole Caldera case, in that light, is about whether Microsoft legitimately had a reason to glue MS-DOS and Windows 95 together, or whether they introduced some artificial dependencies/incompatabilities in order to destroy the market for other *-DOS operating systems. A similar theme was brought up in the DOJ trial with regards to IE, and whether IE has a legitimate reason to be so heavily "integrated" with Windows, or whether Microsoft resorted to some cheap hacks and word games to get a leg up on Netscape.
I used revolutionary because Win 2K is a *total* rewrite. It doesn't include ANY code from the old days. 95/NT4 was revolutionary as it isn't really a fancy dos shell, ala WFW. New paradigm, in a way.
W2K is not a total rewrite. NT is still the skeleton around which W2K is based. Yes, there have been architectural changes to the underlying structure, but it's still recognizably Windows NT. Even Microsoft admits it, as one of the marketing gimmicks for W2K is the slogan of, "Based On NT Technology!" There are going to be stickers on the W2K boxes hyping that fact.
W2K may be a great evolutionary step, but it is indeed only evolutionary. Added features and/or new code don't make something evolutionary; evolution, by definition, incorporates new features and ideas.
I surmise it's actually a little too complex for you.
While you are correct in stating that if you give away the source, you give away the program, there are a large proportion of people who don't wish to compile every single program on their own. If that was the case, distributions wouldn't exist as they do today.
People want the ability to download precompiled binaries and have them execute, and for many they personally have no desire to read the source code. Towards that end, the GPL gives no mandate that the source code *necessarily* be *included* along with the application being distributed in binary form. All it requires of the author/distributor is that they make the modified source available within reasonable limits (and those limits are enumerated in the GPL).
Red Hat could quite easily turn off their FTP servers. All that FTP provides is an efficient method of source distribution, more so than posting it via a hyperlink or requiring the user to send a request via email and receive a large attachment in return. But Red Hat is under no license-enforced obligation to distribute their code for free via FTP, nor do they have to include it on every CD they ship - all they have to do is provide free access to it via a reasonable method to anyone who asks for the source.
Maybe what you meant is that Red Hat can't *charge* for the source. In that case, I believe the GPL is in your favour.
The software doesn't have to be made available for free, as a careful reading of the GPL would enlighten you to. The GPL only states that if you write a piece of software and place it under the GPL, you have to release your source code, or if you make modifications to a piece of sotware that was GPL'd, then you have to release the source to your modifications.
The GPL says nothing about monetary value or having to provide it for free. Red Hat could shut down their FTP servers as long as they provided a method for people to get the source for the GPL components of Red Hat (which includes providing an email address to send requests for source). That Red Hat provides FTP servers is only out of loyalty and benevolence to the Linux community; they're under no legal obligation to make their distribution available for free.
Please, if I'm incorrect, point me to the relevant portion of the GPL which contradicts my statements.
According to the specifications, Java *always* uses big-endian byte ordering. It is in the documents; it's simpy buried somewhere (I don't remember where I found it, unfortunately, otherwise I'd provide a link).
I discovered this when trying to read a file formatted little-endian, and had to come up with (what I considered) to be some very clunky hacks to read the little-endian data correctly (reading it in backwards, then doing byte swaps). Doing that also pointed out how few unsigned types there actually are in Java. I wound up having to use the next size up (ints for short ints, etc.) in order to not have Java mangle the value of the data I was reading in.
I suspect it was done this way because of the platforms they developed on (I'm assuming Sparcs or similar) being big-endian as well. Then again, you know what they say about assumptions ("When you assume, you make an ass out of 'u' and 'me'.")
Why do you consider everything written on MSN (even marked OPINION) to be Microsoft's position, rather than a personal opinion?
It's not marked opinion, and it's not put forth as one person's perspective. It's Microsoft's web site (hence the 'MS' portion of 'MSN'), and if it was not representing an official position of Microsoft Corp. it should have some sort of disclaimer in the article that the author does not speak for them.
That such statements are blatantly lacking from the page implies that the author is, in some capacity, representing Microsoft's views.
After stealthily hacking into Microsoft's computer network, I was able to download a file containing some of the PR documents Microsoft is working on:
BeOS Myths ("Despite claims by advocates that BeOS is a superior multimedia platform, Windows offers preemptive multithreading, a real microkernel architecture, and a commercial-quality file system with true journalling capability.")
OS/2 Warp Myths ("Some people think OS/2 Warp is still viable on servers, despite it's reliance on five-year-old technology.")
PalmOS Myths ("Just because it has six times the market doesn't make it any less inferior to WinCE!")
GCC Myths ("GCC lacks an extensive GUI and RAD capability to design high-powered graphical business applications.")
Non-Microsoft Application Compatability Myths ("Just because a program says it's compatible doesn't mean it is. We go to great lengths to make sure our applications support our file formats completely, and you can only be guaranteed to read Microsoft files correctly if you use Microsoft products.")
Open Standards Myths ("Proprietary standards actually increase competition, and we all know how important competition is in today's ultra-competitive marketplace that is focused on competition, especially given that Microsoft is locked in competitive battles with competing competitors.")
Hardware Support Myths ("The great prevalence of drivers for Windows-based operating systems is because Windows allows for the most complete and active use of the hardware device; Linux' command line, for example, only uses a fraction of the power of modern video cards.")
Great, fine, wonderful. The statement isn't directly attributable to Microsoft.
It's still absurdly wrong. Take two brand new computer users, assign one a manual for Windows NT and the task to install and configure the latest stable iteration of Windows NT, and the other a manual for BeOS and the task to install and configure the latest stable iteration of BeOS.
What the author of the article is trying to tell me is that the latter user (installing BeOS) will have a more difficult time configuring his operating system than the former user (installing NT), and that contention is just plain wrong.
As for the jabbiness of Slashdot, this is ostensibly a Linux forum, even if it preaches simply "news for nerds". If you frequented a die-hard Windows NT chat forum, I'm sure not of us would be surprised at the amount of pro-Windows, anti-Linux advocacy.
Please. BeOS configures at least twice as easy as Windows NT does now. Networking configuration is abysmal in NT, whereas in BeOS it's pretty much one-stop shopping. It took me about fifteen minutes to install drivers and set them up correctly for my system under Windows; under BeOS, it took me all of two not only to set it up, but to get my dynamic IP and visit a couple of my favorite web sites.
As for the Linux crack, NT may be "easier," but what it lacks in luser-friendliness it makes up for in raw power and flexibility. Further, for any competent computer user, Linux actually isn't significantly harder than NT to set up.
I suppose I should be grateful; now the PR arm of Microsoft is viewing not only Linux, but BeOS, an operating system not even competing with Windows NT, never mind competing in the same class, and OS/2, an arguably dead operating system that is no longer supported by their parent, as targets for their FUD.
The obvious answer, then, is to genetically breed-out the genes that correlate between insanity and intelligence.
We're talking arbitrarily infinite knowledge here, in other words. At some unspecified point in time, humanity will have completely mapped out and understood the human genome, and that includes such things as linking factors between insanity, intelligence, obesity, laziness, athleticism, and so on. Given that arbitrary knowledge, not only can you install someone with any of those traits, but you can also unlink them with the negative factors.
How arbitrary "arbitrary" actually is, however, is the real crapshoot. How long until we possess enough knowledge that we can understand the interaction between genes? Mapping things out ("Hey, xy124ht controls I.Q.!") is going to be the easy part compared to trying to untangle the weave of interactions between genes.
Slashdot Mirror
One of the features that's been discussed as a long-term objective is for multiple user support in BeOS. The foundation for it exists (for instance, BeOS does possess RWX and file ownership concepts from the standard UNIX worlds, although given that it is currently a single user OS, most applications freely ignore it and the operating system isn't too heavy in enforcing them).
Whether the multiuser support will be more similar to Linux, in that multiple users can use multiple shells, or NT, where only a single user can be on the machine at a given time, but the OS supports multiple user contexts, is unknown (and the level of multiple user support that exists doesn't really give much of an indication).
Ages ago, Be had a pre-release version of a JDK running on their operating system using native threads. Apparently, Java ran *extremely* well using the BeOS native threading.
Assuming they aren't bound by NDAs, and that BeOS still possessed the code, documentation, know-how, or all three, there isn't any reason they simply couldn't adapt the techniques they used from the old JDK to the new one (with appropriate modifications, yes).
Upgrading works like this:
To upgrade from a dot-oh release (4.0) to a point release (4.x) = $0
To upgrade from a point release (4.x) to a point release (4.y) = $0
To upgrade from a dot-oh release (4.0) or a point release (4.x) to the next doh-oh release (5.0) = $25
I think you may have read his statement incorrectly - I know I did the first time through.
He says, "For example, BeOS could have little messages showing up saying that their OS won't work if Windows is installed on another partition (even though it certainly does, as win3.1 did with DR-DOS)..."
What he means is, "For example, BeOS could have little messages showing up saying that their OS won't work if Windows is installed on another partition (even though it certainly does work, as win3.1 did with DR-DOS)..."
Does that make things clearer?
*IIRC*, and I may not, I believe it was that legal briefs have to be a certain size or less, including footnotes. Wordperfect's word counter does include footnotes in the total count, but Microsoft Word does not.
In layman's terms, lawyers could abuse MS Word to file briefings that are actually longer than the law allows, putting others of their profession at a disadvantage. Rather than up the word count, the legal community has taken a rather dim view of those who try and circumvent (or perhaps just push a little) the limit.
From the article I read, apparently MS has been contacted about this issue. Whether they've since changed things, I don't know (this was before Office 2K was released, so it's possible this may be a moot point by now).
Three things spring immediately to mind:
- venetian blinds
- window shade
- curtains
:)
Ack!
(I think I'll just stop, now that I've made an ass out of myself...)
Pardon the boldface, please.
I must've accidentally typed a instead of a
when doing some formatting.
*sigh*
Well, I've always gotten the part about why it was a crummy (and not very intelligent) thing to do; I was just wondering why it's illegal, or at least civilly vulnerable.
Disclaimer: IANAL
I believe the reason it's illegal is because it was artificial. Windows didn't require anything MS-DOS-specific in order to run; Microsoft merely implemented some checks and utilized some poorly-/un-documented tricks in order to prevent Windows 95 from operating on DR-DOS. Similarly, IE was hacked into the existing codebase of Windows for the "inseperable" defense they used in the DOJ trial.
If a product legitimately requires some underlying product to run, then I don't believe it's tying.
The whole Caldera case, in that light, is about whether Microsoft legitimately had a reason to glue MS-DOS and Windows 95 together, or whether they introduced some artificial dependencies/incompatabilities in order to destroy the market for other *-DOS operating systems. A similar theme was brought up in the DOJ trial with regards to IE, and whether IE has a legitimate reason to be so heavily "integrated" with Windows, or whether Microsoft resorted to some cheap hacks and word games to get a leg up on Netscape.
I used revolutionary because Win 2K is a *total* rewrite. It doesn't include ANY code from the old days. 95/NT4 was revolutionary as it isn't really a fancy dos shell, ala WFW. New paradigm, in a way.
W2K is not a total rewrite. NT is still the skeleton around which W2K is based. Yes, there have been architectural changes to the underlying structure, but it's still recognizably Windows NT. Even Microsoft admits it, as one of the marketing gimmicks for W2K is the slogan of, "Based On NT Technology!" There are going to be stickers on the W2K boxes hyping that fact.
W2K may be a great evolutionary step, but it is indeed only evolutionary. Added features and/or new code don't make something evolutionary; evolution, by definition, incorporates new features and ideas.
By that logic, I count for... eight Linux installations:
:)
- Slackware 7.0 (preordered, but ordered nonetheless)
- Slackware 4.0
- Slackware 3.6
- Slackware 95
- Stampede 0.86
- Debian 2.0
- Red Hat 5.1
- Red Hat 4.2
Woohoo!
I surmise it's actually a little too complex for you.
While you are correct in stating that if you give away the source, you give away the program, there are a large proportion of people who don't wish to compile every single program on their own. If that was the case, distributions wouldn't exist as they do today.
People want the ability to download precompiled binaries and have them execute, and for many they personally have no desire to read the source code. Towards that end, the GPL gives no mandate that the source code *necessarily* be *included* along with the application being distributed in binary form. All it requires of the author/distributor is that they make the modified source available within reasonable limits (and those limits are enumerated in the GPL).
Red Hat could quite easily turn off their FTP servers. All that FTP provides is an efficient method of source distribution, more so than posting it via a hyperlink or requiring the user to send a request via email and receive a large attachment in return. But Red Hat is under no license-enforced obligation to distribute their code for free via FTP, nor do they have to include it on every CD they ship - all they have to do is provide free access to it via a reasonable method to anyone who asks for the source.
Maybe what you meant is that Red Hat can't *charge* for the source. In that case, I believe the GPL is in your favour.
The software doesn't have to be made available for free, as a careful reading of the GPL would enlighten you to. The GPL only states that if you write a piece of software and place it under the GPL, you have to release your source code, or if you make modifications to a piece of sotware that was GPL'd, then you have to release the source to your modifications.
The GPL says nothing about monetary value or having to provide it for free. Red Hat could shut down their FTP servers as long as they provided a method for people to get the source for the GPL components of Red Hat (which includes providing an email address to send requests for source). That Red Hat provides FTP servers is only out of loyalty and benevolence to the Linux community; they're under no legal obligation to make their distribution available for free.
Please, if I'm incorrect, point me to the relevant portion of the GPL which contradicts my statements.
>I'll be happy to tell you how to improve performance.
:)
Could you share some of your tidbits, please?
According to the specifications, Java *always* uses big-endian byte ordering. It is in the documents; it's simpy buried somewhere (I don't remember where I found it, unfortunately, otherwise I'd provide a link).
I discovered this when trying to read a file formatted little-endian, and had to come up with (what I considered) to be some very clunky hacks to read the little-endian data correctly (reading it in backwards, then doing byte swaps). Doing that also pointed out how few unsigned types there actually are in Java. I wound up having to use the next size up (ints for short ints, etc.) in order to not have Java mangle the value of the data I was reading in.
I suspect it was done this way because of the platforms they developed on (I'm assuming Sparcs or similar) being big-endian as well. Then again, you know what they say about assumptions ("When you assume, you make an ass out of 'u' and 'me'.")
IANAE*, but I believe the one-time pad method is damn close to unbreakable.
[*] I Am Not An Expert
Um, whoever marked this 'Insightful,' I think it's relatively clear that the author was indeed being sarcastic.
Why do you consider everything written on MSN (even marked OPINION) to be Microsoft's position, rather than a personal opinion?
It's not marked opinion, and it's not put forth as one person's perspective. It's Microsoft's web site (hence the 'MS' portion of 'MSN'), and if it was not representing an official position of Microsoft Corp. it should have some sort of disclaimer in the article that the author does not speak for them.
That such statements are blatantly lacking from the page implies that the author is, in some capacity, representing Microsoft's views.
After stealthily hacking into Microsoft's computer network, I was able to download a file containing some of the PR documents Microsoft is working on:
Prepare for the FUD-storm, people!
Great, fine, wonderful. The statement isn't directly attributable to Microsoft.
It's still absurdly wrong. Take two brand new computer users, assign one a manual for Windows NT and the task to install and configure the latest stable iteration of Windows NT, and the other a manual for BeOS and the task to install and configure the latest stable iteration of BeOS.
What the author of the article is trying to tell me is that the latter user (installing BeOS) will have a more difficult time configuring his operating system than the former user (installing NT), and that contention is just plain wrong.
As for the jabbiness of Slashdot, this is ostensibly a Linux forum, even if it preaches simply "news for nerds". If you frequented a die-hard Windows NT chat forum, I'm sure not of us would be surprised at the amount of pro-Windows, anti-Linux advocacy.
Maybe it's a /really/ good password? :)
The one that's already been configured for them? ("Wow, I didn't have to touch a thing!")
Please. BeOS configures at least twice as easy as Windows NT does now. Networking configuration is abysmal in NT, whereas in BeOS it's pretty much one-stop shopping. It took me about fifteen minutes to install drivers and set them up correctly for my system under Windows; under BeOS, it took me all of two not only to set it up, but to get my dynamic IP and visit a couple of my favorite web sites.
As for the Linux crack, NT may be "easier," but what it lacks in luser-friendliness it makes up for in raw power and flexibility. Further, for any competent computer user, Linux actually isn't significantly harder than NT to set up.
I suppose I should be grateful; now the PR arm of Microsoft is viewing not only Linux, but BeOS, an operating system not even competing with Windows NT, never mind competing in the same class, and OS/2, an arguably dead operating system that is no longer supported by their parent, as targets for their FUD.
Microsoft must be running scared.
Nitpick: From the child's vantage point, it would indeed by downloading. So perhaps it's more of a perspective issue than a semantic one.
The obvious answer, then, is to genetically breed-out the genes that correlate between insanity and intelligence.
We're talking arbitrarily infinite knowledge here, in other words. At some unspecified point in time, humanity will have completely mapped out and understood the human genome, and that includes such things as linking factors between insanity, intelligence, obesity, laziness, athleticism, and so on. Given that arbitrary knowledge, not only can you install someone with any of those traits, but you can also unlink them with the negative factors.
How arbitrary "arbitrary" actually is, however, is the real crapshoot. How long until we possess enough knowledge that we can understand the interaction between genes? Mapping things out ("Hey, xy124ht controls I.Q.!") is going to be the easy part compared to trying to untangle the weave of interactions between genes.