Yes, cable might only be available in the capitals, but, if you can get it, it's very good quality connection (reliable 128kbit upstream/400kbit downstream, ~20 ms pings to most Australian sites, and ~300ms pings to the US), and by world standards it's not badly priced either (60 AUD/month - that's about 35 USD).
I don't like Telstra much either, but this isn't a bad service.
For one thing, the data was medical telemetry. If this goes down, it's no big deal - in fact, if the book about Apollo 13 is anything to go by, astronauts regard the things as an instrusive and irrelevant distraction. For another thing, the data was still received through "alternate methods".
Not that NASA shouldn't be concerned - any denial of service is always a concern, particularly for anything related to flight control - but claiming this was an "emergency" is debasing the term.
konstant works for Microsoft (and, when making pro-Microsoft posts in comments includes that fact in his.sig).
This is not a criticism, it's just something to keep in mind when you read his take on things - just as it is when any employee of a company publicly comments on something directly involving that company's business).
OK, so there is Debian. So I happen to be one of the "unfortunate" KDE users. I have read the Slashdot article the other day, I fully understand the legal issues, but I happen to still don't understand why an inclusion in non-free seems to be a non-option.
It's not an issue of freeness or non-freeness. It's an issue that, in the Debian's opinion, that the unadorned GPL and the QPL make it illegal to legally distribute KDE binaries linked to Qt. Free or non-free is irrelevant.
Oh, and if you want KDE for Debian, another poster has pointed out how to get it. It's extremely simple to do.
the man is talking about formally-specified and verified systems, where everything is specified in some mathematically-based specification language, and then the implementation is demonstrated to have desirable properties in relation to the specification. Chaotic, bazaar-style development doesn't match with this style of development at all.
However, this kind of formally verified system is extremely costly to develop, extremely difficult to adapt to changing circumstances (and retain the verified properties), and still doesn't guarantee that it does what you want it to do - mistakes in the specification or mistakes in the verification process are just as likely as mistakes in coding.
Frankly, for 99.9% of the software written in the world, this kind of thing is utterly impractical and will remain so. I don't mind consigning the remaining 0.1% to cathedral-style approaches (though open source can still help spot bugs that the verification doesn't catch).
Re:Address space is going to kill off the x86
on
Is The x86 Obsolete?
·
· Score: 2
Yes, the x86 might survive as a backwards compatibility mode, but any *new* 64-bit is not going to run natively on old machines.
If you're going to have to change ISA, and cope with all the nuisances that entails, why wouldn't you swap to the one offering the very best price/performance compromise? As far as backwards compatibility goes, you can run x86 code on the Intel/HP IA-64, and, if it comes down to it, on the PPC and Alpha through emulation. What's the special attraction of the Sledgehammer?
Re:Address space is going to kill off the x86
on
Is The x86 Obsolete?
·
· Score: 2
The 4004 was a 4-bit microprocessor which was not binary compatible with *anything* else. The 8008 and 8080 were all 8-bit processors which were *not* binary compatible with the 8086.
The 286 ISA was a superset of the 8086 - any code that used protected mode was *not* backwards compatible. Ditto the 386 - it added new modes that were not backwards compatible. However, the 386 ISA has stayed mostly unchanged (notable exceptions include MMX, 3DNOW, and whatever Intel's latest hack is called) through the days of the 486, Pentium, PPro, PII, and PIII, as well as the Cyrix and AMD equivalents.
Yes, the weird-ass segmented addressing modes exist, but I haven't seen anybody show any enthusiasm for trying to *use* them.
AMD's Sledgehammer proposal might successfully extend the x86 architecture to support a true 64-bit address space. It might be a horrible flop. Whichever way it goes, Sledgehammer code will *not* run on anything other than a Sledgehammer processor. Ergo, Sledgehammer is a new ISA, related though it may be to the original x86 one.
Patterson and Hennessy make the point in their seminal book - no architecture has survived an address space crunch, which arrives on x86 around 2-4 GB of memory (and is already biting with the Linux limitations on file size). Typical desktops are still a few generations away from this amount of virtual memory, let alone physical memory, but servers are getting to the point where this poses a serious limitation. Ergo, Intel is providing the IA-64, which will remove the limitation (modulo some PCI problems with >2^32 byte address space, IIRC - have they been fixed yet?).
Once this becomes a more widespread problem, the x86 architecture, in its present form, is doomed. At that point, what the industry will converge on (and whether it will converge at all) is an open question.
Bubble sort actually kicks butt for mostly-sorted lists. If you have a list which you
know is almost in order but couldn't be bothered to apply one of the better general algorithms to it, use bubble sort.
Bubble sort is better than a general-purpose sort on a mostly-sorted list, so, if you care about performance, you would always use bubble sort on mostly-sorted lists in preference to quicksort, mergesort or radix sort. Actually, though, insertion sort is better again in this kind of situation, so use it instead of bubble sort.
A guy on my department has written about these - the basic problem is that if somebody does figure out a way to imitate your hand/signature/retinal characteristics (and, remember, they can get access to the data because the whole data has to be stored for comparison, not just a signature) you're in trouble. It's rather difficult to get a new, non-compromised retina:)
I have seen a theory mentioned previously (can't remember where) that went something like this:
Microsoft doesn't really care about the specific result of the trial, it's just trying to delay a resolution until George Bush Jr. gets elected. Bush will intervene and stop the DOJ's pursuit of Microsoft.
Could somebody comment on the
Likelihood that this is Microsoft's strategy
That, assuming that Bush is elected (which, from an outsider's perspective, looks at least as likely as not), that Bush will a) want to, and b) be able to, achieve a resolution that keeps Microsoft together and "innovating".
The QPL is open-source (and RMS-free, I believe), but it's licence is considerably more restrictive than the GPL.
KDE is licenced under plain GPL.
Amongst other things, the GPL says that unless it's a system library, if you link something to GPL'd code you've got to make it available under the GPL.
Therefore, the KDE people would have to make Qt available under the terms of the GPL
This can't happen, as the GPL permits things that are expressly forbidden under the QPL
Therefore, KDE programs (particularly KDE binaries) areillegal to distribute at all!
There are several potential solutions:
Stick your head in the sand and hope the problem stays away as noone is likely to sue. This is indeed a practical solution, and it's one that most of the commercial distros have chosen. Debian, however, takes its free software ideals extremely seriously, so they simply won't distribute KDE.
Troll Tech release Qt under the GPL, LGPL, or the BSD licence. If it was going to happen, it would have already happened, and, from a business perspective, the Trolls do have sound commercial reasons for not doing so (they sell a different version that has different licencing terms and runs on Windows, and the practical effect of the QPL is making a Windows port of the existing code very difficult). The only way this is likely to happen is if some good samaritan buys out Troll Tech (IBM, some other benevolent multinational . . . are you listening - it'd be pocket change to you guys!)
Finally, the solution that the good doctor is advocating - that KDE changes its licence to specifically allow linking against Qt without Qt being affected by the GPL. This requires them to get permission from everyone who has contributed a non-trivial patch, as well as any code that they have borrowed from any other GPL software - as well as acknowledge that there is a problem, which they really don't want to do (this is understandable, it's human nature)
Or do you think there will ever be a truly unbiased, trusted source (perhaps like the way the media should be) where specific information about tax cut proposals and so forth will be located?
Do you really think such a thing is possible, even with the best of intentions? How many trusted, unbiased sources of advice for choosing a Linux distribution do you know:-)
What exactly are KDE and Gnome? Are they a combination of the GUI toolkit and a collection of programs like the toolbar and Gnumeric and such?
Pretty much. They consist of a set of libraries, including a widget library, inter-application communication, and other useful utilities like printing, a set of utilities like file managers, and applications which make use of their facilities.
Where do they stand in relation to X itself and a window manager?
They stand above the X protocol, and therefore machines using either have all the nice networking abilities of X. All existing X applications run fine under either system.
Any window manager will work with either KDE or GNOME, but there are some window managers which have extra capabilities which make them work better with the systems. KDE ships with its own window manager, kwm, and GNOME ships with Sawfish, but GNOME has shipped with Enlightenment in the past.
Is there any real difference between KDE and Gnome, or are they just two different products filling the same niche (like IE vs. Netscape)?
From a user's perspective, there's not a great deal. Frankly, until the office suites get closer to release, the decision isn't all that important.
From an ideological perspective, Qt (the base toolkit on which KDE is derived) is under a licence which is free enough to be regarded as open-source and RMS-free, but is a real pain as it is viewed by many (including the Debian project) as being GPL-incompatible. Therefore, in the view of Debian, KDE violates its own licence! The KDE people, however, disagree.
From a programmer's perspective, the toolkits are different - Qt/KDE is based around C++, and Gtk+/GNOME are based around C. Up until recently, the C++ support for GNOME has been regarded as pretty bad, and the C support from KDE equally so. They have made some other differing architectural decisions - GNOME is using CORBA to support object embedding, where KDE is using their own protocol.
Finally, I'd just like to point out that GNOME apps run fine on a KDE desktop and vice versa. So, provided you have enough memory to keep both widget sets at once without thrashing, choosing one doesn't mean you've lost the other.
Disclaimer: I use and develop with GNOME, but have no experience with programming KDE apps.
You're absolutely right - this bug is juvenile. As I understand it, that was the point - Heinlein wrote it as a plea for the young men of the time to understand the need for people to place their lives on the line in defence of their fellow citizens, presumably concerned about the early signs of the social changes of the 60's.
However, in Paul Verhoeven's (the movie director) opinion, and my own, along the way he advocates fascist military government, and consequently the movie was a brutal satirisation of the book.
I had an interesting discussion with another Slashdot reader about the book a few months ago on whether Heinlein really intended to advocate a fascist political system in the book or not. After an exchange of fascinating e-mails, we agreed to disagree on the book.
I still think it's well worth a read now, just to decide for yourself what he was really on about.
He's been investigating the possibility of "information agents" to to more intelligent searching of the WWW (for instance, a search tool specifically for finding university staff homepages). They have figured out ways to use the structure of typical web sites to improve the search over keyword matching.
Check out the Agents Group for this and other projects.
If you really want to go up against a government, you're going to need some modern weaponry - even guerilla armies need mines, mortars, grenades, and semtex - none of which are available to the general public. Are you seriously arguing that everyone should be allowed to own this stuff? How about tanks, warplanes, missiles - for that matter what about a nuke?
However, if you want to organise a bit of civil disobedience, what you need are enough bodies on the street - armed or unarmed, it makes little difference. If the government concerned doesn't fold, weapons smuggling ain't that difficult.
Oh, it's just you that feels the need to defend yourself from the government in a bad mood? All an assault rifle is likely to do is make the cops more likely to kill you in the process of arresting you.
So, IMHO, anyone who believes that widespread posession of guns makes any difference to the potential for governmental tyranny, or prevents resistance of that tyranny, is kidding themselves.
Not many people here seem to realize it, but out on the water extra GPS accuracy could well save lives, particularly in "man overboard" situations. It's quite disconcerting to be looking at a GPS chartplotter and watching your GPS plot your position as 100 yards or so east of a reef, when you know perfectly well you're west of it, and you see the little "SA" graphic flashing off in the corner of the GPS set.
Thank you, whomever was responsible for this decision.
I'm really glad to see this. In my experience, the great flaw in the OSS model is the quality of the code. Can we be honest? The vast majority of it is complete crap, developed by amateurs with absolutely no clue how develop to professional standards.
I disagree. There is crap open-source stuff out there, but overall the coding standard is really not too bad, and some is excellent. The important stuff tends towards the excellent (at least from my experience).
By contrast, go and have a play with arbitrary pieces of Windows shareware . . .
However, even (relatively) low-powered components still generate heat, and packing components in more and more densely creates a double whammy by concentrating the heat from all the components in a smaller space and leaving less room for air circulation to cool them. Water-cooling might just be a space-saving measure, not a way to run hyper-fast CPUs.
Total floating point performance: 6.2GFlops at 300Mhz, or roughly equivalent to a 1.5Ghz Athlon.
Yeah, so it does kick-arse floating point, but the performance less than a factor of two better than what a bog-standard Athlon - provided you can write code that uses the chip to its full potential (and, given the brief description above, that's probasbly quite a challenge).
In any case, floating point is totally irrelevant for code-cracking, which is the basic reason governments restrict supercomputers.
Send polite letters to each of your congressional representatives, and to each of your local school board members.
That's a fair first step, but what if the system is still introduced and geeks, goths, homosexuals, chess-club members, and other different-but-harmless students start getting harassed or suspended due to the WAVE program, and the school board still wasn't listening?
I don't know about you, but I'd be prepared to do something wrong if I believed that the alternative would be a greater wrong.
So, yes, I suppose it's possible that some guy at the NSA invented a way to factor the numbers, but then again, are your communications something that the NSA would really be interested in? Somehow, I doubt mine are.
No, but if you're a European company doing export business with Asia, or a NGO like Amnesty International, the NSA almost certainly do try to read your mail. There's plenty of people who have real, concrete and current needs to make their mail NSA-proof (or at least NSA-resistant).
Slashdot Mirror
I don't like Telstra much either, but this isn't a bad service.
Not that NASA shouldn't be concerned - any denial of service is always a concern, particularly for anything related to flight control - but claiming this was an "emergency" is debasing the term.
This is not a criticism, it's just something to keep in mind when you read his take on things - just as it is when any employee of a company publicly comments on something directly involving that company's business).
It's not an issue of freeness or non-freeness. It's an issue that, in the Debian's opinion, that the unadorned GPL and the QPL make it illegal to legally distribute KDE binaries linked to Qt. Free or non-free is irrelevant.
Oh, and if you want KDE for Debian, another poster has pointed out how to get it. It's extremely simple to do.
However, this kind of formally verified system is extremely costly to develop, extremely difficult to adapt to changing circumstances (and retain the verified properties), and still doesn't guarantee that it does what you want it to do - mistakes in the specification or mistakes in the verification process are just as likely as mistakes in coding.
Frankly, for 99.9% of the software written in the world, this kind of thing is utterly impractical and will remain so. I don't mind consigning the remaining 0.1% to cathedral-style approaches (though open source can still help spot bugs that the verification doesn't catch).
If you're going to have to change ISA, and cope with all the nuisances that entails, why wouldn't you swap to the one offering the very best price/performance compromise? As far as backwards compatibility goes, you can run x86 code on the Intel/HP IA-64, and, if it comes down to it, on the PPC and Alpha through emulation. What's the special attraction of the Sledgehammer?
The 286 ISA was a superset of the 8086 - any code that used protected mode was *not* backwards compatible. Ditto the 386 - it added new modes that were not backwards compatible. However, the 386 ISA has stayed mostly unchanged (notable exceptions include MMX, 3DNOW, and whatever Intel's latest hack is called) through the days of the 486, Pentium, PPro, PII, and PIII, as well as the Cyrix and AMD equivalents.
Yes, the weird-ass segmented addressing modes exist, but I haven't seen anybody show any enthusiasm for trying to *use* them.
AMD's Sledgehammer proposal might successfully extend the x86 architecture to support a true 64-bit address space. It might be a horrible flop. Whichever way it goes, Sledgehammer code will *not* run on anything other than a Sledgehammer processor. Ergo, Sledgehammer is a new ISA, related though it may be to the original x86 one.
Once this becomes a more widespread problem, the x86 architecture, in its present form, is doomed. At that point, what the industry will converge on (and whether it will converge at all) is an open question.
Bubble sort is better than a general-purpose sort on a mostly-sorted list, so, if you care about performance, you would always use bubble sort on mostly-sorted lists in preference to quicksort, mergesort or radix sort. Actually, though, insertion sort is better again in this kind of situation, so use it instead of bubble sort.
A guy on my department has written about these - the basic problem is that if somebody does figure out a way to imitate your hand/signature/retinal characteristics (and, remember, they can get access to the data because the whole data has to be stored for comparison, not just a signature) you're in trouble. It's rather difficult to get a new, non-compromised retina :)
I have seen a theory mentioned previously (can't remember where) that went something like this:
Could somebody comment on the
There are several potential solutions:
Do you really think such a thing is possible, even with the best of intentions? How many trusted, unbiased sources of advice for choosing a Linux distribution do you know :-)
Pretty much. They consist of a set of libraries, including a widget library, inter-application communication, and other useful utilities like printing, a set of utilities like file managers, and applications which make use of their facilities.
Where do they stand in relation to X itself and a window manager?They stand above the X protocol, and therefore machines using either have all the nice networking abilities of X. All existing X applications run fine under either system.
Any window manager will work with either KDE or GNOME, but there are some window managers which have extra capabilities which make them work better with the systems. KDE ships with its own window manager, kwm, and GNOME ships with Sawfish, but GNOME has shipped with Enlightenment in the past.
Is there any real difference between KDE and Gnome, or are they just two different products filling the same niche (like IE vs. Netscape)?From a user's perspective, there's not a great deal. Frankly, until the office suites get closer to release, the decision isn't all that important.
From an ideological perspective, Qt (the base toolkit on which KDE is derived) is under a licence which is free enough to be regarded as open-source and RMS-free, but is a real pain as it is viewed by many (including the Debian project) as being GPL-incompatible. Therefore, in the view of Debian, KDE violates its own licence! The KDE people, however, disagree.
From a programmer's perspective, the toolkits are different - Qt/KDE is based around C++, and Gtk+/GNOME are based around C. Up until recently, the C++ support for GNOME has been regarded as pretty bad, and the C support from KDE equally so. They have made some other differing architectural decisions - GNOME is using CORBA to support object embedding, where KDE is using their own protocol.
Finally, I'd just like to point out that GNOME apps run fine on a KDE desktop and vice versa. So, provided you have enough memory to keep both widget sets at once without thrashing, choosing one doesn't mean you've lost the other.
Disclaimer: I use and develop with GNOME, but have no experience with programming KDE apps.
However, in Paul Verhoeven's (the movie director) opinion, and my own, along the way he advocates fascist military government, and consequently the movie was a brutal satirisation of the book.
I had an interesting discussion with another Slashdot reader about the book a few months ago on whether Heinlein really intended to advocate a fascist political system in the book or not. After an exchange of fascinating e-mails, we agreed to disagree on the book.
I still think it's well worth a read now, just to decide for yourself what he was really on about.
Check out the Agents Group for this and other projects.
However, if you want to organise a bit of civil disobedience, what you need are enough bodies on the street - armed or unarmed, it makes little difference. If the government concerned doesn't fold, weapons smuggling ain't that difficult.
Oh, it's just you that feels the need to defend yourself from the government in a bad mood? All an assault rifle is likely to do is make the cops more likely to kill you in the process of arresting you.
So, IMHO, anyone who believes that widespread posession of guns makes any difference to the potential for governmental tyranny, or prevents resistance of that tyranny, is kidding themselves.
xacc is now gnucash. Development ceased many months ago on the Mo/Lesstif version (bugfixes are continuing, though)- the developers switched to GTK+.
Thank you, whomever was responsible for this decision.
I disagree. There is crap open-source stuff out there, but overall the coding standard is really not too bad, and some is excellent. The important stuff tends towards the excellent (at least from my experience).
By contrast, go and have a play with arbitrary pieces of Windows shareware . . .
However, even (relatively) low-powered components still generate heat, and packing components in more and more densely creates a double whammy by concentrating the heat from all the components in a smaller space and leaving less room for air circulation to cool them. Water-cooling might just be a space-saving measure, not a way to run hyper-fast CPUs.
Yeah, so it does kick-arse floating point, but the performance less than a factor of two better than what a bog-standard Athlon - provided you can write code that uses the chip to its full potential (and, given the brief description above, that's probasbly quite a challenge).
In any case, floating point is totally irrelevant for code-cracking, which is the basic reason governments restrict supercomputers.
That's a fair first step, but what if the system is still introduced and geeks, goths, homosexuals, chess-club members, and other different-but-harmless students start getting harassed or suspended due to the WAVE program, and the school board still wasn't listening?
I don't know about you, but I'd be prepared to do something wrong if I believed that the alternative would be a greater wrong.
No, but if you're a European company doing export business with Asia, or a NGO like Amnesty International, the NSA almost certainly do try to read your mail. There's plenty of people who have real, concrete and current needs to make their mail NSA-proof (or at least NSA-resistant).
So you understand that Australians and New Zealanders are from different nations, do you? You're obviously not really American then . . . :)