Typical Apple weenie... Let's see...for the inflated price of one G4/500, I can buy two P3/~800 and a GeForce...closer floating point performance and much faster everywhere else.
Sure, AltiVec is nice...but if it's just slapped onto an otherwise slow processor, then big deal. SETI@Home scores will be lower, but I'll stick with a machine where I can get decent performance on *every* operation...
Wow, that got moderated up really fast...goes to show how most moderators are really just pattern-matchers. (Try it sometime: write a comment with no content but in the form of the ones that tend to get moderated up, and watch...)
The point is to bring big non-free apps over to Linux, right? And why aren't they here already? Because companies tend not to feel comfortable without having a reliable vendor who you can yell at if something goes wrong. Troll Tech provides that, and they even give their stuff away for open-source projects.
So we've got a company that's helping to bring heavy hitters over while allowing current open-source projects to develop, and making a buck in the process. Why are you complaining again?
You're living in the days of VAX as well, apparently...
Believe it or not, someone (actually, quite a lot of people) already noticed that little problem! And that's why modern X implementations (go XFree86!) are hacked to take as much advantage of locality as possible without breaking the abstraction too badly.
Even if there were a significant hit, though, abstraction would be sufficient generalization.
At first glance, the name bears a striking similarity to NT Embedded...but then I realized it was Qt. If you want a real clone of NT, go for embedded GNOME: comparatively as large and just as buggy.
Yes, I want to run KDE on Embedded Linux on my fridge. Go Qt!
...because they're already monitoring in other places. Fort Meade, anyone?
There's likely not any terribly specialized monitoring equipment in there, and if there is, it likely wouldn't be sufficiently hardened or sophisticated for the use people expect or fear. Besides, most of it's probably nearly 18 months old, anyway.
And though it's likely to be in every post, I feel compelled to add that they can always ship some to my place if they run out of ideas.
Glad to see that they're putting Netscape Corp's crypto export license to good use!
But, (and this probably will be asked six times while I'm writing this) when will the new export regulations take effect, so you don't have to be blessed by a major corporation to ship?
BSD: Companies can use it, but you don't get any money or credit. Realistically, though, this isn't much of a loss -- if the BSDies I've run into are any indication, IBM would never consider hiring you anyway. End result: "Big corporations are taking over. They're doing exactly what the license intended them to do, but that still doesn't make me happy. Bitch bitch moan moan."
Linux: Companies that want to make extensive modifications don't want the GPL to show their secrets to the world. So they use the hated BSD or the even more irrationally hated closed source solution. End result: "Big corporations don't care about Linux. They don't understand how superior we are. Bitch bitch moan moan."
The point: The point of open source software is to improve the world, not to turn everyone into little soldiers for your personal OS/license jihad. You should be celebrating the fact that IBM, with its army of programmers and massive code base, has been impressed with open source software.
Game programmers may not mind writing in assembly for speed. Fine -- they're paid to be concerned about speed. However I, and most other programmers, are paid to write working, maintainable programs in a reasonable amount of time, not to write slightly faster and partially unreadable programs in excessive amounts of time.
I tend to write two kinds of programs: simple user utilities (most of which can be done in Perl to ignore this garbage) and security-critical, low-level daemons. The former are not speed-sensitive enough to write in assembly. In the latter, I am more concerned about being able to instantly determine program flow and implications than having to grok assembly.
I have no problem with people using assembly in very specialized applications. But I think anyone considering using assembly for any reason but as an absolute last resort should be sent to work on Microsoft BOB.
This was hilarious. >Its short so porting isn't really an issue. Correct me if I'm wrong, but to *port*, you first have to *parse*...Try porting from CISC to RISC. It's significantly more of a pain than playing a little with C code and recompiling. >ASM is just as readable as C these days. I think someone else went after this one, but assembly is still assembly. It's not like they renamed the instructions or something. >Its fast as hell... Have you heard of optimizing compilers? They've gotten pretty good the past few years or so... So, yes, sometimes you need assembly (like interrupt and page fault handlers). But if you're like that group that wrote an entire base OS directly in assembly, you're insane...
Actually, the "problem" with functional languages is that they've been more research-minded than anything else, so they haven't been tuned and optimized as much as, say, C++.
Well, let's see...there's room to speed up functional languages. You can make them implicitly parallel (try that in C). They're easier to implement garbage collection in.
If we want things to be "DAMN fast", why don't we just write everything in assembly? Let's hand-tune all our code! Yeah! (David Cutler complained about this when first implementing NT at Microsoft, btw -- he wanted to develop for multiple platforms to force cross-platform compatibility, but MS programmers of the time would automatically code in x86 assembly to make things faster...)
Garbage collection improves the memory management model, which is the point in the first place. Which is better: slightly slower code that reduces the number of memory errors and security holes, or faster code with bugs?
(For comparison, which is better: portably written, slower C or fast, nonportable, nonreadable x86 assembly?)
Transparent garbage collection has always been a good thing...it's just been a slow good thing. Hopefully, hardware support for garbage collection would simplify matters in the way that hardware threads could simplify the problems of multithreaded code.
(Of course, both problems become easier if one programs in an appropriately intelligent functional language, not C -- a language that facilitates integrating such things in a transparent way -- but that's most definitely another rant...)
When I hear questions like "but will it be flexible?" I can't help thinking that it sounds a little like whining from folks comfortable with their current programming model. Considering the number of problems that memory management gets programmers into, you'd think that any move towards a fast, automatic system (think George W. -- Java on cocaine) would be welcomed...
So, I guess I see two paths here: 1) Someone breaks the stupid patent, takes Amazon to court, and wins. Maybe it contributes to a review of patent law; maybe not.
2) Amazon realizes it's a stupid patent and doesn't ever bother to enforce it. (They may go on to get more patents, until it is discovered that Bezos uses Emacs, and resigns in shame...)
Option (2) seems more likely...in which case, does that herald the coming of stupid patents? Will we see e-whatever companies getting patents for the same reasons that they hoard Nerf (and, appropriately enough, just as effective)?
It will get disassembled, yes. but two problems: 1) it's probably written in VLIW assembly. (just a learning curve, not a huge block.) 2) ever heard of encryption? authentication? if you don't get the morons who wrote Windows DVD drivers on the job, you can easily keep wannabes like yourself away from the problem.
Actually, the technology exists today -- the blocking factor has really been the networking.
For example: A coworker of mine showed me a flyer for a digital VCR a little while back. One of the features of this model was that it would download the TV listings periodically to aid timer recording. To accomplish this, they had a modem built into the VCR and a special 800 number set up to dial into -- a very expensive proposition. With Bluetooth, they could instead insert a very cheap interface card and tell the VCR to look at xml.tvguide.com. In this case, the technology is hobbled by the network interface.
Socket Pentium II/IIIs allow lots of computing power to be crammed into a small space. StrongARM embedded units are prevalent, as will be Crusoe boxes. This market already has cheap, small, efficient horsepower -- they just need a cost-effective way to network it.
Well, which battle is easier: implementing a private short-range solution that plugs into the telco and is independent of the feed to your home, or taking on the telco full force?
Besides, technologies like Bluetooth, in the long run, only help to improve telco service. As more and more people move to a connectivity model more like the electric companies (i.e. always on, everything in the home plugs into it), the market expands greatly, bringing competition, lower prices, and better service.
And remember that the telco that the public sees and the one that the Internet sees are quite different. (You never saw Candace Bergen hawking OC-3's, did you?) The customer service complaints that one hears about with the various telco companies do not carry over to their backbone divisions.
But hey, if you fix the telcos, I won't complain:)
*WE* put people on the moon? No, *Americans* put *Americans* on the moon. Canadians watched it on TV, when they weren't too busy watching the "All Moose Porn, All The Time" channel.
What's wrong with Canada? Let's see...there's the grand triumph of socialism, nationalized health care -- which bankrupts the country and forces people to come to the US for surgery. Then there's Quebec. Just go smack ol' Frenchie in the face a few times. Or better yet, let them go independent and turn into a Third World country.
Crappy weather, high taxes, poor health care, tiny industrial base...who wouldn't want to come, eh?
okay, I have to give the British credit for the Battle of Britain. Half-credit, though, because there's no chance in hell that the RAF alone could have stood against the Luftwaffe.
The more I read, the more of a fruitcake you sound like...Mars is safe because England is. Hmmm. Last I checked, you can shoot a rocket to both...
We have a place in my country for people like you: it's called Montana.
Slashdot Mirror
Typical Apple weenie...
Let's see...for the inflated price of one G4/500, I can buy two P3/~800 and a GeForce...closer floating point performance and much faster everywhere else.
Sure, AltiVec is nice...but if it's just slapped onto an otherwise slow processor, then big deal. SETI@Home scores will be lower, but I'll stick with a machine where I can get decent performance on *every* operation...
Wow, that got moderated up really fast...goes to show how most moderators are really just pattern-matchers. (Try it sometime: write a comment with no content but in the form of the ones that tend to get moderated up, and watch...)
The point is to bring big non-free apps over to Linux, right? And why aren't they here already? Because companies tend not to feel comfortable without having a reliable vendor who you can yell at if something goes wrong. Troll Tech provides that, and they even give their stuff away for open-source projects.
So we've got a company that's helping to bring heavy hitters over while allowing current open-source projects to develop, and making a buck in the process. Why are you complaining again?
You're living in the days of VAX as well, apparently...
Believe it or not, someone (actually, quite a lot of people) already noticed that little problem! And that's why modern X implementations (go XFree86!) are hacked to take as much advantage of locality as possible without breaking the abstraction too badly.
Even if there were a significant hit, though, abstraction would be sufficient generalization.
>--The knowledge that you are an idiot, is what
>distinguishes you from one.
After a comment like that, I know it; do you?
At first glance, the name bears a striking similarity to NT Embedded...but then I realized it was Qt. If you want a real clone of NT, go for embedded GNOME: comparatively as large and just as buggy.
Yes, I want to run KDE on Embedded Linux on my fridge. Go Qt!
...because they're already monitoring in other places. Fort Meade, anyone?
There's likely not any terribly specialized monitoring equipment in there, and if there is, it likely wouldn't be sufficiently hardened or sophisticated for the use people expect or fear. Besides, most of it's probably nearly 18 months old, anyway.
And though it's likely to be in every post, I feel compelled to add that they can always ship some to my place if they run out of ideas.
That little rant showed that you don't like Microsoft's way of doing business. It did not show that Microsoft is violating any legal contracts.
It may surprise you (and relieves me) to know that the force of your will is not yet law.
Glad to see that they're putting Netscape Corp's crypto export license to good use!
But, (and this probably will be asked six times while I'm writing this) when will the new export regulations take effect, so you don't have to be blessed by a major corporation to ship?
So what do these announcements entail?
BSD: Companies can use it, but you don't get any money or credit. Realistically, though, this isn't much of a loss -- if the BSDies I've run into are any indication, IBM would never consider hiring you anyway. End result: "Big corporations are taking over. They're doing exactly what the license intended them to do, but that still doesn't make me happy. Bitch bitch moan moan."
Linux: Companies that want to make extensive modifications don't want the GPL to show their secrets to the world. So they use the hated BSD or the even more irrationally hated closed source solution. End result: "Big corporations don't care about Linux. They don't understand how superior we are. Bitch bitch moan moan."
The point: The point of open source software is to improve the world, not to turn everyone into little soldiers for your personal OS/license jihad. You should be celebrating the fact that IBM, with its army of programmers and massive code base, has been impressed with open source software.
Game programmers may not mind writing in assembly for speed. Fine -- they're paid to be concerned about speed. However I, and most other programmers, are paid to write working, maintainable programs in a reasonable amount of time, not to write slightly faster and partially unreadable programs in excessive amounts of time.
I tend to write two kinds of programs: simple user utilities (most of which can be done in Perl to ignore this garbage) and security-critical, low-level daemons. The former are not speed-sensitive enough to write in assembly. In the latter, I am more concerned about being able to instantly determine program flow and implications than having to grok assembly.
I have no problem with people using assembly in very specialized applications. But I think anyone considering using assembly for any reason but as an absolute last resort should be sent to work on Microsoft BOB.
This was hilarious. >Its short so porting isn't really an issue. Correct me if I'm wrong, but to *port*, you first have to *parse*...Try porting from CISC to RISC. It's significantly more of a pain than playing a little with C code and recompiling. >ASM is just as readable as C these days. I think someone else went after this one, but assembly is still assembly. It's not like they renamed the instructions or something. >Its fast as hell... Have you heard of optimizing compilers? They've gotten pretty good the past few years or so... So, yes, sometimes you need assembly (like interrupt and page fault handlers). But if you're like that group that wrote an entire base OS directly in assembly, you're insane...
Actually, the "problem" with functional languages is that they've been more research-minded than anything else, so they haven't been tuned and optimized as much as, say, C++.
Well, let's see...there's room to speed up functional languages. You can make them implicitly parallel (try that in C). They're easier to implement garbage collection in.
So which one would you choose?
And you miss the point...
If we want things to be "DAMN fast", why don't we just write everything in assembly? Let's hand-tune all our code! Yeah! (David Cutler complained about this when first implementing NT at Microsoft, btw -- he wanted to develop for multiple platforms to force cross-platform compatibility, but MS programmers of the time would automatically code in x86 assembly to make things faster...)
Garbage collection improves the memory management model, which is the point in the first place. Which is better: slightly slower code that reduces the number of memory errors and security holes, or faster code with bugs?
(For comparison, which is better: portably written, slower C or fast, nonportable, nonreadable x86 assembly?)
Why do I care what JWZ thinks again?
He wrote a lot of code. Big deal. He doesn't have any theoretical training that would come into play here, just programmer feedback.
Transparent garbage collection has always been a good thing...it's just been a slow good thing. Hopefully, hardware support for garbage collection would simplify matters in the way that hardware threads could simplify the problems of multithreaded code.
(Of course, both problems become easier if one programs in an appropriately intelligent functional language, not C -- a language that facilitates integrating such things in a transparent way -- but that's most definitely another rant...)
When I hear questions like "but will it be flexible?" I can't help thinking that it sounds a little like whining from folks comfortable with their current programming model. Considering the number of problems that memory management gets programmers into, you'd think that any move towards a fast, automatic system (think George W. -- Java on cocaine) would be welcomed...
So, I guess I see two paths here:
1) Someone breaks the stupid patent, takes Amazon to court, and wins. Maybe it contributes to a review of patent law; maybe not.
2) Amazon realizes it's a stupid patent and doesn't ever bother to enforce it. (They may go on to get more patents, until it is discovered that Bezos uses Emacs, and resigns in shame...)
Option (2) seems more likely...in which case, does that herald the coming of stupid patents? Will we see e-whatever companies getting patents for the same reasons that they hoard Nerf (and, appropriately enough, just as effective)?
It will get disassembled, yes.
but two problems:
1) it's probably written in VLIW assembly. (just a learning curve, not a huge block.)
2) ever heard of encryption? authentication? if you don't get the morons who wrote Windows DVD drivers on the job, you can easily keep wannabes like yourself away from the problem.
Actually, the technology exists today -- the blocking factor has really been the networking.
For example: A coworker of mine showed me a flyer for a digital VCR a little while back. One of the features of this model was that it would download the TV listings periodically to aid timer recording. To accomplish this, they had a modem built into the VCR and a special 800 number set up to dial into -- a very expensive proposition. With Bluetooth, they could instead insert a very cheap interface card and tell the VCR to look at xml.tvguide.com. In this case, the technology is hobbled by the network interface.
Socket Pentium II/IIIs allow lots of computing power to be crammed into a small space. StrongARM embedded units are prevalent, as will be Crusoe boxes. This market already has cheap, small, efficient horsepower -- they just need a cost-effective way to network it.
Well, which battle is easier: implementing a private short-range solution that plugs into the telco and is independent of the feed to your home, or taking on the telco full force?
:)
Besides, technologies like Bluetooth, in the long run, only help to improve telco service. As more and more people move to a connectivity model more like the electric companies (i.e. always on, everything in the home plugs into it), the market expands greatly, bringing competition, lower prices, and better service.
And remember that the telco that the public sees and the one that the Internet sees are quite different. (You never saw Candace Bergen hawking OC-3's, did you?) The customer service complaints that one hears about with the various telco companies do not carry over to their backbone divisions.
But hey, if you fix the telcos, I won't complain
...it still doesn't compare to the Great Bosnian Vowel Airlift. (If you don't know, buy the Second Best of Car Talk CD to hear a reading.)
backspace over VMS. you know, that's not a bad philosophy. :)
whew.
I thought for a second you were that damned "Here's a tarball of Transmeta's code morpher!" that's really a "Don Knotts plays a hacker" page.
thanks!
>Well, at least you guys are free to own guns.
But I'd gladly trade that right in a second if the US could only become a member of the noble and successful EU.
(Thanks for the sarcasm tag, btw. That refined European humour would be too subtle to detect otherwise.)
*WE* put people on the moon? No, *Americans* put *Americans* on the moon. Canadians watched it on TV, when they weren't too busy watching the "All Moose Porn, All The Time" channel.
What's wrong with Canada? Let's see...there's the grand triumph of socialism, nationalized health care -- which bankrupts the country and forces people to come to the US for surgery. Then there's Quebec. Just go smack ol' Frenchie in the face a few times. Or better yet, let them go independent and turn into a Third World country.
Crappy weather, high taxes, poor health care, tiny industrial base...who wouldn't want to come, eh?
okay, I have to give the British credit for the Battle of Britain. Half-credit, though, because there's no chance in hell that the RAF alone could have stood against the Luftwaffe.
The more I read, the more of a fruitcake you sound like...Mars is safe because England is. Hmmm. Last I checked, you can shoot a rocket to both...
We have a place in my country for people like you: it's called Montana.