... I wouldn't be surprised to see "Judge grants IBM motion to bend SCO over and spank 'em."
If you've followed the case at all (such as the occasional glance at Groklaw), you'll be glad to see that it's been an utterly disastrous string of defeats for SCO.
Even if Linus's place on the list was meant as a compliment, I still think it's incorrect. Linus definitely DOES matter. He may NOT be writing all the code these days, he may not be the one coming up with all the innovations, but he *IS* still the one putting the Linux kernel together into a coherent whole.
Just read the Linux kernel mailing list and you'll see that Linus has an amazing grasp of all the major kernel subsystems, a clear sense of goals and direction for the kernel, as well as things to avoid, and a good ability to delegate tasks to the other kernel developers.
Basically, Linus remains "benevolent dictator" for the Linux kernel, and I'd say he's doing a highly effective job in that role. I'd put him in a top 10 list of tech people who do matter.
Also, this particular analyst concludes that Blu-ray and HD-DVD will "not be a repeat of VHS vs. Beta" and that a stalemate is the likely outcome.
Yeah, that's just retarded:-) Basically they're saying:
"History be damned! No one will win *this* format war. The merits of these products are similar, and these things are always won on technological merit."
the 10560x10560 format will probably get professional digital camera users drooling.... I imagine the memory card vendors, hard drive vendors, backpack vendors, and chiropractors will be drooling at this as well:-)
Well, I've just gotten on fiber, and so far so good. I'm in a rental house and we just got Verizon fiber last week. My computer connects through 802.11b, and the 11 Mbps connection is *definitely* saturated with the fiber connection upstream:-)
The good things about Verizon fiber so far:
They have a promo where we pay $30/month instead of the $40/month that we were paying for DSL. Lasts a year.
They gave us a new 802.11g router.
The guy who installed it forgot to take his box of Cat 5E cable when he left, and didn't come back when we called!!!! So now I have a free 1000 ft box of Cat 5E cable, woohoo that's like $60 worth at least.
Granted, those are 8-bit processors with no more than 64k or so of Flash, and no more than a few Kb of SRAM, but they're surprisingly powerful and useful. You can get 'em in clock rates of 50 MHz or more and all kinds of nifty on-chip peripherals are available... my favorites are the on-chip ADC and USB controllers! Plus the whole thing can be had for about $3, and Microchip (which makes the PIC) is *amazingly* generous in giving away free samples to anyone who asks (I've gotten so many of them I actually feel a bit guilty!).
You *ARE* correct that having external address and data lines does take up a lot of pins on a microprocessor. Having the memory onboard with a microcontroller is very nice, greatly reduces pin count. My favorite AVR microcontroller is available in a 28-pin DIP package, which is great for prototyping, and up to 25 of those pins can be used for general purpose digital I/O. Handy that!
Frankly, I don't see a great advantage to combining something like a 32-bit ARM CPU with a few megs of Flash and SDRAM in one package. The development of those products is inevitably more expensive and complex than for simpler 8-bit embedded systems. And as you say, different products are required to produce them. Most of the integrated applications seem to go in a slightly different direction: for example, many companies today sell a wireless-router-on-a-chip which combines a MIPS or ARM CPU, ethernet switch, DRAM controller, and 802.11g transceiver. Just add Flash, DRAM, and baseband wireless:-)
I have no idea about what that alarmist warning is about. Normally, if you do upgrade rather than dist-upgrade, it'll just not upgrade all of the packages, because it will only upgrade packages which are straightforward changes from version X to version Y. When you do dist-upgrade rather than upgrade, it will also rearrange dependencies so that if package A formerly depended on package B, and now A depends on C and B is no longer needed, it will do The Right Thing.
To make a long story short, upgrade is basically just more conservative and "stupid"... I've never had a problem doing an upgrade rather than a dist-upgrade.
Well, my server is my laptop, so ya, I carry it all the time;-) I do carry tools quite frequently, though not in huge quantities. I go grocery shopping very often with my bike. A couple of panniers can hold quite a bit, althought there's one hill that's a bit tough with a full load. If I wanted to carry even more stuff, I could buy a bike trailer for when I needed it... fortunately I'm lucky to own a car for those situations where it's truly more efficient.
MOD PARENT UP! Good advice, honestly. I'm amazed at how many people in congested urban environments (or otherwise) don't consider riding a bike to work. It's a great breath of fresh air and with a little planning you can, hopefully, find a route that takes you off busy streets and through residential areas and trails and such. I have only a 10 mile round trip commute myself, but often do a 20 mile version to and from my girlfriend's house. Always refreshing and enjoyable, especially this time of year.
Sometimes I think that cars have become so embedded in American life that people are unable to imagine any other form of transportation. This seems to be more true the older people get, my dad who's otherwise a perfectly rational person wouldn't consider riding a bike for his 2 mile commute, even though we live in a quiet flat town and he could use more exercise.
Why Intel? Notebooks, reliability, availability, price, loyalty, and even performance.
AMD has been great on notebooks for *at least* 5 years (when I got my Athlon notebook from HP, which I love).
Since when have AMD procs ever been unreliable??? I've read about a few Intel CPUs that got too hot and fried recently, and lots of overclocker problems, but other than that it seems that reliability isn't an issue with processors.
Apparently, the AMD availability issues have been greatly exaggerated.
AMD processors *are cheaper*... so how would Intel win on price???
Performance: AMD processors are better performing these days. Just google for Pentium 4 vs Athlon 64 benchmarks, you'll find the latter to be better performing, lower power consumption, and cheaper.
Loyalty is the only plausible reason you've given to stick with Intel. And I mean, come on, a knowledgeable IT buyer has had 10+ *years* since AMD became competitive on value, and about 3 *years* since they overtook Intel in terms of high-performance.
Because when Americans are give the quick, cheap, or good (pick two) choice, good is a distant third choice.
Okay... but as far as I can tell, AMD has all three: quick, cheap, and good! Intel-based systems have been more expensive in every segment I've ever considered over the past 10 years at least (laptop, budget desktop, workstation), and every single benchmark I've read for 2+ years says that the AMD proc stomps the Intel in both speed and power consumption (e.g. Athlon 64 vs Pentium 4).
Why does anybody use Intel CPUs for anything anymore?
As far as I can tell, AMD CPUs these days are cheaper, more innovative, faster, and more power-efficient than Intel CPUs, in the budget, desktop, and server markets. I won't buy Intel anymore, my last 5 computers have been Athlon 64, Athlon XP, K6-3, K6-2, and Cyrix.
I'm aware that AMD has sometimes had problems supplying enough volume to customers, but is there any other downside? I mean, at work we have all these ugly-as-sin black Dell minitowers in terribly-designed cases with the front USB ports facing the wrong way almost to the point of uselessness. I don't get it... why do so many people use Dell and Intel?
Why does anybody use Intel CPUs for anything anymore?
As far as I can tell, AMD CPUs these days are cheaper, more innovative, faster, and more power-efficient than Intel CPUs, in the budget, desktop, and server markets. I won't buy Intel anymore, my last 5 computers have been Athlon 64, Athlon XP, K6-3, K6-2, and Cyrix.
I'm aware that AMD has sometimes had problems supplying enough volume to customers, but is there any other downside? I mean, at work we have all these ugly-as-sin black Dell minitowers in terribly-designed cases with the front USB ports facing the wrong way almost to the point of uselessness. I don't get it... why do so many people love Dell and Intel?
I didn't think OS X was anything great to begin with. I didn't like how an amorphous, ever-changing subset of it was open source and the rest closed. I didn't like the graphics when it came to actually doing stuff.
I don't want to run OS X on my white box x86, but I sure *do* enjoy watching others do so despite Apple's misguided attempts to prevent it:-)
My original post was meant to be half-joking. The point I was trying to make is that Slashdot readers in generally are a lot more enthusiastic about messing around with operating systems and such that average people. Which I think explains why the promise of a small increase in reliability might not have impressed the readers of this article too much.
Although your points indicate why this increased operating system reliability and flexibility would actually be just what a lot of hackers are looking for! So ya, good post!
Does anyone on slashdot actually WANT to have uncrashable hardware? NO! Most of us love tweaking our hardware and puttering around with the BIOS settings and rebuilding our kernel just for fun, etc. My computer has never kernel panicked (that I can remember), but I reboot it every few days with a new kernel.
Linux is very reliable for me, even on newer hardware with a bleeding edge kernel. Why should I care whether it has a microkernel or monolithic kernel? Everything I deal with is user space. If it runs GNOME, is POSIX-like, and supports some kind of automatic package management, I'll be happy as a clam.
Will hardware drivers be developed faster and more reliably with a microkernel? That seems to be the biggest hurdle in reliable OS development these days... Anyone have a good answer for that, I honestly don't know.
Okay, so many are in another countries. But how many countries DON'T have laws against this?
Aren't *all* phishes coming from other countries?? I have *never* received a phish that didn't have at least a grammatical or spelling error, or some use of a phrase that sounded stilted or off. As far as I can tell, no one has *ever* written a phish that uses proper English in a style that's suitable for a business communication.
My advice for how to avoid phishing is: read the complete email!! Does it sound like a professional native English speaker at a bank wrote it? If not, it's phishing. If it *seems* okay so far, *then* start checking the links and things.
Thanks again for that helpful post. A few more questions:
(1) It seems like IBM totally dominates mainframe hardware and operating systems. Doesn't this lack of competition lead to extortionate prices? I've heard about their anti-competitive behavior a few decades back. Are they more interoperable now, I mean do other companies make processors, storage systems, etc. that can interoperate with IBM mainframes?
(2) Secondly, it seems like mainframes are very legacy-oriented. EBCDIC, green-screen terminals, proprietary everything, etc. I understand that the high throughput and reliability makes continued use a reasonable proposition for a lot of companies that have a significant investment in mainframes. But what about *new* companies with burgeoning data storage requirements? Are *new* banks and *new* airlines and *new* insurance companies using mainframes?
(3) From everything I've read about mainframes, it seems like they're ripe for "reinventing": keep the high throughput and high reliability, lose the proprietary hardware. As things like Xen mature, and as Intel and AMD provide explicit CPU-level support for virtualization, it seems like open-source virtualization and resource management running on commodity hardware could really give mainframes a run for the money. Any thoughts on that?
Was every contributer a blind proponent of the ideas? Were the contrary views edited out over time?
Um, maybe... but if you're skeptical you can look at the history and talk pages and see ALL the modifications that have EVER been made to that page. It's one of wikipedia's greatest features when you're trying to make sense of a controversial or fast-changing article.
I doubt that Baidupedia is going to leave censored pages sitting around under "History":-)
Slashdot Mirror
... I wouldn't be surprised to see "Judge grants IBM motion to bend SCO over and spank 'em."
If you've followed the case at all (such as the occasional glance at Groklaw), you'll be glad to see that it's been an utterly disastrous string of defeats for SCO.
I think we pretty much agree... but I still maintain that even if someone is replaceable in theory, it doesn't mean that they don't matter :-)
Even if Linus's place on the list was meant as a compliment, I still think it's incorrect. Linus definitely DOES matter. He may NOT be writing all the code these days, he may not be the one coming up with all the innovations, but he *IS* still the one putting the Linux kernel together into a coherent whole.
Just read the Linux kernel mailing list and you'll see that Linus has an amazing grasp of all the major kernel subsystems, a clear sense of goals and direction for the kernel, as well as things to avoid, and a good ability to delegate tasks to the other kernel developers.
Basically, Linus remains "benevolent dictator" for the Linux kernel, and I'd say he's doing a highly effective job in that role. I'd put him in a top 10 list of tech people who do matter.
Also, this particular analyst concludes that Blu-ray and HD-DVD will "not be a repeat of VHS vs. Beta" and that a stalemate is the likely outcome.
:-) Basically they're saying:
Yeah, that's just retarded
"History be damned! No one will win *this* format war. The merits of these products are similar, and these things are always won on technological merit."
the 10560x10560 format will probably get professional digital camera users drooling. ... I imagine the memory card vendors, hard drive vendors, backpack vendors, and chiropractors will be drooling at this as well :-)
BSD confirms it. Microsoft is dead.
It'd be fun to watch my dog chase a $4,000 cat around though :-)
The good things about Verizon fiber so far:
Eh... what??? It's already been done, been available for years. Nearly all the popular PIC microcontrollers and the AVR microcontrollers have Flash, CPU, and static RAM on chip.
:-)
Granted, those are 8-bit processors with no more than 64k or so of Flash, and no more than a few Kb of SRAM, but they're surprisingly powerful and useful. You can get 'em in clock rates of 50 MHz or more and all kinds of nifty on-chip peripherals are available... my favorites are the on-chip ADC and USB controllers! Plus the whole thing can be had for about $3, and Microchip (which makes the PIC) is *amazingly* generous in giving away free samples to anyone who asks (I've gotten so many of them I actually feel a bit guilty!).
You *ARE* correct that having external address and data lines does take up a lot of pins on a microprocessor. Having the memory onboard with a microcontroller is very nice, greatly reduces pin count. My favorite AVR microcontroller is available in a 28-pin DIP package, which is great for prototyping, and up to 25 of those pins can be used for general purpose digital I/O. Handy that!
Frankly, I don't see a great advantage to combining something like a 32-bit ARM CPU with a few megs of Flash and SDRAM in one package. The development of those products is inevitably more expensive and complex than for simpler 8-bit embedded systems. And as you say, different products are required to produce them. Most of the integrated applications seem to go in a slightly different direction: for example, many companies today sell a wireless-router-on-a-chip which combines a MIPS or ARM CPU, ethernet switch, DRAM controller, and 802.11g transceiver. Just add Flash, DRAM, and baseband wireless
I have no idea about what that alarmist warning is about. Normally, if you do upgrade rather than dist-upgrade, it'll just not upgrade all of the packages, because it will only upgrade packages which are straightforward changes from version X to version Y. When you do dist-upgrade rather than upgrade, it will also rearrange dependencies so that if package A formerly depended on package B, and now A depends on C and B is no longer needed, it will do The Right Thing.
To make a long story short, upgrade is basically just more conservative and "stupid"... I've never had a problem doing an upgrade rather than a dist-upgrade.
Can you carry a server on a bike?
;-) I do carry tools quite frequently, though not in huge quantities. I go grocery shopping very often with my bike. A couple of panniers can hold quite a bit, althought there's one hill that's a bit tough with a full load. If I wanted to carry even more stuff, I could buy a bike trailer for when I needed it... fortunately I'm lucky to own a car for those situations where it's truly more efficient.
Well, my server is my laptop, so ya, I carry it all the time
MOD PARENT UP! Good advice, honestly. I'm amazed at how many people in congested urban environments (or otherwise) don't consider riding a bike to work. It's a great breath of fresh air and with a little planning you can, hopefully, find a route that takes you off busy streets and through residential areas and trails and such. I have only a 10 mile round trip commute myself, but often do a 20 mile version to and from my girlfriend's house. Always refreshing and enjoyable, especially this time of year.
Sometimes I think that cars have become so embedded in American life that people are unable to imagine any other form of transportation. This seems to be more true the older people get, my dad who's otherwise a perfectly rational person wouldn't consider riding a bike for his 2 mile commute, even though we live in a quiet flat town and he could use more exercise.
Why Intel? Notebooks, reliability, availability, price, loyalty, and even performance.
... so how would Intel win on price???
AMD has been great on notebooks for *at least* 5 years (when I got my Athlon notebook from HP, which I love).
Since when have AMD procs ever been unreliable??? I've read about a few Intel CPUs that got too hot and fried recently, and lots of overclocker problems, but other than that it seems that reliability isn't an issue with processors.
Apparently, the AMD availability issues have been greatly exaggerated.
AMD processors *are cheaper*
Performance: AMD processors are better performing these days. Just google for Pentium 4 vs Athlon 64 benchmarks, you'll find the latter to be better performing, lower power consumption, and cheaper.
Loyalty is the only plausible reason you've given to stick with Intel. And I mean, come on, a knowledgeable IT buyer has had 10+ *years* since AMD became competitive on value, and about 3 *years* since they overtook Intel in terms of high-performance.
Okay... but as far as I can tell, AMD has all three: quick, cheap, and good! Intel-based systems have been more expensive in every segment I've ever considered over the past 10 years at least (laptop, budget desktop, workstation), and every single benchmark I've read for 2+ years says that the AMD proc stomps the Intel in both speed and power consumption (e.g. Athlon 64 vs Pentium 4).
Okay, perhaps you'd prefer if I state that "I have succumbed to the FUD that AMD chip shortages are a frequent problem." :-)
Why does anybody use Intel CPUs for anything anymore?
As far as I can tell, AMD CPUs these days are cheaper, more innovative, faster, and more power-efficient than Intel CPUs, in the budget, desktop, and server markets. I won't buy Intel anymore, my last 5 computers have been Athlon 64, Athlon XP, K6-3, K6-2, and Cyrix.
I'm aware that AMD has sometimes had problems supplying enough volume to customers, but is there any other downside? I mean, at work we have all these ugly-as-sin black Dell minitowers in terribly-designed cases with the front USB ports facing the wrong way almost to the point of uselessness. I don't get it... why do so many people use Dell and Intel?
Why does anybody use Intel CPUs for anything anymore?
As far as I can tell, AMD CPUs these days are cheaper, more innovative, faster, and more power-efficient than Intel CPUs, in the budget, desktop, and server markets. I won't buy Intel anymore, my last 5 computers have been Athlon 64, Athlon XP, K6-3, K6-2, and Cyrix.
I'm aware that AMD has sometimes had problems supplying enough volume to customers, but is there any other downside? I mean, at work we have all these ugly-as-sin black Dell minitowers in terribly-designed cases with the front USB ports facing the wrong way almost to the point of uselessness. I don't get it... why do so many people love Dell and Intel?
My thoughts exactly!
:-)
I didn't think OS X was anything great to begin with. I didn't like how an amorphous, ever-changing subset of it was open source and the rest closed. I didn't like the graphics when it came to actually doing stuff.
I don't want to run OS X on my white box x86, but I sure *do* enjoy watching others do so despite Apple's misguided attempts to prevent it
I agree on all points :-) Well said!
My original post was meant to be half-joking. The point I was trying to make is that Slashdot readers in generally are a lot more enthusiastic about messing around with operating systems and such that average people. Which I think explains why the promise of a small increase in reliability might not have impressed the readers of this article too much.
Although your points indicate why this increased operating system reliability and flexibility would actually be just what a lot of hackers are looking for! So ya, good post!
Does anyone on slashdot actually WANT to have uncrashable hardware? NO! Most of us love tweaking our hardware and puttering around with the BIOS settings and rebuilding our kernel just for fun, etc. My computer has never kernel panicked (that I can remember), but I reboot it every few days with a new kernel.
And allowing proprietary stuff into the kernel solves *what* problems exactly? :-P
Linux is very reliable for me, even on newer hardware with a bleeding edge kernel. Why should I care whether it has a microkernel or monolithic kernel? Everything I deal with is user space. If it runs GNOME, is POSIX-like, and supports some kind of automatic package management, I'll be happy as a clam.
Will hardware drivers be developed faster and more reliably with a microkernel? That seems to be the biggest hurdle in reliable OS development these days... Anyone have a good answer for that, I honestly don't know.
Thanks again for that helpful post. A few more questions:
(1)
It seems like IBM totally dominates mainframe hardware and operating systems. Doesn't this lack of competition lead to extortionate prices? I've heard about their anti-competitive behavior a few decades back. Are they more interoperable now, I mean do other companies make processors, storage systems, etc. that can interoperate with IBM mainframes?
(2)
Secondly, it seems like mainframes are very legacy-oriented. EBCDIC, green-screen terminals, proprietary everything, etc. I understand that the high throughput and reliability makes continued use a reasonable proposition for a lot of companies that have a significant investment in mainframes. But what about *new* companies with burgeoning data storage requirements? Are *new* banks and *new* airlines and *new* insurance companies using mainframes?
(3)
From everything I've read about mainframes, it seems like they're ripe for "reinventing": keep the high throughput and high reliability, lose the proprietary hardware. As things like Xen mature, and as Intel and AMD provide explicit CPU-level support for virtualization, it seems like open-source virtualization and resource management running on commodity hardware could really give mainframes a run for the money. Any thoughts on that?
Um, maybe... but if you're skeptical you can look at the history and talk pages and see ALL the modifications that have EVER been made to that page. It's one of wikipedia's greatest features when you're trying to make sense of a controversial or fast-changing article.
I doubt that Baidupedia is going to leave censored pages sitting around under "History"