Dell is huge in North America, but in the vast Asia: India, Pakistan, Indonesia, China, its all custom-made PCs with the cheapest Taiwanese motherboards. The CPU of course is always AMD, a small price/performance difference is more important than any other factor.
AMD is winning both in the lowest end (non-western markets), and in the highest end (Think of rack servers purchased in volumes with the exception of IBM xSeries). But I think Intel is still selling a greater number of chips due to their name alone. Something that is very likely to change.
AMD is in a great position now, but do you even remember the good old K5 chips with the bad AMD logo print on top? Those chips sometimes just died, and windows froze here and there. The K5 singlehandedly gave AMD a bad name which has persisted in a few circles.
Another problem with AMD was the early Athlons which drank more than 80W, the slot Athlons usually. No chip had ever needed so much power, and cpu fans weren't always as good as the ones on Pentium MMX chips, and with enough dust around, Athlons sometimes died after ~3 years of use.
And I'm just talking about the earlier Athlons. Later ones really hit the spot and lifted AMD. I remember replacing my custom Athlon case with a PentiumIII system from IBM, it was much more cool and quite. See how the world has changed? The P4 uses more power than the Athlon64, AMD took the lead in the move to x64, their chips use less power and clearly have more price/performance ratio.
And the newer P4s now require fans, the average of which gives you 34 dB or more of noise. We've lost the last of reasons to buy an Intel chip.
I guess Duke Nukem Forever will come out too?
on
Sarge is Now Frozen
·
· Score: 1
Debian is one major cottage industry. It is the foundation of many distros, its a superdistro. An updated debian will produce great distributions with all updated software...
But theyll have to do more than post a slashdot story to convince me... (ranting without actually helping)
I disagree about the speed of the train being an issue in a large country. Trains are simply much less hassle, especially if run like subways, than airplanes. Airports are usually large, far away, require advanced booking and the schedule isnt nearly frequent enough. On a train, you can also change your mind and continue the journey, or disembark early. Depending on the ticket system, you can get off and get on a train heading the opposite way. Try than on airplanes.
The reason why USA doesnt have highspeed trains is feasibility. Subways only exist in big cities and not between cities for the same reason; not enough people will use it not frequently enough, so the per-mile running costs will be excessive. So they'll use busses which can have a much slower schedule; or even traditional trains which do not even require roads, and if run maybe once a week, is the cheapest way, per mile, to transport people.
Many security people think the whole aim of tightening up is to completely disallow a hack. 100% security is impossible, and the scene becomes a competition between the security guys and the finance/management people about how much to spend on it.
My idea of security is to clearly make it unworthwhile to hack. To gain $10, the hacker/cracker should have to spend $20 in cracking effort. An OpenBSD router is fine for a ~50-employee company. 14-character passwords, smartcards and biometrics would be excessive.
Their existance is important, but for the most part, the tails of the bell curve constantly lose out, which is why they're tails and not the hump. The best is to be at the top of the hump, while making sure the tails stretch wide enough, for the curve to survive anything. Stay atop the hump. Push others to the tails.
They had a 'list' of whitelisted mailservers of which our company wasnt a part. We called their admins, they asked us to contact OUR isps to signup their forms so the whole block becomes whitelisted. They wouldnt accept the mailserver admin. They will only accept the ISP itself filling out their coveted forms.
I called the ISP and they went huh?!? The ISP had nothing to do with our mailservers for which we are admins.
I was hoping something like this would happen, so we can tell our AOL customers we're on their blacklist and they're on ours. They should really get out and get a yahoo or gmail email account... where life is normal and people dont have to call their ISPs to fill out forms to get on lists just to send a damn email.
I have a little network card that isnt supported in other BSDs. So I'll fork the entire BSD OS, and put my card's driver into it.
Maybe Linus should fork the kernel, to have a special kernel just to run his new bitkeeper replacement.
A brand new OS sounds great, both for a kid developer, and for a major country like China. But really what they do is copy over most of the (free) code and rebrand it, adding their few improvements. Ideally people would release their packages as projects under sourceforge or something, but everyone likes to sound big. China would love to say they have an OS rather than creating locales for their language; and a kid in a garage would love to say he's programmed up an OS. Puts him in the same league as Linus, Tannenbaum and K&R, even though those guys didnt program up their whole OSes.
Its much better than 10 machines because if only one person is using it, she has 100% of the server's power, much like cable internet or shared hosting.
But VNC is a better idea. We do provide ~20 users with terminal services desktop on a windows2000 server (dual PIII 1.4GHz), and it has been very snappy and impressive for 3 odd years now. No plans to upgrade it. At one time we had ~20 Pentium1 workstations using the server; savings and easy administration.
The only issue appears with games and movies. Movies can be streamed, outside of VNC, and games... well...go buy an Athlon64.
If you've been collecting emails from a long time, reading your oldest emails are really interesting, a bit like time travel. I checked mail from ~9 years ago, was surprised how immature some subjects were, but was impressed with the writing, I used to write better...
I'd really be interested in my current emails 30 years from now. I wonder if the email companies can 'hide' older mail, and sell them to you years later at a high cost, or to your relatives when you die.
Motherboards are supposed to only be cpu socket, chipset, and other sockets and connectors. You're supposed to add/remove cards and devices in a custom fashion.
Apart from the obvious non-dual nature of Athlon FX55, you'll find motherboards that meet your demand.
I personally am looking for a simple nforce3 or nforce4 motherboard with large number of memory and pci slots. I mean more than 5 slots, if 10 slots are possible in that space, great, else a pci brige chip/riser card could be used.
And would be awesome, and with the price tag of current boards too.
Tried it once for a few users, they grew completely dependent on it, next everyone wanted it. There are so many ways to fax: email, drop folder, print to fax, and all faxes can be saved in an ODBC location. Has its own user administration, per-user cover pages and great scheduling features.
I'd do anything to have a linux version. They did apparently once have a solaris version, dont know where that went
Linux, NetBSD require at least 4MB ram for anything usable (yeah heard of linux-tiny uclibc etc) and are aimed at larger mobile devices that require lots of functionality and expandability. You can put perl, java apps, develop QT apps etc on these, and theyre heavily used in routers, PC104 industrial and medical machines and other custom embedded systems that arent quite mobile.
Then theres QNX, Palm and Symbian. Palm is entirely focused on PDAs and PDA-like personal devices. It has great graphics and personal-apps, mostly on ARMs and dragonballs. Symbian runs on ARMs, and is focused on even smaller devices, specifically cellphones. Both palm and symbian are very user-oriented, and cant support a great deal of multitasking since they dont support MMUs, except the latest palmos.
QNX and vxworks are non-user-oriented. vxworks was used in the mars rovers for instance, and is commonly used in ethernet switches and routers.
So you cant use symbian in routers, cant use palmos in mars rovers(efficiently), cant use linux in the smallest and simplest cellphones without the memory overhead, cant use 32-bit OSes in 8-bit microcontrollers and so on.
You can either be secure by design and implementation.. (OpenBSD), or you can be secure by piling up difficulties in the way... (Windows XP SP2 + Norton Antivirus etc).
Not using root is in the second category. Even OpenBSD has all services disabled by default and many Linux distros are enabling iptables to close most ports by default, both of which fall in the second security category. Ideally, the first category would be enough. In real life, you need both the good design and implementation, and the second security layers because nothing can be 100% secure. Even if it was 100% secure, the human error always exists, which is undenyable.
So in theory root is secure, as long as you're perfect, your passwords are extremely difficult, and your OS something like OpenBSD, or like OpenVMS for the VAX, which very few hackers would even want to learn to hack. But in real life, stay away from root.
That would get the world on their feet, and focused on the asteroid than the wars and political squabbling. The Human race will finally have a purpose... to save Earth.... unless its aimed at Africa in which case the powers to be will make a point of ignoring it.
But even a near miss will almost certainly alter the Earth's orbit, possibly fixing global warming.
Heck we could use nukes to make sure it readjusts global warming, maybe CAUSE the impending ice age and end of the neocene.
Cant make a brand name even if you package it and slap a pricey sticker?
Arent suse and redhat doing that and succeeding at it?
At work we handle ~70 computers, and will switch to linux if it werent for the binary apps we depend on. The binary market is slowly increasing for linux, so all those points are invalid.
And why cant you make a brand name with a pricey sticker? Why cant you make a consistent look and feel? Apple did that with FreeBSD, and there are plenty of linux companies out there...
(2) Hire people, keep them away from each other. Do not let them access to work theyve already done, and try to induce amnesia all the time. Assign a security guard to each person, and track their off-hour work to make sure they dont steal anything. And make SURE theyre scanned as they leave the building, and confescate all data-carrying media. Like SCO and Microsoft, keep a good legal team and sue people around who seem to do what you do.
Tough choices? Well in IT you have to make tough business choices, and the results will stick with you forever.
At a few terahertz, I think the frequency of the electromagnet field around these transistors will be the same as red light, so the chips will glow. I dont know if they'll emit laser light.
Its funny how far we've come. We can use the individual waves of red light as a clock signal for our chips.... in the near future.
I like the idea of 'free' software. Not so much when its license is designed to take more and more of the 'freedom' from people because theyre making money. Free should be Free, not as in free... but...
I'd like some projects to remain GPL, others to be BSD. Projects that are a good idea, but its uptake and improvement by companies can damage its market, should remain GPL. Other really good pieces of software should remain BSD, since a great deal of software is later based on it, and people who'd like to see such improvements in free, will just improve the original code and try to beat the proprietary ones.
But such a GPL3 idea is silly, and most authors wouldnt like their software to be forbidding. An OSS programmer wants recognition and wide use of their software. Noone wants to make another man rich; for free. Thats the essence of the BSD GPL difference, and the GPL3 idea is an extremist's view
Theyre not the first company who said oh crap, we could've done that. No they couldn't have.
I submit articles to wikipedia because I know I'm not making someone else rich, and the information there remains pretty much unbiased. I'd love to read Linux articles and articles on GPL, software patents etc, on Encarta. If I submit 'Linux is more stable', will it get approved?
So heres another company that says we'll look cool to opensource people, let them work on our projects and we'll sell the stuff. Sun has been an expert there, and IBM has really nailed it down. Microsoft is the antagonist, the ultimate enemy, the boss in the final stage. Not only does the slashdot crowd (for one) loves to hate Microsoft, the'd love to inflict any level of harm to them.
Binary driver vendors only distribute binary drivers for certain kernel versions of certain distros, mostly redhat suse and mandrake. The nVidia drivers are an example, but they can also recompile for vanilla kernels, but what about say a binary driver compiled for the stock 2.4 kernel that comes with redhat 9 shrike? Will it work seamlessly with WBEL?
I'd imagine all kernels were recompiled, at least to remove the word 'redhat'. I know I could download RHES kernels from their installation floppies and use those... but is that required to run precompiled kernel modules?
I'd agree that the largest part of a startup's expenditure is paychecks for a minimum amount of time. Cheap office, ebay machines and OSS software arent a big fraction, paychecks and marketing costs are.
For this reason, if you can start it up alone, or with a relative/friend, costs will be minimum, or else do a partnership. Another option is to have customers already lined up, in case youre splitting from a company and stealing their customers, or you know a company in need of such services/products. Different businsses work in different ways...
Slashdot Mirror
Dell is huge in North America, but in the vast Asia: India, Pakistan, Indonesia, China, its all custom-made PCs with the cheapest Taiwanese motherboards. The CPU of course is always AMD, a small price/performance difference is more important than any other factor.
AMD is winning both in the lowest end (non-western markets), and in the highest end (Think of rack servers purchased in volumes with the exception of IBM xSeries). But I think Intel is still selling a greater number of chips due to their name alone. Something that is very likely to change.
AMD is in a great position now, but do you even remember the good old K5 chips with the bad AMD logo print on top? Those chips sometimes just died, and windows froze here and there. The K5 singlehandedly gave AMD a bad name which has persisted in a few circles.
Another problem with AMD was the early Athlons which drank more than 80W, the slot Athlons usually. No chip had ever needed so much power, and cpu fans weren't always as good as the ones on Pentium MMX chips, and with enough dust around, Athlons sometimes died after ~3 years of use.
And I'm just talking about the earlier Athlons. Later ones really hit the spot and lifted AMD. I remember replacing my custom Athlon case with a PentiumIII system from IBM, it was much more cool and quite. See how the world has changed? The P4 uses more power than the Athlon64, AMD took the lead in the move to x64, their chips use less power and clearly have more price/performance ratio.
And the newer P4s now require fans, the average of which gives you 34 dB or more of noise. We've lost the last of reasons to buy an Intel chip.
Like the Athlon days?
Debian is one major cottage industry. It is the foundation of many distros, its a superdistro. An updated debian will produce great distributions with all updated software...
But theyll have to do more than post a slashdot story to convince me... (ranting without actually helping)
I disagree about the speed of the train being an issue in a large country. Trains are simply much less hassle, especially if run like subways, than airplanes. Airports are usually large, far away, require advanced booking and the schedule isnt nearly frequent enough. On a train, you can also change your mind and continue the journey, or disembark early. Depending on the ticket system, you can get off and get on a train heading the opposite way. Try than on airplanes.
The reason why USA doesnt have highspeed trains is feasibility. Subways only exist in big cities and not between cities for the same reason; not enough people will use it not frequently enough, so the per-mile running costs will be excessive. So they'll use busses which can have a much slower schedule; or even traditional trains which do not even require roads, and if run maybe once a week, is the cheapest way, per mile, to transport people.
Many security people think the whole aim of tightening up is to completely disallow a hack. 100% security is impossible, and the scene becomes a competition between the security guys and the finance/management people about how much to spend on it.
My idea of security is to clearly make it unworthwhile to hack. To gain $10, the hacker/cracker should have to spend $20 in cracking effort. An OpenBSD router is fine for a ~50-employee company. 14-character passwords, smartcards and biometrics would be excessive.
Their existance is important, but for the most part, the tails of the bell curve constantly lose out, which is why they're tails and not the hump. The best is to be at the top of the hump, while making sure the tails stretch wide enough, for the curve to survive anything. Stay atop the hump. Push others to the tails.
They had a 'list' of whitelisted mailservers of which our company wasnt a part. We called their admins, they asked us to contact OUR isps to signup their forms so the whole block becomes whitelisted. They wouldnt accept the mailserver admin. They will only accept the ISP itself filling out their coveted forms.
I called the ISP and they went huh?!? The ISP had nothing to do with our mailservers for which we are admins.
I was hoping something like this would happen, so we can tell our AOL customers we're on their blacklist and they're on ours. They should really get out and get a yahoo or gmail email account... where life is normal and people dont have to call their ISPs to fill out forms to get on lists just to send a damn email.
I have a little network card that isnt supported in other BSDs. So I'll fork the entire BSD OS, and put my card's driver into it.
Maybe Linus should fork the kernel, to have a special kernel just to run his new bitkeeper replacement.
A brand new OS sounds great, both for a kid developer, and for a major country like China. But really what they do is copy over most of the (free) code and rebrand it, adding their few improvements. Ideally people would release their packages as projects under sourceforge or something, but everyone likes to sound big. China would love to say they have an OS rather than creating locales for their language; and a kid in a garage would love to say he's programmed up an OS. Puts him in the same league as Linus, Tannenbaum and K&R, even though those guys didnt program up their whole OSes.
Its much better than 10 machines because if only one person is using it, she has 100% of the server's power, much like cable internet or shared hosting.
But VNC is a better idea. We do provide ~20 users with terminal services desktop on a windows2000 server (dual PIII 1.4GHz), and it has been very snappy and impressive for 3 odd years now. No plans to upgrade it. At one time we had ~20 Pentium1 workstations using the server; savings and easy administration.
The only issue appears with games and movies. Movies can be streamed, outside of VNC, and games... well...go buy an Athlon64.
If you've been collecting emails from a long time, reading your oldest emails are really interesting, a bit like time travel. I checked mail from ~9 years ago, was surprised how immature some subjects were, but was impressed with the writing, I used to write better...
I'd really be interested in my current emails 30 years from now. I wonder if the email companies can 'hide' older mail, and sell them to you years later at a high cost, or to your relatives when you die.
Motherboards are supposed to only be cpu socket, chipset, and other sockets and connectors. You're supposed to add/remove cards and devices in a custom fashion.
Apart from the obvious non-dual nature of Athlon FX55, you'll find motherboards that meet your demand.
I personally am looking for a simple nforce3 or nforce4 motherboard with large number of memory and pci slots. I mean more than 5 slots, if 10 slots are possible in that space, great, else a pci brige chip/riser card could be used.
And would be awesome, and with the price tag of current boards too.
www.actfax.com.
Tried it once for a few users, they grew completely dependent on it, next everyone wanted it. There are so many ways to fax: email, drop folder, print to fax, and all faxes can be saved in an ODBC location. Has its own user administration, per-user cover pages and great scheduling features.
I'd do anything to have a linux version. They did apparently once have a solaris version, dont know where that went
You mention Palm and Symbian. You mention Linux.
They generally tend to different niche markets.
Linux, NetBSD require at least 4MB ram for anything usable (yeah heard of linux-tiny uclibc etc) and are aimed at larger mobile devices that require lots of functionality and expandability. You can put perl, java apps, develop QT apps etc on these, and theyre heavily used in routers, PC104 industrial and medical machines and other custom embedded systems that arent quite mobile.
Then theres QNX, Palm and Symbian. Palm is entirely focused on PDAs and PDA-like personal devices. It has great graphics and personal-apps, mostly on ARMs and dragonballs. Symbian runs on ARMs, and is focused on even smaller devices, specifically cellphones. Both palm and symbian are very user-oriented, and cant support a great deal of multitasking since they dont support MMUs, except the latest palmos.
QNX and vxworks are non-user-oriented. vxworks was used in the mars rovers for instance, and is commonly used in ethernet switches and routers.
So you cant use symbian in routers, cant use palmos in mars rovers(efficiently), cant use linux in the smallest and simplest cellphones without the memory overhead, cant use 32-bit OSes in 8-bit microcontrollers and so on.
So please ask a more specific question.
You can either be secure by design and implementation.. (OpenBSD), or you can be secure by piling up difficulties in the way... (Windows XP SP2 + Norton Antivirus etc).
Not using root is in the second category. Even OpenBSD has all services disabled by default and many Linux distros are enabling iptables to close most ports by default, both of which fall in the second security category. Ideally, the first category would be enough. In real life, you need both the good design and implementation, and the second security layers because nothing can be 100% secure. Even if it was 100% secure, the human error always exists, which is undenyable.
So in theory root is secure, as long as you're perfect, your passwords are extremely difficult, and your OS something like OpenBSD, or like OpenVMS for the VAX, which very few hackers would even want to learn to hack. But in real life, stay away from root.
That would get the world on their feet, and focused on the asteroid than the wars and political squabbling. The Human race will finally have a purpose... to save Earth.... unless its aimed at Africa in which case the powers to be will make a point of ignoring it.
But even a near miss will almost certainly alter the Earth's orbit, possibly fixing global warming.
Heck we could use nukes to make sure it readjusts global warming, maybe CAUSE the impending ice age and end of the neocene.
Cant make a brand name even if you package it and slap a pricey sticker?
Arent suse and redhat doing that and succeeding at it?
At work we handle ~70 computers, and will switch to linux if it werent for the binary apps we depend on. The binary market is slowly increasing for linux, so all those points are invalid.
And why cant you make a brand name with a pricey sticker? Why cant you make a consistent look and feel? Apple did that with FreeBSD, and there are plenty of linux companies out there...
No justification given for that statement...
(1) Hire trustworthy people
(2) Hire people, keep them away from each other. Do not let them access to work theyve already done, and try to induce amnesia all the time. Assign a security guard to each person, and track their off-hour work to make sure they dont steal anything. And make SURE theyre scanned as they leave the building, and confescate all data-carrying media. Like SCO and Microsoft, keep a good legal team and sue people around who seem to do what you do.
Tough choices? Well in IT you have to make tough business choices, and the results will stick with you forever.
At a few terahertz, I think the frequency of the electromagnet field around these transistors will be the same as red light, so the chips will glow. I dont know if they'll emit laser light.
Its funny how far we've come. We can use the individual waves of red light as a clock signal for our chips.... in the near future.
I like the idea of 'free' software. Not so much when its license is designed to take more and more of the 'freedom' from people because theyre making money. Free should be Free, not as in free... but...
I'd like some projects to remain GPL, others to be BSD. Projects that are a good idea, but its uptake and improvement by companies can damage its market, should remain GPL. Other really good pieces of software should remain BSD, since a great deal of software is later based on it, and people who'd like to see such improvements in free, will just improve the original code and try to beat the proprietary ones.
But such a GPL3 idea is silly, and most authors wouldnt like their software to be forbidding. An OSS programmer wants recognition and wide use of their software. Noone wants to make another man rich; for free. Thats the essence of the BSD GPL difference, and the GPL3 idea is an extremist's view
Theyre not the first company who said oh crap, we could've done that. No they couldn't have.
I submit articles to wikipedia because I know I'm not making someone else rich, and the information there remains pretty much unbiased. I'd love to read Linux articles and articles on GPL, software patents etc, on Encarta. If I submit 'Linux is more stable', will it get approved?
So heres another company that says we'll look cool to opensource people, let them work on our projects and we'll sell the stuff. Sun has been an expert there, and IBM has really nailed it down. Microsoft is the antagonist, the ultimate enemy, the boss in the final stage. Not only does the slashdot crowd (for one) loves to hate Microsoft, the'd love to inflict any level of harm to them.
In Soviet Russia... ...ummm, I dont think I'll visit.
Binary driver vendors only distribute binary drivers for certain kernel versions of certain distros, mostly redhat suse and mandrake. The nVidia drivers are an example, but they can also recompile for vanilla kernels, but what about say a binary driver compiled for the stock 2.4 kernel that comes with redhat 9 shrike? Will it work seamlessly with WBEL?
I'd imagine all kernels were recompiled, at least to remove the word 'redhat'. I know I could download RHES kernels from their installation floppies and use those... but is that required to run precompiled kernel modules?
And certainly wouldnt want to be a n00b in the game.
I'll bet everyone will be camping all the time.
I'd agree that the largest part of a startup's expenditure is paychecks for a minimum amount of time. Cheap office, ebay machines and OSS software arent a big fraction, paychecks and marketing costs are.
For this reason, if you can start it up alone, or with a relative/friend, costs will be minimum, or else do a partnership. Another option is to have customers already lined up, in case youre splitting from a company and stealing their customers, or you know a company in need of such services/products. Different businsses work in different ways...