Just because some games don't use that other 256MB doesn't mean that no apps use it. The "pro" cards have been at 512MB to 640MB for a while, now. They wouldn't even bother selling them if no one knew what to do with them.
The unity of Slashdotters above is quite beautiful. I've never seen anything quite like it, where an entire discussion can be moderated redundant once and be completely correct. It is the most amazing thing to have happened since user #1 signed on...to be honest, it brings a tear to my eye.:'-)
You know what, Microsoft is a perfect fit for an American car company. They've improved their cars a lot recently, and it's about time for a setback, kinda like the early 1980s.
Do you have any links as to why SPARC64 is faster than UltraSPARC?
I don't know a whole lot about these chips. It looks like SPARC64 might draw a little more power, and it could be Sun and Fujitsu focus on slightly different strategies, where Sun is more power efficient, and Fujitsu goes all-out. A quick look at SPECJBB2000 seems to indicate that the 1.2GHz UltraSPARC IV systems are about 18% behind the 1.8GHz SPARC64 systems on a per-socket basis. I'd also wonder what the price difference is between comparable servers from each company, it could be they are very similar in price/performance.
Yet people will be queueing up to buy this next one.
They probably won't. The ramp-up to Longhorn will be slower than that for Windows XP, simply because XP was a whole new generation of "good enough". Even my family is still split between Win 98 and XP. I probably will never get XP, unless I needed it for testing software. No one has the latest version of Office. It's expensive and Office 97 is "good enough" (and I've moved to StarOffice).
For this next round, Microsoft will most certainly be competing against themselves, possible more than they will be against Linux and Apple.
Trying to communicate legitimately with mass e-mail is sort of like trying to talk to someone at a rock concert. Your lucky if they receive even one word of it.
Ugh, slavery is the new Nazism regarding Godwin's Law. I'll add stock prices, too. Okay, the new Godwin's Law is: If you mention slavery, Nazism, or compare stock prices in a thread, that thread is immediately terminated and no further argument can commence.
I'm also being subtly given the impression that the above is actually deliberate, "come to us, we're free and open! You can't read or understand the source anyway but that doesn't matter!" then smack them with the support fees and subscriptions for software they haven't got a hope in hell of understanding without dedicated support.
Having some experience with Solaris, I had no trouble installing Solaris 10 (free download), working around a couple minor issues, and installing new software even with new icons in GNOME. If I were in an environment where a few hundred dollars a year in support would pay for itself in time savings, then that's not so bad, either. Sun's documentation is generally very good, but occasionally getting a second opinion from support is better (e.g., for commercial money-making installations).
Linux will always be getting better, certainly, but Sun has shown what their big R&D budget can do with Solaris. The big bullet-points listed for Solaris 10 all work out of the box with minimal administration, and, when ZFS rolls around, that'll be another biggie.
Also, on the flip side, Sun can reduce Linux' differentiation by adopting things like GNOME and integrating it. When I installed Solaris 10, I just pointed the login screen to GNOME, and there it was, complete with StarOffice, GIMP 2, Evolution, etc. Having worked with Solaris for close to a decade dealing with CDE, this is a really big improvement.
In the long term, I think Solaris and Linux will reinforce eachother using a common GNOME/KDE desktop system. This is a _really_good_thing_, and will be the biggest threat to Microsoft. Imagine all Linux users and all Solaris users having a 100% compatible desktop on top of the most robust kernels ever made.
I said OSS. RMS tries to bring it all under an umbrella of morality and politics, which is more baggage than most people really need. I think there is a good reason why the Libertarian nutcases tended to focus on OSI over FSF, because they understand that people are free to live life on their own terms, with or without a manifesto.
They are now propriatary software developers, and it is immoral to support ubunto because of it, unfortunatly.
You need to understand that OSS licensing is merely a set of terms in a transaction. If the terms are suitable for you, fine, if not, fine, but it isn't a moral dillema at all.
It is entirely possible to argue against closed standards without using the morality card (which really makes you look immature, BTW). What about investment protection? What about risk mitigation? What about cost savings? What about interoperability?
I could rant for days on end against Microsoft, for example, and not once say they are immoral. While they certainly have very poor ethics, I'm not convinced it comes down to fire-and-brimstone morality, yet.
Imagine if a future version of Congress got a hold of time travel. If that ever happened, then future generations of Bushies would have come back and given Saddam a good phase pistol whipping. Therefore time travel is impossible. QED
Voila! Horizontal and Vertical scaling with 100% upward compatibility. Not bad...Who else offers this?
The problem is that we're arguing with people who think a $900 Dell box with Best Buy-grade IDE disks is 'enterprise' and that Linux is as good as Solaris because with kernel patches, recompiles, and non-standard configurations it is possible to emulate Solaris 10's feature list.
That would be terrible. All it would do is take Sun right back out of the low-end markets, because no one buying $3000 1RU Linux servers really wants to spend another $10,000+ on Oracle per server. Oracle does some really neat stuff, but they are slowly facing the same problems Microsoft is: free competition are eroding profit margins rapidly. Hell, Microsoft can't even give Windows away in some places.
That might be enough for getting an internship or, if you're lucky, a first job. However, after that, being able to speak confidently from work experience carries 10x the weight of "familiar with". If a person really did help to install a network, that means a lot, IMO. Being able to recite all the OSI network layers from a book is nice, but it lacks a component of real responsibility.
This also extends to non-job things, too. There's a reason why joining a club or taking up a sport is significant, esepcially if the person actually participates in that club. It sort of separates the coasters from the do-ers. When I was in college, I tried to lead a club and it was awful. No one wanted to do anything--they just wanted the bullet point on their resume.
Unless someone is a real super-star (Ph.D. written all over them), there is something to be said for striving for a 3.XX GPA rather than a 4.0 and finding other things to do with that time. I had an A- average, and every single interviewer seemed to find that perfectly adequate. A B+ would probably be just fine backed up with a solid resume. What always lit up the interviewers was when I pulled out a CAD drawing I had done or started talking about an interesting project I did over the summer.
There was a news report on PBS about the Acela trains. Something about how every one of them is basically an untested prototype that had late engineering changes that increased their weight. The weight stressed the brakes too much causing the problems.
It seems like the typical government BS project to me.
Making data executable out of convenience. Leaving their software completely full of exploitable buffers. Not displaying warning messages. Having no useful firewall for nearly a decade. Trying to integrate everything and the kitchen sink regardless of utility. Providing ActiveX. Promoting web standards that work against sysadmins' security efforts. Hiding bugs because they can. Avoiding making a compartmentalized multi-user system until the last moment.
Sun are saying that Niagara will be _at_least_ 15 times faster than the UltraSPARC IIIi. Even considering the USIIIi lags Xeon in SPECint, 15X is way beyond Xeon, and it's all at 56 watts. That's a pretty kick-ass improvement, IMO. BTW, floating point was never a goal for Niagara, as it's meant to be extremely well suited to web servers and J2EE servers.
I think Sun has already said that future "big-iron" CPUs will come from Fujitsu, with Sun focusing on Niagara and Rock. Given that SPARC64 has always kept pace with Alpha, POWER, etc. this shouldn't be a bad thing at all.
A 256-way SMP Opteron box would be pretty neat, but wouldn't it have to be implemented in 16-core chunks?
These are things, if you have a technical mind, that you can learn fairly easily from a book.
Reading a book can't go on a resume. A summer internship is the best place to learn the basics, beefs up the resume, and can provide valuable references (or even a job) for the future.
Slashdot Mirror
Just because some games don't use that other 256MB doesn't mean that no apps use it. The "pro" cards have been at 512MB to 640MB for a while, now. They wouldn't even bother selling them if no one knew what to do with them.
The unity of Slashdotters above is quite beautiful. I've never seen anything quite like it, where an entire discussion can be moderated redundant once and be completely correct. It is the most amazing thing to have happened since user #1 signed on...to be honest, it brings a tear to my eye.
You know what, Microsoft is a perfect fit for an American car company. They've improved their cars a lot recently, and it's about time for a setback, kinda like the early 1980s.
"Microsoft: Drivin' like its 1979."
"Blackcomb" is the rumoured project name for the OS after Longhorn.
Microsoft has already bet the farm on Longhorn, and it'll be ho-hum to the max. Odds are Longhorn will be Microsoft's last major OS effort.
Do you have any links as to why SPARC64 is faster than UltraSPARC?
I don't know a whole lot about these chips. It looks like SPARC64 might draw a little more power, and it could be Sun and Fujitsu focus on slightly different strategies, where Sun is more power efficient, and Fujitsu goes all-out. A quick look at SPECJBB2000 seems to indicate that the 1.2GHz UltraSPARC IV systems are about 18% behind the 1.8GHz SPARC64 systems on a per-socket basis. I'd also wonder what the price difference is between comparable servers from each company, it could be they are very similar in price/performance.
Yet people will be queueing up to buy this next one.
They probably won't. The ramp-up to Longhorn will be slower than that for Windows XP, simply because XP was a whole new generation of "good enough". Even my family is still split between Win 98 and XP. I probably will never get XP, unless I needed it for testing software. No one has the latest version of Office. It's expensive and Office 97 is "good enough" (and I've moved to StarOffice).
For this next round, Microsoft will most certainly be competing against themselves, possible more than they will be against Linux and Apple.
Trying to communicate legitimately with mass e-mail is sort of like trying to talk to someone at a rock concert. Your lucky if they receive even one word of it.
Ugh, slavery is the new Nazism regarding Godwin's Law. I'll add stock prices, too. Okay, the new Godwin's Law is: If you mention slavery, Nazism, or compare stock prices in a thread, that thread is immediately terminated and no further argument can commence.
I'm also being subtly given the impression that the above is actually deliberate, "come to us, we're free and open! You can't read or understand the source anyway but that doesn't matter!" then smack them with the support fees and subscriptions for software they haven't got a hope in hell of understanding without dedicated support.
Having some experience with Solaris, I had no trouble installing Solaris 10 (free download), working around a couple minor issues, and installing new software even with new icons in GNOME. If I were in an environment where a few hundred dollars a year in support would pay for itself in time savings, then that's not so bad, either. Sun's documentation is generally very good, but occasionally getting a second opinion from support is better (e.g., for commercial money-making installations).
Linux will always be getting better, certainly, but Sun has shown what their big R&D budget can do with Solaris. The big bullet-points listed for Solaris 10 all work out of the box with minimal administration, and, when ZFS rolls around, that'll be another biggie.
Also, on the flip side, Sun can reduce Linux' differentiation by adopting things like GNOME and integrating it. When I installed Solaris 10, I just pointed the login screen to GNOME, and there it was, complete with StarOffice, GIMP 2, Evolution, etc. Having worked with Solaris for close to a decade dealing with CDE, this is a really big improvement.
In the long term, I think Solaris and Linux will reinforce eachother using a common GNOME/KDE desktop system. This is a _really_good_thing_, and will be the biggest threat to Microsoft. Imagine all Linux users and all Solaris users having a 100% compatible desktop on top of the most robust kernels ever made.
I said OSS. RMS tries to bring it all under an umbrella of morality and politics, which is more baggage than most people really need. I think there is a good reason why the Libertarian nutcases tended to focus on OSI over FSF, because they understand that people are free to live life on their own terms, with or without a manifesto.
They are now propriatary software developers, and it is immoral to support ubunto because of it, unfortunatly.
You need to understand that OSS licensing is merely a set of terms in a transaction. If the terms are suitable for you, fine, if not, fine, but it isn't a moral dillema at all.
It is entirely possible to argue against closed standards without using the morality card (which really makes you look immature, BTW). What about investment protection? What about risk mitigation? What about cost savings? What about interoperability?
I could rant for days on end against Microsoft, for example, and not once say they are immoral. While they certainly have very poor ethics, I'm not convinced it comes down to fire-and-brimstone morality, yet.
I don't know, but the pillow talk certainly is much better, now.
- Anonymous Sheep Herder
Imagine if a future version of Congress got a hold of time travel. If that ever happened, then future generations of Bushies would have come back and given Saddam a good phase pistol whipping. Therefore time travel is impossible. QED
Voila! Horizontal and Vertical scaling with 100% upward compatibility. Not bad...Who else offers this?
The problem is that we're arguing with people who think a $900 Dell box with Best Buy-grade IDE disks is 'enterprise' and that Linux is as good as Solaris because with kernel patches, recompiles, and non-standard configurations it is possible to emulate Solaris 10's feature list.
That would be terrible. All it would do is take Sun right back out of the low-end markets, because no one buying $3000 1RU Linux servers really wants to spend another $10,000+ on Oracle per server. Oracle does some really neat stuff, but they are slowly facing the same problems Microsoft is: free competition are eroding profit margins rapidly. Hell, Microsoft can't even give Windows away in some places.
That might be enough for getting an internship or, if you're lucky, a first job. However, after that, being able to speak confidently from work experience carries 10x the weight of "familiar with". If a person really did help to install a network, that means a lot, IMO. Being able to recite all the OSI network layers from a book is nice, but it lacks a component of real responsibility.
This also extends to non-job things, too. There's a reason why joining a club or taking up a sport is significant, esepcially if the person actually participates in that club. It sort of separates the coasters from the do-ers. When I was in college, I tried to lead a club and it was awful. No one wanted to do anything--they just wanted the bullet point on their resume.
Unless someone is a real super-star (Ph.D. written all over them), there is something to be said for striving for a 3.XX GPA rather than a 4.0 and finding other things to do with that time. I had an A- average, and every single interviewer seemed to find that perfectly adequate. A B+ would probably be just fine backed up with a solid resume. What always lit up the interviewers was when I pulled out a CAD drawing I had done or started talking about an interesting project I did over the summer.
There was a news report on PBS about the Acela trains. Something about how every one of them is basically an untested prototype that had late engineering changes that increased their weight. The weight stressed the brakes too much causing the problems.
It seems like the typical government BS project to me.
How can you blame Microsoft for that?
Making data executable out of convenience. Leaving their software completely full of exploitable buffers. Not displaying warning messages. Having no useful firewall for nearly a decade. Trying to integrate everything and the kitchen sink regardless of utility. Providing ActiveX. Promoting web standards that work against sysadmins' security efforts. Hiding bugs because they can. Avoiding making a compartmentalized multi-user system until the last moment.
I guess that's a start.
Sun are saying that Niagara will be _at_least_ 15 times faster than the UltraSPARC IIIi. Even considering the USIIIi lags Xeon in SPECint, 15X is way beyond Xeon, and it's all at 56 watts. That's a pretty kick-ass improvement, IMO. BTW, floating point was never a goal for Niagara, as it's meant to be extremely well suited to web servers and J2EE servers.
I think Sun has already said that future "big-iron" CPUs will come from Fujitsu, with Sun focusing on Niagara and Rock. Given that SPARC64 has always kept pace with Alpha, POWER, etc. this shouldn't be a bad thing at all.
A 256-way SMP Opteron box would be pretty neat, but wouldn't it have to be implemented in 16-core chunks?
How are their Opteron servers and UltraSPARC IIIi servers "custom, 'high-end'"? They aren't even expensive. Linux is fully supported on the Opterons.
Well, given Microsoft already mastered unsecuring HTML, the next challenge, obviously, is plain text.
These are things, if you have a technical mind, that you can learn fairly easily from a book.
Reading a book can't go on a resume. A summer internship is the best place to learn the basics, beefs up the resume, and can provide valuable references (or even a job) for the future.
My CC company uses a rotating set of questions whenever I call. Occasionally, they even catch me off guard.
Ewww, it tastes like grandma!