If the chipset was designed at the same time as the Itanium, PC100 was probably expected to be the state-of-the-art at the time systems were available. Or it may be that Intel only intended Itanium systems to use RDRAM, and just threw in token SDRAM support. Either way, I expect that when McKinley comes out, you'll see support for faster speeds.
Re:Linux advocacy: VR3 framework for the Desktop?
on
Agenda, Not Hidden
·
· Score: 1
The truth of the matter is that programmers often don't THINK like users. It's often not a pride (high horse) issue, but more of a "well, _I_ understand how that works. It's intuitive for ME." thing.
And also the converse: "It's intuitive for me, what parts don't you understand?" It's difficult to code a good interface when you have no idea what a newbie considers a good interface.
Really? When I was over there last year, they had racks and racks of iMode phones for ~$30. I don't know if these were the "latest & greatest", though. Maybe they were the "remaindered" selection.
I'm not sure how you'd organise your documents though.
Possibly through a set of filters/queries? With no filter, *everything* on your system would be presented. You could create filters to specify files by date (creation/modification), "subject" (whatever you decide that should be), type, or a variety of user-specified attributes. Filters would have the ability to be combined with AND and OR operators. NTFS has some of these capabilities, but no support as far as I can tell, I understand BeOS (or its file browser) has something like it.
Re:What else can these guys hack?
on
Home Improvement
·
· Score: 2
Hmmmm, I have this vision of a torous filled with water, fitted to a hollow axle of some sort, with a bowl mounted to another tube on the spin axis with a pipe extending into the water. Some kind of tube leads to the hollow axle. Spin the wheel, light the bowl and party on! Gyroscopic effect might make it kind of hard to pass though...
I was a salaried employee at one place, and as part of downsizing they decided that developers should take part in tech support. So we took turns carrying a beeper. No compensation, just the laudable privilege of helping out our employer. Now I always make sure to make that a negotiating point -- if they won't put a "no beepers" clause in the contract, it means extra money.
compilers still don't produce code as good as hand-coded assembly language and please don't quote me the "myth of the magic compiler" that is supposed to produce code better than humans because you can always code whatever tricks humans would do into the compiler
Actually, for almost all contemporary RISC processors, compilers will always outdo a human on any real problem, because you have to take register & pipeline scheduling into account. Humans can do it, but it's more time-intensive than CISC assembly coding, and the code produced by the compiler is always better than that produced by a naive coder. I love writing assembler code, but once you have to start worrying about prefetch queue lengths and pipeline stalls, the warm fuzzy fades pretty quickly.
Is there some way you can put a bandwidth cap in place? Say NMb/day, after that you tell people to try again later? Sure, it's not as nice an arrangement, but there's no reason to martyr yourself for your hobby.
I wonder how much stress non-P4-optimized code places on the processor. If one or two of the integer units aren't active because the code wasn't scheduled properly, you'll see 100% CPU utilization, but the CPU won't really be functioning at 100% of potential.
Anybody have any numbers on instruction scheduling efficiency for the P4 on non-optimized code?
That was my first response to this, too. If you look at Sprint's coverage areas, they tend to be just along major highways. Maybe it's just me, but the thought of people on the road enjoying streaming video makes me twitchy.
I heard a similar story back in the mid-80s, only that time it was a pair of IBM Series/1s that ran the traffic signaling system for a large chunk of downtown. Downtown where, I don't remember -- New York? Chicago?
Because they haven't gotten their time's worth out of Java yet. If you spend a couple of years learning & practicing Java (or any language/utility), you don't want to just throw away that knowledge and move on for no good reason. Plus, Java was/is a better solution than most of the other tools available at the time; it doesn't look (at this point) like this offers that much of an advantage over Java to make the move compelling.
Visual SourceSafe is much more reliable and easier to use than CVS
Yeah, for users maybe. Try quickly extracting the source tree for V1.1.8 of your project when you're working on V2.5.1. Even if you've got it labelled properly, it's still a major headache. And don't think that CVS is the only product with gnomes; check out this error message from VSS:
There is a diff chain size mismatch in file 'DocumentPaginator.java' (hfncaaaa) at version 10 (versions earlier than that version can no longer be retrieved from the database).
VSS is OK if you're working with MS development tools, but if you're working with anything else, I'd go with CVS.
Actually, I heard (may be apocryphal) that Nokia moved into a new glass-walled office building. It turned out that the tint applied to the windows effectively blocked the frequencies that cell phones use.
Which is fine with me. I can crank out Q&D VB apps in a couple of hours that would take a couple of days to write in C++. Most of these wouldn't get written if they took that long. So who cares if the VB message board app the secretary uses takes three times as long to update the screen as it would written in C++? On her 450MHz PIII, it's still under a second. And if the user can't tell the difference, what's more important -- your time, or the computer's time?
The currently-envisioned solution would be to pair each unit with another unit, termed the "girlfriend". The unit will then only need reference the girlfriend to find out exactly what flaws it has and how bad they are...
Slashdot Mirror
You must always have plan B
Actually, it's plan B (the backup system) that's failed. Plan A (the primary system) is OK.
If the chipset was designed at the same time as the Itanium, PC100 was probably expected to be the state-of-the-art at the time systems were available. Or it may be that Intel only intended Itanium systems to use RDRAM, and just threw in token SDRAM support. Either way, I expect that when McKinley comes out, you'll see support for faster speeds.
The truth of the matter is that programmers often don't THINK like users. It's often not a pride (high horse) issue, but more of a "well, _I_ understand how that works. It's intuitive for ME." thing.
And also the converse: "It's intuitive for me, what parts don't you understand?" It's difficult to code a good interface when you have no idea what a newbie considers a good interface.
the price is a far more reasonable $300
Really? When I was over there last year, they had racks and racks of iMode phones for ~$30. I don't know if these were the "latest & greatest", though. Maybe they were the "remaindered" selection.
lets say there is already a website that says you torture puppies
That's not analogous. This is more akin to registering itorturecutepuppies.com and making it point to his website.
I'm not sure how you'd organise your documents though.
Possibly through a set of filters/queries? With no filter, *everything* on your system would be presented. You could create filters to specify files by date (creation/modification), "subject" (whatever you decide that should be), type, or a variety of user-specified attributes. Filters would have the ability to be combined with AND and OR operators. NTFS has some of these capabilities, but no support as far as I can tell, I understand BeOS (or its file browser) has something like it.
Hmmmm, I have this vision of a torous filled with water, fitted to a hollow axle of some sort, with a bowl mounted to another tube on the spin axis with a pipe extending into the water. Some kind of tube leads to the hollow axle. Spin the wheel, light the bowl and party on! Gyroscopic effect might make it kind of hard to pass though...
if the bottom falls out of the space program
Hey, it won't matter -- in 0G, if the bottom falls out, you'll just float there...
He's dead, Alan.
I was a salaried employee at one place, and as part of downsizing they decided that developers should take part in tech support. So we took turns carrying a beeper. No compensation, just the laudable privilege of helping out our employer. Now I always make sure to make that a negotiating point -- if they won't put a "no beepers" clause in the contract, it means extra money.
compilers still don't produce code as good as hand-coded assembly language and please don't quote me the "myth of the magic compiler" that is supposed to produce code better than humans because you can always code whatever tricks humans would do into the compiler
Actually, for almost all contemporary RISC processors, compilers will always outdo a human on any real problem, because you have to take register & pipeline scheduling into account. Humans can do it, but it's more time-intensive than CISC assembly coding, and the code produced by the compiler is always better than that produced by a naive coder. I love writing assembler code, but once you have to start worrying about prefetch queue lengths and pipeline stalls, the warm fuzzy fades pretty quickly.
Is there some way you can put a bandwidth cap in place? Say NMb/day, after that you tell people to try again later? Sure, it's not as nice an arrangement, but there's no reason to martyr yourself for your hobby.
...I just like the concept of "conculting"
he is STRESSING his cpu
I wonder how much stress non-P4-optimized code places on the processor. If one or two of the integer units aren't active because the code wasn't scheduled properly, you'll see 100% CPU utilization, but the CPU won't really be functioning at 100% of potential.
Anybody have any numbers on instruction scheduling efficiency for the P4 on non-optimized code?
That was my first response to this, too. If you look at Sprint's coverage areas, they tend to be just along major highways. Maybe it's just me, but the thought of people on the road enjoying streaming video makes me twitchy.
I heard a similar story back in the mid-80s, only that time it was a pair of IBM Series/1s that ran the traffic signaling system for a large chunk of downtown. Downtown where, I don't remember -- New York? Chicago?
Because they haven't gotten their time's worth out of Java yet. If you spend a couple of years learning & practicing Java (or any language/utility), you don't want to just throw away that knowledge and move on for no good reason. Plus, Java was/is a better solution than most of the other tools available at the time; it doesn't look (at this point) like this offers that much of an advantage over Java to make the move compelling.
I would guess it would take to AMD, but PowerPC systems should be OK.
Visual SourceSafe is much more reliable and easier to use than CVS
Yeah, for users maybe. Try quickly extracting the source tree for V1.1.8 of your project when you're working on V2.5.1. Even if you've got it labelled properly, it's still a major headache. And don't think that CVS is the only product with gnomes; check out this error message from VSS:
There is a diff chain size mismatch in file 'DocumentPaginator.java' (hfncaaaa) at version 10 (versions earlier than that version can no longer be retrieved from the database).
VSS is OK if you're working with MS development tools, but if you're working with anything else, I'd go with CVS.
Guess you missed the display PostScript that the original NeXT machines used. Better luck next time.
Actually, I heard (may be apocryphal) that Nokia moved into a new glass-walled office building. It turned out that the tint applied to the windows effectively blocked the frequencies that cell phones use.
it's PROGRAM efficiency that's down.
Which is fine with me. I can crank out Q&D VB apps in a couple of hours that would take a couple of days to write in C++. Most of these wouldn't get written if they took that long. So who cares if the VB message board app the secretary uses takes three times as long to update the screen as it would written in C++? On her 450MHz PIII, it's still under a second. And if the user can't tell the difference, what's more important -- your time, or the computer's time?
Palms do have the ability to require a password from the user. My guess is this would suffice.
Will the storage unit "know" where its flaws are?
The currently-envisioned solution would be to pair each unit with another unit, termed the "girlfriend". The unit will then only need reference the girlfriend to find out exactly what flaws it has and how bad they are...
Imagine a beowulf cluster of those things!
Of what, Canadians?