I'm very much aware of the AMD chipsets being reference designs - it gets mentioned in half the reviews. I'm sorry to hear about your problems with the S2462. I've merely had very good experience with AMD 760 chipset compared to the Via KT133. Quite a few people have said the KT266A is stable - maybe I've misjudged Via.
Its a pity that stability seems to be such a difficult to measure thing - it seems that no single component causes a system to be unstable, but rather a particular combination (Via KT133 and SB Live! spring to mind). I'd love to see some formal test for stability, but reviews rarely mention stability at all, and even less frequently in any quantifiable way.
The VIA board in question myself and my flatmate worked on, and its an Abit KT7A. Admittadely, having got it up and running, it seems to be okay, but we still have a lot of problems at the start.
Yes, we were using a SB Live! That was part of the problem, and playing with the settings in the BIOS did eventually help with that. The much more serious problem was that putting significant load on the IDE system (copying large files from CD-ROM to HD for example) caused the system to hang. There was a third problem involving AGP driving values. Yes, with the help of various web sites we finally solved all the problems, but compared to the two AMD 761 motherboards I've worked with, it was a helluva hassle.
I agree that any system requires a lot of work to get it to perform optimally, but we were hitting problems just trying to install Windows, let alone actually being able to get to the configuration stage!
Yes - personally I'm looking forward to seeing what nVidia come up with next - I bought a new machine in August, so I won't be upgrading quite yet . I'm guessing we will see continuing product releases that have certainly very acceptable, if not top of the range, graphics/sound, at the time. So 6-12 months from now, we'll be looking at a nForce board with GeForce 3 Ti200 level graphics.
They may also, and I'm personally quite enthusiastic about this, release boards without integrated graphics/sound. Or perhaps just without integrated graphics - I seem to recall that the sound matches almost anything currently available anyway. Given this is their first attempt, future motherboards may well show significant speed benefits over the competition, and I'd certainly consider using an nForce board for my next machine.
Although interestingly enough it appears that the ASUS board is using a seperate network controller, which would suggest that they have significant problems using the onboard one.
As to cost - yes, although I beleive in the test both were using GeForce 3s - a straight nForce wouldn't be even close to a match for a KT266A with GeForce 3, but I do feel it would be interesting to see a plain nForce compared to a KT266A with a GeForce 2 MX card.
Yes, the KT266A does show an approximate 5-10% lead on the nForce, however:
This is using an external graphics card - the GeForce 3. The nForce has a GeForce 2 MX equivalent, and I imagine that comparing the speed of the nForce to a KT266A with GeForce 2 MX would prove insightful, too.
The nForce, IMHO, is aimed at the OEM market. It has not just graphics, but sound integrated onto the motherboard, at a significant cost saving compared to buying them seperately.
I cannot find any reference to stability, and my experience of Via chipsets, compared to Intel and AMD chipsets, is that they are less stable and more likely to have problems (the last Via based computer I had to set up took two people 5 days to get working correctly, compared to AMD and Intel based computers which have worked perfectly from first boot up).
If you're looking for raw speed, over all other concerns, yes the KT266A is probably for you, and would go well with a Creative Audigy and GeForce 3 Ti500. However if you're looking for a less powerful system, with a still respectable specification, the nForce is likely to work out a lot cheaper.
2D performance depends on the card. I looked very carefully into the available cards, and ended up with a Gainward GeForce 3 Powerpack!!! (yes, that abuse of punctuation is part of the name). It may not be the absolute fastest card out there, but in the same area, has all the features I'm looking for, and in terms of image quality I cannot tell the difference between it and a Matrox card.
Yes, I'm sure. My friend group is scarily intelligent (I'm around 130, for reference) - I know two further people with IQs over 180. 140 isn't genius level though - well, not on the test my friend was doing.
Okay, quick intro to IQ. It isn't actually a standard, IQs are only really comparable if the test is the same (I've done *pause* 3 official tests, with results varying by about 15 points in total - 130 is an average). IQ is meant to indicate your IQ, relative to the rest of the population. At 100, you have 50%, and 50% below you. At 50, you have 25% below you, 75% above. At 150, you have 75% below you, 25% above. At 194, then, you have 97% of the population below you, and 3% above. Theoretically
You do get instances of IQ results coming out at over 200, which indicates an incredible result in a test where most people do badly. This basically involves statistics, bell curves, normal distributions and that sort of thing, which I'm not going into here. The point is though, the numbers don't mean a helluva lot if you're not doing the same test.
No, I meant as in the technology society, the role playing society, the rock music society... how's your profile doing, BTW?:)
I know what you mean about the IQ thing, though. It seems more a measurement of a person's ability to think in particular ways - one flatmate, with an IQ about equal to mine (and just think about that whole average thing a while) is _incredible_ at logically thinking through problems.
No one person is lowering the average that much. And having checked it, the average is actually 150-ish. My fault, should actually done the maths the first time.
I think everyone is right about this - IT types are both naturally solitary, and not naturally solitary.
Most IT types are reasonably intelligent, right? Above average, typically. That puts them in the same group that is frequently picked on and excluded in school. As a result of that, they are generally going to have had to become capable of being their own company, and will be nervous around groups.
However, the same people tend to group together, as they understand each other well. Okay, lets take my friend group as an example:
I live in a house with 3 people. One person is as interested in computers as I am, and the other two both use them frequently. To make the point about intelligence, one of the people in the house has an IQ of over 190 (no, that's not a typo), and average for the house is about 140.
We share a relatively common friend group - we tend to all be members of the same socities (university, although we're no longer students), which tend to contain a pretty similar membership group. Average IQ for that group is probably 140-150.
Obviously, we do have friends outside that group, and people in that group may not be particularly intelligent/computer orientated, but as a general rule, intelligent people hang around intelligent people, and there aren't that many of us.
A better system would seem to be either the current idea of subscriptions, or people could pay on a per-article basis (that may not cover every eventuality, but I beleive it covers most).
The problem with both ideas is that people will download the data, and mirror it for anyone else to read for free. We may even end up with some crazy situation involving webpages being bounced around Gnutella or similar!
This tends to be done, actually, to stop people from spamming. After all, if people can't connect to other mail servers, they can't spam through them, and an ISP can monitor their own mail servers.
Why exactly is it a problem for you that you can only send e-mail through your ISP, anyway? If you wanted to receive mail, I'm sure you'd find POP3 (port 110) a much more useful?
Sorry, but the guys who actually sit around with stopwatches (and, occasionally, EEGs and CAT scanners) to quantify this sort of thing disagree with you
His results state that people take 2 seconds to find any given key combination. Sit there, get a stopwatch, find out just how long 2 seconds really is. Can you honestly tell me you've ever witnessed a power user pausing their input for two seconds before hitting a key combination they're familiar with (cut & paste spring to mind).
It's worth mentioning that if the sound is kept solely digital (MP3 decoded to digital sound, and then sent from a digital output), this shouldn't happen, but that requires hardware to decode the digital output when it arrives.
Also, AFAIK only the nForce sound system can actually encode into digital, as opposed to simply outputting pre-encoded digital sound.
Personally, I'd have felt that a combination was better; for example, the system I'm working on, currently, logs you in via SSL (and because we have the luxury of a nicely controlled user base, we're our own certificate authority), and gives you a session ID. This ID cannot be reused once you logout, and times out automatically after about 10 minutes.
This means that cookies never store personal information, and hijacking the session is relatively pointless (the application is loosely equivalent to a bulletin board in terms of security needs).
What on earth are you talking about? How do certificates come into this? Have you ever really looked into web application development?
Point one, cookies are anonymous, unless you supply personal data to the site setting the cookie, so that they can put it in the cookie. They are not some magic trick that can scan your name and address straight from your brain!
Certificates are good for proving you're a specific person, which if you're looking for anonymity, is a bit counter-intuitive.
Web sites have no state maintenance method inherent to them. Unless cookies are available, the only way of keeping track of trivial details like your login, shopping basket etc. is by encoding every single URL the site sends you to, to include that data. This is horrifically inefficient, and tricky to ensure works correctly.
This law would mean that almost every e-commerce site in the EU would have to be rewritten. Those sites would also increase significantly in complexity, as every page would have to become dynamic so they can ensure your data is in every single URL the sites gives you.
I wish people would actually research technologies, rather than assuming everything they've ever heard about it is true!
MSI motherboards, in my experience, don't get the temperature reading too accurate. Additionally, there seems to be a large increase in heat output relative to speed, in Athlons, which would explain some of the rest of this.
On the other hand, I still wouldn't have expected it to read over 50C./P.
Democracy means that the majority of the people in your country beleive in a government which restricts the use of marijuana.
In theory at least, this means that the majority beleive that marijuana are wrong, and probability indicates that the larger group is likely to be right.
I'd suggest that trying to persuade the majority to agree with you is your best bet.
"QToolBar::QToolBar main window cannot be 0." - Looks like a problem caused by an interface generator to me. This means that a program somewhere is just plain wrong, not that its error handling has problems.
"assertion `queue != NULL' failed." - Incorrect code means that an assert() test is failing.
"Unknown type `calendar' - Okay, not sure what these messages are indicating - probably code trying to reference some type of something, that does not exist.
I'm skipping the Aethera messages - they're too cryptic for even me!
"QTime::setHMS Invalid time -1:-1:-1.000" - Looks like an object is being initialised, but with invalid values, and those values are not being updated before the system access them.
Balse/Gnumeric error messages - does the author have any idea how complex code would become if every single time you accessed a file, you offered the possibility of requesting a new filename, then writing over the original pathname, and retrying until it either worked or the user gave up? Messages giving hints about how to fix the problems they indicate are good, adding a massive amount of complexity to your code is not.
Basically, almost all the messages he mentions are indicative of broken code. It may work, but that doesn't mean it's correct by a long way, and that these problems exist is just poor workmanship on the part of the coder responsible (assuming these are all meant to be stable releases - if not, someone please shoot the writer for missing the point of alpha/beta tests.
The remaining messages indicate minor problem - none of this has anything to do with error handling, and everything to do with testing your code properly before releasing it!
I'm using 2 60GXP 61gb drives, at home and at work, with no problems after 2 months. The one at work runs for approx. 8 hours a day, in a case whose average temperature is 35C. The one at home runs for about 6 hours a day, in a case whose average temperature is 32C.
It is a mechanical part, it will fail. Do not expect it not to fail. Expect it not to fail in the first few years (5 should be a good number, IMHO).
Re:Until you want a system that doesn't suck
on
The New Athlons
·
· Score: 1
There is no point to having a single MP CPU - you're paying a fortune for a feature you aren't using.
If you desperately want SCSI, MSI do a nice board with onboard SCSI, that won't break the balance like the Tyan will. Personally though I think an IDE RAID is going to better than a single SCSI drive, for a lot less.
Why do you want ECC RAM? Are you doing applications so incredibly critical that you cannot run the risk of memory corruption affecting it? Especially as ECC slows down the system!
Slashdot Mirror
I'm very much aware of the AMD chipsets being reference designs - it gets mentioned in half the reviews. I'm sorry to hear about your problems with the S2462. I've merely had very good experience with AMD 760 chipset compared to the Via KT133. Quite a few people have said the KT266A is stable - maybe I've misjudged Via.
Its a pity that stability seems to be such a difficult to measure thing - it seems that no single component causes a system to be unstable, but rather a particular combination (Via KT133 and SB Live! spring to mind). I'd love to see some formal test for stability, but reviews rarely mention stability at all, and even less frequently in any quantifiable way.
The VIA board in question myself and my flatmate worked on, and its an Abit KT7A. Admittadely, having got it up and running, it seems to be okay, but we still have a lot of problems at the start.
Yes, we were using a SB Live! That was part of the problem, and playing with the settings in the BIOS did eventually help with that. The much more serious problem was that putting significant load on the IDE system (copying large files from CD-ROM to HD for example) caused the system to hang. There was a third problem involving AGP driving values. Yes, with the help of various web sites we finally solved all the problems, but compared to the two AMD 761 motherboards I've worked with, it was a helluva hassle.
I agree that any system requires a lot of work to get it to perform optimally, but we were hitting problems just trying to install Windows, let alone actually being able to get to the configuration stage!
Yes - personally I'm looking forward to seeing what nVidia come up with next - I bought a new machine in August, so I won't be upgrading quite yet . I'm guessing we will see continuing product releases that have certainly very acceptable, if not top of the range, graphics/sound, at the time. So 6-12 months from now, we'll be looking at a nForce board with GeForce 3 Ti200 level graphics.
They may also, and I'm personally quite enthusiastic about this, release boards without integrated graphics/sound. Or perhaps just without integrated graphics - I seem to recall that the sound matches almost anything currently available anyway. Given this is their first attempt, future motherboards may well show significant speed benefits over the competition, and I'd certainly consider using an nForce board for my next machine.
Although interestingly enough it appears that the ASUS board is using a seperate network controller, which would suggest that they have significant problems using the onboard one.
As to cost - yes, although I beleive in the test both were using GeForce 3s - a straight nForce wouldn't be even close to a match for a KT266A with GeForce 3, but I do feel it would be interesting to see a plain nForce compared to a KT266A with a GeForce 2 MX card.
Yes, the KT266A does show an approximate 5-10% lead on the nForce, however:
If you're looking for raw speed, over all other concerns, yes the KT266A is probably for you, and would go well with a Creative Audigy and GeForce 3 Ti500. However if you're looking for a less powerful system, with a still respectable specification, the nForce is likely to work out a lot cheaper.
2D performance depends on the card. I looked very carefully into the available cards, and ended up with a Gainward GeForce 3 Powerpack!!! (yes, that abuse of punctuation is part of the name). It may not be the absolute fastest card out there, but in the same area, has all the features I'm looking for, and in terms of image quality I cannot tell the difference between it and a Matrox card.
Alien vs Predator 2 is the only game I've found that does actually only request your CD key if you access the multiplayer menu. Nice touch, IMHO.
Yes, I'm sure. My friend group is scarily intelligent (I'm around 130, for reference) - I know two further people with IQs over 180. 140 isn't genius level though - well, not on the test my friend was doing.
Okay, quick intro to IQ. It isn't actually a standard, IQs are only really comparable if the test is the same (I've done *pause* 3 official tests, with results varying by about 15 points in total - 130 is an average). IQ is meant to indicate your IQ, relative to the rest of the population. At 100, you have 50%, and 50% below you. At 50, you have 25% below you, 75% above. At 150, you have 75% below you, 25% above. At 194, then, you have 97% of the population below you, and 3% above. Theoretically
You do get instances of IQ results coming out at over 200, which indicates an incredible result in a test where most people do badly. This basically involves statistics, bell curves, normal distributions and that sort of thing, which I'm not going into here. The point is though, the numbers don't mean a helluva lot if you're not doing the same test.
No, I meant as in the technology society, the role playing society, the rock music society... how's your profile doing, BTW? :)
I know what you mean about the IQ thing, though. It seems more a measurement of a person's ability to think in particular ways - one flatmate, with an IQ about equal to mine (and just think about that whole average thing a while) is _incredible_ at logically thinking through problems.
No one person is lowering the average that much. And having checked it, the average is actually 150-ish. My fault, should actually done the maths the first time.
I think everyone is right about this - IT types are both naturally solitary, and not naturally solitary.
Most IT types are reasonably intelligent, right? Above average, typically. That puts them in the same group that is frequently picked on and excluded in school. As a result of that, they are generally going to have had to become capable of being their own company, and will be nervous around groups.
However, the same people tend to group together, as they understand each other well. Okay, lets take my friend group as an example:
I live in a house with 3 people. One person is as interested in computers as I am, and the other two both use them frequently. To make the point about intelligence, one of the people in the house has an IQ of over 190 (no, that's not a typo), and average for the house is about 140.
We share a relatively common friend group - we tend to all be members of the same socities (university, although we're no longer students), which tend to contain a pretty similar membership group. Average IQ for that group is probably 140-150.
Obviously, we do have friends outside that group, and people in that group may not be particularly intelligent/computer orientated, but as a general rule, intelligent people hang around intelligent people, and there aren't that many of us.
A better system would seem to be either the current idea of subscriptions, or people could pay on a per-article basis (that may not cover every eventuality, but I beleive it covers most).
The problem with both ideas is that people will download the data, and mirror it for anyone else to read for free. We may even end up with some crazy situation involving webpages being bounced around Gnutella or similar!
This tends to be done, actually, to stop people from spamming. After all, if people can't connect to other mail servers, they can't spam through them, and an ISP can monitor their own mail servers.
Why exactly is it a problem for you that you can only send e-mail through your ISP, anyway? If you wanted to receive mail, I'm sure you'd find POP3 (port 110) a much more useful?
Sorry, but the guys who actually sit around with stopwatches (and, occasionally, EEGs and CAT scanners) to quantify this sort of thing disagree with you
His results state that people take 2 seconds to find any given key combination. Sit there, get a stopwatch, find out just how long 2 seconds really is. Can you honestly tell me you've ever witnessed a power user pausing their input for two seconds before hitting a key combination they're familiar with (cut & paste spring to mind).
It's worth mentioning that if the sound is kept solely digital (MP3 decoded to digital sound, and then sent from a digital output), this shouldn't happen, but that requires hardware to decode the digital output when it arrives.
Also, AFAIK only the nForce sound system can actually encode into digital, as opposed to simply outputting pre-encoded digital sound.
First hint on the battery. You need to combine an acid and an alkali ("base" is the term used in the game).
"Could it be... Chris?"
Ah, right, got it. I think.
Personally, I'd have felt that a combination was better; for example, the system I'm working on, currently, logs you in via SSL (and because we have the luxury of a nicely controlled user base, we're our own certificate authority), and gives you a session ID. This ID cannot be reused once you logout, and times out automatically after about 10 minutes.
This means that cookies never store personal information, and hijacking the session is relatively pointless (the application is loosely equivalent to a bulletin board in terms of security needs).
Thoughts?
What on earth are you talking about? How do certificates come into this? Have you ever really looked into web application development?
Point one, cookies are anonymous, unless you supply personal data to the site setting the cookie, so that they can put it in the cookie. They are not some magic trick that can scan your name and address straight from your brain!
Certificates are good for proving you're a specific person, which if you're looking for anonymity, is a bit counter-intuitive.
Web sites have no state maintenance method inherent to them. Unless cookies are available, the only way of keeping track of trivial details like your login, shopping basket etc. is by encoding every single URL the site sends you to, to include that data. This is horrifically inefficient, and tricky to ensure works correctly.
This law would mean that almost every e-commerce site in the EU would have to be rewritten. Those sites would also increase significantly in complexity, as every page would have to become dynamic so they can ensure your data is in every single URL the sites gives you.
I wish people would actually research technologies, rather than assuming everything they've ever heard about it is true!
MSI motherboards, in my experience, don't get the temperature reading too accurate. Additionally, there seems to be a large increase in heat output relative to speed, in Athlons, which would explain some of the rest of this.
On the other hand, I still wouldn't have expected it to read over 50C./P.
Democracy means that the majority of the people in your country beleive in a government which restricts the use of marijuana.
In theory at least, this means that the majority beleive that marijuana are wrong, and probability indicates that the larger group is likely to be right.
I'd suggest that trying to persuade the majority to agree with you is your best bet.
In sequence:
Basically, almost all the messages he mentions are indicative of broken code. It may work, but that doesn't mean it's correct by a long way, and that these problems exist is just poor workmanship on the part of the coder responsible (assuming these are all meant to be stable releases - if not, someone please shoot the writer for missing the point of alpha/beta tests.
The remaining messages indicate minor problem - none of this has anything to do with error handling, and everything to do with testing your code properly before releasing it!
I'm using 2 60GXP 61gb drives, at home and at work, with no problems after 2 months. The one at work runs for approx. 8 hours a day, in a case whose average temperature is 35C. The one at home runs for about 6 hours a day, in a case whose average temperature is 32C.
It is a mechanical part, it will fail. Do not expect it not to fail. Expect it not to fail in the first few years (5 should be a good number, IMHO).
There is no point to having a single MP CPU - you're paying a fortune for a feature you aren't using.
If you desperately want SCSI, MSI do a nice board with onboard SCSI, that won't break the balance like the Tyan will. Personally though I think an IDE RAID is going to better than a single SCSI drive, for a lot less.
Why do you want ECC RAM? Are you doing applications so incredibly critical that you cannot run the risk of memory corruption affecting it? Especially as ECC slows down the system!