Teaching specific apps in schools is a poor idea, since by the time these kids venture into the workplace the apps they used in school will be horrendously obsolete and no longer used... I used wordperfect (for dos) in school, and before that we had acorn archimedes systems running riscos and some apps i forget the name of.
What the schools use isn't important, because whatever they use will be different from what people have to use after school. How they teach is what matters, teach general concepts rather than specific apps.
Interesting... I would have used AMD for server and higher end systems, where the hypertransport interconnect and numa system would come into play (ie multiple physical cpus) , and intel for lower end systems with just 1 cpu...
Haven't looked much at i7, but it supposedly catches up with hypertransport.
The onboard sound is often poorly shielded, and will suffer EM interference from other components in the system... Tho if you use the digital output it's often much better.
But you're so right about the cheap speakers, most "computer speakers" are complete garbage and will sound like total crap regardless of what you connect them to. So if you're just gonna use speakers like that, don't waste money on a decent sound card.
ECC also causes a slight performance hit... That's why it's typically found in servers where stability and correctness are more important than outright performance. Gamers will probably favor non ECC.
That's the whole idea... Force people to move from xp to vista, and make it more difficult to use any other OS.
This is the microsoft business model, because making vista a desirable product that people would actually want to switch to with everything else being equal would cost them more. It just goes to show the utter disdain they have for everyone else.
I hate how proprietary vendors put in artificial restrictions to prevent you doing things... Older versions and other systems don't have these restrictions... A technical limitation is one thing, but to intentionally cripple a feature that already worked perfectly well is completely underhanded.
They did the same with IDE and a whole heap of other things...
Design something that's just a quick fix and barely adequate for today, and tomorrow you can sell another one.
But if you look at SCSI, I don't believe it has any practical size limits, and you could easily connect a huge disk array to an ancient SCSI-1 controller.
Yes, the biggest problem is that MS would simply refuse to support any filesystem they didn't control... They only grudgingly support the filesystems used on optical media because they had no choice...
Many filesystems have reference implementations available which are BSD licensed, such as UFS and EXT2... UFS is actually supported out of the box by pretty much everything except windows.
Yes, exactly, so qt still offers more freedom than a typical proprietary library... You are free to write GPL code, and you can also write proprietary code but you must buy a license to do so. Proprietary code may have the second option, but not always, and is unlikely to have the first or at least not to the same level.
So instead of bigger more bloated software and ever faster hardware... As you pointed out, the average user can get on perfectly well with a 2.4ghz and 1gb ram, and thats probably overkill anyway. Instead, work on making more efficient software instead of adding features people don't need and fancy graphical interface effects... So instead of that 2.4ghz power hungry p4 with 1gb of ram, use an atom or arm based machine with 512mb ram which consumes a fraction of the power and is small and quiet. This is how things will end up sooner or later, the average user doing everything they need to on a small quiet power efficient machine, and only certain niches having a big noisy computer.
Yes, small laptops none of which have more than 2gb of ram, and none of which have high end videocards with several gb of memory... Intended to compete with Arm, which are also 32bit and low power.
Plenty of projects die, and their source sits on sourceforge untouched for years... But that source is there for anyone to see and use in the future if they so desire... Plenty of proprietary apps die too, after which they're no use to anyone and only the original vendor can make use of the code.
There are also quite a lot of projects which haven't seen development for years because they do everything they're supposed to. How many applications have you used which reached a peak at some point, and all subsequent versions have been slower, more bloated and more annoying? Lots of people swear by win2k or xp and hate vista.... With open source, older functional versions can still be used instead of being forced to use the latest version against your will.
For what reason? Was the original USB code taken from somewhere else and they're not allowed to release it, or are they trying to intentionally differentiate the versions? What's to stop third parties adding in some new open source usb support code, such code already exists in qemu and may be reusable?
And windows is somehow less restrictive than the GPL? Windows restricts your personal use of the software, does not provide source code at all and prohibits redistribution of the binaries you do have...
You won't pay for a qt license, but were perfectly happy to pay for a windows license? That's quite a double standard right there...
The GPL does not restrict usage, it only restricts distribution... That's why you as an end user don't care, because the restrictions don't affect you.
Restricts isn't even the wrong word, it actually grants you some rights you wouldn't have otherwise (under plain copyright), but doesn't go so far in it's granting of rights as other licenses such as BSD. The GPL is also a lot less restrictive than virtually all proprietary licenses, but anti-gpl trolls often seem to overlook that fact.
The pdf import plugin is what does it for me... But aside from that, i agree... Software like this reached it's peak for the average user some time ago, and new versions offer nothing that people need...
I would like to see OO become more modular, like the linux kernel where... Feature modules are not loaded until you use them (faster startup times, less memory usage in typical cases).. Features can be disabled by removing the module..
OO as it stands is bloated, but it needs to be... people will refuse to switch to oo because it lacks some obscure feature that ms has, and yet adding all these features to oo makes the app bigger and slower for everyone.
word and powerpoint aren't available in 64bit versions, so i have no idea... firefox and openoffice perform a bit better however, when compiled as 64bit (on linux).
Ofcourse linux needs 64bit drivers too, but the vast majority of those drivers are open source... A lot of them needed nothing more than a recompile, many had already been tested on 64bit hardware such as sparc or alpha which has been around a lot longer.
Most windows drivers are made by the hardware supplier, and they have very little incentive to port drivers for non current hardware, or to make drivers for new hardware while the potential market is small (and it will never grow if very little hardware is supported).
The only people with motive to make 64bit drivers aren't able to for windows, but are able to for linux.
So proprietary code stifles the advancement of hardware.... Meanwhile, open source drivers for linux and bsd have been ported to 64bit hardware years ago.
Some of the performance benefit comes from being able to use more modern cpu features like sse2 and better code scheduling for superscalar cpus, as all 64bit x86-64 cpus have these features... This advantage can also be gained for 32bit code by compiling it for the specific cpu, which is obviously not feasible for a binary distributed system.
Some also comes from the extra registers...
On an architecture like sparc or powerpc, which was actually designed for future 64bit expansion in the first place, 64bit code actually tends to run slower, as a 64bit program will use more memory (larger pointers at least)... These processors will read pairs of 32bit values at a time, among other things, when running 32bit code.
The performance improvement with x86 is due to the generally crufty design of x86, and the fact that 64bit required a clean break from it and therefore a new lowest common denominator platform for 64bit code. It's due to this, rather than any inherent benefits of being 64bit.
He ends up with 2.5gb likely due to his videocard, it could be the motherboard design or other peripherals but is most likely to be the videocard... You can get videocards with several gigs of memory onboard now.
Teaching specific apps in schools is a poor idea, since by the time these kids venture into the workplace the apps they used in school will be horrendously obsolete and no longer used... I used wordperfect (for dos) in school, and before that we had acorn archimedes systems running riscos and some apps i forget the name of.
What the schools use isn't important, because whatever they use will be different from what people have to use after school. How they teach is what matters, teach general concepts rather than specific apps.
Interesting...
I would have used AMD for server and higher end systems, where the hypertransport interconnect and numa system would come into play (ie multiple physical cpus) , and intel for lower end systems with just 1 cpu...
Haven't looked much at i7, but it supposedly catches up with hypertransport.
The onboard sound is often poorly shielded, and will suffer EM interference from other components in the system... Tho if you use the digital output it's often much better.
But you're so right about the cheap speakers, most "computer speakers" are complete garbage and will sound like total crap regardless of what you connect them to. So if you're just gonna use speakers like that, don't waste money on a decent sound card.
ECC also causes a slight performance hit... That's why it's typically found in servers where stability and correctness are more important than outright performance.
Gamers will probably favor non ECC.
That's the whole idea...
Force people to move from xp to vista, and make it more difficult to use any other OS.
This is the microsoft business model, because making vista a desirable product that people would actually want to switch to with everything else being equal would cost them more. It just goes to show the utter disdain they have for everyone else.
I hate how proprietary vendors put in artificial restrictions to prevent you doing things...
Older versions and other systems don't have these restrictions...
A technical limitation is one thing, but to intentionally cripple a feature that already worked perfectly well is completely underhanded.
They did the same with IDE and a whole heap of other things...
Design something that's just a quick fix and barely adequate for today, and tomorrow you can sell another one.
But if you look at SCSI, I don't believe it has any practical size limits, and you could easily connect a huge disk array to an ancient SCSI-1 controller.
Yes, the biggest problem is that MS would simply refuse to support any filesystem they didn't control... They only grudgingly support the filesystems used on optical media because they had no choice...
Many filesystems have reference implementations available which are BSD licensed, such as UFS and EXT2... UFS is actually supported out of the box by pretty much everything except windows.
The 1700 series was just end of life'd last year and are very cheap...
As are the lower end 8xx series (non modular)...
Yes, exactly, so qt still offers more freedom than a typical proprietary library...
You are free to write GPL code, and you can also write proprietary code but you must buy a license to do so.
Proprietary code may have the second option, but not always, and is unlikely to have the first or at least not to the same level.
So instead of bigger more bloated software and ever faster hardware...
As you pointed out, the average user can get on perfectly well with a 2.4ghz and 1gb ram, and thats probably overkill anyway.
Instead, work on making more efficient software instead of adding features people don't need and fancy graphical interface effects...
So instead of that 2.4ghz power hungry p4 with 1gb of ram, use an atom or arm based machine with 512mb ram which consumes a fraction of the power and is small and quiet.
This is how things will end up sooner or later, the average user doing everything they need to on a small quiet power efficient machine, and only certain niches having a big noisy computer.
Yes, small laptops none of which have more than 2gb of ram, and none of which have high end videocards with several gb of memory...
Intended to compete with Arm, which are also 32bit and low power.
Plenty of projects die, and their source sits on sourceforge untouched for years... But that source is there for anyone to see and use in the future if they so desire...
Plenty of proprietary apps die too, after which they're no use to anyone and only the original vendor can make use of the code.
There are also quite a lot of projects which haven't seen development for years because they do everything they're supposed to. How many applications have you used which reached a peak at some point, and all subsequent versions have been slower, more bloated and more annoying? Lots of people swear by win2k or xp and hate vista.... With open source, older functional versions can still be used instead of being forced to use the latest version against your will.
It seems most people don't need such features, and so they have a fast database that suits their needs...
Don't like it? use postgres.
For what reason? Was the original USB code taken from somewhere else and they're not allowed to release it, or are they trying to intentionally differentiate the versions?
What's to stop third parties adding in some new open source usb support code, such code already exists in qemu and may be reusable?
And windows is somehow less restrictive than the GPL?
Windows restricts your personal use of the software, does not provide source code at all and prohibits redistribution of the binaries you do have...
You won't pay for a qt license, but were perfectly happy to pay for a windows license? That's quite a double standard right there...
Also take a look at:
http://doc.trolltech.com/4.4/license-gpl-exceptions.html
It seems you can write code under various licenses and link it to qt...
The GPL does not restrict usage, it only restricts distribution...
That's why you as an end user don't care, because the restrictions don't affect you.
Restricts isn't even the wrong word, it actually grants you some rights you wouldn't have otherwise (under plain copyright), but doesn't go so far in it's granting of rights as other licenses such as BSD.
The GPL is also a lot less restrictive than virtually all proprietary licenses, but anti-gpl trolls often seem to overlook that fact.
The pdf import plugin is what does it for me...
But aside from that, i agree... Software like this reached it's peak for the average user some time ago, and new versions offer nothing that people need...
I would like to see OO become more modular, like the linux kernel where...
Feature modules are not loaded until you use them (faster startup times, less memory usage in typical cases)..
Features can be disabled by removing the module..
OO as it stands is bloated, but it needs to be... people will refuse to switch to oo because it lacks some obscure feature that ms has, and yet adding all these features to oo makes the app bigger and slower for everyone.
Can i have your unused ram? Seems a shame to waste it.
word and powerpoint aren't available in 64bit versions, so i have no idea...
firefox and openoffice perform a bit better however, when compiled as 64bit (on linux).
Atom cpus are targeted at a different market...
Ofcourse linux needs 64bit drivers too, but the vast majority of those drivers are open source... A lot of them needed nothing more than a recompile, many had already been tested on 64bit hardware such as sparc or alpha which has been around a lot longer.
Most windows drivers are made by the hardware supplier, and they have very little incentive to port drivers for non current hardware, or to make drivers for new hardware while the potential market is small (and it will never grow if very little hardware is supported).
The only people with motive to make 64bit drivers aren't able to for windows, but are able to for linux.
So proprietary code stifles the advancement of hardware....
Meanwhile, open source drivers for linux and bsd have been ported to 64bit hardware years ago.
Some of the performance benefit comes from being able to use more modern cpu features like sse2 and better code scheduling for superscalar cpus, as all 64bit x86-64 cpus have these features... This advantage can also be gained for 32bit code by compiling it for the specific cpu, which is obviously not feasible for a binary distributed system.
Some also comes from the extra registers...
On an architecture like sparc or powerpc, which was actually designed for future 64bit expansion in the first place, 64bit code actually tends to run slower, as a 64bit program will use more memory (larger pointers at least)... These processors will read pairs of 32bit values at a time, among other things, when running 32bit code.
The performance improvement with x86 is due to the generally crufty design of x86, and the fact that 64bit required a clean break from it and therefore a new lowest common denominator platform for 64bit code. It's due to this, rather than any inherent benefits of being 64bit.
He ends up with 2.5gb likely due to his videocard, it could be the motherboard design or other peripherals but is most likely to be the videocard...
You can get videocards with several gigs of memory onboard now.