Exactly. TCO of Windows XP is about to go through the roof when exploits are no longer patched. If you're running XP everywhere you are going to be wide open to a an enterprise scale disaster.
Given the XP holdouts clearly don't like Microsoft's current offerings, and Mac is growing faster in percentage terms, and Linux appears to be finally getting somewhere - i don't think these XP holdouts will be migrating to another Windows box any time soon.
If you're losing data on any mainstream OS, it is a user failure, not an OS failure. All of them provide easy ways to back up. I haven't lost any data since 1996. The last data I lost was actually on a Linux box.
Safari supposedly uses 30% of the power of Firefox - at least if the presentation yesterday is to be believed. Still, running 10.9 here right now and it seems pleasant enough. battery seems to be holding up...
Consoles (and gaming in general) are an unusual niche. Most software that does real world stuff has to survive across multiple generations of hardware, where the low level details change. The compiler, OS scheduler, etc know better (and certainly will know better in future) than you with regards to what resources are available and how to use them for most of your program's lifetime.
In a fairly counter-intuitive way - hardware evolves to run the software faster. This is why we're still running on x86 lookalike machines - because intel designs hardware to run existing software.
Don't get me wrong - i'm not saying you won't get better performance in the short term. But over the long term, all that work you did will be invalidated next hardware cycle. And if you chose an intelligent algorithm in the first place, chances are the hardware will evolve to run it faster.
Unless you're in a high performance critical niche industry - Spend your time making the lower level details of your code correct, safe and secure instead.
That. At the end of the day it doesn't matter what programming language you use if your algorithm is shit.
Get the algorithm right. If performance is not good enough, profile and optimize hot spot(s).
CPU architecture can and will change. Unless you're a hardware guy or in a pretty specific niche field, trying to get too clever with low level optimizations is only going to be shooting yourself in the foot when the new architecture changes the rules on you next year. Also - if you picked the correct algorithm, the compiler/cpu instruction set will likely be optimized for it sooner or later.
Unless you're in a very niche field, software is generally written to survive multiple generations of hardware advances.
As I said, if your billable rate is $250/hr (and if you're a professional, you're likely billing that out or even higher) and the machine causes you 4 hours of being unusable when you need it for work per year (during business hours - virus infection, screwing around getting some peripheral to work on a PC, chasing down crashes due to a windows bluetooth driver etc - all of which I've dealt with on PCs in the past 12 months) then after 5 years as far as the company is concerned, it has paid for itself.
This is why Linux or in some cases Windows (for example) is a non-starter for many people. If you consider your time to be worthless, great. Some people don't.
Saved time. If your billable rate is $250/hr and it saves you 4 hours per year, it adds up. Having owned (and built) both PCs and Macs, i'm not going back.
100% that. Even recently with a bunch of xeon servers (2007 spec), it was cheaper to buy new machines (2012 spec Cisco UCS) with more RAM than it was to upgrade the RAM, as the RAM standard changed and high capacity DIMMs in the old format were super expensive.
OMG i agree with a jedidiah post. As above, compare thunderbolt with 10GBe Cisco cables or even 1GBe copper SFPs (the equivalent of which is included on each end of a TB cable) and you're talking 10% of the cost.
I'm here to tell you that hardware is replaced periodically due to failure and continuation of support contracts anyway. In the real world, many companies simply lease their hardware and upgrade every 3 years. Or depreciate and re-purchase every 3 years. In the real world, it is a tax write off.
Why on earth do you think Linus would make a ton of work for himself in order so that you (not him) can buy some cheap piece of hardware less carefully. His job is not to enable laziness on your part.
And this is why Linux will never make it on the desktop.
And plenty of people with current hardware have no linux driver support at all, or shitty support that doesn't fully utilise their hardware? What's your point?
I'd much rather my brand new machine worked, than some 8 year old box I pilfered out of the garbage.
The increased resource requirements DO in fact do more with your computer. It means apps get developed in man-weeks or man-months rather than years or decades. It means we have a standard set of libraries shared between apps.
When RAM is under $10/GB, I'd much rather spend a couple of gigabytes on the OS platform to get apps that actually get released in a timely manner, include more features and plain do cool stuff that I want to do than bitch that they use more ram than my router built in 1992.
By programmer efficiency, I meant programmer hrs vs. real world problems solved. As I mentioned earlier, it is a trade-off. If you want an OS written in 2/3 of the resources for the same actual problem set, you will be waiting 2-3x as long for it or more. Or it will get "too hard" and never come out. Ever.
If you want to run wm2 or fluxbox or whatever, go for it. Other people have real world tasks they want to do and are prepared to spend say, 50 bucks on RAM and not have to fuck around wasting their time to do it.
You are in the minority. How up to date are your PCI slots, RAM slots, SATA/SAS ports and external buses? Upgrading CPU and video card won't upgrade them, and bandwidth is the major issue to overcome in the PC architecture.
So, you know... 120 gigabit of external connectivity, 12 GB of video memory and a 12 core CPU with built in PCIe SSD as standard is "nothing pro about it"?
Are you on crack? No you won't be able to upgrade it continually for 8 years. People who need this sort of power will replace machines every 3-5 years to get best bang for buck. Expandability via internal busses is over-rated - by the time you go to replace that video card, the PCIe standard will be outdated, the CPU socket will be outdated, the memory slot will be outdated, etc.
Yes, it is different to the previous Mac Pro. But to claim this isn't a high end machine is just being deliberately retarded about it.
Uhhhh - no, not exactly. The E libraries work differently. Everything works differently. I've run exactly the same applications side by side on an E installation, and a Gnome installation, and then a K installation. E uses more resources than something like OpenBox, but considerably less than G or K, and certainly less than Unity.
Because it doesn't do as much. If you compare E to QT or GTK I'm sure you'll find that it has nowhere near the same feature set.
E's problem is that some things simply remain unfinished.
Considered why that may be?
I'll grant some slim possibility that you are partly right. Let's suppose that to finish all the unfinished stuff, E ends up using 10% more resources than it does now. It will STILL be significantly more efficient than any of the heavy weights in use today. People who are carefully conserving every single miliwatt may still opt for OpenBox or similar, but the differences aren't going to be deal breakers for all of us.
I think you'll find it will end up being somewhere around the same size as KDE or Gnome if the feature set is comparable.
I've seen it happen with Linux vs. Windows in fact. Windows 95 copped a heap of shit for being so large and requriing so many resources when it was released.
Fact is: despite being a hacked 32 bit environment on top of a 16 bit boot loader, in terms of what it actually DOES, it is comparable with a modern Linux distribution of today. Except Linux still doesn't have standard high level libraries for sound, graphics, a standard widget toolkit, etc.
Now obviously Windows 95 is a pile of shit and fails for many other reasons, but in terms of what it supports vs. size, it makes a typical linux distribution look pretty bloated. Linux used so much less back in the day partially because it did so much less... It's smaller than Linux is now because it had a company behind it who made the call one picking ONE widget set, ONE sound library, etc.
Besides, a driver ABI doesn't exclude open source drivers. If that was to happen, someoen reverse engineers it and we're back to where we are today. Except we have a binary blob to reverse engineer which did work with A VERSION of the kernel.
Nobody cares that your router runs in 128 megs of RAM, when it does not provide the ease of use or rapid development provided by high level application frameworks. And yes I've built plenty of Linux boxes (with as little as 4 MEG - your 32 meg router was 4x the size of my first X11 box) myself. Thing is - they aren't desktops.
As far as a desktop environment goes - It isn't a "waste" of resources, it is a trade-off for programmer efficiency..
As any computer scientist will tell you - computing power scales, human processing power does not. In other words, processor time/RAM is cheap. Programmer time is not.
This is why we do not code everything in assembly language - it is only used if absolutely necessary.
Slashdot Mirror
Exactly. TCO of Windows XP is about to go through the roof when exploits are no longer patched. If you're running XP everywhere you are going to be wide open to a an enterprise scale disaster.
Given the XP holdouts clearly don't like Microsoft's current offerings, and Mac is growing faster in percentage terms, and Linux appears to be finally getting somewhere - i don't think these XP holdouts will be migrating to another Windows box any time soon.
If you're losing data on any mainstream OS, it is a user failure, not an OS failure. All of them provide easy ways to back up. I haven't lost any data since 1996. The last data I lost was actually on a Linux box.
point. missed. for what its worth, i was billing out 270/hr (as in, company rate) 10 years ago for my services.
Safari supposedly uses 30% of the power of Firefox - at least if the presentation yesterday is to be believed. Still, running 10.9 here right now and it seems pleasant enough. battery seems to be holding up...
So maybe you should learn java instead.
Consoles (and gaming in general) are an unusual niche. Most software that does real world stuff has to survive across multiple generations of hardware, where the low level details change. The compiler, OS scheduler, etc know better (and certainly will know better in future) than you with regards to what resources are available and how to use them for most of your program's lifetime.
In a fairly counter-intuitive way - hardware evolves to run the software faster. This is why we're still running on x86 lookalike machines - because intel designs hardware to run existing software.
Don't get me wrong - i'm not saying you won't get better performance in the short term. But over the long term, all that work you did will be invalidated next hardware cycle. And if you chose an intelligent algorithm in the first place, chances are the hardware will evolve to run it faster.
Unless you're in a high performance critical niche industry - Spend your time making the lower level details of your code correct, safe and secure instead.
That. At the end of the day it doesn't matter what programming language you use if your algorithm is shit.
Get the algorithm right. If performance is not good enough, profile and optimize hot spot(s).
CPU architecture can and will change. Unless you're a hardware guy or in a pretty specific niche field, trying to get too clever with low level optimizations is only going to be shooting yourself in the foot when the new architecture changes the rules on you next year. Also - if you picked the correct algorithm, the compiler/cpu instruction set will likely be optimized for it sooner or later.
Unless you're in a very niche field, software is generally written to survive multiple generations of hardware advances.
As I said, if your billable rate is $250/hr (and if you're a professional, you're likely billing that out or even higher) and the machine causes you 4 hours of being unusable when you need it for work per year (during business hours - virus infection, screwing around getting some peripheral to work on a PC, chasing down crashes due to a windows bluetooth driver etc - all of which I've dealt with on PCs in the past 12 months) then after 5 years as far as the company is concerned, it has paid for itself.
This is why Linux or in some cases Windows (for example) is a non-starter for many people. If you consider your time to be worthless, great. Some people don't.
Saved time. If your billable rate is $250/hr and it saves you 4 hours per year, it adds up. Having owned (and built) both PCs and Macs, i'm not going back.
100% that. Even recently with a bunch of xeon servers (2007 spec), it was cheaper to buy new machines (2012 spec Cisco UCS) with more RAM than it was to upgrade the RAM, as the RAM standard changed and high capacity DIMMs in the old format were super expensive.
That was meant to be "10GbE copper SFPs" not 1GbE.
OMG i agree with a jedidiah post. As above, compare thunderbolt with 10GBe Cisco cables or even 1GBe copper SFPs (the equivalent of which is included on each end of a TB cable) and you're talking 10% of the cost.
I'm here to tell you that hardware is replaced periodically due to failure and continuation of support contracts anyway. In the real world, many companies simply lease their hardware and upgrade every 3 years. Or depreciate and re-purchase every 3 years. In the real world, it is a tax write off.
And this is why Linux will never make it on the desktop.
And plenty of people with current hardware have no linux driver support at all, or shitty support that doesn't fully utilise their hardware? What's your point?
I'd much rather my brand new machine worked, than some 8 year old box I pilfered out of the garbage.
The increased resource requirements DO in fact do more with your computer. It means apps get developed in man-weeks or man-months rather than years or decades. It means we have a standard set of libraries shared between apps.
When RAM is under $10/GB, I'd much rather spend a couple of gigabytes on the OS platform to get apps that actually get released in a timely manner, include more features and plain do cool stuff that I want to do than bitch that they use more ram than my router built in 1992.
By programmer efficiency, I meant programmer hrs vs. real world problems solved. As I mentioned earlier, it is a trade-off. If you want an OS written in 2/3 of the resources for the same actual problem set, you will be waiting 2-3x as long for it or more. Or it will get "too hard" and never come out. Ever.
If you want to run wm2 or fluxbox or whatever, go for it. Other people have real world tasks they want to do and are prepared to spend say, 50 bucks on RAM and not have to fuck around wasting their time to do it.
You are in the minority. How up to date are your PCI slots, RAM slots, SATA/SAS ports and external buses? Upgrading CPU and video card won't upgrade them, and bandwidth is the major issue to overcome in the PC architecture.
So, you know... 120 gigabit of external connectivity, 12 GB of video memory and a 12 core CPU with built in PCIe SSD as standard is "nothing pro about it"?
Are you on crack? No you won't be able to upgrade it continually for 8 years. People who need this sort of power will replace machines every 3-5 years to get best bang for buck. Expandability via internal busses is over-rated - by the time you go to replace that video card, the PCIe standard will be outdated, the CPU socket will be outdated, the memory slot will be outdated, etc.
Yes, it is different to the previous Mac Pro. But to claim this isn't a high end machine is just being deliberately retarded about it.
If you're spending 5 grand on a machine, 40 bucks on a cable isn't a big deal.
Cooling without noise.
Next!
Because it doesn't do as much. If you compare E to QT or GTK I'm sure you'll find that it has nowhere near the same feature set.
Considered why that may be?
I think you'll find it will end up being somewhere around the same size as KDE or Gnome if the feature set is comparable.
I've seen it happen with Linux vs. Windows in fact. Windows 95 copped a heap of shit for being so large and requriing so many resources when it was released.
Fact is: despite being a hacked 32 bit environment on top of a 16 bit boot loader, in terms of what it actually DOES, it is comparable with a modern Linux distribution of today. Except Linux still doesn't have standard high level libraries for sound, graphics, a standard widget toolkit, etc.
Now obviously Windows 95 is a pile of shit and fails for many other reasons, but in terms of what it supports vs. size, it makes a typical linux distribution look pretty bloated. Linux used so much less back in the day partially because it did so much less... It's smaller than Linux is now because it had a company behind it who made the call one picking ONE widget set, ONE sound library, etc.
Besides, a driver ABI doesn't exclude open source drivers. If that was to happen, someoen reverse engineers it and we're back to where we are today. Except we have a binary blob to reverse engineer which did work with A VERSION of the kernel.
If you want to be negative about it, sure. I'd rather have a driver that works today than no driver at all.
Nobody cares that your router runs in 128 megs of RAM, when it does not provide the ease of use or rapid development provided by high level application frameworks. And yes I've built plenty of Linux boxes (with as little as 4 MEG - your 32 meg router was 4x the size of my first X11 box) myself. Thing is - they aren't desktops.
As far as a desktop environment goes - It isn't a "waste" of resources, it is a trade-off for programmer efficiency..
As any computer scientist will tell you - computing power scales, human processing power does not. In other words, processor time/RAM is cheap. Programmer time is not.
This is why we do not code everything in assembly language - it is only used if absolutely necessary.