Guys, get a grip! This is nothing new - the ReadMe.txt file for WHS says the same thing:
It's being "nothing new," and already documented in a readme makes the shortcomings of WHS more outrageous, not less so.
Don't *store* frequently modified critical files such as Outlook (.ost) files on the server. Keep them local, and let WHS back them up like any other file. This excellent advice applies to all operating systems[...]
No, it doesn't apply to all operating systems. In fact, if it applies at all, it only applies to Windows out of all the operating systems I've ever used (Digital Unix, VMS, Linux, Windows). At work (different jobs) for almost two decades now, all my files for all applications under VMS/Unix/Linux have been stored over the network--the local drive holds the operating system only. (And on VMS clusters the local drives of client systems only held swap space, the OS itself booted over the network.)
Given that: 1) the CEO, all of current management, sales and computer programmers who kept their mouths shut, remain in place, 2) the CEO being the same person who pledged to bring the elections over to the Republicans,
Um, no.
Walden O'Dell was the CEO who was committed to "help Ohio deliver its electoral votes to the President."
He was canned in 2005. Thomas Swidarski replaced him and then canned 5 of the 7 most senior Diebold execs.
"By my estimates, OO 2.0 is probably a suitable MS-Office
replacement for about 95-98% of users out there"
Well, that's a reflection of your experience.
And a reflection of mine as well. I reckon about 70-80% of the Excel spreadsheets
I see are purely documentation--without a single formula containing more than a single number or date.
I recently received directions to a hotel in a spreadsheet!
There are those for whom Excel is the right tool, but most
can do all their spreadsheeting in any spreadsheet program at all.
There is an excellent article on this in the May 1979 issue of Scientific American. Some notes from the article:
They credit Sir John Herschel for suggesting in 1828 that time should be the same in large regions, and claim that GMT had been widely adopted in England, Scotland, and Wales by mid-century.
However, one nation-wide time zone clearly wouldn't work over a continental nation like the U.S. So apparently C. F. Dowd, a principal of a seminary for girls, was the first to suggest zones in 1869, and Canadian Sandford Fleming then suggested making such a system world-wide. Railroads in the U.S. and Canada began operating on Standard Railway Time on 18 Nov. 1883, reducing the number of railroad times from at least 56 to 4.
Also interesting is the changing of the time zone boundaries over the years. People seem to prefer the sun setting later, and there was an advantage to being in the same zone as the east-coast population centers, so the western boundary of the Eastern Time Zone has moved from it's original location through the middle of Ohio to the western edge of Indiana.
I ran a quick, one-datapoint test. Opening a Word document was as fast in OOo2.0 as in Word, but opening a large Excel spreadsheet in (an already running) OOo2.0 was very slow. Most of the time I had a progress bar labeled "Calculating..." at the bottom of the screen, so I don't think XML parsing was the culprit.
On my Fedora system, the OOo calc package lists i386 as the architecture, so I'm wondering if SSE/SSE2 floating point is enabled. The old stack-based floating point is slow.
"fork an incompatible version that breaks backwards compatibility"
You mean like Itanium? [*] Backwards compatibility is the monopoly. A Windows that wouldn't run any of the existing Windows software out there wouldn't sell. People and businesses would continue buying the 'Win32 legacy' product. An Office that wouldn't read the old.DOC formats wouldn't sell. Et cetera.
* Strictly speaking, Itanium doesn't break backwards compatibility--it'll run X86 code. But if you intend to run X86 code Itanium is a poor choice. That's why AMD's X86-64 was such a great idea: substantial improvements and real backwards compatibility.
In my experience (big companies) can and do 'tinker with distributions', whether Linux or Windows.
There is a standard corporate build that gets installed on all machines. (And in the case of Linux, many of the apps are likely to be 'installed' on a fileserver, and running them queries a networked license server.)
But I imagine (not my experience) that medium-sized companies can't afford to build their own customized setup, and then your comments apply.
So if I go buy a Windows XP license today, do those 5 years of mainstream support
and 5 years of extended support start today, or from the introduction of XP?
My last employer switched to a policy of clearing the harddrives of old machines they donated to schools. The clearing was accomplished by writing random data over the whole drive seven times. They understandably didn't want who-knows-what internal information just being sent off to the outside. I've got to believe that everyone's doing this nowdays.
And in any case, larger companies lease licenses in bulk, and cannot transfer them. So the receiving charity would lack the license to use the installed Windows (or Office, etc) anyways.
" The main performance gain from going to x86-64 does not come from larger operands and larger addressing space. It comes from a cleaned-up instruction set architecture and, most importantly, from a larger set of registers."
Yep, it's the registers. But in the case of AMD, it's also that memory latency is lower due to the integrated northbridge. On memory intensive jobs, it matters.
"Well, I am a paleontologist, and I can *definitely* tell you that *nobody* in this profession makes hypotheses on sounds, mating patterns etc. based on a toenail of an extinct animal. That kind of BS allegations is reserved to strawman-building creationists."
First, grab a dictionary and look up 'hyperbole'. Then, and this is the tough part, consider that what the original poster was expressing is true, and frequently so.
I subscribe to National Geographic, and read a fair share of the science articles in newspapers, and can attest that an enormous amount of inference about the behavior of long-extinct animals is made from a few fossilized bone fragments. I understand the difference between Nat. Geo. and a scientific journal, but the scientists are being interviewed and quoted for these articles.
I see this, and I think of the hyena. When I was a child, the hyena was a scavenger. Hardy lions hunted and brought down live game, and only afterwards did the hyenas move in to eat what was left. By the time I was a young adult, it had been discovered that hyenas hunt at night. Lions were found to have driven hyenas away from their kill, and the hyenas--powerless to stop it--had to wait until the lions had their fill. But now, years after that, I'm told that hyenas both hunt and scavenge.
This is great. We're learning more about hyenas. We see how our earlier observations were accurate, but incomplete. We refine our theories as we learn more. But consider that the hyena is an animal known to us historically (i.e., in writing) for thousands of years. We can go out with cameras and observe the behavior of living hyenas, and still get it wrong. How much more likely are we to be wrong when our observations are limited to a few fossils? There is no justification for the hubris so rampant in the study of fossils.
Starter Edition: AKA Checkbox Edition 1. Indeed aimed at the "two-thirds world", but not as a product. Its purpose is to have an argument against the governments of nations with mostly poor citizens who look the other way with piracy. With this, MS can say "but there is an affordable alternative."
Home Basic Edition: AKA Checkbox Edition 2. Aimed at Europe, with no bundled media apps. Purpose is to fight sanctions.
Home Premium Edition: This is the version most home users will buy, but now MS gets to charge even more because it is demonstrably more capable (media) than HBE.
Professional Edition: Perhaps the most aptly named. Aimed at the professional who buys his own laptop. Many are predicting that companies will stop issuing laptops to employees in the future, who will be expected to just have their own. Much as has happened with cell phones. But lacks the fancy multi-media apps to force more customers up to Ultimate.
Small Business Edition: Pro with some added business features, see article.
Enterprise Edition: Pro with all business features, see article. Virtualization will allow employees to run their personal stuff on a different virtual machine, protecting the corporate network.
Ultimate Edition: Pro with media, and as many bells and whistles as possible so that the price can be jacked as high as possible. Professionals who want the media features will have to pay through the nose to get them, for Windows Pro lacks them and HPE won't connect to the corporate network. Students are caught, too, for HPE won't connect to the campus network (same with XP Home today).
CPUs should not be overclocked. First, you do not have the extensive set of tests that manufacturers use. So while an overclocked CPU might seem to work, it may well be that you simply haven't exercised the appropriate critical circuit path. Nor have you likely tested the CPU at the extremes of temperature. And even if the CPU does work at the speed you're clocking it, you may have pushed the current it draws high enough that electromigration of metal traces will lead to a premature failure.
CPU manufacturers particularly dislike overclocking because it leads to CPU failure that makes the manufacturer look like it didn't make a quality product. But since it did, that's not fair.
Lord only knows how pricing is done, but your part should be running as fast as it reliably can. There is no incentive for a chip maker to cripple clock speeds; the faster the part runs the more they can--and do--charge.
Now, if it ever came to be that there were only one CPU manufacturer, the game would change. The lone manufacturer might under-clock some parts to spread pricing options.
The USA accepted the metric system in 1866, but has never rejected the customary units. But there are a lot of caveats (metric-only labeling is rare, for example). But Americans seldom use metric units outside of medicine, engineering, and science (and anything electrical, since there are no customary electrical units).
(I say "customary" instead of "imperial" because the USA never adopted the British imperial standard of 1824. The difference between the two systems is substantial in the the units of capacity (e.g. the gallon).
But units of measure are perhaps the perfect example of open standards. And historically, they were enacted to protect consumers.
"As long as all this is optional (which it will be, because of laptops)"
It may be optional, but not because of (new) laptops. Laptops will mostly be UMA with the frame buffer in main memory (which should easily slide into that 2GB:-). Of course, the GPU will require a high-bandwidth path to memory.
"Businesses already have almost -no- incentive to switch"
True.
"Now, instead of just buying expensive licences, they have to upgrade the graphics cards on their vanilla work PCs??"
But license fees are not the only issue. Businesses don't want to switch from W2K (or won't want to switch from XP) because those systems are known to work with all their apps and software systems. In many cases, switching the operating system will entail a fair bit of expense and a great deal of risk. Unless there is enticing value on the the other side, why switch?
If there is a desire to switch, the cost of the required graphics card (likely in new PCs) will not influence the decision.
"I don't see how this equates to 2GB of DDR3, though."
Nor do I. Going from 512M to 2G is 4x, so it's more than what's required for a doubling of pointer size. Does the page size increase in a 64-bit system?
"MS continually drives up hardware requirements as time
progresses"
Indeed. But certainly not to spitefully inflict
hardship on customers. Arguably, an ideal world for MS would be
one in which people spent none of their PC dollars on hardware,
and all of them on MS software. And in fact, the business world
is much like that--large customers usually buy licenses which
are periodically renewed. But consumers need a reason to go buy
new software, and MS's office suite has reached a point of
maturity from where it's hard to increase the value, and thus
the desirability, of the next version.
MS likely believes that the path to new consumer value is via
multimedia. The better the GPU hardware, the greater the
potential for MS to sell multimedia applications or services.
So they create an OS requirement for top-of-the-line graphics
hardware. PC makers comply, for they will not sell many
machines if Windows isn't properly supported. And then MS can
sell multimedia to properly equipped consumers. It's a theory
anyway. (And of course they don't like being behind Apple in
the eye candy dept.)
But let us concoct an even wilder theory. MS has
called
for a 100-dollar PC. Of course, they want the 100-dollar PC to go
"down-market" in "some of these contries" where they see
piracy. However, nobody _wants_ a down-market anything: if the
price difference between a 100-dollar PC and a 250-dollar PC is
due to the price difference between Windows and
Windows--Starter Edition, the citizens of poor countries will
continue to purchase black-market copies of Windows. But if the
price difference is also due to hardware that cannot
be cheaply obtained, and the cheap hardware does not support the
flagship version of Windows, then the poor will have to be
content with the "down-market" Windows product, which will be
sold affordably.
With Evince on Gnome 2.10 I haven't been able to get print
to work at all. I need to use xpdf for that.
Rotating documents is another problem. Both of these
just _have_ to get fixed, and given that they
already work reasonably well in xpdf, I don't
understand why they're a problem in Evince/Popplar.
"configuring minor things should not be readily available to users"
I strongly disagree. Largely because what the developer considers a minor issue might be major to me, the regular user. Say rather that "configuring minor things should not be necessary" (because the system is designed so well). But it should still be possible and available.
It takes remarkable skill for a developer to listen to user comments of the form, "I want feature X, and feature Y,....", realize that features X and Y have merit but integrate poorly, and come back with a reconfiguration of the design incorporating the merits of X and Y but in a consistent and well-thought-out way. In my experience, I've only really seen this with in-house CAD developers who work very closely with their users.
Slashdot Mirror
It's being "nothing new," and already documented in a readme makes the shortcomings of WHS more outrageous, not less so.
Don't *store* frequently modified critical files such as Outlook (.ost) files on the server. Keep them local, and let WHS back them up like any other file. This excellent advice applies to all operating systems[...]No, it doesn't apply to all operating systems. In fact, if it applies at all, it only applies to Windows out of all the operating systems I've ever used (Digital Unix, VMS, Linux, Windows). At work (different jobs) for almost two decades now, all my files for all applications under VMS/Unix/Linux have been stored over the network--the local drive holds the operating system only. (And on VMS clusters the local drives of client systems only held swap space, the OS itself booted over the network.)
Given that:
1) the CEO, all of current management, sales and computer programmers who kept their mouths shut, remain in place,
2) the CEO being the same person who pledged to bring the elections over to the Republicans,
Um, no.
Walden O'Dell was the CEO who was committed to "help Ohio deliver its electoral votes to the President."
He was canned in 2005. Thomas Swidarski replaced him and then canned 5 of the 7 most senior Diebold execs.
And a reflection of mine as well. I reckon about 70-80% of the Excel spreadsheets I see are purely documentation--without a single formula containing more than a single number or date. I recently received directions to a hotel in a spreadsheet! There are those for whom Excel is the right tool, but most can do all their spreadsheeting in any spreadsheet program at all.
They credit Sir John Herschel for suggesting in 1828 that time should be the same in large regions, and claim that GMT had been widely adopted in England, Scotland, and Wales by mid-century.
However, one nation-wide time zone clearly wouldn't work over a continental nation like the U.S. So apparently C. F. Dowd, a principal of a seminary for girls, was the first to suggest zones in 1869, and Canadian Sandford Fleming then suggested making such a system world-wide. Railroads in the U.S. and Canada began operating on Standard Railway Time on 18 Nov. 1883, reducing the number of railroad times from at least 56 to 4.
Also interesting is the changing of the time zone boundaries over the years. People seem to prefer the sun setting later, and there was an advantage to being in the same zone as the east-coast population centers, so the western boundary of the Eastern Time Zone has moved from it's original location through the middle of Ohio to the western edge of Indiana.
On my Fedora system, the OOo calc package lists i386 as the architecture, so I'm wondering if SSE/SSE2 floating point is enabled. The old stack-based floating point is slow.
You mean like Itanium? [*] Backwards compatibility is the monopoly. A Windows that wouldn't run any of the existing Windows software out there wouldn't sell. People and businesses would continue buying the 'Win32 legacy' product. An Office that wouldn't read the old .DOC formats wouldn't sell. Et cetera.
* Strictly speaking, Itanium doesn't break backwards compatibility--it'll run X86 code. But if you intend to run X86 code Itanium is a poor choice. That's why AMD's X86-64 was such a great idea: substantial improvements and real backwards compatibility.
But I imagine (not my experience) that medium-sized companies can't afford to build their own customized setup, and then your comments apply.
So if I go buy a Windows XP license today, do those 5 years of mainstream support and 5 years of extended support start today, or from the introduction of XP?
harddrives of old machines they donated to schools.
The clearing was accomplished by writing random
data over the whole drive seven times. They
understandably didn't want who-knows-what internal information just being sent off to the outside. I've got to believe that everyone's doing this nowdays.
And in any case, larger companies lease licenses in bulk, and cannot transfer them. So the receiving charity
would lack the license to use the installed Windows (or Office, etc) anyways.
Yep, it's the registers. But in the case of AMD, it's also that memory latency is lower due to the integrated northbridge. On memory intensive jobs, it matters.
First, grab a dictionary and look up 'hyperbole'. Then, and this is the tough part, consider that what the original poster was expressing is true, and frequently so.
I subscribe to National Geographic, and read a fair share of the science articles in newspapers, and can attest that an enormous amount of inference about the behavior of long-extinct animals is made from a few fossilized bone fragments. I understand the difference between Nat. Geo. and a scientific journal, but the scientists are being interviewed and quoted for these articles.
I see this, and I think of the hyena. When I was a child, the hyena was a scavenger. Hardy lions hunted and brought down live game, and only afterwards did the hyenas move in to eat what was left. By the time I was a young adult, it had been discovered that hyenas hunt at night. Lions were found to have driven hyenas away from their kill, and the hyenas--powerless to stop it--had to wait until the lions had their fill. But now, years after that, I'm told that hyenas both hunt and scavenge.
This is great. We're learning more about hyenas. We see how our earlier observations were accurate, but incomplete. We refine our theories as we learn more. But consider that the hyena is an animal known to us historically (i.e., in writing) for thousands of years. We can go out with cameras and observe the behavior of living hyenas, and still get it wrong. How much more likely are we to be wrong when our observations are limited to a few fossils? There is no justification for the hubris so rampant in the study of fossils.
Starter Edition: AKA Checkbox Edition 1. Indeed aimed at the "two-thirds world", but not as a product. Its purpose is to have an argument against the governments of nations with mostly poor citizens who look the other way with piracy. With this, MS can say "but there is an affordable alternative."
Home Basic Edition: AKA Checkbox Edition 2. Aimed at Europe, with no bundled media apps. Purpose is to fight sanctions.
Home Premium Edition: This is the version most home users will buy, but now MS gets to charge even more because it is demonstrably more capable (media) than HBE.
Professional Edition: Perhaps the most aptly named. Aimed at the professional who buys his own laptop. Many are predicting that companies will stop issuing laptops to employees in the future, who will be expected to just have their own. Much as has happened with cell phones. But lacks the fancy multi-media apps to force more customers up to Ultimate.
Small Business Edition: Pro with some added business features, see article.
Enterprise Edition: Pro with all business features, see article. Virtualization will allow employees to run their personal stuff on a different virtual machine, protecting the corporate network.
Ultimate Edition: Pro with media, and as many bells and whistles as possible so that the price can be jacked as high as possible. Professionals who want the media features will have to pay through the nose to get them, for Windows Pro lacks them and HPE won't connect to the corporate network. Students are caught, too, for HPE won't connect to the campus network (same with XP Home today).
The old government General Accounting Office is now the Government Accountability Office.
CPU manufacturers particularly dislike overclocking because it leads to CPU failure that makes the manufacturer look like it didn't make a quality product. But since it did, that's not fair.
Lord only knows how pricing is done, but your part should be running as fast as it reliably can. There is no incentive for a chip maker to cripple clock speeds; the faster the part runs the more they can--and do--charge.
Now, if it ever came to be that there were only one CPU manufacturer, the game would change. The lone manufacturer might under-clock some parts to spread pricing options.
And I was using 'phone' on VMS in 1984. VMS dates to 1978 or so, don't know if phone goes back that far or not.
The USA accepted the metric system in 1866, but has never rejected the customary units. But there are a lot of caveats (metric-only labeling is rare, for example). But Americans seldom use metric units outside of medicine, engineering, and science (and anything electrical, since there are no customary electrical units).
(I say "customary" instead of "imperial" because the USA never adopted the British imperial standard of 1824. The difference between the two systems is substantial in the the units of capacity (e.g. the gallon).
But units of measure are perhaps the perfect example of open standards. And historically, they were enacted to protect consumers.
"good measure, pressed down, and shaken together"
It may be optional, but not because of (new) laptops. Laptops will mostly be UMA with the frame buffer in main memory (which should easily slide into that 2GB :-). Of course, the GPU will require a high-bandwidth path to memory.
True.
"Now, instead of just buying expensive licences, they have to upgrade the graphics cards on their vanilla work PCs??"
But license fees are not the only issue. Businesses don't want to switch from W2K (or won't want to switch from XP) because those systems are known to work with all their apps and software systems. In many cases, switching the operating system will entail a fair bit of expense and a great deal of risk. Unless there is enticing value on the the other side, why switch?
If there is a desire to switch, the cost of the required graphics card (likely in new PCs) will not influence the decision.
Nor do I. Going from 512M to 2G is 4x, so it's more than what's required for a doubling of pointer size. Does the page size increase in a 64-bit system?
Indeed. But certainly not to spitefully inflict hardship on customers. Arguably, an ideal world for MS would be one in which people spent none of their PC dollars on hardware, and all of them on MS software. And in fact, the business world is much like that--large customers usually buy licenses which are periodically renewed. But consumers need a reason to go buy new software, and MS's office suite has reached a point of maturity from where it's hard to increase the value, and thus the desirability, of the next version.
MS likely believes that the path to new consumer value is via multimedia. The better the GPU hardware, the greater the potential for MS to sell multimedia applications or services. So they create an OS requirement for top-of-the-line graphics hardware. PC makers comply, for they will not sell many machines if Windows isn't properly supported. And then MS can sell multimedia to properly equipped consumers. It's a theory anyway. (And of course they don't like being behind Apple in the eye candy dept.)
But let us concoct an even wilder theory. MS has called for a 100-dollar PC. Of course, they want the 100-dollar PC to go "down-market" in "some of these contries" where they see piracy. However, nobody _wants_ a down-market anything: if the price difference between a 100-dollar PC and a 250-dollar PC is due to the price difference between Windows and Windows--Starter Edition, the citizens of poor countries will continue to purchase black-market copies of Windows. But if the price difference is also due to hardware that cannot be cheaply obtained, and the cheap hardware does not support the flagship version of Windows, then the poor will have to be content with the "down-market" Windows product, which will be sold affordably.
With Evince on Gnome 2.10 I haven't been able to get print to work at all. I need to use xpdf for that. Rotating documents is another problem. Both of these just _have_ to get fixed, and given that they already work reasonably well in xpdf, I don't understand why they're a problem in Evince/Popplar.
I strongly disagree. Largely because what the developer considers a minor issue might be major to me, the regular user. Say rather that "configuring minor things should not be necessary" (because the system is designed so well). But it should still be possible and available.
It takes remarkable skill for a developer to listen to user comments of the form, "I want feature X, and feature Y, ....", realize that features X and Y have merit but integrate poorly, and come back with a reconfiguration of the design incorporating the merits of X and Y but in a consistent and well-thought-out way. In my experience, I've only really seen this with in-house CAD developers who work very closely with their users.