Slashdot Mirror
"Windows 7 Compatible" PCs Must Be 64-bit
Barence writes "Microsoft has started certifying PCs as 'compatible with Windows 7' — and is looking to avoid the mistakes that dogged the Vista-Capable scheme. Whereas Microsoft certified PCs that could only run Vista Home Basic last time around, this time PCs will have to work with all versions of Windows 7 to qualify for the sticker, including 64-bit versions of the OS. Microsoft also claims, 'products that receive the logo are checked for common issues to minimize the number of crashes, hangs, and reboots experienced by the user.'"
Because if they didn't release a 32-bit edition of the OS, it would piss off too many people. You'd have noticeable faction of people up in arms. I'm all for 64-bit computing, I'm not looking back. But there's enough people out there with 1GB of RAM or less that would complain. 64-bit OSes and 64-bit applications have a slightly larger memory footprint because pointers, offsets, and certain kernel object handles are suddenly 64-bits in length instead of 32-bit.
What? They're not ending support of the 32-bit installer. There's no "abandoning" occurring. It figures the trolls are the ones who read the summary backwards and upside down.
The thing that is going to cause havoc with the VT requirement is that it intel went through a period, I'm not sure if they are still in it, where they disabled it on a seemingly arbitrary subset of their CPUs, with only minor differences in model name. Then, of course, vendors worked their BIOS magic. Just look at this list. You have an E7400, do you have VT? Well, do you have an E7400-SLGQ8 or an E7400-SLGW3? It's nothing that your IT department couldn't slog through for you(and if you are really lucky, they've been speccing for it for some time now); but I pity the plight of the adventurous but dubiously detail oriented guy who learns that XP mode isn't going to happen because he has the Q8300-SLB5W rather than the Q8300-SLGUR.
If it were something like "You need a Xeon for it to work", that'd be annoying; but it wouldn't really confuse anybody. As it is, though, there are going to be a whole lot of confused people out there.
The sticker in question (Windows 7 Compatible) is not intended for use on a computer -- it's intended for peripherals and add-ons. Mice, keyboards, graphics cards, network cards, routers, etc. etc.
.
What the hell is wrong this site? Are the editors becoming so lazy that they don't stop for two seconds to understand the stupidity of their headlines? You would think that Win7 isn't being offered in 32-bit mode from reading it. Instead, what it means is that any device you buy with that sticker will work with 32-bit windows and 64-bit windows.
this time PCs will have to work with all versions of Windows 7 to qualify for the sticker
Nonsense, there are lot's of systems out there, particularly Netbooks, which will not. Certainly will not necessarily be 64-bit.
If it only ran on 64-bit-capable systems, why is there a 32-bit version of Win 7 at all?
What exactly are you not understanding? This has exactly zero to do with a machine's ability to run Windows 7. This has everything to do with whether or not the manufacturer gets to put a little sticker on the case. The lack of the sticker does not mean that the computer is not capable of running any version of Windows 7, it simply means that the computer has not been certified to run every version of Windows 7.
"Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
Cue the Linux fanbois... ...screaming about how Bill is abandoning their customers after YEARS of support, whilst the Penguin does the same with 2 years of a kernel release.
Note the silence of the Mac Jihad.
I guess you read the summary backwards and didn't even consider clicking on the article.
I'm no Microsoft fan (Linux purist of 6 years now) but they are merely requiring hardware makers to provide stable 32-bit and 64-bit drivers in order to get a "Works with Windows 7 Certification."
This is a good thing for every day people.
Just recently I tried to help out a friend with a Vista 64bit computer to get his Hauppauge WinTV PVR 150 to work. Apparently it does not support any more than 3GB of RAM and is basically unusable (he has 8GB of RAM). It causes programs to crash and flat-out will not work with Pinnacle Studio 9.
Hauppauge claims it has something to do with the 64bit memory allocation or something. I can't quite remember what it was.
Maybe this will require them to revisit their drivers and make it "Just Work" like it should.
There is free support for Linux via the newsgroups, forums, Wiki sites, HOWTOs, Man pages, and many other things. But beware of the trolls that like to bite the n00bz and say RTFM. You need to have actually read the Linux manual before asking questions which consists of man pages.
Man ls
For example will display a man page for the "ls" command which functions like the MS-DOS "dir" command and some Linux distros will have a "dir" batch file to help DOS users adapt.
Judging from you attitude this sort of thing happened to you.
Remember, Slashdot does not have a -1 disagree moderation, and no, troll, flamebait, and overrated are not substitutes.
It may run fine, but there are interoperability problems with win64 and Outlook - serious ones.
The Exchange management applets for mailbox moves and such use mapi functions from Outlook.
Because of shitty planning, you can't run these applets on a win64 machine. You have to run them from a 32 bit machine with the tools installed.
ExMerge is only an option if you have old ansi psts - mine are all unicode.
The point is that there *still* are major issues with 64bit systems and interoperability of productivity software, not to mention hardware support.
That's absolute bullshit (or fud). You expressly do not have to pay Microsoft to get their approval. You just need to sign your application / device driver using a certificate from a bunch of trusted CAs.
.
See here for a list of trusted CAs: http://msdn.microsoft.com/en-us/library/ms995347.aspx
> You don't think firefox uses a huge amount of pointers?
No. Most of the data is page content in various forms and stages of processing. The pointers would only be pointing to it, and would be comparatively few in number.
> The last time I saw a comparison by someone who was advocating that 64bit was always better
> showed a 20-30% RAM increase, and then tried to pretend that "didn't matter".
Funny, I can't /find/ a decent comparison anywhere. All I see is blanket statements of "64bit is bigger", and all the examples usually only measure executable size. Yes, executable size is larger, but the reason is not obvious. If you do a size -A comparison of both, you'll see that the size increase comes entirely from the .eh_frame section, which is needed on x64 because code normally does not contain frame pointers. This section is loadable, but is not paged in unless you throw an exception (by definition, an exceptional event), or get a backtrace for debugging. The actual loaded code is 10-20% smaller for what I have tested, so what you get here is a disk size penalty (which doesn't matter because the extra data isn't read until an exception), and smaller code size in RAM (which DOES matter).
> if for some reason you can't get 8GB of RAM, then you should seriously consider only using 32bit, IMNSHO.
Man, you really need a reality check. My Linux system (x64, of course) is currently using only 308M total, with the KDE beast and whatever crap it thinks it needs, and firefox. There's absolutely no reason to require ungodly amounts of RAM for normal operation, whether on x32 or x64.