We're going to have to give up on recognition and put more effort into partitioning. We need setups where each web page renders in its own jail, and it doesn't matter if the browser is insecure - when the page closes, a program exits and any corrupted info goes away.
Of course, this will break Active-X, toolbars, downloads, etc. Then again, on business systems, you want those things broken.
Once the browser is locked down like that, you need a "guard" program. When you want to move a file out of a browser's jail, it has to go through a program that "sanitizes" it. Often, a translation to a well-documented format that doesn't contain execution capability will do the job.
Yeah, that's a very good idea... and that's why Microsoft has implemented almost exactly that with Internet Explorer 7+ in Vista. It's called "Protected Mode". Here's a good starting point to read about the technology.
ActiveX controls and toolbars can still operate inside this jailed environment (keep in mind that conceptually, binary code and toolbars aren't the -problem-... it's what they're able to access that has been the problem), but they don't have access to anything outside of it. If, due to the nature of the plug-in, it really should be able to work with the user's files, there is the Protected Mode API which gives plug-in developers their only way out of the jail, and it's extremely limited... basically, you can find out where you're allowed to write to on the file system or registry (for storing configuration information), and you can trigger a save dialog that is executed in the context of the parent IE process (the one that isn't jailed), and you can trigger the file save... and that's about it.
When the file save happens, Internet Explorer makes use of the OS's Attachment Execution Service interface to save the file. This is something that was introduced with XP SP2, and it offers a couple of things: it lets anti-virus, anti-malware etc. applications hook in and scan the file and give the thumbs-up before allowing the save. In other words, the file doesn't even touch the non-jailed environment (ie. your regular user profile) before it gets scanned. Another thing the AES does is attach the source URL of the file (if appropriate), and store it in an NTFS alternate data stream. AES is also used by Outlook Express (Windows Mail in Vista) and Microsoft's various IM programs.
As for sanitising data, go ahead and put an application out onto the market that will automatically translate Office document formats into equivalent OpenOffice formats if you like. I'm sure some people would be interested in it, though bear in mind that the Office formats contain a bunch of features that ODF doesn't support yet (ODF doesn't have a facility for highlighting text in a way that's distinct from background colours, for example)...
Uhhh.... hate to say it, but you're pretty badly misinformed.
Microsoft never talked about a ground-up rewrite of Windows for Longhorn/Vista -- ever. Their intention all along has been to build on the existing NT 5.x kernel and core architecture. You're welcome to provide links to prove me wrong on that, but it's pretty unlikely you'll find anything more than speculation.
Internet Explorer has never been part of the NT kernel -- ever. Where'd you get that idea from, Slashdot? IE's rendering engine (MSHTML) is a core system component, yes, but it runs entirely in user-mode. Same with the scripting components, the user interface, and all the protocol networking stuff (WinInet, etc.) Again, you're welcome to provide links to prove me wrong on that, but it's pretty unlikely you'll find anything. (If it aids in your understanding, think of it like this: saying that IE is part of the NT kernel is like saying that bash is part of the Linux kernel)
"Aero's.NET setup" is completely meaningless; Aero is a visual theme and a UI ethos, and nothing more. It has nothing to do with.NET. The Desktop Window Manager (which makes Aero possible) is written almost entirely in C++, and is built on DirectX 9. The.NET APIs are Windows Forms (which give you backwards compat prior Windows versions down to Win98), and Windows Presentation Foundation (which is for XP SP2, 2K3 SP1, and Vista only)
Also, there is absolutely no feasible way to do emulation of Win32 in a 'Mac OS Classic'-type fashion. Three major problems come to mind immediately: First, you have a serious problem in that you suddenly have no API to use when implementing device drivers. Vista does add the capability to write device drivers in.NET using the User Mode Driver Framework, but... what's the expectation here? That every hardware manufacturer on the planet is going to have to completely rewrite/all/ their drivers simply because Microsoft chose to dump Win32 with zero transition period? That's sheer madness; Vista would be a certain failure without good device driver compatibility. Second, you're now imposing a performance penalty on gamers, on every single game ever written. How is Direct3D supposed to be performant if you're putting a virtualisation wall in front of it? How is DirectSound supposed to work with the audio device if direct hardware access has been taken away? Third, what's supposed to happen to everything from Cygwin to Delphi? How about Java? Without Win32, every non-Microsoft language would be terminally fucked. What, you don't seriously think Sun is going to reimplement their JRE in the.NET Framework, do you? That's sheer madness!
Also, in your virtualisation candyland, that WMF flaw would have still existed, and it could've still corrupted other applications and data if properly exploited... after all, that virtualization environment *must* have access to your data to operate correctly, right? You can sandbox code all you want, but you can't sandbox user data in any sensible fashion.
I could name several more serious problems with the idea, too. It's a complete non-starter.
It's a lot harder than a "re-compiling with larger values". The "OLE container"-based Excel file format (XLS) stores row numbers in a two-byte field, thus creating a hard limitation of 65,536 possible values. In-memory representations of large spreadsheets are easy, of course, but you do have to be able to save these files... that's why Excel 2007 has an entirely new XML-based file format (XLSX), so they don't get stuck in this kind of situation again.
To see what I mean about this limitation (and others) with the binary-based format, you can read the Excel file format specification that the OpenOffice folks published. It's here: http://sc.openoffice.org/excelfileformat.pdf
To be clear about this, SMB 2.0 is only used when it determines that the machine on the other end supports this new, more efficient, version of the protocol. The original SMB is still used when communicating with all previous versions of Windows, as well as Samba. It says so right in the article. Furthermore, I know it works because I myself have had no trouble at all connecting the latest Vista beta to my Linux-based Samba server. Look, we all know the SMB protocol basically sucks and is inefficient; an upgrade which fixes its most common problems has been long overdue, and Microsoft is the only company in a position to make it happen. What we should be hoping for at this point is that theose EU lawsuits will result in full documentation of the SMB 2.0 protocol being produced by Microsoft so that Tridge & co. can implement it quickly. If that happens, great, if not, we still have SMB 1.0.
As for Services for Macintosh... is it really needed these days, considering that OS X's Windows networking support (built on Samba) is pretty good? Probably the biggest loss is that Vista won't be able to directly mount AFP-only machines like Mac OS 9, but there are Samba implementations available for OS 9 anyhow... SFM is pretty horrific anyhow -- they never really got it right. What do you trust more: that Microsoft will get an implementation of Mac networking right? Or that Apple (who rides on Trige & co.'s coattails) will get an implementation of Windows networking right? Personally, I opt for the latter.
http://en.wikipedia.org/wiki/Features_new_to_Windo ws_Vista... read the whole thing (yes, it's long and wordy), and then explain to me how Monad and WinFS are collectively better than the hundreds of improvements listed there. Be sure you understand what Vista *is*, as opposed to what it *isn't*.
It's actually pretty easy for a company to prevent employees from writing to mass storage devices with XP SP2: Change one registry key on every machine... simple stuff with an Active Directory environment.
More significantly though, this kind of thing really makes a case for Microsoft's Rights Management Services technology... even if you were able to copy the physical documents onto an iPod, they'd be completely useless to you outside the organization because they're encrypted, and only by talking to the RMS server (located internally) can they be unlocked.
No, actually, MSHTML doesn't go "deeper". It's just a library that (like Webkit) renders HTML and a few other standards. You really don't understand this on a deep technical level, do you? MSHTML isn't tied to the "kernel" by any stretch of the imagination... however, certain parts of Windows, such as the help system, use MSHTML as their rendering engine, so that they don't have to go inventing another markup language to present the same fundamental kind of hyperlinked content that the web itself is based on. That's called component reuse.
The reason MSHTML has been such a major source of security vulnerabilities over the years is because it allowed many arbitrary OS components to be instantiated inside of it, with the same permissions as the user. That's not really a failing of the OS -- that's a design failure of MSHTML! There is an important difference here, and you'd do well to understand it before levelling blame.
That's a common misconception. Sure, "Safari" can be easily removed, just like "Internet Explorer" can, but the real meat of the rendering and parsing work is done by WebKit, which is also used for Dashboard, Mail, and a number of third-party applications (notably Adium, SubEthaEdit, Colloquy, Xcode and NetNewsWire)... in this sense it's EXACTLY like Windows' MSHTML and MSXML, in that it is a required, integrated component of the operating system.
A lot of the money that the Gates Foundation doles out ends up supporting people who are REALLY poor. This isn't about people who care about whether or not their computer software is "free" or "pirated" here.
This is about people who don't have enough money for *clothes* or *food*. This is about children who won't survive the first two years of their life because of diseases. This is about fighting AIDS. This is about whole regions where electricty and water are scarce.
The fact that Linux is free, or even *exists* for that matter, is simply not a concern for the starving. Between Gates, Stallman, Jobs, and Torvalds, Gates is the/only/ one of them that has taken serious time to learn the problems of these regions, see it for himself, and try to help. Heck... Steve Jobs calls up Bono when he wants to sell a product to well-to-do urban people in first-world countries. Bill Gates calls up Bono to talk about the needs of the poor, and what they can do to help.
How about games? I'm sure I'm not the only Mac user who would love to be able to run *any* PC game on a MacBook or iMac without having to deal with compatibility issues, or have a separate "gaming PC" around.
Plus, there are many people who use a single laptop at work & home. A lot of them are required by work to be running Windows (for domain functionality, line of business applications, etc.)... a MacBook running dual-boot Windows and OS X would be perfect for Mac users who need to work in this environment. Windows for work; OS X for personal stuff.
Hahaha. Allright, bro, you bookmark this posting and if Vista comes out without the functionality I've described, get in touch with me and I will Paypal you $20 USD. I think this stupid Windows Update website crap is going to be a thing of the past, and I'm willing to bet money on it.
It's not. I checked. There is no ActiveX control in Vista for Windows Update. Those pages are part of a control panel applet now; it gets its information about available updated by calling a web service, exactly the same way that the Automatic Updates client in Windows XP gets its information.
It's changed, okay? Save yourself the embarassment and come around to this line of thinking -- it's the way it is n Vista. Posting screenshots of Windows XP is completely irrelevant to a discussion about how the technology works in Vista. COMPLETELY irrelevant.
You are, of course, welcome to try it out for yourself and confirm the technical accuracy of what I'm saying instead of jumping to unfounded conclu.... oh, shit, this is Slashdot, never mind that idea.
Microsoft *has* changed Windows Update in Vista so that it is no longer a web application. Here, I took a screenshot of this for you.
Microsoft *has* disabled the execution of any ActiveX by default in Internet Explorer 7, too. Do some basic research (hint: google "ActiveX IE7") and you can learn more about it.
XmlHttpRequest is a genuine Microsoft innovation. It's been around since 1999, but nobody really knew about it until Google created Gmail and people started disassembling Gmail's code wondering "how did they make it so f'ing fast?" Every other browser out there now implements a variation of this particular Microsoft technology; even Apple credits Microsoft for this.
Microsoft has spent the last year working almost solely on improving Entourage to work better with Exchange. Last week they released Office 2004 Service Pack 2, which contains improvements to everything you've noted as being problems: Public folder support; sharing of mailboxes, calendars, contacts; complete global address list support; ability to do delegation... and so on and so forth.
More information on MS's website.
Excuse me, but what makes you qualified to talk on this subject?
First, YOU CAN'T EVEN GET THE NAME OF THE FILE CORRECT. It's NOT called a "swap" file -- it's called a "paging" file. These are two different things... yes, it used to be called a swapfile in the old, old days of Windows, but it's not any more. This fundamental oversight alone invalidates every other statement you make about Windows's virtual memory subsystem.
Second, the PAGING FILE is a single, contiguous file on a fresh OS install. Try it sometime, you'll see that I'm right. The paging file is created to a default size of (physical memory * 1.5), and will remain at that size, and at its original unfragmented location on the HDD, until you attempt to use more than 250% of your machine's physical memory. You will be notified by Windows when this happens -- your machine will also slow to a crawl given how far you are overextending it.
Third, a fixed min/max PAGING FILE size opens you up to a very clear and obvious problem (well, at least to people who think about it): If you fill up your physical memory and pagefile, Windows cannot make use of your HDD to provide additional memory for your applications. End result? Memory allocation requests start failing! How could this possibly be a good thing?
Fourth, the ONLY time that the PAGING FILE resizes, is if additional memory is required. It never, ever happens at any other time. Ever.
Quite simply, "value added", you do NOT understand how Windows manages memory. I realise it's a Slashdot mantra to propogate inaccurate and incorrect information about Windows, but you should be ashamed of yourself for doing so. There is some good reading out there on the subject, like Microsoft Windows Internals, Fourth Edition, which explains how the underlying Windows kernel technologies -- including virtual memory -- actually work.
Yeah, and Microsoft also has a free app which does really fast local searching, even with search-as-you-type capability -- but unlike Google, it can also index network drives.
I also tried to get my 53-year old mother to switch to Firefox as well, recently... she seemed okay with it at first, but after a few days she switched back to Internet Explorer. She simply didn't perceive any fundamental benefit of Firefox over IE6 (especially after SP2), and really, to a lot of end-users there isn't anything really noteworthy.
I also got her to switch from Winamp over to iTunes recently... and that stuck REALLY fast. She found a lot of benefit in being able to buy music online through iTMS, burn and rip CD's, and learn more about artists.
Moral of the story: If you mom's smart enough to raise you to be a decent adult, then she's also smart enough to perceive when learning something new is going to give her a measurable benefit.
"C:\Program Files\Messenger" contains the 'Windows Messenger' product, which is the IM client that is part of Windows XP... MSN Messenger is a different program altogether, and is installed in a different directory. You can run Windows Messenger and MSN Messenger at the same time.
Wouldn't it be better if Microsoft was to fix their bloody insecure software instead??
Yeah, and they've got a pretty effective solution called NGSCB in the works, but I bet you'll be ready to shit on that simply because we've been told so many times by anti-Microsoft zealots how evil it will be!
Remember, viruses aren't typically spread via exploitable vulnerabilities in Windows... they are attached to executable files, or are themselves complete executables, and are spread by way of social engineering (Anna Kournikova's tits, e.g.) and lax security on file shares and the like. A virus can be as benign-seeming as a screen saver.
When's the last time we heard about a virus spreading by way of an unpatched Windows security vulnerability, anyhow? It's been a long time.
Slashdot Mirror
We're going to have to give up on recognition and put more effort into partitioning. We need setups where each web page renders in its own jail, and it doesn't matter if the browser is insecure - when the page closes, a program exits and any corrupted info goes away.
... it's what they're able to access that has been the problem), but they don't have access to anything outside of it. If, due to the nature of the plug-in, it really should be able to work with the user's files, there is the Protected Mode API which gives plug-in developers their only way out of the jail, and it's extremely limited... basically, you can find out where you're allowed to write to on the file system or registry (for storing configuration information), and you can trigger a save dialog that is executed in the context of the parent IE process (the one that isn't jailed), and you can trigger the file save... and that's about it.
Of course, this will break Active-X, toolbars, downloads, etc. Then again, on business systems, you want those things broken.
Once the browser is locked down like that, you need a "guard" program. When you want to move a file out of a browser's jail, it has to go through a program that "sanitizes" it. Often, a translation to a well-documented format that doesn't contain execution capability will do the job.
Yeah, that's a very good idea... and that's why Microsoft has implemented almost exactly that with Internet Explorer 7+ in Vista. It's called "Protected Mode".
Here's a good starting point to read about the technology.
ActiveX controls and toolbars can still operate inside this jailed environment (keep in mind that conceptually, binary code and toolbars aren't the -problem-
When the file save happens, Internet Explorer makes use of the OS's Attachment Execution Service interface to save the file. This is something that was introduced with XP SP2, and it offers a couple of things: it lets anti-virus, anti-malware etc. applications hook in and scan the file and give the thumbs-up before allowing the save. In other words, the file doesn't even touch the non-jailed environment (ie. your regular user profile) before it gets scanned. Another thing the AES does is attach the source URL of the file (if appropriate), and store it in an NTFS alternate data stream. AES is also used by Outlook Express (Windows Mail in Vista) and Microsoft's various IM programs.
As for sanitising data, go ahead and put an application out onto the market that will automatically translate Office document formats into equivalent OpenOffice formats if you like. I'm sure some people would be interested in it, though bear in mind that the Office formats contain a bunch of features that ODF doesn't support yet (ODF doesn't have a facility for highlighting text in a way that's distinct from background colours, for example)...
Uhhh.... hate to say it, but you're pretty badly misinformed.
.NET setup" is completely meaningless; Aero is a visual theme and a UI ethos, and nothing more. It has nothing to do with .NET. The Desktop Window Manager (which makes Aero possible) is written almost entirely in C++, and is built on DirectX 9. The .NET APIs are Windows Forms (which give you backwards compat prior Windows versions down to Win98), and Windows Presentation Foundation (which is for XP SP2, 2K3 SP1, and Vista only)
.NET using the User Mode Driver Framework, but... what's the expectation here? That every hardware manufacturer on the planet is going to have to completely rewrite /all/ their drivers simply because Microsoft chose to dump Win32 with zero transition period? That's sheer madness; Vista would be a certain failure without good device driver compatibility. Second, you're now imposing a performance penalty on gamers, on every single game ever written. How is Direct3D supposed to be performant if you're putting a virtualisation wall in front of it? How is DirectSound supposed to work with the audio device if direct hardware access has been taken away? Third, what's supposed to happen to everything from Cygwin to Delphi? How about Java? Without Win32, every non-Microsoft language would be terminally fucked. What, you don't seriously think Sun is going to reimplement their JRE in the .NET Framework, do you? That's sheer madness!
Microsoft never talked about a ground-up rewrite of Windows for Longhorn/Vista -- ever. Their intention all along has been to build on the existing NT 5.x kernel and core architecture. You're welcome to provide links to prove me wrong on that, but it's pretty unlikely you'll find anything more than speculation.
Internet Explorer has never been part of the NT kernel -- ever. Where'd you get that idea from, Slashdot? IE's rendering engine (MSHTML) is a core system component, yes, but it runs entirely in user-mode. Same with the scripting components, the user interface, and all the protocol networking stuff (WinInet, etc.) Again, you're welcome to provide links to prove me wrong on that, but it's pretty unlikely you'll find anything. (If it aids in your understanding, think of it like this: saying that IE is part of the NT kernel is like saying that bash is part of the Linux kernel)
"Aero's
Also, there is absolutely no feasible way to do emulation of Win32 in a 'Mac OS Classic'-type fashion. Three major problems come to mind immediately: First, you have a serious problem in that you suddenly have no API to use when implementing device drivers. Vista does add the capability to write device drivers in
Also, in your virtualisation candyland, that WMF flaw would have still existed, and it could've still corrupted other applications and data if properly exploited... after all, that virtualization environment *must* have access to your data to operate correctly, right? You can sandbox code all you want, but you can't sandbox user data in any sensible fashion.
I could name several more serious problems with the idea, too. It's a complete non-starter.
It's so that Slashdot posters can beta-test their latest anti-Microsoft talking points!
It's a lot harder than a "re-compiling with larger values". The "OLE container"-based Excel file format (XLS) stores row numbers in a two-byte field, thus creating a hard limitation of 65,536 possible values. In-memory representations of large spreadsheets are easy, of course, but you do have to be able to save these files... that's why Excel 2007 has an entirely new XML-based file format (XLSX), so they don't get stuck in this kind of situation again.
To see what I mean about this limitation (and others) with the binary-based format, you can read the Excel file format specification that the OpenOffice folks published. It's here: http://sc.openoffice.org/excelfileformat.pdf
To be clear about this, SMB 2.0 is only used when it determines that the machine on the other end supports this new, more efficient, version of the protocol. The original SMB is still used when communicating with all previous versions of Windows, as well as Samba. It says so right in the article. Furthermore, I know it works because I myself have had no trouble at all connecting the latest Vista beta to my Linux-based Samba server. Look, we all know the SMB protocol basically sucks and is inefficient; an upgrade which fixes its most common problems has been long overdue, and Microsoft is the only company in a position to make it happen. What we should be hoping for at this point is that theose EU lawsuits will result in full documentation of the SMB 2.0 protocol being produced by Microsoft so that Tridge & co. can implement it quickly. If that happens, great, if not, we still have SMB 1.0.
As for Services for Macintosh... is it really needed these days, considering that OS X's Windows networking support (built on Samba) is pretty good? Probably the biggest loss is that Vista won't be able to directly mount AFP-only machines like Mac OS 9, but there are Samba implementations available for OS 9 anyhow... SFM is pretty horrific anyhow -- they never really got it right. What do you trust more: that Microsoft will get an implementation of Mac networking right? Or that Apple (who rides on Trige & co.'s coattails) will get an implementation of Windows networking right? Personally, I opt for the latter.
http://en.wikipedia.org/wiki/Features_new_to_Wind
It's actually pretty easy for a company to prevent employees from writing to mass storage devices with XP SP2: Change one registry key on every machine... simple stuff with an Active Directory environment.
More significantly though, this kind of thing really makes a case for Microsoft's Rights Management Services technology... even if you were able to copy the physical documents onto an iPod, they'd be completely useless to you outside the organization because they're encrypted, and only by talking to the RMS server (located internally) can they be unlocked.
No, actually, MSHTML doesn't go "deeper". It's just a library that (like Webkit) renders HTML and a few other standards. You really don't understand this on a deep technical level, do you? MSHTML isn't tied to the "kernel" by any stretch of the imagination... however, certain parts of Windows, such as the help system, use MSHTML as their rendering engine, so that they don't have to go inventing another markup language to present the same fundamental kind of hyperlinked content that the web itself is based on. That's called component reuse.
The reason MSHTML has been such a major source of security vulnerabilities over the years is because it allowed many arbitrary OS components to be instantiated inside of it, with the same permissions as the user. That's not really a failing of the OS -- that's a design failure of MSHTML! There is an important difference here, and you'd do well to understand it before levelling blame.
That's a common misconception. Sure, "Safari" can be easily removed, just like "Internet Explorer" can, but the real meat of the rendering and parsing work is done by WebKit, which is also used for Dashboard, Mail, and a number of third-party applications (notably Adium, SubEthaEdit, Colloquy, Xcode and NetNewsWire)... in this sense it's EXACTLY like Windows' MSHTML and MSXML, in that it is a required, integrated component of the operating system.
A lot of the money that the Gates Foundation doles out ends up supporting people who are REALLY poor. This isn't about people who care about whether or not their computer software is "free" or "pirated" here.
/only/ one of them that has taken serious time to learn the problems of these regions, see it for himself, and try to help. Heck... Steve Jobs calls up Bono when he wants to sell a product to well-to-do urban people in first-world countries. Bill Gates calls up Bono to talk about the needs of the poor, and what they can do to help.
This is about people who don't have enough money for *clothes* or *food*. This is about children who won't survive the first two years of their life because of diseases. This is about fighting AIDS. This is about whole regions where electricty and water are scarce.
The fact that Linux is free, or even *exists* for that matter, is simply not a concern for the starving. Between Gates, Stallman, Jobs, and Torvalds, Gates is the
How about games? I'm sure I'm not the only Mac user who would love to be able to run *any* PC game on a MacBook or iMac without having to deal with compatibility issues, or have a separate "gaming PC" around.
Plus, there are many people who use a single laptop at work & home. A lot of them are required by work to be running Windows (for domain functionality, line of business applications, etc.)... a MacBook running dual-boot Windows and OS X would be perfect for Mac users who need to work in this environment. Windows for work; OS X for personal stuff.
FYI, Vista's firewall is a two-way firewall.
s /cableguy/cg0106.mspx
More details: http://www.microsoft.com/technet/community/column
Hahaha. Allright, bro, you bookmark this posting and if Vista comes out without the functionality I've described, get in touch with me and I will Paypal you $20 USD. I think this stupid Windows Update website crap is going to be a thing of the past, and I'm willing to bet money on it.
It's not. I checked. There is no ActiveX control in Vista for Windows Update. Those pages are part of a control panel applet now; it gets its information about available updated by calling a web service, exactly the same way that the Automatic Updates client in Windows XP gets its information.
It's changed, okay? Save yourself the embarassment and come around to this line of thinking -- it's the way it is n Vista. Posting screenshots of Windows XP is completely irrelevant to a discussion about how the technology works in Vista. COMPLETELY irrelevant.
You are, of course, welcome to try it out for yourself and confirm the technical accuracy of what I'm saying instead of jumping to unfounded conclu.... oh, shit, this is Slashdot, never mind that idea.
Microsoft *has* changed Windows Update in Vista so that it is no longer a web application. Here, I took a screenshot of this for you.
Microsoft *has* disabled the execution of any ActiveX by default in Internet Explorer 7, too. Do some basic research (hint: google "ActiveX IE7") and you can learn more about it.
Does this satisfactorily address your concerns?
The Evil Bit is inherently evil! :-)
Ahem, "embarrassing" is spelled with TWO R's, not one.
XmlHttpRequest is a genuine Microsoft innovation. It's been around since 1999, but nobody really knew about it until Google created Gmail and people started disassembling Gmail's code wondering "how did they make it so f'ing fast?" Every other browser out there now implements a variation of this particular Microsoft technology; even Apple credits Microsoft for this.
Microsoft has spent the last year working almost solely on improving Entourage to work better with Exchange. Last week they released Office 2004 Service Pack 2, which contains improvements to everything you've noted as being problems: Public folder support; sharing of mailboxes, calendars, contacts; complete global address list support; ability to do delegation... and so on and so forth. More information on MS's website.
I have fifteen Phish concerts shared out over Bittorrent, you insensitive clod!
Excuse me, but what makes you qualified to talk on this subject?
First, YOU CAN'T EVEN GET THE NAME OF THE FILE CORRECT. It's NOT called a "swap" file -- it's called a "paging" file. These are two different things... yes, it used to be called a swapfile in the old, old days of Windows, but it's not any more. This fundamental oversight alone invalidates every other statement you make about Windows's virtual memory subsystem.
Second, the PAGING FILE is a single, contiguous file on a fresh OS install. Try it sometime, you'll see that I'm right. The paging file is created to a default size of (physical memory * 1.5), and will remain at that size, and at its original unfragmented location on the HDD, until you attempt to use more than 250% of your machine's physical memory. You will be notified by Windows when this happens -- your machine will also slow to a crawl given how far you are overextending it.
Third, a fixed min/max PAGING FILE size opens you up to a very clear and obvious problem (well, at least to people who think about it): If you fill up your physical memory and pagefile, Windows cannot make use of your HDD to provide additional memory for your applications. End result? Memory allocation requests start failing! How could this possibly be a good thing?
Fourth, the ONLY time that the PAGING FILE resizes, is if additional memory is required. It never, ever happens at any other time. Ever.
Quite simply, "value added", you do NOT understand how Windows manages memory. I realise it's a Slashdot mantra to propogate inaccurate and incorrect information about Windows, but you should be ashamed of yourself for doing so. There is some good reading out there on the subject, like Microsoft Windows Internals, Fourth Edition, which explains how the underlying Windows kernel technologies -- including virtual memory -- actually work.
Yeah, and Microsoft also has a free app which does really fast local searching, even with search-as-you-type capability -- but unlike Google, it can also index network drives.
It's called MSN Desktop Search.
I also tried to get my 53-year old mother to switch to Firefox as well, recently... she seemed okay with it at first, but after a few days she switched back to Internet Explorer. She simply didn't perceive any fundamental benefit of Firefox over IE6 (especially after SP2), and really, to a lot of end-users there isn't anything really noteworthy.
I also got her to switch from Winamp over to iTunes recently... and that stuck REALLY fast. She found a lot of benefit in being able to buy music online through iTMS, burn and rip CD's, and learn more about artists.
Moral of the story: If you mom's smart enough to raise you to be a decent adult, then she's also smart enough to perceive when learning something new is going to give her a measurable benefit.
"C:\Program Files\Messenger" contains the 'Windows Messenger' product, which is the IM client that is part of Windows XP... MSN Messenger is a different program altogether, and is installed in a different directory. You can run Windows Messenger and MSN Messenger at the same time.
Wouldn't it be better if Microsoft was to fix their bloody insecure software instead??
Yeah, and they've got a pretty effective solution called NGSCB in the works, but I bet you'll be ready to shit on that simply because we've been told so many times by anti-Microsoft zealots how evil it will be!
Remember, viruses aren't typically spread via exploitable vulnerabilities in Windows... they are attached to executable files, or are themselves complete executables, and are spread by way of social engineering (Anna Kournikova's tits, e.g.) and lax security on file shares and the like. A virus can be as benign-seeming as a screen saver.
When's the last time we heard about a virus spreading by way of an unpatched Windows security vulnerability, anyhow? It's been a long time.