Slashdot Mirror

Windows uses WHQL signed drivers - which means that Microsoft have said that the driver is OK.

Sadly, there is a limit to what even Microsoft can test.

If the driver is not OK, WHQL means nothing (ie. MS don't care about stability). If the driver is OK, and the hardware is OK (works under the other OS) that leaves only one option - that Windows isn't 100% perfect! Sorry fanboys, the truth hurts sometimes.

Your reasoning is so faulty it's hard to know where to start. But let's try this: have you considered the possibility that Linux's usage patterns of the particular bit of hardware happen to mask the hardware problem, whereas Windows' usage patterns cause the problem to surface? Or vice versa, in other situations. I've experienced both scenarios (and some others including OS X). Perhaps you have only experienced one?

I've never understood why this is technically a difficult or impractical solution.

This is not a technical issue. I'll repeat that - it is not a technical issue.

Microsoft could indeed do that, but then, say, when you upgraded to XP, half your hardware would stop working*. Add-on hardware/peripherals are highly commoditised, which means they don't spend a whole bunch of time achieving quality bars for drivers. If it runs, they ship it (and sometimes if it doesn't).

Compatibility is one of the reasons Windows is so successful - just read a few of Raymond Chen's posts on the subject, for example. I mean, here's one where he's talking about supporting undocumented behaviour, and gives reasons why MS does this. And you want them to drop documented, correct behaviour? The Microsoft conspiracists would go mad, for starters.

Here's a choice quote from Raymond from that post:

"I recall a survey taken a few years ago by our Setup/Upgrade team of corporations using Windows. Pretty much every single one has at least one "deal-breaker" program, a program which Windows absolutely must support or they won't upgrade."

Of course, there are other people who believe that this pursuit of compatibility is wholly wrong. I'm not totally in either camp, but I do like it when I upgrade Windows and most stuff still works, so I'm more closely aligned with Raymond's views.

In lieue of a technical explanation, I've assumed that this could be considered good for microsoft: making hardware compatiability an expensive task is good for the market giant, I guess (as the present discussion indicates).

Given that MS want as much hardware as possible to work with their OS, why would this be a good thing? Are you saying they engineered this situation purely so that Apple wouldn't be able to enter the commodity PC OS market?

Like Napoleon, I prefer the explanation of incompetence. Most driver writers suck at their jobs (this is a generalisation of the fact that most people suck at their jobs). They have people shouting at them to finish the software so their employer can ship the hardware, and if they're one of the few who do bring up the subject of quality people look at them funny.

Why do we need a hundred modem drivers? Is there really no other way?

Modem drivers are possibly a bad example, because most modem drivers use the Windows 'Unimodem' driver, which is a standard modem driver that can be tweaked by a text/setup file to indicate which features they support, and where they deviate from the norm. So a lot of modem 'drivers' are really just a short text file. (That doesn't mean modem drivers can't be a pain - I once spent about a day trying to find a driver for the Diamond USB modem I wanted to give someone in my family to use for dial-up internet access - unfortunately, Diamond had sold their modem business off to someone else, who didn't care if you could get the drivers for old modems or not. It was quicker to throw the POS away in the end.)

* Leave it.

I've never understood why this is technically a difficult or impractical solution.

This is not a technical issue. I'll repeat that - it is not a technical issue.

Microsoft could indeed do that, but then, say, when you upgraded to XP, half your hardware would stop working*. Add-on hardware/peripherals are highly commoditised, which means they don't spend a whole bunch of time achieving quality bars for drivers. If it runs, they ship it (and sometimes if it doesn't).

Compatibility is one of the reasons Windows is so successful - just read a few of Raymond Chen's posts on the subject, for example. I mean, here's one where he's talking about supporting undocumented behaviour, and gives reasons why MS does this. And you want them to drop documented, correct behaviour? The Microsoft conspiracists would go mad, for starters.

Here's a choice quote from Raymond from that post:

"I recall a survey taken a few years ago by our Setup/Upgrade team of corporations using Windows. Pretty much every single one has at least one "deal-breaker" program, a program which Windows absolutely must support or they won't upgrade."

Of course, there are other people who believe that this pursuit of compatibility is wholly wrong. I'm not totally in either camp, but I do like it when I upgrade Windows and most stuff still works, so I'm more closely aligned with Raymond's views.

In lieue of a technical explanation, I've assumed that this could be considered good for microsoft: making hardware compatiability an expensive task is good for the market giant, I guess (as the present discussion indicates).

Given that MS want as much hardware as possible to work with their OS, why would this be a good thing? Are you saying they engineered this situation purely so that Apple wouldn't be able to enter the commodity PC OS market?

Like Napoleon, I prefer the explanation of incompetence. Most driver writers suck at their jobs (this is a generalisation of the fact that most people suck at their jobs). They have people shouting at them to finish the software so their employer can ship the hardware, and if they're one of the few who do bring up the subject of quality people look at them funny.

Why do we need a hundred modem drivers? Is there really no other way?

Modem drivers are possibly a bad example, because most modem drivers use the Windows 'Unimodem' driver, which is a standard modem driver that can be tweaked by a text/setup file to indicate which features they support, and where they deviate from the norm. So a lot of modem 'drivers' are really just a short text file. (That doesn't mean modem drivers can't be a pain - I once spent about a day trying to find a driver for the Diamond USB modem I wanted to give someone in my family to use for dial-up internet access - unfortunately, Diamond had sold their modem business off to someone else, who didn't care if you could get the drivers for old modems or not. It was quicker to throw the POS away in the end.)

* Leave it.

Keyboard accelerators are a standard Windows thing. Before Win2k they were always visible, but after that they are hidden until you press Alt.

http://blogs.msdn.com/oldnewthing/archive/2005/05/ 03/414317.aspx

Office has always had shortcuts as far as I know, certainly Office 97 has them. Office does it's own UI though, but I think it follows the OS standard at the time it was released. Probably they were hidden until you press Alt in Office 2000. It's hard to tell as the latest copy I have is the 97 one.

Tell the immoral Microsoft dumbasses responsible what you think of this disgusting behavior directly at:

http://blogs.msdn.com/fontblog/

Actually, MS is changing their liscensing in support of virtual operating systems.

New virtualization use rights for Windows Server(TM) 2003 R2 Enterprise Edition and Windows Server "Longhorn" Datacenter Edition enable cost-effective consolidation. Licenses for the upcoming Windows Server 2003 R2 Enterprise Edition will allow customers to run up to four virtual instances on one physical server at no additional cost. Licenses for the Datacenter Edition of the version of Windows Server, code-named "Longhorn," will give customers the right to run an unlimited number of virtual instances on one physical server.

http://blogs.msdn.com/virtual_pc_guy/archive/2005/ 10/10/479186.aspx

Apparently the current release (updated on March 20) is "layout complete" -- i.e. all the new features intended for IE7 are present -- so aside from bugs that they decide to fix, websites should appear the same in the current release as they will in the final.

Also, my understanding is that IE7 is still scheduled for sometime this year, which means it'll arrive before Vista.

I recommend checking out the IE Team Blog. They regularly post on new features, changes to the rendering engine, etc.

As for hacks specifically, a few months ago, they started recommending a shift away from using CSS hacks and toward using conditional comments. The latter can be used to target specific IE versions with intended functionality, rather than side effects.

Even better, follow the link. It turns out that ALL the main browser people, MS, Mozilla, Opera and KDE got together and agreeded on colours and padlock information and layout for the address bars. It won't just be MS colouring the bar in IE7, everyone will, and in a common, standard way.

So if every browser is going to do it, in the same way, then how come only MS is being dismissed here?

I highly doubt MS is going to shun their own proprietary technology ...

They have a new proprietary technology: WPF/E
http://blogs.msdn.com/mharsh/archive/2006/03/23/55 9106.aspx

"Neither Windows nor Linux uses kernel audio mixing -- they rely on hardware mixing instead."

You're simply incorrect. Others have covered ALSA on Linux so I'll not touch that here. Windows 2000 shipped with something called KMixer which did software audio mixing six years ago. Windows Vista is getting rid of this and going with something new called the Audio Engine that does the mixing is user mode. See http://channel9.msdn.com/ShowPost.aspx?PostID=1460 29 for some video discussing the new solution.

Windows x64 has the same codebase of Windows Server 2003 32bit, just ported to AMD64 architecture. The only missing portion is the 16bit compatibility layer, but you can be sure that there is code in x64 which has no reason to exist if not to support 16bit programs, even if it explicitely doesn't support them.

Having too different codebases is too hard.

If you want to read about how a mess Windows is because of requirement of backwards compatibility with crap software, read Raymond Chen's blog at http://blogs.msdn.com/oldnewthing/ . He's an asshole, but you'll get an idea.

Microsoft have told people about this

http://blogs.msdn.com/oldnewthing/archive/2004/02/ 13/72476.aspx

and App Verifier does a test (HighVersionLie IIRC) to detect it.

And they test new Windows versions with a vast amount of software and make Windows lie to it without user interaction. Even if they missed it you can get Windows to lie to programs thay the testing missed -

(found via a 1 minute Google search)

http://www.microsoft.com/windowsxp/using/helpandsu pport/learnmore/appcompat.mspx

The old new thing, describes some of the hacks win32 uses to stay compatible with badly written applications. Things like dummy events, hidden windows, duplicate event stacks, etc.

Well, here's all you need to know: http://channel9.msdn.com/ShowPost.aspx?PostID=1488 20

This is the Windows Kernel team telling us what they've been up to (in terms).

WinFS is still being developed. Beta code was released in Aug 2005 and Dec 2005.
http://blogs.msdn.com/winfs/archive/2005/08/29/457 624.aspx
http://blogs.msdn.com/winfs/archive/2005/12/01/499 042.aspx

The final version will be added in a service pack in the future.

WinFS is still being developed. Beta code was released in Aug 2005 and Dec 2005.
http://blogs.msdn.com/winfs/archive/2005/08/29/457 624.aspx
http://blogs.msdn.com/winfs/archive/2005/12/01/499 042.aspx

The final version will be added in a service pack in the future.

Here's an older video looking at a build of the shell UI:http://channel9.msdn.com/Showpost.aspx?postid=1 14694

Here's an older video looking at a build of the shell UI:http://channel9.msdn.com/Showpost.aspx?postid=1 14694

But let's also be honest about where it isn't. For instance, it uses two rings of protection rather than the four available, which translates into more possibilities for errant drivers and such to bring down the kernel.

Also, it was designed for a single-user environment, and although improved over the years, lacks certain security facilities that other operating systems such as OpenVMS have.

Wow, two completely false MS bashing Slashdot posts in one day? You guys are getting good!

Office is not delayed.

The new menu replacements are called ribbons, and the background coloring is a part of the conditional formatting. The latter seems very usable!

Thos this one will be a big shift

Plenty of it is going multi-threaded for dual core / HT

I guess this is what is causing the delay, concurrency is hard.

Multi-threaded Excel blog entry

IE 7, when run on Windows Vista, would not have fallen victim to this or any other exploit of this nature. The reason for this is the fact that IE 7 on Vista runs as a user with virtually no privileges, regardless of privileges of the user using IE 7.

Essentially all actions that require higher privileges, such as writing to non-temp locations on the file system, executing applications, installing plugins, changing settings, etc, will be done through the use of a broker.

The broker is very small, perhaps only a few thousand lines of code. This makes auditing the broker far easier than auditing the hundreds of thousands of lines in IE 7.

When IE 7 wants to save a file to the user's desktop, for instance, it must first "ask" the broker if it can do this. The broker is written in such a way that all actions require the user to confirm this is OK via a dialog box. If the user says it's OK the broker completes the action on behalf of IE 7.

If IE 7 has a buffer overflow or exploit of some kind and tries to do something nasty it will always fail because it is running as a user with basically no privileges on the system.

There is a video that describes this in detail on Microsoft's Channel 9 web site.

Today, Microsoft's announcement indicated that Windows XP users would be able to upgrade to IE7. Thus, this is a "new feature" for IE7 that IE7 did not have before today - backward compatibility with older operating systems.

May I be the first to welcome you to yesteryear: http://blogs.msdn.com/ie/archive/2005/02/28/382054 .aspx

Assuming you don't keep up with IE news beyond Slashdot, you should still have caught one of the many dupes that talks about IE 7 being available for Windows XP.

The web components built into GNOME are only good for rendering simple HTML and XML and do not have any active scripting features/plugins etc. Think of it more as a preview for web documents, like image thumbnailing.

Unless those web documents need ECMAScript in order to render correctly. Before you flame me about principles that ECMAScript should never be necessary, think about documents written in languages not supported by major operating systems.[1] They need to be transliterated at view time from an ad-hoc character encoding into a stream of <img> elements that refer to glyphs representing the characters of the language, and it's a lot more bandwidth-efficient to do this using a script at the client side than at the server side, even with gzip transfer encoding.

[1] "I want to make sure everyone understands that we (Microsoft) don't add characters to Windows unless they are in characters that are in Unicode." (Citation) Unfortunately, not all scripts in existence are encoded.

Windows 2000 is no longer in the windows labelled "mainstream support" so the less they have to deal with it the better for their support teams. On IEBlog, they also cite specifically why it can work for WinXP and not Win2K. It's because of the security upgrades done to XP in service pack 2 which they claim are not easily back-ported into 2K.

Except that thats wrong (or at least dated information). Latest word is OpenGL will be fully supported. See http://blogs.msdn.com/kamvedbrat/archive/2006/02/2 2/537624.aspx for details.

No one I know ever sends that data back, but I'm sure someone must have once.

Plenty of users do. There's a great blog posting by Raymond Chen called There's an awful lot of overclocking out there where he talks about investigating some of these "Watson" crashes.

The crashes were impossible - instructions like

xor eax, eax

Turns out unscrupulous vendors were selling overclocked computers without informing buyers. Pretty cool article.

That said, it reminds me of an interesting story. What happens when a company doesn't want to wait for MS to ship them the final version of an OS (say... Windows 95)? The answer is in this fun little entry in The Old New Thing weblog from a Microsoft employee.

I've read that MS won't support XSLT 2.0. Now I have not tried my hand at a real AJAX program. But what I've heard is that AJAX and XSLT go hand-in-hand. With processing occurs totally/primarily on the client-side.

The first article linked to notes ATLAS being "for AJAX-style development." Won't lack of XSLT 2.0 features, like regular expression, limit what ATLAS can do? And will the limitation be to the point where you're better off writting a standard ASP.NET page?

The key phrases here are 'claim' and 'their platform'. What about cross platform web applications I ask you?!

"which they claim will greatly reduce the effort in developing AJAX style applications on their platform."

How about ACID2 complicance in IE7 or implementing the features developers are asking for:

http://annevankesteren.nl/2005/03/ie7-wishlist
http://news.com.com/Next+Explorer+to+fail+Acid+tes t/2100-1032_3-5813897.html
http://blogs.msdn.com/ie/archive/2005/03/09/391362 .aspx

Also I wouldn't say they are "breaking compatibility" as much a fixing bugs -- if you worked around IE6 Strict Mode, you should be fine

This is not true. There's all kinds of changes to strict mode that can break backwards compatibility. I quote from the ieblog:

It is has been our policy since IE6 that under quirks doctype we will not make any behavioral changes so that existing pages will continue to render unmodified, but under the strict doctype we want to change behavior to be as compliant as possible with the web standards. For IE7, we introduced new CSS functionality (see Chris' blog post for the full list) and cleaned up our parser bugs. This leads now to several CSS hacks failing. If you are using IE7 (you are MSDN subscriber or received a copy at the PDC) you may notice major sites breaking due to the use of CSS hacks and the strict doctype.

Bottom line: if you used quirks, you'll be fine, if you used strict, you might have a lot of work on your hands.

This is exactly what DOCTYPE is for, and exactly how it's used in Firefox.

Absolutely 100% totally incorrect. Just because Firefox does it, that doesn't mean it's right. This isn't "exactly what DOCTYPE is for". That's completely wrong. The doctype line is for referencing an external DTD subset, which defines the markup language's grammar. The use of it for picking rendering modes is a Microsoft-invented hack, and nothing at all to do with what it is designed for.

"Kenneth is working as an intern here at Microsoft for the summer on the Office team as a tester. He uses Visual Studio to find bugs (and to code on his own time). He writes emails in Outlook. Does all the usual stuff that most developers or testers at Microsoft do. With one difference.

He can't see the screen because he's been blind since he was three years old."(Source)

Check out the video at channel9 (click the source link above). There's especially one really good question/answer combo in there, I'll transcript it:

[ Start: 08:50 ] Interviewer: How could the software be improved for you? What would you tell the Visual Studio team, for instance, to do to make the software better.

Kenneth: Visual Studio has been very difficult for me to learn with JAWS. JAWS company itself does not specifically support Visual Studio as of now. And they're really working hard on adding the support to the JAWS software. So I think as of now that's the main thing that's holding back the accessibility levels, that JAWS itself hasn't really fully incorporated it. But I think Visual Studio relies very heavily on colored text, for symbolizing things. So rather than having the text that's related to one thing versus another isolated in different locations on the screen, they're all in a list and the different categories are symbolized for different text colors and background colors. I think that makes it a little bit difficult to sort things out on the screen with a screen reader.

So that would be one thing I could suggest, but I think we're primarily waiting on JAWS. [...] As long as they're continually aware of the accessibility levels of their software and test them for that aspect of the usability. Then I think it'll continue to be usable, as long as the specialized software such as JAWS evolves along with it. And so I think that they're working hard at companies like JAWS and their competitors to stay current. And sometimes they fall behind, and I think for specialized programs like software development that aren't as common as say Word processors that sometimes are not as up to date. [ End: 11:12 ]

define: JAWS
"JAWS (an acronym for Job Access With Speech) is a screen reader for the visually impaired." (Source)

The other problem is that it breaks the case insensitivity of English. As you point out, you could write your post all in lower case, the capitalistion of the first letter of a sentence or proper nouns is redundant. You can write English on a device which only supports upper case, and it's still ok too. This free vs Free distinction stops you doing that. Shouldn't in be free for the normal dictionary definition of free and some new phrase for the new definition. What about "open source" for example?

Seems to me that whoever thought up this free vs Free thing needs to spend time reading Raymond Chen on adding new features without breaking back compatibility.

I'm getting sick of this myth that .NET is so much slower than native code. For one example that completely shatters this theory, click here.

Read this blog posting by Dan Fernandez:

"...For those of you that refuse to believe, here's an estimate of the lines of managed code in Microsoft applications that I got permission to blog about:

• Visual Studio 2005: 7.5 million lines
• SQL Server 2005: 3 million lines
• BizTalk Server: 2 million lines
• Visual Studio Team System: 1.7 million lines
• Windows Presentation Foundation: 900K lines
• Windows Sharepoint Services: 750K lines
• Expression Interactive Designer: 250K lines
• Sharepoint Portal Server: 200K lines
• Content Management Server: 100K lines

Watch this video: http://channel9.msdn.com/ShowPost.aspx?PostID=1578 43 You'll see Expression comes in 3 flavours and is geared at a kind of Flash alternative (working with XAML). Only one of the 3 versions is geared for photoshop type useage. It's more geared towards vector graphics from what I can see.

Because they don't care about standards. Firefox does not completely pass, but it's almost there. However, take a look at a screenshot from IE7. And what does MS say? "To relieve the tension, we'll just tell you now: we don't give a crap about web standards, so IE7 will have to have code especially designed for it, hopefully leading to the extermination of alternate browsers." (paraphrased and interpreted with no bias at all from here)

LOL, facts.... http://support.microsoft.com/kb/303247/EN-US/ Thats funny, there is code in one file, and in two files just like I said, obviously you know what you're talking about. Really no client apps? http://blogs.msdn.com/vsdata/archive/2004/11/01/25 0823.aspx Hey, you know what keep believing everything you read and being a misinformed fool.

Maybe the next major MS release to be named Interrobang or Irony Mark. Rather than spelling the name out, it will be the cool thing to use the wacky obscure characters hence making it impossible to find any thing about it.

Maybe the next major MS release to be named Interrobang or Irony Mark. Rather than spelling the name out, it will be the cool thing to use the wacky obscure characters hence making it impossible to find any thing about it.

There's also MS's free Kid's Programming Language. Based on BASIC, but structured rather than linear.

and raise another one

Not particularly concerned.

Not particularly concerned.

Not particularly concerned.

Yay innovation.

This seems to be the same Niels Ferguson who now works for Microsoft, and over whose dead body Vista will include a goverment backdoor for its cryto file storage feature. That is to say, he's against the idea. Good for him.

http://blogs.msdn.com/si_team/archive/2006/03/02/5 42590.aspx