Why vista will suck!

[HyperHacker]

Its not 10% but its getting there slowly.Typo, I presume? :p

[dragongodz]

this thread got quite funny. its amazing people claiming that just because XP had bugs etc when it was released that it is ok for Vista too and people should still be buying it.

bull.

i know i sure didnt buy Xp when it first came out, didnt buy it until i had to for certain software i use. the same is going to apply to Vista. since i have tried Vista on friends PCs and seen how it really doesnt add anything special i know its really not a must "upgrade".
infact the longest it lasted on my friends PCs was 3 weeks.

anyway for those willing to see the funny side of things have a look at this

http://www.metacafe.com/watch/857874/vista_true_info/

:D

[Dr.Khron]

Thanks for the link, dragongodz, that was very funny, and well produced, too.

If you guys want read a heated debate about the usefulness of Vista, check out this HardOCP thread:

Is Windows Vista finally hitting its stride? (http://www.hardforum.com/showthread.php?t=1237347)

A bunch of corporate IT guys chime in, including this interesting comment:

However, Microsoft *is* being truthful on sales of Vista Business -- every business I deal with has started buying Vista Business instead of XP Pro -- but, as is often the case, Microsoft is only telling PART of the story. The reason the companies are buying it is because Vista Business counts as a legal Windows XP Pro license as well -- so, for the same price, it's a better deal. You get an XP Pro license for now and a Vista Biz license for when (& if) it becomes viable.

The fact that Microsoft seems to be omitting is that even though businesses are BUYING Vista Business -- they are only buying it for licensing reasons and then exercising their downgrade rights and installing XP Professional.

[Sharktooth]

With the initial performance characteristics of Windows Vista leaving much to be desired (see our previous post on the subject (http://exo-blog.blogspot.com/2007/09/what-intel-giveth-microsoft-taketh-away.html)), many IT organizations have put off deploying the new OS until the first service pack (SP1) is released by Microsoft early next year. The thinking goes that SP1 will address all of these early performance issues and somehow bring Windows Vista on par with - or at least closer to - Windows XP in terms of runtime performance.

Unfortunately, this is simply not the case. Extensive testing by the exo.performance.network (www.xpnet.com) research staff shows that SP1 provides no measurable relief to users saddled with sub-par performance under Vista.
http://exo-blog.blogspot.com/2007/11/vista-sp1-performance-dud.html

Research Staff said...

See our previous post on "What Intel Giveth, Microsoft Taketh Away" for numbers from XP SP2 vs. Vista RTM. Basically, Vista is roughly 2x slower than XP on the same hardware across the board.

RS

[Shinigami-Sama]

why does that not surprise me?

are there any numbers for vista server yet?
2008 is it?

[Sharktooth]

dunno... im not interested in M$ crapserver2008...

[Shinigami-Sama]

dunno... im not interested in M$ crapserver2008...

seems no-one I know is
I've asked several people I know if theres any decent info on it and keep coming up blank or some PR fluff on second rate sites

[burfadel]

Readyboost does make a bit of difference, but whether to use it or not comes down to a few things.

First of all, most devices aren't suitable for readyboost. Even the ones that do pass the initial test (it tests the drives small read/write performance automatically when you first use it), they don't provide the low access and transfer times that makes readyboost beneficial.

I have a 2gb Apacer HT203 for my readyboost, its one of the fastest for this purpose. The best readyboost drive I know of is the Patriot Xporter XT. A drives rated speed on the packet is its sequential read/write rate. Readyboost relies solely on random read/writes of small amounts of data. A fast drive, 200x, may actually not pass the requirements, whereas a 120x drive passes quite well.

A cheap readyboost passable drive is not really beneficial, latency is important! Those cheap drives won't really provide any beneficial gain if they work for readyboost.

On 32 bit Vista, readyboost makes sense, since you are effectively limited to 2gb of ram. I know you can run more than that, same with XP, but its not really effective. Anyways, 2gb is ideal, but if you have a 4gb drive thats even better (very very slightly). If you specifically buy a readyboost drive then the extra cost is not worth it for 4gb.

If you're on 64 bit vista, with 2gb or less of ram it would be much more beneficial to spend money on RAM before spending it on a readyboost drive. 4gb or ram compared to 2gb will provide a significant more gain than the readyboost! Even with 4gb of ram, readyboost is beneficial though.

Also remember a 256mb readyboost on a good readyboost drive, such as the Patriot Xporter XT is more beneficial than 4gb on a key (even if its from another good brand) that only just passes the readyboost test.

Readyboost is supposedly going to be improved slightly with SP1.

[[P]ako]

are there any numbers for vista server yet?
2008 is it?

I haven't seen any yet. However, I am running the beta 3 of Server 2008 and as far as encoding goes, the speed is the same as with XP Pro SP2. Also, I should mention that, at least in my test machine, Server 2008 is more stable than XP Pro SP2.

Test machine specs: AMD XP +1500, 512 MB of SDRAM (PC133), ATI Radeon 7000 64 MB VRAM

[Shinigami-Sama]

ako;1067624']I haven't seen any yet. However, I am running the beta 3 of Server 2008 and as far as encoding goes, the speed is the same as with XP Pro SP2. Also, I should mention that, at least in my test machine, Server 2008 is more stable than XP Pro SP2.

Test machine specs: AMD XP +1500, 512 MB of SDRAM (PC133), ATI Radeon 7000 64 MB VRAM

ahh
how well does it function a server though?
or have you tried it out much that way?

[Inventive Software]

I'd imagine it would be a logical step forward in hardware support as much as anything, considering the last server OS was Window Server 2003. 5 years is a long time in computer hardware. But that the encoding speed is the same as XP SP2 is quite remarkable. It goes to show that Vista really isn't suited for main-stream just yet. Needs a couple of years to really crunch out the performance issues. And bugs of which there will be thousands.

[burfadel]

Having the speed of encoding the same as xp sp2 is expected. Installing a new operating system doesn't replace your cpu automatically!

The comparison should be made between xp sp2 and vista x64. Vista x64, even for 32 bit apps, is supposed to be faster. Due to many factors, including memory issues etc I think people should either go xp sp2/(sp3 shortly) or vista x64 (sp1 shortly). The 32 bit Windows is a waste of time and should never have been made. At least that way it would be easier on manufacturer's making drivers, as well as on Microsoft not having to deal with so many updates.

My ultimate opinion is XP should have been revamped with sp3, maybe even include things like readyboost etc, and have its core based on Server 2003 (Windows 5.2 instead of 5.1)... and then release it kind of like they did with 98se. Instead of free+pay for postage, a nominal update fee should have applied, such as say $40? That would mean most people on sp2 wouldn't mind spending the money to upgrade. Microsoft ends up with more profit due to significantly more upgrades to the new 'XP', it reduces the amount of updates they have to compile by 2 (sp3 uses the same updates as server 2003, and vista 32 doesn't exist so no updates there either)!, manufacturers would have to write one less driver, consumers are happy, etc etc. and more time could be spend on getting x64 right!

But in the end its corporate greed as they expected everyone to switch to vista 32 regardless.

[JohnnyMalaria]

The comparison should be made between xp sp2 and vista x64. Vista x64, even for 32 bit apps, is supposed to be faster. Due to many factors, including memory issues etc I think people should either go xp sp2/(sp3 shortly) or vista x64 (sp1 shortly). The 32 bit Windows is a waste of time and should never have been made. At least that way it would be easier on manufacturer's making drivers, as well as on Microsoft not having to deal with so many updates.

It doesn't follow that 64-bit Vista should be faster than 32-bit XP or 32-bit Vista when running 32-bit apps. 32-bit apps run natively on the CPU as, well, 32-bit. For CPU intensive apps, the performance will be almost identical, possibly slightly less (32-bit requests for disk access etc have to be translated to the 64-bit kernel drivers for the hardware). (This has been my experience when running benchmarks on XP SP2, x64 and both flavors of Vista).

Not making 32-bit Vista would be foolish. At a stroke, legacy hardware would be redundant unless the vendors write 64-bit drivers. Expecting all hardware to have 64-bit drivers in time for Vista launch is unrealistic. For new hardware and software, Microsoft provide development tools that can generate 32-bit and 64-bit code from the same source code so it offers no advantage for the vendors to only create 64-bit drivers.

My ultimate opinion is XP should have been revamped with sp3, maybe even include things like readyboost etc, and have its core based on Server 2003 (Windows 5.2 instead of 5.1)... and then release it kind of like they did with 98se. Instead of free+pay for postage, a nominal update fee should have applied, such as say $40? That would mean most people on sp2 wouldn't mind spending the money to upgrade. Microsoft ends up with more profit due to significantly more upgrades to the new 'XP', it reduces the amount of updates they have to compile by 2 (sp3 uses the same updates as server 2003, and vista 32 doesn't exist so no updates there either)!, manufacturers would have to write one less driver, consumers are happy, etc etc. and more time could be spend on getting x64 right!

But in the end its corporate greed as they expected everyone to switch to vista 32 regardless.

Like XP x64, Vista derives much of its codebase from the server versions of Windows. Vista is an opportunistic stepping stone on the way to the next server version - i.e., Microsoft can make money (to pay for ongoing R&D without which we'd still be using DOS). For Microsoft to get sidetracked into extending XP would be expensive and unnecessary. They'd also get a lot of criticism for not delivering a new OS.

The 32-bit vs 64-bit argument is rather moot with Vista since both versions come with the DVD. For retail purchasers (i.e., not OEM), they can make the choice - Microsoft provide the choice. Most are choosing the 32-bit, primarily because of the lack of 64-bit driver for their existing hardware.

[zambelli]

The 32-bit vs 64-bit argument is rather moot with Vista since both versions come with the DVD. For retail purchasers (i.e., not OEM), they can make the choice - Microsoft provide the choice. Most are choosing the 32-bit, primarily because of the lack of 64-bit driver for their existing hardware.
One very good thing that came out of Vista development was the unification of 32-bit and 64-bit build process at Microsoft. Most Vista code just gets built automatically as 64-bit - there's no need for devs to make separate changes. This changes the view of 64-bit computing quite dramatically, as it's no longer viewed as a special project, but an elementary part of compiling code.

[[P]ako]

ahh
how well does it function a server though?
or have you tried it out much that way?

I haven't tried it as a server. My intention was to test it as a desktop OS.

[Shinigami-Sama]

ako;1067678']I haven't tried it as a server. My intention was to test it as a desktop OS.

is it a nicer set up than a 2k3 desktop?
I had some problems getting nlite to cooperate with 2k3 nicely

[foxyshadis]

Not making 32-bit Vista would be foolish. At a stroke, legacy hardware would be redundant unless the vendors write 64-bit drivers. Expecting all hardware to have 64-bit drivers in time for Vista launch is unrealistic. For new hardware and software, Microsoft provide development tools that can generate 32-bit and 64-bit code from the same source code so it offers no advantage for the vendors to only create 64-bit drivers.

That did happen with NT/2000, and to a lesser extent, 95. It was painful and caused a lot of angst, but people either got over it and bought new hardware or stuck with the old OS forever. The popularity of XP eventually drove large-scale releases of new drivers; the Vista launch has been more like 2000's, but surely most devices would at least be supported in 64-bit more quickly this way.

As it is, unless you're a techie you can't install unsigned drivers on Vista anyway, which in my experience about 2/3 of them are.

[[P]ako]

is it a nicer set up than a 2k3 desktop?
I had some problems getting nlite to cooperate with 2k3 nicely

Yes, it is very lean.

[Shinigami-Sama]

That did happen with NT/2000, and to a lesser extent, 95. It was painful and caused a lot of angst, but people either got over it and bought new hardware or stuck with the old OS forever. The popularity of XP eventually drove large-scale releases of new drivers; the Vista launch has been more like 2000's, but surely most devices would at least be supported in 64-bit more quickly this way.

As it is, unless you're a techie you can't install unsigned drivers on Vista anyway, which in my experience about 2/3 of them are.
why does microsoft have such a hard time signing drivers?
its not like demand has decreased...

ako;1067799']Yes, it is very lean.

nice :)

[foxyshadis]

Microsoft doesn't have a hard time signing them, companies have a hard time submitting them, especially the ones that release regular updates. Every sign costs a lot of money, but Vista has an entirely new signing model based on 3rd party testing suites (as well as Microsoft's own), so the cost is potentially much lower now. But still, paying some amount of money for every free maintenance release is a bit of a disincentive to bother.

[Shinigami-Sama]

Microsoft doesn't have a hard time signing them, companies have a hard time submitting them, especially the ones that release regular updates. Every sign costs a lot of money, but Vista has an entirely new signing model based on 3rd party testing suites (as well as Microsoft's own), so the cost is potentially much lower now. But still, paying some amount of money for every free maintenance release is a bit of a disincentive to bother.

I take that as having a hard time signing them
I'd agree with an initial payment, and say 3 re-signs for revisions
then another payment, but if 3rd party can get them signed all the better

[HyperHacker]

Every sign costs a lot of moneyAnd now we see the real reasons behind DRM and "protection" schemes. :p

[Sharktooth]

exactly... MS already gets enough money by signing drivers (just think at ATI that has WHQL certified drivers once per month!) they can release the windows OSes for free...
But the greed attracts even more greed...

[honai]

Microsoft doesn't have a hard time signing them, companies have a hard time submitting them, especially the ones that release regular updates. Every sign costs a lot of money, but Vista has an entirely new signing model based on 3rd party testing suites (as well as Microsoft's own), so the cost is potentially much lower now. But still, paying some amount of money for every free maintenance release is a bit of a disincentive to bother.

MS already gets enough money by signing drivers (just think at ATI that has WHQL certified drivers once per month!)

Do you have links or reference that show if and how much companies are being charged by Microsoft for having their drivers signed? Because I don't believe any of it, and unless you can prove any of the allegations this thread is becoming another pointless MS bashfest based on false rumors, not facts.

And doing a google search for "microsoft driver signing cost" and taking the word of an anonymous commenter at a blog at face value is not research, no matter how many gullible "researchers" link to it. Just because someone is wearing a helmet on a construction site it doesn't make him an architect, you know.

[Sharktooth]

:search: (google :p) and you will find why smaller companies cant afford to have their drivers certified even if they could pass any WHQL tests...

[honai]

here's another smarta$$ fanboy

Just because your post count reads 6,984 it doesn't mean you have a right to insult other visitors. That kind of rude behavior actually diminishes any of the valid points you might have.

And like I said, a google search is not research. Where are the hard facts? Repeating a rumor doesn't make it true. Please give me a reference from anyone involved in WHQL driver signing who can confirm that it costs "a lot of money", if any at all.

[Sharktooth]

Just the Verisign Class 3 signing costs 500$... plus 250$ for each attempt of WHQL testing (that doesnt actually mean certification!).....
http://www.microsoft.com/whdc/getstart/testing.mspx
oh, btw, you started insulting this community intelligence by contraddicting facts before even getting informed...

[honai]

Just the Verisign Class 3 signing costs 500$

The 500$ for the certificate goes to VeriSign, not Microsoft. Microsoft has no control over how much VeriSign charges for it.

plus 250$ for each attempt of WHQL testing

Links? References? Where did you get that information from?

And let me quote again:

plus 250$ for each attempt of WHQL testing

Every sign costs a lot of money

MS already gets enough money by signing drivers

Even with that juxtaposition I'm still eager to see the references that confirm the price tag of "250$ for each attempt of WHQL testing".

[Sharktooth]

http://download.microsoft.com/download/9/c/5/9c5b2167-8017-4bae-9fde-d599bac8184a/WLP_Policies.doc
also keep in mind it's 250$ for EACH OS the driver is certified for...
if the driver package is made up of multiple drivers each one must be also certified...

[honai]

Thanks for the document. Now that wasn't too hard, was it? I hope now we can get back to discussing facts, not rumors.

also keep in mind it's 250$ for EACH OS the driver is certified for...

It's 250$ for each OS family, e.g. testing for both Windows XP x86 and x64 amounts to 250$, not more.

So now that we know the facts do you still want to maintain that a company as big as Microsoft

already gets enough money by signing drivers

when in actual fact it's a mere 250$ per OS family for submitting a driver?

And by the way, given the costs accumulated by an actual Microsoft employee having to at least handle the submission, how is Microsoft supposed to make any profit from this? Fact is that it's a standard industry practice - not invented by Microsoft - to have nominal fees to deter hobbyists from keeping your employees busy when the former have no actual business interest. You'll find this all over the place. Talk to any software vendor who has such a program, and they'll tell you that unless you pay the nominal fee they won't even handle your contact.

Same with conference fees. Many of those don't cost 500$ entrance fee because they want to make a profit from it, but because they want to filter out folks who are not serious about it. Think about it.

The point is that if Microsoft dropped that price tag they'd get swamped with hundreds of thousands of submissions from people who have no actual business in it other than getting something "cool" from Microsoft for free.

[Sharktooth]

yes, considering drivers are submitted multiple times before getting a certification (each attempt costs 250$)... and that usually drivers require multiple submission each times coz they come as multiple drivers package (just think at an audio or a video card with TV output) and each submission costs 250$...
Plus, if you want the same driver certified for 2000/XP/2003 you need to pay 2 submissions at once too (2000 is free if you have it submitted with another OS)....
Do some calculations and look at the costs...
Also take into consideration that a new driver version requires all that crap again... and also signatures expires and you need to look at the mantaining policies...

[honai]

So what is your suggestion to Microsoft? How would you solve this "problem" if you had a say at Microsoft?

And more importantly: Who is the client you are advocating here? The shareware programmer? Who gets hurt by the current practice? I don't get it.

[Sharktooth]

The "client" is free software as defined here: http://www.fsf.org/licensing/essays/free-sw.html
Or at least to pay a reasonable price for something that actually fits the user needs AND works without needing to spend more money by buying new hardware just coz the software contains a whole bunch of useless stuff.
Vista is exactly the opposite...

[honai]

Perhaps I didn't make myself clear enough. You complained that the driver signing cost is too high, right? So what would your suggestion be? That Microsoft made the WHQL process available for free? If not, what's the price tag you'd propose? Please give a specific number.

[Sharktooth]

As it has been proven (again, use google) WHQL certification is useless.

[honai]

You're dodging the questions. What do you propose? Get rid of the WHQL process (because it's "useless"), or lower the price (because Microsoft is "making money" out of it)? Which way do you want it?

[Sharktooth]

Everyone has his own opinion.
Whatever i suggest MS will do what they want... so, it's useless.

[honai]

I see. So you complain about a situation, but when asked what you'd change you decline to give anything at all.

That's all I needed to know.

[moshmothma]

Do you have links or reference that show if and how much companies are being charged by Microsoft for having their drivers signed? Because I don't believe any of it, and unless you can prove any of the allegations this thread is becoming another pointless MS bashfest based on false rumors, not facts.



Thanks for the document. Now that wasn't too hard, was it? I hope now we can get back to discussing facts, not rumors.


Having just stumbled on this thread I really find your comments distracting and detracting. You made some pretty harsh allegations and when proven wrong, you find something else to be harsh about. It's post like these that really make these kinds of forums tedious. You should probably have just started another thread if you wanted to voice your support for MS. I would personally be quite happy if you'd drop it or start another thread.
Happy Thanksgiving

[honai]

It seems your reading was less than cursory.

You made some pretty harsh allegations and when proven wrong

You are wrong. Here's the initial allegations that I questioned:

Every sign costs a lot of money

and

MS already gets enough money by signing drivers

Both allegations were proven wrong. Every sign does *not* cost a lot of money, and Microsoft does *not* get enough money by signing driver. Actually, given the current wages and rates in the IT industry, Microsoft is probably operating the WHQL at a loss.

and when proven wrong, you find something else to be harsh about

No, the topic staid the same: Is Microsoft charging too much for WHQL driver signing? Answer: No.

It's post like these that really make these kinds of forums tedious.

Actually, it's posts your like yours that make these kinds of forums tedious - people who come late to the party, read very carelessly, draw premature conclusions, and steer off the conversation because they totally misunderstood what the discussion is about.

You should probably have just started another thread if you wanted to voice your support for MS.

And it's precisely sentences like those that inflame a discussion. This is not about support for Microsoft, this is about getting facts straight. I don't care if it's Microsoft or any other company.

I would personally be quite happy if you'd drop it or start another thread.

I would personally be quite happy if you made a meaningful contribution to the topic at hand, instead of talking about meta-aspects of the discussion, or the implied motives of posters.

Do you have an opinion on WHQL driver signing?

[Wilbert]

@Sharktooth, i suggest that you edit your disrespectfull and offensive posts, and refrain from doing this in the future.

[Shinigami-Sama]

@Sharktooth, i suggest that you edit your disrespectfull and offensive posts, and refrain from doing this in the future.

I actual in agree with shark, and honai hasn't been to friendly either

anyways
500$ and 250$ doesn't sound that bad on paper
but if you have to submit your driver 4 or 5 times to get it to pass WHCL tests, it starts to add up, for very little return at all on the final product

[[P]ako]

The MSFT authenticode signing digital ID can be $399:

http://www.verisign.com/code-signing/msft-partner-code-signing/index.html

And just in case this is not clear, I would like to point out that the digital ID license is for the whole company regarless of the number of products it wishes to submit to MSFT. It is not a license for each product.

WHQL testing becomes mandatory only with 64-bit server, in all other cases the companies don't need their drivers to pass WHQL testing to submit them (proof of that were the ATI drivers mentioned many posts above). Now, if the companies were to do their job properly, they'd be submitting their drivers for testing only once or maybe twice. Unless said companies think that it is cheaper to have MSFT testing their drivers for them than doing the tests at home. Who would like to waste their developers' precious and expensive time testing drivers?

[HyperHacker]

I see. So you complain about a situation, but when asked what you'd change you decline to give anything at all.

That's all I needed to know.You don't need to know how to solve a problem to know the problem exists. What do you suggest we do about global warming?

[JohnnyMalaria]

Hmm, let's see how MS are out to rip money from the blood and sweat covered hardware and software developers.

My Verisign certificate - $99 a year and I can sign every and any piece of software I care to - even Notepad.

For less than $400 per year, I have had a full MSDN subscription entitling me to unrestricted access to all MS products and (essentially) unlimited installs.

To get Microsoft Certified Partner status, you have to submit an application for testing - not by MS - but Veritest. In my case, the testing will be done in Ireland. It costs $400 for the main test and $400 for "an elective" test. In my case, that's a feature of the application that requires an MS Office application - no big whoop. The application has to pass a variety of important tests to ensure stability, proper behavior as far as not abusing the security of the OS (no, nothing to do with DRM before your twitching DRM gene kicks in) and it has to have gotten positive references from money-paying customers.

Ouch - $800. Well, after a long (1hr) telephone conversation with Microsoft, I received a voucher for $800 to cover the cost of the testing.

And I'm just me - not some huge software company. *Anyone* can get all the same benefits I have received with little effort. (Basically, a web site that says you are an ISV and a promise to release a product of some form within 18 months of signing up).

What's in it for MS? They get to leverage your product into increased sales of their products. Normal, everyday co-promotion. Will that happen with my products? Unlikely. But some ISV's products will and the cost of a dirt cheap MSDN subscription, free product testing and other perks more than pays for itself.

BTW, as far as driver signing for hardware goes, get real. A few hundred dollars for the signing is bugger all compared to the cost of development of the hardware, updates to firmware, circuits etc. Do you really think the likes nVidia, HP, Canon, Intel etc etc wave their arms in outrage? No.

And finally - before you bitch and moan about the big, evil 64-bit driver signing mandate, do some research. Here's a clue:

bcdedit /set loadoptions DDISABLE_INTEGRITY_CHECKS

[JohnnyMalaria]

but if you have to submit your driver 4 or 5 times to get it to pass WHCL tests, it starts to add up

If you have to submit your driver 4 or 5 times then you or your programmers don't know what you/they are doing. The DDK provides all the documentation and tools to develop drivers and test them exactly the same way they are tested by the labs before submitting them.

[Shinigami-Sama]

If you have to submit your driver 4 or 5 times then you or your programmers don't know what you/they are doing. The DDK provides all the documentation and tools to develop drivers and test them exactly the same way they are tested by the labs before submitting them.

have you seen recent drivers?

[foxyshadis]

Actually you can get the WHQL test suites, you don't ever have to submit them 4-5 times unless your team never bothers to test.

[Sharktooth]

succesfully pass WHQL certification is not so easy. it depends on hardware too. a small company cant afford to recreate all possible situations (hardware combinations) to be sure the driver will pass the tests.
The problem is MS cant do that too. So basically, getting the certification is useless but gives the end user a false feeling of safeness (hey there's a nice Designed for Windows logo and drivers are certified by MS! it must work! ... not ... and the game crashes, or looks ugly... or thousands of other problems coz the AGP current is too low or the driver was not meant to work with that software or other crap like that...).
Some companies (read nvidia) even get beta drivers WHQL certified. Those drivers have BIG issues but they still pass the WHQL tests (WTF!?!?) while making the system COMPLETELY unstable!
Now, let's consider how many hardware makers struggle to get the certification just for marketing reasons (hey, we got certified by MS! our product is the best!) and how much software/hardware is submitted for certification and THEN think at how much income has MS just for that... certifications.
Still, that certification doesnt mean that driver/hardware will work on your system AND that the driver/hardware is bug free... but who cares? it has a nice "Designed for Windows" logo and a nice WHQL certification...
It's just another way for MS to earn money in a questionable way (WHQL = Windows Harware Quality Labs... what?!? what labs?!? it's a software that automatically does some tests!).
So basically, MS gets money from both customers and hardware makers and their retail products are way too expensive for the end users while having impossible license restrictions, bugs, security holes and problems (think at malwares, spywares, trojans, viruses, vulnerabilities, Internet Explo"d"er!!!, etc...).
Also, considering the bugs and bloats you get with recent releases of their softwares (starting from the OSes - since this thread is about how much Vista sucks!) and the need for hardware upgrades you need to DO EXACTLY THE SAME THINGS you were able to do with previous releases of the same products (win2000 or XP), as a human being with his own intelligence, i (and you too!) would think twice (or more) before spending a cent on MS products, expecially when there are free and (more) functional alternatives.

@wilbert: done.

@honai: note -- the thread title is "Why Vista will suck!" and not "Suggestions for MS"...

[JohnnyMalaria]

succesfully pass WHQL certification is not so easy. it depends on hardware too. a small company cant afford to recreate all possible situations (hardware combinations) to be sure the driver will pass the tests.

Have you ever written a kernel-mode driver for Windows? A driver is typically written for a specific piece of hardware. It doesn't matter what other hardware is in the system. The OS sends specific requests to a driver at a number of possible priority levels. If the driver makes one mistake, the OS bug-checks (BSOD) and purposely stops. Kernel code has to be right every time - there's no room for error. Bug-checks occur for two reasons: the driver breaks the rules (which the self-test suite of tools in the DDK will detect) or it communicates incorrectly with the hardware it is paired with. That's sloppy programming or bad hardware. It makes no difference if your driver for your widget is running on a system with 3rd party hardware X, Y or Z.

The problem is MS cant do that too. So basically, getting the certification is useless but gives the end user a false feeling of safeness.....

The primary purpose of digital signing (drivers or applications) is to assure the end user that the software comes from where it claims to and that it hasn't been tampered with. e.g., when you download a driver for some hardware from a 3rd party driver repository, you can be confident that it is exactly the same driver as distributed by the original vendor and not some malware-infested hack.

It's just another way for MS to earn money in a questionable way (WHQL = Windows Harware Quality Labs... what?!? what labs?!? it's a software that automatically does some tests!).

Come on, Mr. Shark. Re-read what you wrote and you'll surely see the nonsense in it.

Windows Hardware Quality Labs - it's about certifying hardware, not just the driver(s) for it. So, somebody (or perhaps a robot) has to receive the specimen hardware and drivers from the vendor. Rather difficult to FTP a PCI-X card - maybe when teleporting becomes a reality. Once received, it has to be installed in a cleanly prep'd PC running the OS specified by the vendor. Then the testing software has to run and report the results. If the hardware is designed for multiple OS families, it needs to be tested on each - a reimaging each time.

So basically, MS gets money from both customers and hardware makers........<Completely unrelated but predictable diatribe removed>