Microsoft Concedes Vista Launch Problems 594
notdagreatbrain writes "Maximum PC just posted a lengthy feature looking back at the myriad problems that went into Microsoft's 6 billion dollar failure of the Vista launch. Aside from running benchmarks comparing Vista at launch how its performing now, they also found a Microsoft exec who was willing to speak frankly about Vista. The Microsoft source blamed bad drivers from GPU companies and printer companies for the majority of Vista's early stability problems and described User Account Control as poorly implemented but defended it as necessary for the continued health of the Windows platform. He assailed OEM system builders for including bad, buggy, or just plain useless apps on their machines in exchange for a few bucks on the back end. Finally he conceded that Apple appeals to more and more consumers because the hardware is slick, the price is OK, and Apple doesn't annoy its customers (or allow third parties to)."
The RAM error (Score:5, Informative)
TFA misses a major misstep. Microsoft allowed Vista to be shipped on hardware that just wasn't up to the task. Vista is unusable with less than a gig of memory, but chain stores were flooded with laptops equipped with "only" 512MB. This gave new users a terrible experience. "First boot" of a new laptop took half an hour. No application, not even Solitaire, would run without freezing.
Two of my family members had Vista laptops ... for a few hours, anyway, until I installed Ubuntu. Performance problems all went away after that.
No issues here (Score:2, Informative)
Top of the line? (Score:3, Informative)
The systems they used in the Mojave advertisements were HP Pavilion DV 2000 machines with 2GB of RAM.
http://arstechnica.com/journals/microsoft.ars/2008/07/31/windows-mojave-advertisements-start-to-appear-in-the-wild [arstechnica.com]
Not really top of the line...
Appe Doesn't Annoy Its Customers? (Score:5, Informative)
Apple doesn't annoy its customers (or allow third parties to)
AT&T. Locked iPhones. Can't do anything not officially blessed with your iPhone unless you unlock it. Can't register your iPhone with anyone other than AT&T. iTunes is loaded with DRM, and QuickTime is pretty annoying... I do love Apple, but seriously, they constantly flirt with annoying their customers far more than most companies.
Re:Don't you dare blame the GPU/Printer companies! (Score:5, Informative)
hardly gave them any time
Manufacturers had more than full year before the release date.
But some manufacturers (Creative [gizmodo.com] comes to mind) didn't ship proper Vista drivers for an additional eight months after Vista Business Final was released.
Re:Top of the line? (Score:5, Informative)
I speak from experience, as someone who owns a DV9000. I ordered it with 1GB of ram to max out the CPU specs, knowing I would upgrade the ram later. Vista ran very very slowly. I upped it to 1.5GB and it runs just fine. Ram makes all the difference with Vista, even though it shouldn't. Say all you want about all the ram usage being pre-caching - if it were just pre-caching, it wouldn't have been slow with twice the ram it "needs", and it wouldn't have sped up dramatically just by adding half a gig more space used only for precaching. If that's their idea of speeding up applicaiton loading with pre-caching, I'd rather go back to the XP way of things.
The "Mojave Experiment" method is simple: get a person who says Vista is slow (because it is, on their machine), sit them in front of a computer with twice the ram theirs has, and get them to do a recorded double-take at how fast said machine is running Vista. Then they go home filled with excitement ready to prove themselves wrong on how slow Vista is and realize their copy of Vista isn't as fast as the Vista they used and wonder what the heck the testers did (remember, these are the kinds of people who read "Word for Dummies" - any person with a clue about technology would be filtered out in a screening process or have his response never aired).
I say all this as someone who uses Windows for daily development without much issue. I don't hate Windows. I do, however, hate dishonesty, whether express or implied, as the lawyers would say.
Re:First Godwin! (Score:5, Informative)
Re:So...... (Score:5, Informative)
I'm gonna agree, with the additional statement that Vista is wonderful AFTER I upgrade my RAM from 2GB to 4GB. Vista Ultimate (x64) is a pretty decent upgrade from XP, and I wouldn't consider going back.
(disclaimer: I also have a macbook and a linux server. All have their drawbacks, all have their strengths. I'm just comparing Vista to XP here.)
Re:They're playing the vista commerical now.. (Score:4, Informative)
As someone who primarily works off of a Mac Pro with 10GB of RAM, I'll call BS on that one. With the exception of the completely broken nVidia drivers, Vista runs fine performance-wise on my old faithful desktop with about a fifth of the specs (now those drivers forced the system in question to go back to XP, but they were causing crashes, not slowdowns).
I'll agree that the ads suck (and I can't imagine why Hulu thinks that I'd switch from OS X to Windows... it just doesn't happen), but not because they're giving the people an unfair representation of Vista.
The UAC needs improvement; other than that and the above-mentioned driver issue, I'd run it on all of my vaguely-current non-Apple hardware.
Re:Top of the line? (Score:2, Informative)
Re:So...... (Score:5, Informative)
I am not usually one for hacking into MS, I have worked with a number of businesses that have been quite successful using their software. On this occasion though, I have to call bullshit. The fact of the matter remains that a number of these drivers did pass WHQL (Windows Hardware Quality Labs) testing. The fact that MS *SIGNED* the steaming brown stuff de-values the program considerably.
Going back to NT4 and previous versions, there was MAJOR issues with quality of drivers, I know certain Xircom cards which would only work with very specific driver configurations as we found out places where the driver was hard coded to use specific ports and IRQs. This all changed with Windows 2000 and signed drivers.
Microsoft should have been harder in their WHQL driver certification for Vista. I know Microsoft only has a finite amount of resources, but they should have known that people were having trouble getting drivers signed and my outsider (un-informed) view is that they should have made more of an effort to work with the OEMs for drivers to work on day 1.
Re:So...... (Score:1, Informative)
"I frequently get crashes that either lock the system up so badly that I can only control the mouse and everything else is frozen..."
Most likely only X Window is frozen. SSH in to your box from another one and reboot it or restart the X Window server.
Re:They DO have some control... (Score:5, Informative)
Read the whole post, please.
I clearly state that I used the included Anytime Upgrade DVD to perform a full, clean install. No OEM disc, I can use that same disc to reinstall Vista Ultimate on my desktop, by simply entering the correct license key.
Yes, I've tried it. Yes, it worked.
Another case of HUAS.
network performance (Score:3, Informative)
Re:So...... (Score:4, Informative)
I... just... don't know where to begin...
Microsoft did, in fact, dump their entire driver model and replace it with something better, with support for full user-mode drivers for certain classes of devices (such as printers, eh?). Hardware vendors responded by wrapping their existing drivers in a compatibility layer.
The same thing could happen on a Mac.
I will admit that Apple has stricter certification standards. This is a Good Thing(TM).
Re:First Godwin! (Score:5, Informative)
Specifically, it says the SS uniforms and the Hitler Youth uniforms.
Re:So...... (Score:2, Informative)
I would suggest that either something is misconfigured or miscompiled. I can't tell you if it is Ubuntu, since I don't really like it much, but I've been led to understand that Debian tends to adopt a conservative enough approach to ensure stability.
But I have used Slackware boxes for the last 16 years, with none of the symptoms you describe. More recently, I have been using Arch, which reflects much of the philosophy of Slackware-type distros, and that seems equally stable. I'm equally sure that the issue doesn't lie with Gnome, which has been stable for a long time.
Re:So...... (Score:3, Informative)
This doesn't necessarily address the problem as far as usability is concerned and wouldn't be satisfactory for a typical desktop user, but do you know if it's the Linux kernel that's crashing or is it X, or the desktop manager?
Have you tried Ctrl-Alt-Backspace? (Usually resets X.) Or you could try Ctrl-Alt-F1 to get a text console, and use commands like 'top' to see which process is clogging things up and kill it.... (Ctrl-Alt-F6/F7/something will switch back to X), or just try something like 'sudo /etc/init.d/gdm restart'. If you're in a position to, you could use another computer to ssh into the machine and do something similar. If you can type a command somewhere within the gui, like an xterm or a konsole, you could try running 'xkill' and then click the mouse on whatever window appears to be causing the problem. (This will kill the process in control of that window.)
This is actually one of the things I really like about Linux, and Unix-like systems in general, because when things do go down it's not generally the underlying system. At most it's probably an application, and it's usually easy to drop back to a text mode and fix the problem with that application without affecting other things.
Windows is definitely getting better at doing this (I can nearly always invoke the task manager to kill things these days without many problems), but it's taken longer for Microsoft to get there given their attachment to doing everything with a GUI as the primary method. Until relatively recently with Windows, it used to be that one thing breaking would take the whole system down.
Reason 8 that Vista Sux (Score:3, Informative)
Re:Don't you dare blame the GPU/Printer companies! (Score:3, Informative)
The vast majority of non-professional printers are brain dead USB pieces of shit that require everything to be software rendered and the printer internals to be controlled directly through the usb cable. The drivers are large because they tell the CPU how to run the printer, instead of spending a little bit more for a cpu inside the printer that takes PCL or postscript or something and renders it inside the printer and controlls the internals.
Re:They're playing the vista commerical now.. (Score:1, Informative)
It's a setup for Windows 7 late next year. The campaign is so expensive because they'll be running with the concept longer than they usually do (18+ months).
Re:So...... (Score:1, Informative)
You know, there is really no reason whatsoever that you can't use an open source driver with windows. The driver API is well published, and there is nothing stopping the community from stepping up and writing its own drivers.
Yes there is. MS effectively only allows certified drivers for 64-bit Windows; presumably they will phase out 32-bit windows and that will be it. The following is speculation but I expect the certification process is for a given binary. Presumably some expense is involved. This effectively means MS is not going to allow FOSS drivers.
Re:Appe Doesn't Annoy Its Customers? (Score:5, Informative)
iTunes is loaded with DRM
On Windows at least, it also installs:
1) an auto-start AppleMobileDeviceService service
2) an auto-start iPodService service
3) an auto-start iTunesHelper application
4) Bonjour as an auto-start service
5) MobileMe
Additionally, it defaults to installing the AppleUpdate service, which - when it finds updates - defaults to installing Safari alongside the updates.
Just to listen to music?
Re:Mac os x is so much better that hack it to run (Score:3, Informative)
No. Only that he was German.
WHQL Testing (Score:1, Informative)
Microsoft has no control over the shit quality of drivers released by hardware manufacturers.
Yes, they do, it's called WHQL Testing [wikipedia.org]. If the driver isn't good you don't approve it, and you don't sign the file.
If the driver is crappy, and it is signed, it means that Microsoft didn't do testing well enough.
Re:So...... (Score:4, Informative)
Except this is mostly untrue untrue for Windows and occasionally untrue for Linux.
There's been very few times on Windows when I couldn't get to the taskmasker to kill something, and there have been times on Linux where I had to power cycle the machine because some combination of make or g++ put it into a catatonic state.
And why the hell were you modded to +5?