Apple, not Gear, deserves the blame for iTunes crashes
Summary: It’s now been a week since Apple’s botched release of iTunes 8, which caused a tidal wave of Vista crashes before it was hastily rolled back. Judging from traffic on Apple’s support forum, pulling the new Apple USB driver and replacing it with the file from iTunes 7.7 succeeded in quieting most of the complaints from most Windows users, although a handful of customers report that they’re still having problems. After looking more closely at the other driver, from Gear Software, I've concluded that it was unrelated to these crashes and might even be an innocent bystander in another iTunes support headache involving missing CD and DVD drive letters.
It’s now been a week since Apple’s botched release of iTunes 8, which caused a tidal wave of Vista crashes before it was hastily rolled back. Judging from traffic on Apple’s support forum, pulling the new Apple USB driver and replacing it with the file from iTunes 7.7 succeeded in quieting most of the complaints from most Windows users, although a handful of customers report that they’re still having problems.
In my analysis last week, I confirmed that Apple was sneaking a couple of driver updates onto the system along with the five installer packages that make up iTunes 8. The USB controller update, which is apparently used when connecting an iPod or iPhone to the system, was tagged as the one appearing on the STOP error page in most of the trouble reports on the Apple forum. That’s the one that got rolled back.
In last week’s post, I also noted that an update to the GEARAspiWDM.sys driver was installed as well. I speculated that this driver might be involved in some of the problem reports, but that turned out not to be the case. This driver had nothing to do with any of the Blue Screen of Death errors I read about. In fact, it turns out that Gear’s driver might actually be an innocent bystander in another iTunes –related support issue.
First, a little background. Gear Software sells the GEARAspi driver set (a DLL and a storage filter driver) with its GEARWorks SDK, which allows third-party developers to build apps that can communicate with CD/DVD recorders, tape drives, media jukeboxes, and other such storage devices. Apple is not the only big customer that uses these drivers; Symantec, Kodak, Cakewalk, and Siemens also use Gear’s tools to enable ripping, formatting, authoring and burning features in their software. In all, there are probably tens of millions of Windows (and Mac and Linux) users running software that incorporates the Gear driver set.
Earlier this year, reports began appearing on Gear’s forum and the Apple support boards that iTunes users were encountering a well-known (but fortunately rare) issue where the CD or DVD drive icon disappeared from the Computer window. This problem has been around in Windows XP for as long as I can remember, and even has its own Knowledge Base article, which explains the Registry tweaks that have to be performed to restore access to the MIA optical drive.
The iTunes problem, as far as I can tell, first began cropping up in March and April, and by late August 26 Gear Software had posted a plea for help in tracking down the cause of the issue, naming a pair of driver files (afs.sys and afs2k.sys) as the most likely culprits. The AFS in those file names refers to Audio File System, and the drivers in question were part of a much older CD-burning engine that, like Gear's software, was sold to third-party developers. It wound up in some widely distributed products, including Broderbund Print Shop, HP Memories Disk Creator, Liquid Audio, and Symantec Ghost. An HP support document specifically notes the likelihood that the AFS driver can cause crashes and explains how to delete or replace it. Oak Technology, which developed the older, apparently flawed drivers, was swallowed up several years ago by Zoran. Downloadable copies of a September 2003 release of the AFS drivers are available at Simpli Software; HP has a newer version, released in 2004.
I looked long and hard to find any signs of other recent problems involving Gear’s software and couldn’t find any, so I’m inclined to give them the benefit of the doubt here. It’s also a perfect example of the complexities of testing software on the Windows platform. I suspect most of those driver conflicts occur because users install printer and scanner software from CDs that are five years old (or older). It’s hard to imagine testing for conflicts with outdated software that was patched long ago, and yet those programs persist in the real world.
And this historical footnote: If the name Oak Technology sounds familiar, then you qualify as an old-timer. MS-DOS boot disks (created by formatting a floppy with the /s switch) typically include a real-mode CD driver called Oakcdrom.sys, which was used in conjunction with Mscdex.exe to allow access to CD drives by floppy-based installers. Same company.
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
FUD mode
You recently accused security software vendors of over egging the problem to help sell their products. So given that this issue affects only a small portion of Vista users why the big fuss.
One story with a few updates would have covered the issues. Three looks a bit like spreading FUD now that the problem has been resolved.
"The problem has been resolved"?
If you think I'm writing too much on a particular topic, please, please, please feel free to ignore those posts and go do something else.
Seriously.
On XP, I uninstalled
Fortunately, MS had an automated tool that fixed it without requiring the usual reg hacks.
Uh
I also wonder what your motives are Ed.
I'm also sure you would rather I did not comment on your blogs. However its your option to proselytise for Microsoft, its mine to be agnostic.
I have no specific interest, or even knowledge, of ipods or itunes. It does however interest me that you felt the need to post 3 stories on the issue. I do not believe a single corporate customer is worried about the problem or even the majority of home users. I therefore have no idea why you feel it so important.
I'm sure if I did own an ipod I would be cross Apple messed up but would simply have delayed whilst someone worked out what ever teething problem they have is sorted. Or have I missed something does itunes 8 contain some key function everyone needs to download today.
Think again
Wrong. IT folks hate Apple software because they know users will install it on their company-issued PCs if allowed to, and it represents a profound security problem.
DING DING DING
better yet
Actually, most folks aren't supposed to have installation rights
It baffles me how our folks in charge of that can't get it right...sigh. Someone's messing up Active Directory rights badly there, it's pretty basic functionality. I really wish they'd fix that once and for all. Folks would also have much less My Web Search and Smiley Central and crud too...
re: better (@beoz)
But then again, our PC IT dept is lacking. I've seen brand new core-2 laptops that take a minute to open a word document (if you click on the document....very fast if you open word/excel and the do a file/open).
Can they fix it? Apparently not.
Profound????????
I think that has to be one of the most ridiculous and inaccurate statements I have heard in years. Your claim is that Apple software is a "profound security problem". Strangely I have not met a corporate client who takes that view, the "IT folk" I know have many worries apple is not one of them. I do not recall any ever saying "I'm worried about Quicktime I hear it gives a profound security problem"
To suggest I think your views a little rabid is more than obvious. Todays Microsoft marketing FUD statement. Avoid using Apple software it represents a "profound security problem". Use only the wonderful safe Microsoft alternatives now. Oh and don't use Real Player either that also pollutes the essence of Microsoftness.
profound
Yup, Apple represents a profound security problem.
"Apple represents a profound security problem
The fact that iTunes sneaks in so many other apps during it's install and update routines, qualifies Apple as a Malware distributor.
Oh yes...Profound!!!!!!
I'm not sure what world you live in, but as an IT Manager, I tell my folks to get rid of Apple's software as soon as we see it around here. Safari, iTunes, Quicktime, whatever -- they've all been implicated in very real, very widespread, very dangerous security holes on a Windows platform. And Apple's annoying tactic of sneaking in extra programs during an update is something I can't tolerate. Microsoft used to do the same until everyone called foul on it; now they're smart and they don't anymore (also Sony).
I can understand Apple trying to play "catch-up" and seeding the market with these illicit downloads, but that only marks them as desperate and afraid. They'll grow up someday...
On no
You miss my point on two grounds. Firstly I would hope there are not that many corporates who even allow users to download their own applications so the issue if not as Ed suggests one that is likley to be profound.
The second is that you seem to be a client of FUD. Whilst your statement has some truth I could make the same claim if not worse for the equivalent MS product. IE in particular has ofren been implicated in "dangerous security holes on a Windows platform". To what extent one set of products more prone to risk that other is open to debate but to see things in such a black and white way may be being determined by preference rather than just facts?
The Accountability Gap
Windows platform is much more complex than what Apple
is used to, but that's no excuse. Apple is capable of
getting a bad record for it's PC software unless it does
more testing and gets more experience.
Microsoft meanwhile is responsible for the proliferation of
open architecture through broad licensing. They have
encouraged lax development standards to speed adoption.
They have not seen fit to manage the exponential increase
of complexity. To the point where their platform bursts it's
boundaries and relies on it's teams of cognoscenti, it's
OEM partners, and it's users to accept the offloaded work.
So who's to blame? It's still Apple. No, really, I'm dead
serious. Apple still has to cobble together something out of
junkyard parts, for a function that is fundamental to a
system and should have been a built-in services for years.
Apple is, and will be accountable. It's refreshing. They are
responsible to fix it, and it's their software that should be
ditched if it's not acceptable.
If it will never be Microsoft's fault for nurturing complexity,
then it is not their "fault" for nurturing "open" either.
The double standards here are yards thick. In the midst of
calling for a "strategically limited" hardware reference spec
from Microsoft, our author is still prepared to ask us
whether we are "for freedom" (and Microsoft), or "against
it" (and Apple).
Just as Vista's rejection is about the cumulative effect of
Windows grief, any given problem is capable of now
becoming the last straw. Users see iTunes work on the
Mac. They see it fail on the PC. If there is a suspicion of
conspiracy or "software malpractice", it would be better if it
came from a more informed place. One that better
understands the virtues of the vertical integration that is
now seen as the better direction, even if it is not to be the
destination.
So does this have everything to do with the divergent
models of MS and Apple, or does it have nothing to do
with them? We'll certainly never know if the focus is always
going to be about this weeks problem, or driver files
(afs.sys and afs2k.sys). We've gone so far down the rabbit
hole that redirecting blame away from Microsoft is a
thriving cottage industry. Users meanwhile are looking for
accountability, good management, rewards rather than
punishments for license compliance. "Openness" was
always the pseudonym for "closed but huge". That shell
game has gotten old.
Doesn't it hurt?
Okay
I just extend the arc of your argument. A principled stance
on software malpractice sees Apple's flawed Windows-
based jukebox, resented, patched, or ditched off PCs in
favor of alternatives. As it should be.
It helps explain Vista resentment. So Vista is working now,
why the illogical "Mojave" response? People are crazy? No,
people are complex, infinitely more so than computers in
fact. The irony? For years folks have had bad drivers. Take
a bow.
Like I say, you've been diligently chronicling the abuse of
Windows users by Developers, Hackers, OEMs, and
Microsoft themselves. There have been weekly calls for forbearance. Apple is this week's flavor. Evidence that
Windows may be inevitable. There will never be an
indication that the platform is the larger and endemic
problem. There will never be a concession that another
model might provide a better experience, or is even worth
looking into. I'm just thrilled that Apple is on the radar to
some small extent, and look forward to the dissection of
any other failures they have on the PC. One thing I can be
sure of we won't hear of any successes.
You sink to ever lower lows!
Missed a Bit
fault.
It's Apples fault.
There it is again! I suppose if you're at odds with any of my
past declarative statements, you'll be at odds with this one
on principle, but at least take it under consideration.
Speaking of principles, ditch the jukebox if you're
principled, then apply the same principled demands for accountability to all the places where Apple isn't, and that's
a good many places. Let me know how that goes. Nothing
makes me angrier than software that sporks itself or your
machine after installing or after deinstalling. Do you need
further clarification?
Do driver issues ring any other bells with you, or are we to
continue with context-free workshops on this week's PC
hassle? If ever, there has been a precedent for driver
related problems on a Windows box, let's put it in context.
That context is... that paragon of virtue... open
architecture. Sorry if that blows your mind.
You need to go into politics.
What, seriously, were you trying to say? I get the gist of one point was Vista is too complicated? Well, if it's too complicated for Apple to write and test on, then don't. Seems like a simple solution.
TripleII