iTunes 2.0 Installer Deletes Hard Drives

Cheviot writes: "It seems Apple's new iTunes 2 installer deletes the contents of users' hard drives if the drives have been partitioned. I personally lost more than 100gb of data. More information is available at Apples Discussions board. (registration required). Apple has pulled the installer, but for hundreds, if not thousands, the damage is already done." The iTunes download page has a nice warning about the problem. Ouch.

Comments

How the hell does this happen?

[dimator]

Well, here's the pseudo-code:


if(installDrive->hasEnoughSpace()){
return startInstall(instalDrive);
} else {
installDrive->formatRecklessly();
return startInstall(installDrive);
}


Hard-to-spot bug, actually.

Re:How the hell does this happen?

[RiffRafff]

Doesn't anyone test their code anymore??? I mean, how hard can this be?

Re:How the hell does this happen?

[Jace of Fuse!]

No.

Sorry.

Extensive testing went out of style around the mid 90's.

Oh -- and actually writting REALLY good code, that's been out of style since at LEAST the 80's.

I blame colleges for letting so many graduates think that they are instantly coders.

Real coders are born, not made.

And anybody who got into computers for the money, and not the thrill of writing code -- well -- they're worse than lame. And they probably use AOL.

Re:How the hell does this happen?

[Apotsy]

It happens because Apple uses the .pax archive format inside of its "pkg" files. This format is known to be a risky choice because it can do very "destructive" updates if not used very, very, very carefully as seen in this case.....

Yes, pax is a complete piece of crap. The good folks over Stepwise warned about problems with it a long time ago (scroll down to the heading "Installer.app"). Apple blithely went on using pax anyway.

Perhaps after this disaster, Apple will finally realize they should maybe start using something more robust for their packaging system on X.

It is worth noting that the Mac OS 9 version of the iTunes 2.0 installer (yes, they are still updating the Mac OS 9 version of iTunes) did not exhibit the problem, as it doesn't use pax. It uses Apple's old "tome" installer archive format instead.

Re:How the hell does this happen?

[Apotsy]

Eh, I take that back. I see now that it wasn't a pax thing after all. It's something else.

But pax still sucks. :-)

Re:How the hell does this happen?

[cshotton]

Apple has had a greatly diminished QA organization over the years. I think this is due in large part to the serious turnover the company experienced in late '97 - early '98. Nearly every engineer that worked on the original Mac OS left during this time. The Developer Technical Support organization was gutted, and quality assurance became an exercise in million monkey button pounding.

Having been the author of a 3rd party product bundled and shipped on Apple hardware, I can tell you that the extent of their QA process doesn't go much beyond making sure the software installs and runs on an out of the box system, followed by some mediocre mashing of buttons and menus. They really don't understand or implement the concept of actually testing on live, deployed, end user (like) systems. They have racks of off the shelf machines with standard software loads. If they install and run and stay up over the weekend, it's shippable.

We would get reams and reams of complaints about how dialog boxes weren't formatted just so, etc., but their QA department never caught a single defect that most would consider a bug in the code. And there were certainly bugs to catch.

This is a chronic problem that most commercial software houses have. They tend to put junior people with little product experience in the QA organizations and assume that by acting like reasonably competent users, they will somehow uncover logic flaws, data errors, and other engineering foibles. The only time I ever saw QA done right was on a NASA project with life critical software systems. The project was staffed with the very most senior engineers running the QA department and all of the junior engineers were slinging code.

It was up to the gray beards to make sure the junior guys wrote code that was to spec, integrated properly, handled all of the possible input scenarios, and actually performed in a live environment. These senior guys were also the architects of the system, so they knew what the software was supposed to do, how it was supposed to be constructed, and what it should take to break it. I doubt that 1 in 100 commercial shops today have an engineer working in the QA department that actually understands the code they are testing down to the module level. When was the las time you saw a QA guy in a design session, learning about how the system he's going to test is going to be architected?

This is so far from the current practice in commercial industry today as to almost have the flavor of a fairy tale. Apple's no different than any number of other companies who are rushing to ship software on a too short schedule. They pay lip service to QA and rely on their early adopter users to find any lingering problems. In this case, they totally dicked over their customers by not doing their job. However, they're only partially to blame since I think the development of iTunes is still done by Casady and Greene under contract to Apple. I'd be surprised if they weren't ultimately responsible for creating everything, including the installer. Regardless, Apple should have tested this before sticking it on-line on a Saturday night.

Re:How the hell does this happen?

[Anonymous Coward]

You raise some interesting points about testing in industry.

I would like to offer some comments as a tester, myself.

Yes, testing is not generally done correctly or particularly rigorously. There have been some common factors in my experience:
1) Code not developed to Specification. This is by far the most common. I can't count how many times I've been told "But this is a better way to implement XXXX."
2) Poor unit testing of software.
3) Delay in testing due to: delays in development, delays in configuration/delivery of environment. (Don't get me wrong, this is really a management issue. I have NEVER been on a project that has let the testing deadline slip even though the entry criteria were slipped - sometimes even by 3 months! It is ridiculous to not adjust testing periods.)
4) Unrealistic expectations or deadlines by PM/managers.
5) Downright incompetent testers. ("Oh, we don't need to test that.")

But, I would tend to disagree about your point concerning QA participation in code reviews of module level code. Frankly, module testing is the responsibility of development. The finest granularity that testing function should be concerned with is the interface level. (unless, of course, your org uses QA people to do module testing.) This is because testing is required to have an understanding of *more* components/modules than any given developer. To expect them to know the intricate workings of the interior of each module is unrealistic. To expect them to know the interface of those modules, how they behave together, the data flows, etc is quite reasonable. This reminds me actually of the biggest problem that I've encountered (addendum to above): poor or non-existent requirement documents. How the hell are developers supposed to do a reasonable job, and testers test it if there are no frickin' requirements?

Re:How the hell does this happen?

[billcopc]

Real coders are born, not made

Amen!

(and @#%& that stupid 20 second reply timer)

Re:How the hell does this happen?

[Phil Wherry]

Hmmm. I'm just expecting Apple to issue a press release soon that says something to the effect of, "in retrospect, perhaps we shouldn't have subcontracted the installation script to RIAA after all."

Re:How the hell does this happen?

[clancey]

extensive testing is no longer cost effective.

Re:How the hell does this happen?

[clancey]

Actually, this is not just a problem in the software industry. It's a problem, in general, across all industries in the USA, at least. QA is an expense, cuts into profits, and slows down production. Most of the QA Departments I've seen are run by people working outside of their specialty. At my site, a mechanical technician is running the entire QA site and determining the QA requirements for electronic testings. He's a good guy and tries hard, but is way outside of his field.

Re:How the hell does this happen?

[ChuyMatt]

And this is why Open source is much better for the end user. It is not said to be perfect, but you get updates that fix all the problems because of such a huge user base. It would be absurdly simple for the closed source people to emulate this in a little way, have automated bug reports (ala bugzilla, kinda) and do the "release often" thing. Hm... I feel like i am quoting Eric S. Raymond right now...

Anyway, Apple has a new installer that fixes this, less than a day after the original installer was released. that is actually rather good for a large company, I think.

Re:How the hell does this happen?

[whereiswaldo]

> Real coders are born, not made.

I'd also add that real coders probably have a certain type of upbringing... need to have an inquisitive mind, want to find out how things work, want to do things better, and only want to be limited by their imagination.

> And anybody who got into computers for the money, and not the thrill of writing code -- well -- they're worse than lame. And they probably use AOL.

Yeah! You said it, man.

Re:How the hell does this happen?

[whereiswaldo]

The sad part is, the users end up being the beta testers here. Companies can put out crappy products and let the end users find the problems. It saves a lot of money in QA. Plus, a lot of people seem to think this is acceptable.
Having used open source and commercial software, I'd have to say there's no clear distinction between which is more or less buggy. It comes down to companies or individuals. But, with open source, you CAN fix the bugs if you want, and you WILL get more frequent fixes. In the meantime, in either case, you're left sitting there with an unusable product until the fix comes in. Plenty of time to look for alternatives.

Re:How the hell does this happen?

[aberkvam]

However, they're only partially to blame since I think the development of iTunes is still done by Casady and Greene under contract to Apple. I'd be surprised if they weren't ultimately responsible for creating everything, including the installer.

Casady & Greene, Inc. was the publisher for SoundJam. The development team didn't work for them directly. When Apple decided that they needed an MP3 player they look a look at the Mac MP3 market, decided that SoundJam was the best, bought the rights to SoundJam, and hired the development team.

If you compare SoundJam to iTunes (especially earlier versions of iTunes) it's not hard to find places where the SoundJam code sticks out a little bit.

But the end result of all of this is that Casady & Greene, Inc. no longer has anything to do with the iTunes code. It's all done in-house by Apple so Apple has to take the full blame. (It's a shame that we'll never see the Mac OS X production version of SoundJam.) There are some more details on the official SoundJam website.

Oh, and the installer used appears to be Apple's own PackageMaker (included with the free Developer Tools) which allows the running of pre-process and post-process scripts.

Re:How the hell does this happen?

[benedict]

I know plenty of people who program "for the thrill of writing code" and *still* can't code their way out of a paper bag.

It takes more than a love of computing to make a good programmer. In my humble opinion, it takes a fair amount of education. A brilliant but naive programmer can screw things up pretty impressively. I'd prefer to use code written by someone who isn't such hot shit but takes the time to learn APIs, read documentation, and familiarize himself with the idiom of the language and/or operating system in question.

Re:How the hell does this happen?

[RoscoHead]

They pay lip service to QA and rely on their early adopter users to find any lingering problems.


Sounds like most Linux drivers to me...hands up those who've never had similar problems with a Linux system?

Seems to me they came up with a fix pretty quick, which flies in the face of those who say open source always provides better service.

Re:How the hell does this happen?

[demon]

I don't think the person you're replying to meant that education isn't important, but without at least some raw talent, and an interest in programming, education by itself just isn't going to make a person into a programmer. Just as with anything - most anybody can be taught to do most tasks, but if they don't have an interest in the task at hand, odds are they're not going to do it _well_.

Re:How the hell does this happen?

[bradintheusa]

We have a rule at work, never hire anyone with an AOL address on their resume.

Re:How the hell does this happen?

[gray code]

Real coders are born, not made.
Pretentious cunt.
(Score:-1)

Heh, truth modded down to oblivion in favor of conceit and masturbatory ego-stroking... Too bad pointing it out doesn't make it go away.

New Apple Slogan

[Anonymous Coward]

Rip. Mix. Burn. Format. Reinstall.

Re:New Apple Slogan

[yesthatguy]

Rinse. Repeat.

Re:New Apple Slogan

[Vonatar]

At least it doesn't install Windows afterwards...

hrm..

[AntiTuX]

Gee, maybe it's an omen....

*Cough*Nomad*Cough*Cough*

Copy Protection

[Tachys]

Guess their copy protection code gets a little over-zealous.

Note to Apple Corp:

[Travoltus]

A budget for a Quality Assurance tester team is 100% NOT WASTEFUL spending.

Liability

[jeti]

I really wonder about the legal foundation of:
"You should've backuped. We're not responsible
for any damage that erasing all your data caused."

(Yes, it's in the license. But can it be valid?)

Re:Liability

[JohnHegarty]

That's like saying we are not responsible that the car you bough has a design flaw, which causes it to blow up ... but were not responsible .... really , i swear...

Re:Liability

[Moridineas]

What do you mean can it be valid? You didn't HAVE to install the software, nobody made you. Indeed, you had to agree to said license before you could download or install it. Sure, it may not be cool, but if it's in the license agreement and doesn't break any laws, i don't see how it can NOT be valid.

Scott

Re:Liability

[jeti]

Well I don't know too much about US law.
But in Germany, if you cause damage by
being grossly careless (grob fahrlaessig),
you're generaly held responsible for the
damage you do. I wonder if a license can/
should be able to override such laws.

The other question is: Can you really expect
people to do a daily backup? What's the
percetage of computer users (and companies)
doing this?

Re:Liability

[Anonymous Coward]

I really wonder about the legal foundation of:
"You should've backuped. "

And more interestingly: if your backup software erases all your data and has such an EULA, what happens?

Re:Liability

[cthugha]

In common law countries (UK, US, Canada, Australia, NZ, etc) parties can be excluded from liability in negligence as part of a contract/legally binding agreement, but that doesn't apply to a third party, and also won't apply in cases of spectacularly gross negligence, IIRC.

Re:Liability

[QuickFox]

there is not a state in the union that will find myself responsible for this if I include the proper EULA

By clicking on this attachment you have accepted this license agreement: IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY SPECIAL, INCIDENTAL, INDIRECT, OR CONSEQUENTIAL DAMAGES WHATSOEVER ARISING OUT OF OR IN ANY WAY RELATED TO THIS VIRUS.

Re:Liability

[gig]

> you're generaly held responsible for
> the damage you do.

But in this case, the "you" is the user. The user downloaded and installed a software package that happened to have a bug in it, and in so doing, they may have lost data and have to restore from backups. End of story.

Software bugs are a Fact of Life. They weren't invented yesterday.

The ironic thing here is that the Mac OS X installer is rarely used ... iTunes 2 is maybe the third app I've ever had to run the installer for. Most apps just come in a disk image, and you mount the image and drag and drop the app where you want it to live. Or they come on a CD or DVD and you drag them to the hard drive in the same way. I prefer that over running the installer and letting it have free reign over the system.

Re:Liability

[jeti]

Here's a quote from the discussion at macnn:

<quote>
Aha! It seems that this is not quite correct for some jurisdictions. They could be liable per the EULA, last line of implied warranty section:

"SOME JURISDICTIONS DO NOT ALLOW THE EXCLUSION OF IMPLIED WARRANTIES OR LIMITATIONS ON APPLICABLE STATUTORY RIGHTS OF A CONSUMER, SO THE ABOVE EXCLUSION AND LIMITATIONS MAY NOT APPLY TO YOU. "
</quote>

(BTW: You definitely got the uppercase right. Obviously it is used to make the EULAs even more unreadable.)

Re:Liability

[AndroidCat]

Contracts themselves have no inate power

Maybe the contracts you sign! Mine generally involve blood and elder signs...

Meanwhile, back in Dunwich

Re:Liability

[lobsterGun]

I'd like to believe that they can be held liable wether or not they had malicious intent. This isnt just a bug, this is gross negligence.

Re:Liability

[wrt]

The EULA doesn't come up until after you install itunes. You accept the license agreement when you run it for the first time.

I didn't lose any data, that would have sucked.

Re:Liability

[42forty-two42]

They say that you should've backuped. ;)

Re:Liability

[whereiswaldo]

IMO, this is a basic problem with the justice system we are using.

Money = Power > Justice

Too bad the scales of justice can be made heavier on the side with the most money and power.

Re:Liability

[Galvatron]

I sure HOPE it's valid, because it's a critical underpinning of the GPL. No one would write Free software if they could be held financially liable for any damage the bugs cause.

Re:Liability

[Sloppy]

If they're liable, then nobody can afford software anymore.

Re:Liability

[ichimunki]

I'm not certain, but I recall reading somewhere that if you get something for free, that certain warranty rights are not binding. Here is some support for that idea: http://www4.law.cornell.edu/uscode/15/2302.html. This posting is not legal advice, and is for entertainment purposes only.

Re:Liability

[geekoid]

This is a forum not to ask that question.
Many people will talk about whats "obvious" even though they have no legal reference.
EULA has never been tested to this degree, so the answer is really unknown.
Personally I'd like to see a class action suit against apple. I would like to get ever consumer advocacy group on the planet involved.
I think the public outcry may have a profound effect on EULA's.
Juyst because it's in a contract, doesn't make it binding.since the EULA is the worse kind of contract, vague, untested, overly broad, and doesn't apply to minor who may install software, it really need to be beat up in court.

Oh, come on...

[CrayBeast]

Don't they test these things, anymore?

Really, in the current economic climate, all the monkeys should have been thrown out of the high-tech jobs, leaving only clueful people.

How does a bug like this occur?

Re:Oh, come on...

[Pathetic Coward]

In the current economic climate, all the clueful people have been laid off. Testing is considered a wasted expense by the MBAs interested only in their personal profit.

Re:Oh, come on...

[GISboy]

Really, in the current economic climate, all the monkeys should have been thrown out of the high-tech jobs, leaving only clueful people.

Well, what you said is the working theory, anyway.

Having worked in the corporate world and the academic world this is the furthest from the truth. The people with a clue, ethics, responsability, talent, skills or value customers are usually the first on the chopping block.
After all, the managers making those 5 and 6 figure salaries have to remain employed so they can continue the (vicous) cycle.

Cynical? Oh, yeah, been there, been IT, seen it happen too many times.

Could apple be any different? That is a tough one to answer. I would have to say no, but to a lesser extent, perhaps.

Why to a lesser extent? For the simple reason that Steve Jobs and Lee Iacocoa (sp?) understood two things about running a company/taking over one:
First get everybody on board with a plan to succede/improve morale.
Second (and this is the kickass part) when you clean house *never, ever* get rid of your workers.
Clean up/fire your middle and upper management levels.

This solves 2 problems (imagine a pyramid):
1) when most layoffs happen they happen to the "base of the pyramid". What happens when you weaken the "foundation" of a company/structure.
Yeah, it falls down or does irrepairable damage.
2)Wiping out the middle section brings those "at the top" closer to the base. Most executive understand the "how and what" of a business, but understanding the "who and why" is what keeps thing "moving forward".

If I remember correctly, Lee I was first, and Jobs subscribed to the idea...it may have come from a /. link when Jobs returned to Apple.

Very good interview.

Of course I've always said a "Phd/manager saying 'in theory' is akin to a used car salesman saying 'trust me' ".

I guess in my snide cynicism I found humor in your altruistic logic

Re:Oh, come on...

[Greyfox]

Clueful people cost more. We recently interviewed a few people for an open position and I reccomended the guy who could actually have helped fix our project up. Our team-lead went for the second most competant person we interviewed on the basis that the guy I wanted would be bored and leave quickly. Our manager went for the wet-behind the ears college graduate. The money involved was the largest factor.

Ask a bunch of hiring managers right now and I think they'll tell you that they'd prefer someone adequate for a job over someone perfect for a job if it meant a salary difference of $10K - $20K. This recession isn't going to lead to a concentration of clueful people in our industry. It's going to lead to a concentration of monkeys.

Re:Oh, come on...

[lowflying]

Really, in the current economic climate, all the monkeys should have been thrown out of the high-tech jobs, leaving only clueful people.

You are dangerously close to questioning the notion that we live in a meritocracy. Take the blue pill and you will feel much better tomorrow.

Dave

Re:Oh, come on...

[CordMeyer]

prove osama's guilt and rig the aus election at the same time. pretty smooth!

Re:Oh, come on...

[aiabx]

The counter-argument is provided by the companies that purged their middle management in the recession of the 80's. They saved a lot of money, there was less management, and everyone was happy(er), but now the senior managers are due to retire, and there's a critical shortage of people available to be promoted to the senior positions. So you end up importing people who don't understand the business, and you're heading for a brand new disaster.
-aiabx

Good Read

[Kira-Baka]

http://newforums.macnn.com/cgi-bin/ultimatebb.cgi? ubb=get_topic&f=46&t=000865

It has some info about causes and solutions...

Re:Good Read

[batobin]

Actually, your URL has a space too many. Here's one that works:

Link, or if you're scared I'm trying to show you porn:
http://newforums.macnn.com/cgi-bin/ultimatebb.cg i? ubb=get_topic&f=46&t=000865

Re:Good Read

[akc]

http://newforums.macnn.com/cgi-bin/ultimatebb.cgi? ubb=get_topic&f=46&t=000865

You need to loose the space after the "?"

Re:Good Read

[Jburkholder]

>Actually, your URL has a space too many

so does yours, courtesy of /.

Re:Good Read

[cygnus]

Actually, your URL has a space too many.

spaces in disk names are what causes this problem in the first place. so if i view it on my Mac, will it erase my hard drive, too? or maybe it'll delete the Internet. damn, Macs are dangerous.

Re:Good Read

[batobin]

Oh, so that explains it...

Just click on my handy dandy link then.

Already updated

[Anonymous Coward]

Apple has already put iTunes 2.0.1 that purportedly takes care of the problem:

http://www.apple.com/itunes/download/

Re:Already updated

[clueless idiot]

They got it right this time. iTunes 2.0.1 installer now only deletes the contents of users' hard drives if the drives have NOT been partitioned.

A bit late now.......

[JohnHegarty]

"An important note for those who have downloaded iTunes 2.0 for Mac OS X:
Apple has identified an installer issue with iTunes 2.0 for Mac OS X that affects a limited number of systems running Mac OS X with multiple volumes (drives or partitions) mounted. For those systems, running the iTunes 2.0 installer can result in loss of user data. While this error is highly unlikely to affect most users, Apple strongly advises that anyone who has downloaded the 2.0 version of iTunes for Mac OS X, as well as anyone who has a beta version of iTunes 2.0 for Mac OS X, immediately remove the iTunes.pkg installer file from their system. A new version that corrects this issue, iTunes 2.0.1 for Mac OS X, is now available from this page. Users who have already installed iTunes 2.0 without incident do not need to reinstall iTunes 2.0.1, but they should still immediately remove the 2.0 installer file from their system. This issue does not affect users of iTunes 2.0 for Mac OS 9."

Its like they are saying , its out fault , honest....

*sigh* Amateurs...

[cthugha]

Now that's not very tight code, is it? Instead, try:

if(!installDrive->hasEnoughSpace())
installDrive->formatRecklessly();
return startInstall(installDrive);

A much superior algorithm for nuking innocent users' hard drives, I'm sure you'll agree.

Re:*sigh* Amateurs...

[dimator]

Much cleaner. I stand corrected.

Re:*sigh* Amateurs...

[B'Trey]

Given today's optimizing compilers, the two are probably functionally equivalent.

Re:*sigh* Amateurs...

[whereiswaldo]

Yes... and that is all that matters. Besides, installer code isn't known to be the most efficient code ever written anyway. Who gives a shit.
Save your talent for the kernel... lol.

Corrected version 2.01 already posted by Apple.

[sakusha]

Come on now, Apple jumped on this one, it was only reported by a couple of people, and they corrected the problem almost immediately. This problem only came to light today, and they have a fix out the same day. I downloaded the new 2.01 version, installed with no problems.

Re:Corrected version 2.01 already posted by Apple.

[JohnHegarty]

Great if you haven't just lost 100gb of data.

Re:Corrected version 2.01 already posted by Apple.

[Spock the Vulcan]

And I salute you, sir, for being brave enough to even think of installing what has clearly been demonstrated as badly written, dangerous software. I somehow don't believe that if a bug like this got through into a release, that it might not have more nice things inside it, that version 2.01 might not have corrected. (Ok, too many negatives in one sentence).

Re:Corrected version 2.01 already posted by Apple.

[wizbit]

This problem only came to light today, and they have a fix out the same day.

Took the words out of my mouth. I concede that this is more severe than a security bug since an otherwise benevolent installer inadvertantly destroyed real data because of a poorly-written shell script.

But the fast turn-around by Apple at least minimized their collateral damage. Had this not turned up within hours, and a fix not been available within a couple more, it could have been a serious problem. But give Apple credit for fixing a problem like this quickly.

iTunes 2 was rushed to market because the iPod requires it. QA slipped, and didn't test this on enough configurations to realize the error before publishing it. But the work that went into iTunes finally brings some features that I've been waiting for to this product (who ever heard of shipping an audio player without an equalizer?) - don't let this cloud the hard work the apple engineers put into iTunes.

Re:Corrected version 2.01 already posted by Apple.

[Lars T.]

What has been "demonstrated as badly written , dangerous software" was the installer shell script. Apple doesn't let the skript kiddies do the real programs.

Re:Corrected version 2.01 already posted by Apple.

[telstar]

Yet in packaging the install script with their product, they are equally responsible ... at least from the moral standpoint. Companies usually cover their asses enough to make sure that 10-page disclaimer you must agree to before installing software includes a clause about them not being legally responsible for any incidental dammage caused by the use of their software.

Re:Corrected version 2.01 already posted by Apple.

[Jeremy Erwin]

While timely bugfixes are to be expected, a certain level of competence should be expected from a software vender. This should have been caught before release.

The error in the installation script was, in part, due to Apple's relaxation of certain unix conventions regarding spaces in filenames--conventions that have proven useful. And while some might dismiss this as merely evidencing Apple's unfamiliarity with UNIX, shouldn't this have been caught by the Nextstep engineers?

Nonetheless, most x.0 releases are buggy.Windows 3.0 was widely regarded as half baked. OSX-10.0x was very slow. The subsequent x.1 release is usually far better.

This even extends into free software. linux-2.4.0 was buggy. gcc-3.0 was buggy. It used to be that there were no hard and fast deadlines for free software, but significant pressure to shorten long development cycles may cause corners to cut. The gcc team even has deadlines...

At least the standard installation scripts (provided by GNU tools) are reliable and trustworthy... (I hope).

Re:Corrected version 2.01 already posted by Apple.

[Lars T.]

Some troll on an other forum claimed there were reports from people who had it happen with (pirated) betas. Now if this is so, they obviously didn't report it to Apple (because their copy was pirated), else Apple would have just fixed it. So it's them who are moraly responsible.

Re:Corrected version 2.01 already posted by Apple.

[PhiloMath]

Not entirely true. This fault was present in the 2F13 installer (not sure if that's the same as 2.0) and it completely wasted my friend's Classic partition. He probably should have reported it, but it was a prerelease copy so he figured it'd be fixed by the final. If Apple did any quality testing, they'd have found the problem, too. That said, it should be pointed out that the flaw is with the installer and not the program itself. iTunes 2 itself seems to be wonderfully coded. This could also explain why the problem wasn't caught, if they were focused on testing the application itself and not the installation process, too.

Its an expensive lesson for Apple to learn. Few can even afford to even begin to offer compensation or be held otherwise accountable for the loss of data -- it's precious stuff. But I'd bet whoever made that installer feels really shitty right now. Karma.

Re:Corrected version 2.01 already posted by Apple.

[firewort]

Your friend shouldn't have been so quick to presume the bug would be found, even in a pre-release.

It would have been far better for him to have reported it, so that it would have gotten some attention-

Remember, the engineers would rather code it and forget it, but if a few users report the problem, they're forced to devote some time to seeing if it's repeatable, and fixing it.

Sure, Apple ought not to need suggestions and bug/issue reports from regular folks, but they give us that feedback page on their website--- Use it!

Re:Corrected version 2.01 already posted by Apple.

[PhiloMath]

Agreed completely. Illegitimate copies breed paranoia, I guess. Another reason why open development is the way to go.

Re:Corrected version 2.01 already posted by Apple.

[firewort]

Apple's typical behavior in this situation (reporting bugs on an illegitimate copy of pre-release software) is _usually_ to quickly contact the bug reporter and get them to sign an NDA, and agree to be a beta-tester.

Your friend could have lucked into being a software tester for Apple, getting legitimately-obtained pre-release software.

This has already been fixed...

[everyplace]

While no-one will disagree that this was a MAJOR problem, it has already been fixed. I have to say that it was pretty bad form for this bug to get out of the door today with the release of iTunes 2, but at the same time, I'm glad that they fixed it within 24 hours.

For a short time this afternoon the OS X installer wasn't available, but it has since been replaced with a version that doesn't randomly reformat other partitions or drives. The new installer can be found at apple's web site, free to download of course, assuming that you can run it, IE you're on a mac.

everyplace

Wide spread?

[panZ]

I'm guessing this has happened to a limited number of people. I used the old iTunes2 installer on a number of machines with multiple partitions as have my friends and none of us lost any data. What conditions cause this "feature" to occur?

Re:Wide spread?

[trash eighty]

yeah worked on my 2 iMacs fine too

did u install under OS9 or X though, 9 might be fine

Re:Wide spread?

[moof1138]

For this to come up you need to have multiple partitions, one of which is named 'Applications.' This is not too common, but it is done.
I knew a guy who did graphic design who did this, I always though it was kind of dumb back then, since apps not running on the boot volume in Mac OS 8.1 - 9.x took a performance/VM hit. It doesn't have that impact on X, but I still don't see much benefit.

Any sort of renumeration?

[tahpot]

I assume they would have had a disclaimer along the lines of "run this installer and we take no responsibility", but surely with an error as bad as this they must take some blame and give their customers adequate compensation. Deleting data is about the worst sin that software can do. The software wasn't marketed at deleting data, so the fact that it does, and irresponsibly, should make them liable.

Late posting?

[ted_rust]

The problem was fixed long before this story was even posted. The bug is not inexcusable, but at least they pulled it quickly, fixed it within a day and posted a new version with a clear explanation of what happened.

--
tsr

[PATCH] Re:How the hell does this happen?

[Anonymous Coward]

There's a bug in you pseudo-code. Attached is a patch which fixes the problem.

--- itunes-install.pseudo-orig Sun Nov 4 01:36:11 2001
+++ itunes-install.pseudo Sun Nov 4 01:36:19 2001
@@ -1,5 +1,5 @@
if(installDrive->hasEnoughSpace()){
- return startInstall(instalDrive);
+ return startInstall(installDrive);
} else {
installDrive->formatRecklessly();
return startInstall(installDrive);

Re:[PATCH] Re:How the hell does this happen?

[jesser]

There's a bug in you pseudo-code. Attached is a patch which fixes the problem.

It's pseudocode. You can't patch pseudocode. You have to pseudopatch it, like this:

- return startInstall(instalDrive);
+ return startInstall(installDrive);

or like this:

Replace instalDrive with installDrive.

Re:[PATCH] Re:How the hell does this happen?

[serial frame]

I don't mean to troll, but...Redundant? Are people THAT humour-impaired?

It seems our moderators need to ease up on their high horses.

On a different note, an installer should NOT remove files...Maybe except /tmp files that are of their own.

The bug

[hysterion]

Summarizing discussions on MacNN and the Apple Forum:

The problem appears to be in two portions of the installer script which could translate into rm -rf /your_drive, if certain paths $1 or $2 contain spaces:

#!/bin/sh

# if current iTunes pkg exists, delete it b/c of Installer bug
if [ -e $1Library/Receipts/iTunes.pkg ] ; then
rm -rf $1Library/Receipts/iTunes.pkg 2> /dev/null
fi

# if iTunes application currently exists, delete it
if [ -e $2Applications/iTunes.app ] ; then
rm -rf $2Applications/iTunes.app 2> /dev/null
fi

Though when I looked, nobody seemed to have found where exactly $1 and $2 are defined; also it might be that disaster only strikes with localized versions of the OS.

Re:The bug

[Lars T.]

Gee, I keep telling people that 99.9% of all "just typed it in and it works" shell skripts fail for filenames with spaces (or othe "funky" characters) in it. This is a typical "UNIX is so damn powerfull" bug.

Re:The bug

[MobyDisk]

I for one am surprised that so many critical pieces of software, particularly installers, are written in loosly-typed non-compiled languages such as perl, or shell scripts. I come from a DOS/Windows world where scripting is not available to the degree it is in *nix, and everything is written in fully compiled languages.

I'm not trying to say that this bug might not have happened if the code were written in C++, but I think it would be more difficult. Languages that allow such convenient string concatenation syntaxes also allow for unseen bugs in the substitution. I've never seen a Windows program install using .vbs files copying to directories. It's not that it couldn't, but it just isn't up to that scale deployment.

Perhaps this practice of using perl/csh for crucial coding should be re-examined?

Re:The bug

[RFC959]

Yeah, but this is why you NEVER EVER use spaces in filenames under Unix if you can possibly avoid it. Whitespaces within tokens clash with a grammar in which whitespaces /separate/ tokens, and if you ignore this, it /will/ bite you at some point. Not that Apple's people are idiots, but they're clearly not used to thinking this sort of thing through.

Re:The bug

[nosferatu-man]

Spaces are perfectly illegal filename characters -- it's just that the laughably primitive Unix tools aren't smart enough to understand them.

Peace,
(jfb)

Re:The bug

[Lars T.]

Which proves that the guy who made the error is a UNIXer, somebody having experience with the Mac would have known that Mac users do use spaces in their file-/volume- names.

Re:The bug

[benedict]

Your post reminds me of a conversation I had with an Apple employee. He shared your opinion of the Unix tools. I challenged him to come up with a replacement syntax for the shell that wouldn't have any problems with spaces.

After I shot down his first half a dozen proposals, he started to gain an appreciation for the difficulty of the problem.

So how would *you* rewrite the shell to get rid of problems like this? Be specific. Remember, in 30 years of unix, no one has found a solution to this problem that doesn't break more things than it fixes. So if you do so, you'll be famous.

I wait with bated breath.

Re:The bug

[Phexro]

"...nobody seemed to have found where exactly $1 and $2 are defined..."

uh... $0 ... $n are the variables that store the command-line arguments to a shell script. awk scripts also wotk this way.

man bash, search for ARGUMENTS.

Re:The bug

[kijiki]

You can use newlines in a filename too. do you think that's a good idea?

Re:The bug

[nosferatu-man]

Who said anything about the shell? I'm including the whole Unix shell concept in the "laughably primitive" category.

Peace,
(jfb)

Re:The bug

[Canyon Rat]

Coming from a Mac background I worked in a Windows NT shop for a week and the techs couldn't figure out why my machine was broken every morning. It turns out that some network task was trying to back up my work at midnight every night. It couldn't cope with the spaces I put in file and directory names.

The real bug there, and in MOSX, is that the system doesn't warn you if the name is going to be a problem. If you try to embed a colon in a Mac OS 9 file name, you get a warning.

Re:The bug

[benedict]

"Unix is passe" is passe.

Re:The bug

[nosferatu-man]

LOL! Needless to say, I disagree; I think that focusing efforts on improving the Unix house of cards is a pointless waste of effort. How anyone can defend (for instance) flattening data to a stream of bytes, the farcical security model, and the reliance on a C api is beyond me. On the gripping hand, plenty of smart people do, and will continue to do so for the forseeable future.

Oh well.

Peace,
(jfb)

Re:The bug

[Evil MarNuke]

this is so bad. I can't believe anyone got paid to write that script. I mean how th hell did it get out the door with out testing all possiable imputs.

Sad.

It just makes no sense

[lambent]

As inexcusable as this is, it's also inexplicable. Why would a music software suite and its installer have the ability to nuke a drive?

Free Format

[starphish]

I've been looking for a good free format utlity with an attractive front end. Is there a PC port?

Re:Free Format

[Colz Grigor]

Yeah, there is, actually:

In Linux (on a PC, right?):

rm -rf /

In a Windows variant command prompt (File|Run|command):

echo y | format C:

or

echo y | del /s /q C:\*.*

If that doesn't do the trick, I'll loan you a large magnet.

::Colz Grigor

--

Re:Free Format

[10.0.0.1]

In a Windows variant command prompt (File|Run|command):

echo y | format C:

This does not work properly. You have to specify the volume label or DOS will format the drive and then ask you for the volume label. The problem here is that DOS won't ever get your input because it is expecting the answer to its question from the echo command which obviously doesn't have the answer. After you press ^c to get out, you will have a drive with an invalid format

You need as follows:

echo y | format /v:infected c:

Then you will get something like:

c:\>dir

Volume in drive C is infected

No files Found

Etc....

Re:Free Format

[archen]

Actually there is an easier way to do this with the undocumented /AUTOTEST switch. Basically it just formats with no furthar prompts.

format C: /AUTOTEST

mmm...

[Wakko Warner]

...smell that? that's a steaming heap of Apple quality.

- A.P.

Need to have a warranty!

[burtonator]

OK.. so. Your hard drive has just been formated by Apple? You have lost months of work and potentially hundreds of thousands of dollars. What will you do now?

Are you going to sue? Did you read your EULA (End User License Agreement)? You probably waved that right when you said "OK".

Apple probably waived all warranty when you installed the software (in some states this isn't legal though)

This is one area where the law needs to be fixed.

With Open Source software at least you have the ability to read the source code.

Imagine if Ford were to wave any warranty with your next Explorer.

Kevin

Re:Need to have a warranty!

[Macka]

Hmm .. I'm not going to waste a moderator point jacking this back up again, but this is NOT a Troll. The man has a point! This IS an Apple article, and MacOS X is supposed to be a click and point friendly face to Unix where you don't have to be a command line jockey to get things done.

Also, the 'read the code and fix it yourself' argument is starting to wear very thin, as both Linux and MacOS X can quite easily be found in the hands of non-techies these days.

To whoever mod'd this down as a troll .. shame on you, for not being more open minded!

Re:Need to have a warranty!

[Detritus]

You have an obligation to take reasonable precautions to protect the data on your computer. That means making backups of any valuable data. Are you going to sue Western Digital if your hard drive fails? What if it gets fried by a lightning strike? Even if Apple was found to be grossly negligent, they shouldn't be held responsible for data that was lost due to the negligence of the computer's owner.

Re:Need to have a warranty!

[Lars T.]

The funny thing is, the buggy code was a shell script, dead easy to read. So it's either not Apples fault, or Open Source software isn't as safe as you say.

Re:Need to have a warranty!

[uchian]

Well by being able to see the shell script, we can see in retrospect that it was just a bug, one that any of us could have made and one you wouldn't have detected unless you happened to have the right conditions for it to have happened. We can _see_ that it wasn't done on purpose, which at least is better than knowing that it was done on purpose.

If it had been closed source, then we would all be very paranoid right now about whether or not there was some hacker at apple with a vendetta.

Re:Need to have a warranty!

[morgus morphus]

Actually, certainly in Europe, but as far as I'm aware also in the US, if Apple was found to have caused this damage through "gross negligence", then they would be responsible for damages, because damages due to gross negligence on the part of the vendor is exactly what can _not_ be excluded.

The legal definition of gross negligence according to lectlaw is:
GROSS NEGLIGENCE - Failure to use even the slightest amount of care in a way that shows Recklessness or willful disregard for the safety of others.

Or more applicable to this case (from here):
Gross negligence
Any action or an omission in reckless disregard of the consequences to the safety or property of another. Sometimes referred to as "very great negligence" and it is more then just neglect of ordinary care towards others or just inadvertence. Also known as the Latin term culpa lata.

Of course, IANAL, but my gut feeling would be that gross negligence would not be applicable in this case... gross negligence is basicaly one step short of intentional damage.

Re:Need to have a warranty!

[SlamMan]

"With Open Source software at least you have the ability to read the source code."

Yeah, right. When was the last time you read the source code to something before you installed it?

Re:Need to have a warranty!

[frleong]

Yeah, it is also a negligence of the computer's owner to have installed Windows 95 or MacOS 9 infested with nasty bugs. MS and Apple shouldn't be held responsible for data that was lost. Am I following correctly your logic?

Re:Need to have a warranty!

[crayz]

You also can just open up the .pkg file, decompress the contents, and install it yourself, if you're so paranoid about a the install script that you in fact are allowed to look at

Re:Need to have a warranty!

[stripes]

With Open Source software at least you have the ability to read the source code.

In this case the bug was in the installer, which is a shell script and quite readable. It was a novice shell script error, which is understandable since Apple is a novice Unix OEM...

Anyway, you can read that part, and even fix it.

Still sucks though.

Re:Need to have a warranty!

[aberkvam]

This IS an Apple article, and MacOS X is supposed to be a click and point friendly face to Unix where you don't have to be a command line jockey to get things done.

I can't tell exactly what you are replying to but I think you are misunderstanding the problem. The bug is in a Bourne shell script that is run by the installer during the install process. The user never sees a command-line or a script. In fact, you really have to go out of your way to even be able to view the script since it's part of the installer "package" which appears as a single file to the user.

This installer is about as point-and-click as you can get. No need to be a command-line jockey.

Re:Need to have a warranty!

[egomaniac]

I love open source as much as the next guy, but I'm getting sick and tired of the "open source fixes all ills" crap.

Are you going to sue? Did you read your EULA (End User License Agreement)? You probably waved that right when you said "OK".

Oh, and you would have legal recourse if this had been open source software? Whom, pray tell, would you sue then? You generally can't sue the responsible party if it's closed source, and you generally can't sue the responsible party if it's open source, and from this you conclude that open source is better is this respect?

With Open Source software at least you have the ability to read the source code.

And, in case you haven't noticed, a dozen people have posted the offending code from this particular problem as well. It was a shell script, so the code was in plain sight. It was also fixed pretty damned quickly, too.

Imagine if Ford were to wave any warranty with your next Explorer.

Or better yet, imagine if your next Explorer were hacked together by hundreds of college-age geeks who had never once met one another, and if you had problems with it the response was "Well, just open up the hood and fix it. You can see how it all works".

I generally use open source in preference to closed source, but I'm getting sick of the attitude that closed source = bad and open source = good. I've used some great closed source products and some truly crappy open source products in my time.

Re:Need to have a warranty!

[geekoid]

Yes I would sue my HD company if it failed in an unreasonable time.
Having seen a company do this successfully, I now it can be done.

Nature of the bug

[HalfFlat]

From the discussion on the Apple discussion web site, the nature of the bug is as follows.

The original installer script has the lines

# if iTunes application currently exists, delete it
if [ -e $2Applications/iTunes.app ] ; then
rm -rf $2Applications/iTunes.app 2> /dev/null
fi

while the replacement (2.0.1) has

# if iTunes application currently exists, delete it
if [ -e "$2Applications/iTunes.app" ] ; then
rm -rf "$2Applications/iTunes.app" 2> /dev/null
fi

In these scripts, $2 corresponds to the volume on which iTunes is to be installed, and will be of the form /Volumes/name-of-volume.

For those unfamiliar with Bourne shell variable expansion, if $2 has spaces in it, the argument to the rm command in the first version of the script will expand to more than one word, and rm will try and delete both of these. The -rf tells rm to delete everything down recursively and not complain about it.

This is particularly a problem on the Mac, where filenames and volume names often have spaces in them., even at the beginning of the name. If one had multiple partitions mounted in /Volumes, and the one on which iTunes was to be installed was called, say, ' OS X', then the rm command would expand to

rm -rf /Volumes/ OS X/Applications/iTunes.app 2> /dev/null

and would then try and delete everything under /Volumes. This is clearly bad.

The second version, by including quotes around the argument, fixes the problem. The quotes force the argument to be treated as a single argument after variable expansion.

Traditionally, people have been super careful about destructive operations and shell expansions. I don't think I've ever seen something like this written in a 3rd party script before, in fact (let alone from the OS vendor!). This could well be an example of programmers new to a Unix-like platform still getting used to the Unix way of doing things, and getting bitten as a result.

Re:Nature of the bug

[Bongo]

Wow. I don't know shell scripts, but I'm taking your post as my first lesson.

I figured I'd learn to write some shell scrips at some point, but I had no idea of the hidden dangers involved.

Re:Nature of the bug

[Florian Weimer]

Traditionally, people have been super careful about destructive operations and shell expansions. I don't think I've ever seen something like this written in a 3rd party script before, in fact (let alone from the OS vendor!). This could well be an example of programmers new to a Unix-like platform still getting used to the Unix way of doing things, and getting bitten as a result.

But even experienced UNIX users have problems to cope with the fact that on a traditional UNIX file system, file names can include all characters except '/' and NUL. For example,

find /some/path -name \*~ | xargs rm

might delete much more than you want, and it's especially dangerous to do this if you are the root user and process directories where unprivileged users have write access. (Yes, I know, the -print0 and -0 options for find and xargs exists, at least in the GNU versions.)

Anyway, it's quite embarrassing to make such a mistake in an installer. It shows how little testing software gets nowadays before it is released. I can't imagine that this couldn't have been noticed by testing the installer on several internal machines (and not just developer ones).

Re:Nature of the bug

[zardie]

This illustrates the importance of non-rooting. While this is an installer script and installer scripts are mostly run as root anyway, we can cut down on the amount of data loss if we test everything without root so we don't lose -everything- in the case of a booboo.

Re:Nature of the bug

[ernst_mulder]

Well, I feel anyone who uses "rm -rf" in scripts running with root privileges should be extremely careful. I feel this is a "unix beginners" error, not something you'd expect from a major company like Apple.

Why didn't they simply use mv to move the old version to the trash?

Welcome to "The Power of Unix".... Now only if we learn how to use it properly...

Maybe this was a scheme to get rid of all those OS 9 partitions people somehoww just keep using? :-)

Re:Nature of the bug

[ceswiedler]

I believe that filenames can even include /. Shells might not like it, but Unix filesystems don't care.

Re:Nature of the bug

[Florian Weimer]

Of course, this depends on the particular file system implementation. For example, the VFS layer of the Linux kernel does not permit slashes in file names.

Re:Nature of the bug

[scrytch]

I believe that filenames can even include /. Shells might not like it, but Unix filesystems don't care

Shells don't care about /, the filesystem most certainly does. Path parsing is not done by the shell, it's done by the filesystem. creat() and open() and company will all fail if you try to have a file with a slash in it. Try it with a C program sometime.

Re:Nature of the bug

[crucini]

Yup. Shell is a beautiful, elegant, but treacherous language. The combination of variable expansion, globbing (*'s and ?'s) and backticks allow one line of code to have nearly infinite meanings. The error in this script is probably present in many scripts on production Unix systems. Sysadmins don't usually expect spaces in file names. Also, adding quotes doesn't protect against the variable containing a '*'. And I'm sure there are other gotchas.

If you want to learn shell, I suggest lurking in comp.unix.shell.

Re:Nature of the bug

[benedict]

Path parsing is not done by the filesystem.
It's above that layer.

quote

[Jotham]

the bug is apparently caused by the lack of quote marks in the install script.

Apparently it only strikes if you 1) havn't uninstalled iTunes first 2) have multiple partitions and 3) have spaces in the name of your partitions

This from MacSlash (posted by Graff as AC):

Well, there is a fixed installer up now. Looks like the following change was made to the "Preflight" file inside the "iTunes.pkg" package:

old version:

#!/bin/sh

# if iTunes application currently exists, delete it
if [ -e $2Applications/iTunes.app ] ; then
rm -rf $2Applications/iTunes.app 2> /dev/null
fi

exit 0

new version:

#!/bin/sh

# if iTunes application currently exists, delete it
if [ -e "$2Applications/iTunes.app" ] ; then
rm -rf "$2Applications/iTunes.app" 2> /dev/null
fi

exit 0

As you can see, they basically placed quotes around the file paths so that any characters such as spaces in path names would not mess up the rm command. So easy, and yet even the best of us forget to do it at times. That's one of the things about the command line - lots of power when used properly, but also many powerful ways to mess everything up.

- Graff

Re:quote

[Skapare]

It also shows one of the weaknesses of a programming language which is based on collections of things (in this case token words on a command) or data structures which can be changed merely by the value of some part of it (e.g. a variable with spaces).

Re:quote

[jerde]

Of course, with the right volume name including a quote, one could still get this same script to erase their drives.

Users are allowed to use quotes in file/volume names.

I just simply don't like variable substitution in a root-run rm -rf command. Gives me the willies.

- Peter

Re:quote

[Gid1]

Apparently it only strikes if you 1) havn't uninstalled iTunes first 2) have multiple partitions and 3) have spaces in the name of your partitions

When I went to download it (before the bugfix) , I did notice that the download page prominently said something to the effect of "If you have a previous version of iTunes installed, you must remove it first". Fortunately, for once, I followed those instructions.

In retrospect, this is a little odd if the installer had the ability to remove the old version. Hrm.

<paranoid>
Maybe this is one of those times when the programmer noticed the bug just before release and the manager cared more about the release deadline than the bug and said "Well, release it for now, and start working on the x.y.1 release." In my experience, not the first time that's happened!
</paranoid>

Hrm.. just noticed another thing. 'if [ -e $2Applications/iTunes.app ];': even if iTunes *had* been de-installed... if '$2' had whitespace in it, -e would still fail, giving an [: xxx/Applications: unexpected operator error (if 'xxx' was the part of the volume name after the whitespace...

Depending on what this script is run on, it might only have effect on people who choose to install it on partitions with whitespace -- so, if your chosen partition didn't have whitespace but your other partitions did, it might not affect it. *shrug*

Massive cock-up on Apple's part, all the same.

Re:quote

[ecampbel]

Quotes get evaluated first and only once, and then variable substitutions happen, so having quotes in your drive name will not cause problems.
For example (in csh):
% setenv foo \"
% echo $foo
"
% echo "some quoted text:$foo"
some quoted text:"

Everything within quotes is passed as one argument to the echo command, and the fact that $foo contains a quote character doesn't cause any problems. If the quotes were evaluated after variable substitutions, an unmatched quote error would have been reported in the last example.

Re: quote

[cjpez]

the bug is apparently caused by the lack of quote marks in the install script.

No, no, no. The bug is caused by having spaces in your directory names. I mean, come on. They're ridiculous. "This is My Hard Drive" is not a good name for your hard drive! :P

The big mistake here...

[jcr]

An app like iTunes shouldn't even HAVE an installer script for OS X. It should just be a disk image file, and installation shouldn't be any more than dragging iTunes.app from the disk image to your /Applications directory.

This problem is a hold-over from iTunes' Mac OS 9 legacy.

-jcr

Re:The big mistake here...

[aberkvam]

It sounds like you aren't familiar with what iTunes does. iTunes isn't just a pretty version of WinAmp. It actually is a CD ripper and CD burner that handles audio and data CDs. It supports the CD-RWs that come with Macs as well as many third party CD-RWs In order to properly support CD burning it has to install drivers "outside its bundle". It has to change things outside of iTunes.app.

I will admit that most applications should just use the "drag and drop" installation and Apple agrees. But there are some cases where the use of an installer can not be avoided and this is one of those cases.

Re:The big mistake here...

[benh57]

Sorry, wrong. The iTunes2 installer updates the Disc Recording framework which is part of the system. Thus it DOES need an installer - and root privs.

Re:The big mistake here...

[Apotsy]

This problem is a hold-over from iTunes' Mac OS 9 legacy.

Oh really? Is that why the X version of the iTunes installer erased hard drives but the 9 version didn't?

Re:The big mistake here...

[Lars T.]

Erm, yes, that's why the OS X version was affected, not the OS 9 one.

Re:The big mistake here...

[fwr]

I'm a little confused, because I'm not an Apple user at all, but I thought that the base OS was similar to Linux as to it's kernel module capabilities. If so, then there shouldn't be any reason to require the modification of "system" directories. You should be able to load a driver for various CD's from any directory. The module loader program may require root privelages, but there shouldn't be any reason that the modules themselves need to be in a particular directory. Or, is OS X not capable of doing this?

Re:The big mistake here...

[aberkvam]

Well, it's similar to Linux in the same way that FreeBSD is similar to Linux. Mac OS X is built on a FreeBSD base with a Mach 3.0 microkernel. Apple actually releases this base as an Open Source project called Darwin. Apple then takes this base and adds the GUI and Apple technologies to create Mac OS X.

The original suggestion was for iTunes to be distributed as a "package" (a directory that appears as a single application file to the user). I would imagine that there would be a way to put drivers in this package that load dynamically. But you mention that the drivers would have to load up with root (Administrator) access. Are you suggesting that a downloadable and user-modifiable application should be allowed root privilages to mess with the hardware?

Mac OS support files live in three main places. /System/Library/ is reserved for the OS itself. Apple is the only one who should be playing around in this location. /Library/ is where third-party stuff is loaded that should be available to all users. Also,each user has their own ~/Library/ folder where they can put things that should only be loaded for that user.

I saw some signs that iTunes also does some things in the system fonts directory. I'm not sure if it's adding new fonts or what, but it's doing something in there. So there may be other reasons for it to be modifying system areas.

Re:The big mistake here...

[uid8472]

It's possible to manually load a kernel extension bundle from anywhere with kextload(8) as root, but for it to be autoloaded when the relevant hardware is detected, or if another kext declares a dependency on it, it has to be in /System/Library/Extensions.

Re:The big mistake here...

[jcr]

It sounds like you aren't familiar with what iTunes does.

If it sounds like that to you, then your hearing is faulty. iTunes should be a self-contained .app, which any user can install in /Applications, ~/Applications, or anywhere else you care to. If it requires a framework or a .kext to talk to the CD/RW drives, then those should be in separate package which require root priveleges to install. In fact, the drivers should be obtained through the Software Update mechanism.

-jcr

Re:The big mistake here...

[aberkvam]

I saw some signs that iTunes may install fonts. It's doing something in the system fonts area anyway. What should be happening here? Can a .app package have self-contained fonts? Should it?

Also, once you are at the point where an application has a .app and a separate package that requires administrator access to install, why not just have the .app be part of the installer, as Apple has done? Why confuse users with a more complex install process?

It is strange that iTunes seems to be responsible for installing system-wide drivers for CD-RWs though...

Re:The big mistake here...

[Apotsy]

Funny you should mention Software Update as a solution to problems you perceive as "legacy" from Mac OS 9. Software Update started out as a 9 feature.

Oh, and I'd love to hear how making the package containing the burning framework a separate item would solve this problem. That would simply move the buggy pkg file to another place.

Whether it's a single install, or two separate install, you still need to install it.

what most people didn't know...

[Festering Leper]

is that they were unknowingly betatesting the new and ultimate file compression system

page out of Roxio's book

[Alcimedes]

you know, the funny part was i was cursing a blue streak when installing roxio cd creator 5 toasted my win2k machine. what are the freakin' odds, i would rant? why the frick is a cd software package set up to kill my machine?

well, i guess it's catching, whatever it is.

lol, i think i'll be waiting a few weeks after the release of software from now on. bleeding edge one to many times.

Re:page out of Roxio's book

[crawling_chaos]

Apparently the school that produced the Pool of Radiance: Ruins of Myth Drannor installer has a few more graduates.

Pool of Radiance also does this.

[juju2112]

The new Pool of Radiance does the same when you try to uninstall it. :] Thankfully there's a patch, though.

Users in OS X

[tunah]

I haven't personally tried OS X, so I don't know about how the multi user stuff fits together, but presumably you need to be running as root to install this (or trash your hd, for that matter). Shouldn't there be a way to install some apps without being root? Seems to me that having the user root by default is a bad strategy in terms of apps going evil, viruses, etc... this OS is not targeted at security-conscious people.

Re:Users in OS X

[headbulb]

your right you have not used os X there are two kind of users in os X admins and regular users the admins can use sudo while the users may not be able to sudo or with certain resrictions

the root acount is disabled as in the password is * meaning none which means you can't login with it

HeadBulb

Re:Users in OS X

[IronChef]

You can enable root login in the NetInfoManager. When you do, it asks you to put in a root password, and then you can log in or su to root if you want.

Re:Users in OS X

[TomRitchford]

I don't believe that the damage was necessarily done by the installer script acting as root. rm -Rf /partition will remove all files that you own and can reach from /partition. If you are the only user on your system, you might well own almost every file on every partition so the damage would be immense.

Also, how would the installer get sudoed? You don't run installers from the command line, you double click a file, right? (And wouldn't anyone be REALLY skeptical of an installer that had to run as root to install?)

ouch, but...

[godawful]

gratefully when coming home from the bar last night i downloaded installed and lost nothing *thank the beer gods* and yep it sucks for those who lost it.. however, from what i've read, tech tool pro is the best app right now for recovering those lost files.. i'm glad i didn't lose anything last night.. but i've seen some accounts of people recovering 50 to 75 % of their lost files.. but truly a bum deal.

Say what you will...

[SonicRED]

But if this didn't happen to you then you have to admit it's pretty fscking hilarious.

Figuratively this is like Steve Jobs deleting his market share, which is basically already a 0 byte file.

Sue them.

[t_allardyce]

Its a virus - it hides behind a legitimate program, performs some sort of check, then delivers a payload. If thats not a virus, then i don't know what is. Just because Apple may 'claim' its a mistake, is no excuse. People have gone to prison for less so i say, give them a big fat law suit.

Re:Sue them.

[damiam]

It's a bug mistakenly left in a legitimate program. It doesn't do a check before it delivers its payload, aka:

if harddrive.multiplepartiotions()
rm -rf /

The code, as an earlier poster stated, is like this, where some values of $1 and $2 could cause the shell to translate the command into rm -rf /:

# if iTunes application currently exists, delete it
if [ -e $2Applications/iTunes.app ] ; then
rm -rf $2Applications/iTunes.app 2> /dev/null fi

If an innocent mistake were a virus, Microsoft would have been sued out of existance already.

Re:Sue them.

[nvainio]

Its a virus - it hides behind a legitimate program, performs some sort of check, then delivers a payload. If thats not a virus, then i don't know what is.

It's not a virus. You mean Trojan horse. Viruses replicate.

Re:Sue them.

[gig]

> Its a virus - it hides behind a legitimate
> program, performs some sort of check, then
> delivers a payload. If thats not a virus,
> then i don't know what is. Just because
> Apple may 'claim' its a mistake, is no
> excuse. People have gone to prison for
> less so i say, give them a big fat law
> suit.

Are you listening to yourself?

Re:Sue them.

[Dyolf Knip]

Its a virus - it hides behind a legitimate program, performs some sort of check, then delivers a payload. If thats not a virus, then i don't know what is. Just because Apple may 'claim' its a mistake, is no excuse. People have gone to prison for less so i say, give them a big fat law suit.

Are you listening to yourself?

He's right to some extent. A bug that causes a program to crash is perfectly ordinary, if annoying. One that 'oops, we just wiped out your entire hard drive, so sorry' is totally unacceptable. Hell, most viruses aren't even designed to be that bad. Software can always be expected to be buggy, but this crossed the line between 'not quite enough debugging' and criminal negligence on Apple's part. Prison is a bit much, but financial liability is quite reasonable.

This might be very good.

[Krapangor]

I usually don't have very much space on my harddisk because it very small and i often can't decide which pictures to delete i downloaded from the internet.
So this installer comes in very handy because it deletes just all data and you don't have to decide whether to delete the picture of all these nice kitties or not.
So you have much more space on your harddisk and can download again much more nice pictures from the internet with cats.
My problem is however that i don't have an MAC and i hope they port it to linux soon so that i have again nice 30 megabytes of free harddisk space.
It is of course very sad that people with important data have lost all important data but you can't have much space and important data on your harddisk all the same time anyway.

Re:This might be very good.

[Tsar cr0bar]

Dammit, when are they gonna add a "Stupid" moderation category?!?

Re:This might be very good.

[Lars T.]

Your Linux distro doesn't have shell scripts? No "rm -rf"?

Hardly anyone actually affected.

[caryw]

I installed iTunes2 and it worked just fine. It seems that all of the people over at MacSlash have had the same results.

Re:Hardly anyone actually affected.

[motherhead]

yep, I have three partitions (9, 10.1, and data) and installed mine friday night. supriseing lack of catastrophy. Oh and even though i detest every proper noun that begins with a lowercase "i", iTunes2 really does kick quite a bit of ass.

reaction from Apple's internal culture?

[dan_bethe]

I'm curious to know the reaction from Apple's internal culture, particularly that of Steve Jobs.

I have read all the Apple books I'm aware of as of before Steve became iCEO, and they're largely a behavioral analysis of Steve's proactive and reactive mind. I haven't read Steve's latest book. Now he's more civilized but is known to spontaneously fire anyone who gets in the way of making extremely critical and improbable things happen, such as when they completely redesigned apple.com and the company store within a few months. In other words, he's still extremely demanding of precision high performance, which is a very good demand.

I wonder what happens behind closed doors after a major FUBAR like this happens. What is said? What are the looks on peoples' faces at the moment? What does Steve do, say, and look like? What chain of the culture panics over what he'll do or over their employment status after hearing the news?

It's a feature!

[wine]

iTunes2 synchronizes data between the iPod and your computer's hd.

So, if you don't have an iPod or your iPod holds no data, everything is erased from your hd in the process of synchronizing.

Seems very obvious this is a feature.

*sigh*

[PhReaKyDMoNKeY]

More ammunition for anti-Apple people. Yay.

Well hey, at least they were quick with the update. I didn't even know the bloody thing was out before I saw the news on the fixed problem. Microsoft _still_ hasn't recalled Windows, and it's been out for, what, nearly twenty years?

Re:iTunes 2.0 problem the exception, not the rule.

[PhReaKyDMoNKeY]

Most likely they just didn't think to test partitioned hard drives. I know it wouldn't have occurred to me. Partitioned hard drives aren't nearly as common among Mac users as they are with Windows/Linux/other users. Before OS X, partitioned hard drives on Apple boxes were all but nonexistent.

It was a very simple, very costly oversight. It's unfortunate, but you can bet your ass they'll test every bloody thing they make on partitioned/multiple drive computers before releasing them from now on.

Ahem

[benh57]

Three hours before this other guy: benh57 Here are your recent submissions to Slashdot, and their status within the system: * 2001-11-04 02:29:02 Apple's iTunes2 Update Wipes Drives (articles,apple) (rejected) Summary: * rejected (1)

Somebody is going to get fired...

[active8or]

We are talking Steve Jobs here, the man who pulled the Geforce 3 cards for a long time since NVidia spoke about the Cube, the man who throws computers after lazy workers...

May Steve have mercy on the poor programmer who did this...

Mvh:
- Knut S.

Re:Somebody is going to get fired...

[Hes Nikke]

the man who {rumors} pulled the Geforce 3 cards for a long time since NVidia spoke about the Cube

thats the Radeon pulled from the cube, and iMac, when ATi blabbed.

why do you think apple is in bed with nvidia now? ;)

the man who throws computers after lazy workers...{/rumors}

it was an $800 digital camera! see it in quicktime :)

Re:Somebody is going to get fired...

[active8or]

I stand corrected... (=

it's not that far-fetched.

[liq-bin]

Although, yes, it's a pretty serious oversight. Apple's installer package of choice seems to be a modified version of the OS installer. While it allows all of the power they need for installing an application, or an OS, or an upgrade, it lacks finesse. In truth, according to some developers who have been told to use that as their installer, it's an incomplete which was custom coded only borely enough to install the OS, and should never be used for apps.

I think this wiil bring to light the serious conceptual flaw of using an OS installer, let alone one renowned for its hackish feel on the underside, for installing applications. This is one good reason that OS departments should be separated from application departments, whether at Apple, M$, or wherever. Maybe Apple shousd be utilizing the drag install which they are trying to sell to the rest of the developer community?

Re:it's not that far-fetched.

[gig]

iTunes is part of Mac OS X. If Apple had an application/OS split, then the iTunes folks would be in the OS camp. It comes with Mac OS X, and parts of it are in the system itself (CD burning stuff) and need to be updated by an installer. That's what the installer is for. Most apps do come on disc or disc image, so the installer is very rarely used.

A lesson to learn?

[Gabhlan]

I installed iTunes 2 nearly two weeks ago (thanks Hotline) and had no problems with it eating my drives. But, had Apple made iTunes 2 generally available previously as a beta with a disclaimer to suit (If it eats your drive, your problem, no ours) this bug would have been picked up long before it reached the final version. If Apple aren't going to spot all the destructive bugs, which generally the have up to now, they need to increase the number of beta testers to everyone on their Apple Developer Connection service. I don't know exactly how many people they dispatch pre-release software to, but I'm on the ADC and never heard a thing.

it is less readable.

[delmoi]

If you're a Moron.

Yet another way in which the Mac is superior to PC

[Colin Bayer]

Look at this innovation!

What Apple did in a few lines of shell script, it takes Windows a complicated installer to do. I personally plan to switch to the Mac platform for its superior file deletion capabilities.

Re:iTunes 2.0 problem the exception, not the rule.

[Edgewize]

Even if they tested partitioned drives, it wouldn't have showed up. The problem only occurs when the target partition has a space in its name, and there is only data loss when the space is the first character. Even 99% of the partitioned Macs out there wouldn't have lost data.

OS 9 Forever!

[darkov]

More evidence that Apple has given up on providing the best user experience. I'm loathed to move to OS X becuase it's not really a Mac OS. OSes 1-9 were a product of careful design, internally and externally. The little interface things (mouse clicks behave completely differently in so many circumstances), the robustness of installing software and fiddling with the system.

Years of experience went into tweaking the OS to be just right. Now it's all gone. They thing making it pretty is enough. Now we've been reduced to being victims of shell scripts missing quotes and zapping our hard disks. Sob. It no better than a Window's box now. Sob.

Re:OS 9 Forever!

[motherhead]

you're just a touch insane. I too have been buying macintoshes since system 6 (by the way, i would love to hear your systems 2 through 5 user experiences...) .

yes, while system 9 is just so terrifically mature that it stuns you with it's depth. i also remember how much I thought that system 7 was an anathma to the Macintosh look a and feel and "useability".

I detested OSX and need my mac to make money with, so have I little time for flakey eye candy. But ever since I Installed OS 10.1 on my drive i have yet to boot into OS 9. Yes it is far less mature and has some goofy little idiosynchosies but it also runs all my DTP/WebDev OS 9 apps perfectly, but then i can also exploit 10's native apps to get grunt work (word proc, mp3 playing, browseing and proofing etc...) all multi-threaded and so on... every day i am amused to find a hidden little nuance to aqua, and i also get to run xdarwin interleaved with aqua so i can run BSD, classic and OSX apps flawlessly, simultainiously.

Also this: Quartz.... mmmm... beatiful display postcript (PDF, i know) love...

So i sympathize with you. and perhaps some lunitic will come up with a nice OS 9.2.1 theme and theme manager to run it over aqua for you. but I think apple is kicking big OS ass with 10.

OS 9 forever? my ass, how many times did you reboot today?

Re:OS 9 Forever!

[trash eighty]

well i run OS9 and my last reboot was 2 weeks ago, and that was to install a new extension

iTunes installs great on OS9 too, good job i didn't try it on my OSX yet ;)

Re:OS 9 Forever!

[darkov]

Granted, System 10 is more stable becuase of memory protection and has better task scheduling, but my 9.2.1 is heaps stable (as long as I don't run MS apps, generally). But I use Macs becuase they're a pleasure to use, and all the little things count. Quartz is loverly and I wish that GX lived on, but they don't improve my work. Pissing around trying to work out how the hell to get 10 to connect to my AFP server and other simple things which seem impossible wastes my time. 10 may have it's own nuances, but they're nothing like 9. What exactly is the big win changing all the subtlities of the UI in an arbitary fashion? Believe it or not, stability and memory protection are incremental improvements for your average user, delivered at great expense of it's usability.

On the subject of the earlier systems - they were fantastic. Why wouldn't they be when the only alternatives were DOS and a very broken early Windows.

Re:OS 9 Forever!

[StorminNorman]

I'm going to give you the benefit of the doubt and assume you're not trolling.

Mac OS X is based on UNIX, a design which, if you'd noticed a previous story today on slashdot, is now 30 years old, much older than Classic Mac OS.

Mac OS X has a carefully redesigned UI that actually works really well. Believe me, I haven't used Classic Mac OS in about 4 weeks now, and even then it was for DVD playback due to an oversight in Mac OS X. The new UI is just as carefully designed, and just as carefully planned as the old one. It has one disadvantage however:

It's new.

That means that there are going to be kinks for a while yet (ever used Mac OS 1.0? I have, and believe me, it has more UI kinks that Aqua does now).

That said, I can't go back to Classic Mac OS for two reasons: Protected Memory and real Multitasking. Between the time that my Centris died, and the acquisition of my iBook Dual USB, my primary computer was an Intel system running Debian Linux. I can happily say that I was able to put up with KDE's quirks (it's nice, but it's got a long way to go before it gets close to Mac OS for usability), simply because the system NEVER BLOODY CRASHED!. Seriously. The only reboots I perform on it are for power outages and kernel upgrades.

Mac OS X is just the same. While I got a few kernel panics on the older 10.0.x series (3... all of them while in public for some reason :( ) I am yet to see one in Mac OS 10.1.

The point is, I would happily live at a commandline if it meant the OS wasn't going to freeze up at least once a day, or if I could happily run as many applications as I wanted without worrying about memory (No more memory size settings! YESSSSS!). And that's not even mentioning the fact that the system actually MULTITASKS. Whoa! Who'd have thought that a Mac would ever be able to multitask and provide memory protection? They've only had an MMU for what? 10 years now?

I love Macs, always have, always will, but the old OS was kludge upon kludge upon kludge. It's old, It's broken, it's been replaced. Stop using it, please. And Aqua takes all of 5 minutes to adjust to from Platinum. It's harder to jump to windows from Platinum than to Aqua.

I mean seriously... would you go on using Windows on an Intel system if you knew that a newer and better operating system was available that didn't have all of Windows bugs?

Re:OS 9 Forever!

[darkov]

You know, you asked no nicely, I thought for a second that I should give it up. But I really do find 10 too hard too use. Maybe I'm too set in my old decrepid technology ways, but I'm just so comfortable with 9 I really don't see any point. The bugs don't bite me, and newer isn't always better.

On the subject of Trolling: I am not now, nor have I ever been a Troll. BUT I have had posts marked as Troll. It's something a bit offensive about /., that an opinion that happens to disagree with what might be commonly agreed is quickly marked as abuse. Have the politically correct been released from the asylums?

Moderators: CONTRARY OPINIONS ARE VALID AS WELL. PLEASE DON'T LEAP TO JUDGEMENT.

Chances are, of course, that this will be marked as a troll post. Sigh.

Re:OS 9 Forever!

[darkov]

What you say is interesting, becuase that's exactly what I thought when I looked at X. It all makes far too much sense not to be true. There is no good reason for thowing out the classic OS. Arrogance is the only answer. I guess one of the negative things about Jobs (along with his good points) is that he lets these sorts of things happen, possibly even encourages them.

I really believe that 10 is the end of the Mac. Not the end of Apple, but the end of the Mac as embodied in my Mac Plus: good ideas implemented with flair and design sensibilities. Now it's all macho my scheduler is better than yours bullshit.

Re:OS 9 Forever!

[stripes]

perhaps some lunitic will come up with a nice OS 9.2.1 theme and theme manager to run it over aqua for you

If you look over on MacNN someone has one fairly far along. I think the only big chunk o' work they have left is the scroll bars. I think it was a Friday headline.

Re:OS 9 Forever!

[StorminNorman]

Just a quick reply to the troll comment... It was rather late at night, and I needed an excuse to rant, so biting at what I thought was a troll was my best option... I noticed after I'd posted that there had been a few other posts in the thread as well and realised that I was wrong (dammit why isn't there an edit comment feature in slashdot yet?) (it took me nearly half an hour to write that, between watching Gundam Wing, IRC and other web boards.)

Sorry if I offended you, but I did think, initially, that you were trolling. (Of course, I thought it was a pretty well written troll too :) )

As to Mac OS X being hard to use... to each his own i guess :)

Stupid, stupid, stupid

[jeti]

Apple is bound to have at least some hundred NDAd beta testers. They overlooked TWO serious bugs in the installer of a final version. I think they must have tested the product, but entirely failed to have the final version of the installer tested.

I think this has to be called gross negligence. I really hope they can own up with a tool to restore rm -rf'ed files. The usual data recovery tools do not seem to work.

Re:Stupid, stupid, stupid

[Lars T.]

How many of those hundreds of NDAd beta testers had a drive with a name that either started with a Space or ended with one (but not in the middle of the name). None. Yup, the error was stupid (like most shell script bugs), but not really easy to find.

Re:Stupid, stupid, stupid

[jeti]

As I read it, it's vice versa. Names with a space in the middle are applied as 2 arguments. Spaces at the beginning and end of the name should be ignored.

Reading the thread at macnn, I get the impression that quite a high percentage of users had such a setup.

Re:Stupid, stupid, stupid

[TheGreek]

Your problem is that you're reading MacNN forums. Those who post there are just slightly less intelligent than those who inhabit ZDNet TalkBack, and they're less likely to understand shell scripting than RMS is to shave his natty beard.

Re:Stupid, stupid, stupid

[LoadStar]

Reading the thread at macnn, I get the impression that quite a high percentage of users had such a setup.

Quite true. The default hard drive name under Mac OS is "Macintosh HD" - including the space. My hard drive is partitioned into "Mac OS X Boot" and "Mac OS 9 Boot" - yes, 9 spaces in there. I don't think that's particularly uncommon either.

Re:Stupid, stupid, stupid

[Lars T.]

Actually, it's a little more complicated. The offending line in the script looks (something) like this:
rm -rf $2Applications/iTunes.app (...)
with $2 holding the name of the disk to install iTunes on, in the form of /Volumes/name-of-volume. The only place where the spaces (and probably other nasty chars like '*') can appear in is name-of-volume, the name given to the drive by the user in the Finder, the way it appears on the desktop.

If the name begins with a space, the rm will delete everything in (and below) /Volumes/, IOW all drives mounted there. If there is one (or more) space(s) in the name (e.g. volume "disk 1" or "some name for a drive"), it will try to delete /Volumes/disk or /Volumes/some - so if you have a volume "disk"/"some", it's toast. It will alo try to delete "1/Applications/iTunes.app" / "name", "for", "a" and "drive/Applications/iTunes.app" in the working directory. I was actually wrong about it always deleting something if there is a space at the end.

Now what happens if there is an asterisk somewher in the name is probably clear - somebody want to try? >:-)

Re:Installer too clever for its own good

[Lars T.]

"rm -rf" is a low-level harddisk call?

bogus shell quoting rules

[mj6798]

This is really a problem in how the Bourne shell handles variables. Lots of UNIX scripts fail when file names contain spaces, which is why most people don't use spaces in file names.

The folks at Bell Labs seem to have realized that this was a mistake, which is why the "rc" shell (also available for Linux) now handles things differently: variable substitution does not result in re-tokenizing.

Re:bogus shell quoting rules

[Fruit]

zsh can optionally disable that feature as well (and reenable it on request for specific operations). I already configured it to do that ;)

Re:bogus shell quoting rules

[uid8472]

zsh also refrains from retokenizing the result of variable substitution. And although zsh is (currently) used on Darwin to emulate sh, in that compatibility mode the SH_WORD_SPLIT option is enabled, restoring the historically correct behavior.

Re:bogus shell quoting rules

[dangermouse]

This is really a problem in how the Bourne shell handles variables.

Okay, there have been multiple posts to this effect, and they're all wrong.

Sorry, but this is not a "problem". The shell operates on lists of words. A variable is assumed to contain a list under most circumstances. List elements (words) are separated by an Internal Field Separater character, defined by $IFS.

Every single one of my Bourne-compatible shells' man pages explicitly state the types of expansion that are employed generally (including word splitting) and provide a short list of exceptions.

This is expected, normal, useful behavior. And it's easily handled, as you can simply wrap any variable in quotes if you don't want it to be tokenized as multiple parameters. If it's something you can't cope with, that's a shame, but don't act as though it's some flaw in the shell... it's there on purpose, and with good reason.

Re:bogus shell quoting rules

[fwr]

There are situations where retokenizing is a desired behavior. I'd rather classify it as inexperience by the programmers who wrote the script with the scripting language. Anyone who has experience with shells knows that this is the way they behave and program accordingly. Blaming the programming language instead of the programmer is like blaming the end-user instead of Apple...

Re:bogus shell quoting rules

[dvdeug]

Very few cases exists where retokenizing is useful, compared to the cases where you have to quote and doublequote and use -print0 and whatever else. It could have been provided by a function instead of as a default behavior.

I don't see why we should blame the programming language. The fact that

#!/bin/sh
rm $1

doesn't work is nonintuitive and painful. If a programming language has a 'feature' that results in a lot of bugs then that feature should have been designed a different way, especially in a language like shell where bugs like these can be so dangerous.

Blaming the programming language is like blaming Apple; the programmer should have double checked his code, but trusted the language not to annihilate everything on a simple typo; the end-user should have double checked the code, but trusted Apple not to annihilate everything on a simple uninstall.

Re:bogus shell quoting rules

[mj6798]

Ah, I see, you live by the "tough man" approach to programming. The fact is, people keep shooting themselves in the foot with this. It's unexpected because it's different from almost any other scripting language. And it's easy to get the same functionality with a different notation. That's why it's a design flaw and that's why the Bell Labs successor to the Bourne shell fixed it.

Re:bogus shell quoting rules

[mj6798]

Oh, I agree: Apple is fully responsible for this, and programmers should get this right if they use the Bourne shell.

But if a system is designed in such a way that people frequently make a certain mistake, and if it is easy to change the system, without limiting its functionality, to keep people from making that mistake, then the system has a design flaw. This applies to the Bourne shell, and that's presumably why its successor at Bell Labs doesn't behave this way anymore.

As for Apple, they should have been using decent package management. If they do need to script, they should have been using a scripting language that gets this right (Perl, Python, etc.). As a last resort, they should at least know their tools.

Re:bogus shell quoting rules

[dangermouse]

Ah, I see, you live by the "tough man" approach to programming. The fact is, people keep shooting themselves in the foot with this. It's unexpected because it's different from almost any other scripting language. And it's easy to get the same functionality with a different notation. That's why it's a design flaw and that's why the Bell Labs successor to the Bourne shell fixed it.

No, I live by the "hammer for a nail, screwdriver for a screw" approach to programming. It's not unexpected behavior if you don't look at an apple and say "hey, that's just like that orange I've been eating."

Bourne is not, nor is any other derivative shell, very much like any other scripting language that was not based largely upon it. You have only to look at the structure of the [ operator, the way control structures are broken up, and the way variables are expanded to recognize that. It's designed primarily for use on a commandline... the fact that you can write your scripts to a file and run them through the shell non-interactively is bonus.

That said, Bourne is usually the right tool for software installation scripts on a Unix machine. You just have to recognize that you're not wielding a screwdriver.

Also, I do realize it's a common mistake. But so is dropping a semicolon or forgetting to close parens in your C-like language of choice. That doesn't make the need for a semicolon or a close-paren a design flaw.

Re:bogus shell quoting rules

[dangermouse]

The fact that


#!/bin/sh
rm $1


doesn't work is nonintuitive and painful.

Erf. It does work, and it is intuitive, if you stop thinking of Bourne as a scripting language that can be used interactively and start thinking of it as a shell that can be scripted.

If I call that script like so:

script something else again

I would expect it to remove 'something', which it does, and leave 'else' and 'again' alone, which it does. If I want it to remove all three, I have to do:

script "something else again"

It will then tokenize the string "something else again" and pass it as a set of three parameters to 'rm'. This is a good thing. It allows me to deal with parameter grouping and pass-through efficiently and intuitively. And since this is a shell, that's very helpful. It is truly scripting and not programming. You're feeding the shell a list of commands to run. It just happens to have a few handy programming features like flow control and conditionals.

If you're in a position where you don't just essentially need a bunch of external commands to run in a given sequence, then for God's sake use a programming language like Perl, Python, Tcl, or even Ksh (which is as much a programming language whose interpreter serves as an interactive shell as Bourne is vice versa).

Re:bogus shell quoting rules

[jesser]

It's very hard to cause data loss with a missing semicolon. In all likelihood, you'll get a syntax error when you try to compile.

Re:bogus shell quoting rules

[jesser]

If variables are supposed to be lists, then why isn't $1foo (without quotes) a syntax error?

Re:bogus shell quoting rules

[dangermouse]

Because bash recognizes $1, substitutes it, and concatenates it with 'foo'. How hard is this?

Re:bogus shell quoting rules

[jesser]

Because bash recognizes $1, substitutes it, and concatenates it with 'foo'.

That's not a justification any more than "Windows crashes because it dereferences dangling pointers" is a justification for Windows crashing.

If variables really are lists (like you said), then $1foo means "concetenate foo to the last term in the list". How often do you want to do that? I'd bet that $1foo is more often typed by someone who didn't realize $1 could contain a space than by someone who actually wants to treat $1 as a list and concatenate foo to the last term. We just saw an example where $1foo not being a syntax error led to severe data loss. Can you give a reason why bash shouldn't issue an error or warning when it sees $1foo outside of quotes?

Re:bogus shell quoting rules

[dangermouse]

If variables really are lists (like you said), then $1foo means "concetenate foo to the last term in the list". How often do you want to do that? I'd bet that $1foo is more often typed by someone who didn't realize $1 could contain a space than by someone who actually wants to treat $1 as a list and concatenate foo to the last term. We just saw an example where $1foo not being a syntax error led to severe data loss. Can you give a reason why bash shouldn't issue an error or warning when it sees $1foo outside of quotes?

No, $1foo means "expand $1 and append foo". You, the intrepid Bourne scripter, should realize that $1 may very well contain IFS characters, recognize what that will mean if the string is retokenized, and quote appropriately.

Look at this:

VARIABLE="$VARIABLE $ELEMENT"

This is a common way of building lists in Bourne. The reason this is common is that it is understood that $VARIABLE will be tokenized when it is expanded, and word splitting on IFS characters will occur. It is also commonly understood that if you do not want a string to be word-split, you quote it so that its IFS characters are ignored. Like I said before, not hard.

I'm sorry, but it takes all of ten seconds to grasp the fact that variables will be expanded and their words split. Occasionally you take it for granted that a list will only have one element, and get away with stuff like $1foo. However, it's not something you do unless you have total control over the contents of $1 (as in, say, a function you've defined that only you will call), and even then it's bad practice.

That's not a justification any more than "Windows crashes because it dereferences dangling pointers" is a justification for Windows crashing.

I couldn't have said it better. The problem there is that someone unfamiliar with basics of pointer manipulation went and dereferenced a null one... the problem is not that the language let him have a null pointer to begin with.

Re:bogus shell quoting rules

[dragonfrog]

Hm! That's really interesting. /bin/csh is the same file as /bin/tcsh too (as you presumably know). Many wrinkles...

Re:bogus shell quoting rules

[spitzak]

I think the suggestion is that '$1foo' should produce an error if $1 contains a space. But '$1 foo' (notice the space after $1) does not produce an error. Also no error is produced if any of this is inside quotes.

I would agree that it seems unlikely that somebody will want to append a word to the last item in a list, and that any occurence of this is an error. If somebody did want the real effect of $1foo they would have to put it in quotes and into a temporary variable, and then insert that temporary variable.

Gotta be said

[Lord_Pall]

So i guess the Ipod/Itunes combo really IS a killer app.

Relative severity vs. a MS flaw

[jht]

Apple posted the initial update either late Friday or early Saturday (I'm not sure exactly when). It was pulled by late in the morning Saturday, they posted a warning shortly afterwards, and when I got up this morning there was a fixed installer online to use.

The Classic version (which most Mac owners are still running) was fine, and the bug seems to have only hit people who didn't follow Apple's instructions that said "remove the old one first" and/or had multi-partitioned drives (multiple partitions aren't nearly as common among Mac users as they are among Windows and Linux users).

So Apple made a gross mistake on one hand, but on the other hand they owned up to it quickly, pulled the offending installer, and fixed/reposted it less than 24 hours later. Most Linux vendors respond about as well, Microsoft usually doesn't (though they were very good about pulling, fixing, and notification with their recent RDP fix that knocked people's Terminal Server systems off the network entirely).

The other mitigating factor was that there aren't that many Mac users relative to the installed base who were affected by the bug - but unfortunately the people who were likeliest to be affected (users who are already running 10.1 as their base OS, have multiple partitions, and don't read the instructions thorougly because - after all - "it's a Mac, who needs instructions?") are exactly the kind of Mac "power users" who swarm Apple's servers constantly looking for new stuff and install it the second it's posted.

I run 10.1 on my TiBook 667, and I downloaded the update. But I deleted the old iTunes version beforehand and only have a single 30GB partition, hence the install went fine..

Re:Relative severity vs. a MS flaw

[Anonymous Coward]

So Apple releases a script that erases your hard drive if certain conditions exist, but somehow this is Microsoft's fault. I love /..

Re:Relative severity vs. a MS flaw

[nycdewd]

Tell it, bro. You have got it down, especially: "the people who were likeliest to be affected (users who are already running 10.1 as their base OS, have multiple partitions, and don't read the instructions thorougly because - after all - "it's a Mac, who needs instructions?") are exactly the kind of Mac "power users" who swarm Apple's servers constantly looking for new stuff and install it the second it's posted." Damn straight! And, fortunately for me, the only difference between said "power users" and myself is that I CAN READ and I DO READ. So, as instructed by the iTunes2 ReadMe , I deleted the previous version of iTunes2 (beta, got it from a Hotline server in Japan) from my OS X partition of my HD which has THREE partitions, BEFORE I installed the official release of iTunes2. Go figure, my partitions are intact. No loss of data. iTunes2 is freakin' great.

Re:Relative severity vs. a MS flaw

[sheldon]

Yeah!

If I recall correctly when SP6 was noticed to affect Lotus Notes it took Microsoft 26 hours to respond!

Those bastards!!!!!!!!

Re:Relative severity vs. a MS flaw

[Anonymous Coward]

Good lord, man. It *ERASES YOUR FUCKING HARD DRIVE*. This isn't vulnerability that'll that the script kiddies a few months to take advantage of, or something that'll cause a performance degradation due memory leak, or something that'll cause the occasional crash.

It *ERASES YOUR FUCKING HARD DRIVE*. This is the *JOKE* bug. This is the kind of thing that when one wants to make a humourous example of Microsoft, one goes "heh, heh, it'll probably erase your hard drive, heh." I mean, the viruses that take advantage of MS vulnerabilities don't even erase your fucking hard drives.

This is a *HUGE* screw up on Apple's part. Trying to compare it to a Microsoft bug-of-the-week is pretty weak - never mind trying to compare it *favorably*.

(As an aside, when Microsoft does take a while to respond (which is rare) it's an obscure bug in a system where it's complicated to fix - not a freakin' forgetting-the-quotes error in a Bourne script. Apple fucked up big time. Deal with it.)

Re:Relative severity vs. a MS flaw

[Knightmare]

Wow... I am amazed at how quick the users of slashdot these days are to try and get a dig at Microsoft. Well mabey Apple did release a product that erases everyting you have on your computer, but remember that time when Microsoft put out a product that had a security hole and it took them at least 2 days to fix it. I am sorry but I feel as if you are using the Chewbacca defense here.
I am not trying to say I agree with everything Microsoft does and that they are perfect, because they are not. But I can't remember the last time I installed a Microsoft product that removed all the contents of my hard drive. Mabey we should be discussing how this could be avoided in the future. Like a sandbox for installers so they have to ask for permission to do things and if they seem loopy to the sandbox it asks you for permission for them.

Re:Relative severity vs. a MS flaw

[jht]

To all the folks who thought I was praising Apple (or blaming Microsoft) - I wasn't! Apple was a bunch of dummies for letting that code escape the dungeons in Cupertino! Really!

Now that I've vented a tad, I'll re-iterate. It was a stupid bug that could have easily been avoided. But OTOH, Apple does not really have a track record of releasing software this broken (or at least, installers this broken), so their response was very good - at least as good as what we've come to see from other vendors if not better than most. That does not, mind you, excuse the severity of the bug - rather, it praises the response.

And if you had followed the CLEAR INSTRUCTIONS to delete the old iTunes first, you never would have encountered the bug in the first place, no matter how many partitions you had. Many (perhaps most) Apple users, unfortunately, do not bother reading instructions.

Every company makes stupid mistakes. The tough part is to handle it well when bad things happen, and Apple did OK (not perfect, because people lost data, but they took appropriate steps quickly) in that regard.

Re:Relative severity vs. a MS flaw

[Erasmus Darwin]

"And if you had followed the CLEAR INSTRUCTIONS to delete the old iTunes first, you never would have encountered the bug in the first place,"

Still, there are certain reasonable expectations and assumptions that people make when it comes to reading the directions. For example, while there are generally instructions on, say, a box of toothpicks, it would be unreasonable for a manufacturer to produce toothpicks that explode if you use them improperly. As long as I don't try and stab myself with it, I expect the tootpick user-error problems to be limited to failing to remove food and the occasional splinter.

With a car or a handgun, on the other hand, it's not unreasonable for a user to read the manual cover to cover (possibly multiple times), reference it again as necessary, even obtain additional instruction and certification for using the device. These are things that're inherently dangerous if misused.

With application-level software, reasonable expectations of the problems of a failed install include:

• The new version doesn't work
• Both versions are installed but unusable
• The user-created data associated with the old version gets wiped out by the installer (generally the worst case scenario and borderline unreasonable)

Personally, I generally backup the data from the previous install to another directory (if it's something I can't live without), then go ahead with the install. At worst, I expect that I'll have to uninstall both versions, maybe manually clean up a few files, and then start over. A few extra minutes of work in exchange for a few minutes of laziness. In the long run, I've found that I win.

Now, if the software were something other than just a standard application, I'd take more care. You can bet that I'd be a lot more paranoid when using software that's supposed to fiddle with partitions, as that's something you'd reasonably expect to be able to destroy your harddrive. But reading for the directions for absolutely everything would just be insane. I don't have that much free time.

Wrong wrong wrong

[The Infamous Grimace]

That code will nuke any partition. The iTunes 2 for OS X installer trashed only the extra partition(s). Hence -

int i = 1;
if(numOfPartitions > 1) {
for( i ; i = numOfPartitions; ++i)
TrashPartition(i) ;
}

I think thats right; I'm new to this whole coding game (sure is fun, though). Remember, the point is to make the sap think that it installed correctly, at least intil he/she tries to access those other partitions.

(tig)

Boycott Apple iTunes

[generic-man]

Dear Readers of the Slashing Dot,

I have been as tolerant as possible of the actions of Apple Computer Incorporated (hereinafter "Apple"). In particular, I am disappointed with their decision not to make their Internet To User Network Experience System (hereinafter "ITUNES") available under the terms of the GNU General Public License (hereinafer "GPL"). As a result, I have decided to use alternative music soft wares, such as Windows Amp (hereinafter "Win-Amp"), the X Multimedia System (hereinafter "X. M. M. S."), and the command-line utility Motion Picture Entertainment Group 123 (hereinafter "mpg123").

I would strongly encourage all readers of this web site, the Slashing Dot, to boycott ITUNES. Please refrain from downloading it, and encourage others from doing the same. I would also recommend that you boycott all of Apple's hard-ware and soft wares.

If we keep a strong will, we will succeed. Thank you.

Re:Boycott Apple iTunes

[alfredo]

But I use a Mac to run Linux? Where can I find non proprietary hardware? If I get a PC I have to pay the MS tax unless I get a naked PC, then I am seen as a pirate. The embedded code in the chips is not open source.

Are pencil and paper GPL?

Re:Boycott Apple iTunes

[dnorman]

> If we keep a strong will, we will succeed. Thank you.

Um, Sorry if I missed something here... succeed at WHAT?

You will definitely succeed in not using iTunes. That's a pretty lofty goal, and I'm sure you'll get a panicked phone call from Sir Jobs himself begging you to reconsider.

People (including myself) use iTunes because it works (VERY well), is easy to use, and scales up very nicely. Do the command line tools do that?

Re:Boycott Apple iTunes

[jrockway]

> Do the command line tools do that?

Well, sure, and probably better (I won't mention anything about shell scripts because that's The Bad Thing Of The Day right now)... but iTunes looks cooler with that spinning nuke thingy...

bug

[Nullsmack]

(stupid) Show stopping bugs....

...Not just for Windows(TM)(R)(C)(NC-17) anymore!

Why is this even being posted?

[Paladeen]

It seems to me that Slashdot really enjoys posing stories about how Apple mess things up while simultaneously ignoring all the things they're doing right.

All the Apple-related news items these days seem to be

"RMS disapproves of Apple Public License"

"Apple sues (someone) over themes"

"Apple does something evil....(bla,bla,bla).

I believe Apple is doing its best to be a good member of the Open Source community while retaining corporate profits. It's certainly a completely different company from what it used to be back in the 90's.

So, hey, give them a break. However evil you regard Apple, it is certainly dwarfed by Microsoft.

Re:Why is this even being posted?

[demon]

Considering that from what I understand, (a) Darwin CVS isn't even usable without a source patch, and (b) some of the more interesting drivers (for ATI, nVidia, Adaptec, etc. hardware) are closed source, I would hesitate to call it true "open source", and unless they finally fixed the APSL, it doesn't meet any official definition of the term either.

And there was a /. story about the iPod - and it had nothing bad to say about Apple, in fact some people were fairly interested in it, even tho it's not going to be a cheap little toy (from what I gather).

Re:Why is this even being posted?

[WildBeast]

yet if the same thing happened with MS software, you'd be the first to complain.

Re:Why is this even being posted?

[unconfused1]

You all are perfectly fine posting whatever you want. That is one of the missions of ./ to let everyone have whatever say they want to have.

I personally am a believer in Apple's OS and applications, but I want to hear when Apple is being evil, just as much as when Microsoft is evil.

As to the posting about people's lost files because of the installer messing up, I think that the lead poster, with the 100GB of missing data is exaggerating. Also why would someone have 100GB of non-trivial data that is not backed up? Seems fishy.

Why I hate the software industry

[Carnage4Life]

You have an obligation to take reasonable precautions to protect the data on your computer. That means making backups of any valuable data. Are you going to sue Western Digital if your hard drive fails?

People regularly sue if hardware is made faultily. Toshiba paid billions to settle a lawsuit with floppy disks that never showed up in the field and couldn't be reproduced. I personally have lost track of the number of class action lawsuits I've seen for faulty computer products.

What if it gets fried by a lightning strike?

Being struck by lightening is an act of nature which is completely different from human negligence. Please get your analogies right.

Even if Apple was found to be grossly negligent, they shouldn't be held responsible for data that was lost due to the negligence of the computer's owner.

Why shouldn't they be held responsible? If attaching your DVD player to your TV blows it up or your fax machine shreds your documents, are you also liable in such situations? Quite frankly I am disgusted with the attitudes of most people in the software industry that assumes that shoddy work is inevitable (all software has bugs? WTF?) and then blames customers when their shittily written software fails to behave as it should.

Programming is less difficult than building a bridge or an airplane and yet software companies have hoodwinked the public into making it seem that badly made software is a fact of life. One day people are going to realize that the software industry has been shamming them all this time and the lawsuits will start to pour in. This is probably when software companies will finally go back to using techniques developed decades ago to improve and measure software quality but by then the damage will be done.

Re:Why I hate the software industry

[Cygnusx12]

.. Quite frankly I am disgusted with the attitudes of most people in the software industry that assumes that shoddy work is inevitable (all software has bugs? WTF?) ...

It's true.

A programmer should never ever assume he/she/it has found all the bugs in their software.IMHO, what they should assume, is even if they have found them all, they'll never know it.

Bugs are a part of software. They always will be. You can't make any assumptions about what some other dumbfuck has done in/with memory.. Whereas in building bridges a similiar analogy would be.. "hey.. these bolts just don't fit.. "

I'm not at all saying some pieces of software aern't poorly tested and implemented.(I swear, I come across one more buffer overflow..) A Bridge is a fixed object that can be watched and maintained, airplanes are actively serviced on an ongoing basis. Whereas a single piece of software can be distributed acrosss a myriad of machines and configurations. (And software IS often patched too)

Although I'm in agreement with you when it comes to liability in some of these cases. I don't see how you can compare building, say a 50 million dollar bridge, or a 10 million dollar airplane to something that's say as common place as a 49.99 piece of software on a cd or dvd.

Re:Why I hate the software industry

[Sentry21]

First of all, if people had read the readme, they wouldn't have had this problem. Secondly...

Programming is less difficult than building a bridge or an airplane and yet software companies have hoodwinked the public into making it seem that badly made software is a fact of life.

Engineers are a very carefully controlled group - you need years of university education to call yourself an 'Engineer', and even then, you have to maintain your skills or you're not an 'engineer' anymore. If you're building a bridge, you have trained engineers that have paid money for this sort of thing. They know what they're doing.

Anyone can be a 'programmer', as demonstrated by more than one piece of software on the market today, and more than one company. They are often inexperienced and overworked, especially in cases where you have to get a product out (anyone read about how the iPod interfaces with iTunes 2? anyone notice how iTunes 2 hadn't been released yet?).

Frankly, I'm not surprised companies disclaim this stuff, because they don't have any guarantee that the person knows what they're doing. If they screw up, the worst that can happen is loss of data on the customer's side, and loss of job on the programmer's side. If an engineer screws up building a bridge or building, she can endanger thousands of lives, thousands of jobs, billions in damages, and if it was proven that they were at fault, they'd probably lose their 'engineer' status for good - which means all their experience is useless, and they have to get a new career field.

Anyone ever notice how bridges and buildings are more often late and over-budget? Compare this to software, which is often rushed out. And you wonder why the quality is different? Companies need to understand that waiting a few more days/weeks isn't as bad as wiping someone's hard drive and then saying 'oops, your fault'.

That said, they did tell you to uninstall first.

--Dan

Re:Why I hate the software industry

[Quasar1999]

Programming is less difficult than building a bridge or an airplane and yet software companies have hoodwinked the public into making it seem that badly made software is a fact of life.

Hold on a second... Building a bridge is pretty simple compared to programming. First, with the bridge, you can design it for 'x' number of cars/people per day, 'x' amount of wind, 'x' amount of earth movement per year, and design and build it based on principles layed out over hundreds of years.

With programming on the other hand, you can design all you want... The end user's hardware configuration alone can be so unique that there is no way you could plan for it. With a bridge, all the parameters aren't changing, hence you can work out anything that can go wrong, and prepare/design for it. Not so with programming, especially when it's going to be an off the shelf software package that is going to be installed in all sorts of configurations.

And I haven't even begun to describe the billions of ways the end user can be a complete idiot and delete half your config files and then run the software, etc... Going back to the bridge, if a person drives a car bomb onto it, and blows it up, the engineers aren't blamed for the bridge's collapse. And yet everyone is blaming the programmer when the end user does the equivelent to the software...

SHEESH!

Re:Why I hate the software industry

[Some Dumbass...]

Quite frankly I am disgusted with the attitudes of most people in the software industry that assumes that shoddy work is inevitable (all software has bugs? WTF?)

Not a programmer, obviously.

You need to understand that not all bugs are actual errors or "shoddy work". Sometimes it's the interaction between multiple pieces of code, each of which works perfectly well, which causes the problem. Sometimes it's people using the code in unexpected ways which causes the problem (the programmer is not omnipotent, remember). That's why programmers say "all software has bugs". It's not that all software has flaws, but rather that it's almost a sure thing that any given piece of software can be used in a way which causes a problem (or "bug") to arise.

Take this iTunes software bug, for example. Even here, it sounds like there would have been no problem if people followed the directions. Given the simplicity of the fix, it should probably still be called a programming bug. However, what happens if some user actually had the old sh shell on there instead of a link to bash? There could be some subtle errors, as bash does not behave 100% like sh. Bug? Probably not, unless you argue that the programmer should have seen it coming. All right then, suppose some user decides that they only want to use csh, so they link sh to csh instead of bash (note: a Bad Idea)? The installer wouldn't work at all. Bug? Hardly.

The point being, the programmer can't plan for every contingency, and this can lead to problems which people call "bugs". That's why "all software has bugs"... because even if the code is flawless, there's no guarantee that it won't fail somehow.

Re:Why I hate the software industry

[dvNull]

Have you ever built a bridge ? I have 2 uncles whoe are civil engineers and yes they have to do a LOT of calculations anticipating every possible situation. For example, like no more than 'x' tonnage on the bridge at a certain time. If a bridge is rated for 200tons of traffic, its built to handle 300 or even higher.

If a programmer makes an installer which deletes the file and in their instructions they say double click this file to install and it deletes all your data, then yes the programmer is at fault. If the instructions clearly say Select /tmp for temp and then install and the user doesnt follow that, then yes its the fault of the user.

We should make software companies liable for the bugs THEY put in and stop blaming the user for everything.

Re:Why I hate the software industry

[anonymous loser]

Have you ever built a bridge ?

Have you ever written a piece of commercial software? The same principles apply. You try to keep the interface as simple as possible, because God knows some stupid luser will find a way to break it, and you want to limit the side effects. You build in all kinds of extra error handling and tolerances for bad input. You create ample instructions (readme, dialogs that pop up before the user can do anything, startup tips, help text, etc.) and people still ignore them.

Yes, sometimes programmers write bugs, and sometimes "hardware" engineers make errors in calculations too. Anyone remember the Mars pathfinder mission?

Errors tend to occur when the engineers are put under pressure to meet a particular deadline, and they aren't given enough time to complete the task. You want it done by Friday? Fine. If it crashes your system, don't say I didn't warn you. However, when Friday comes around, the work is handed over, and the system crashes it's the engineer who takes the blame for shoddy workmanship, and not the customer, manager, etc. who demands the product before it's ready.

If you or anyone else wants to complain about faulty software, feel free. Just don't complain when you have to wait twice as long for the new version, and it doesn't have many new bells and whistles.

Re:Why I hate the software industry

[dvNull]

and the system crashes it's the engineer who takes the blame for shoddy workmanship, and not the customer, manager, etc. who demands the product before it's ready.


Well I believe that the company should take responsibility. If Marketing wants product Backup Exec to ship on Friday and the engineer insists that it isnt ready and at the present all it does is erase your HDD the instant you put the cd in, thats Marketing's fault and in all truth the companie's fault. The Engineer shouldnt have to take the blame, but the mohership (the company) should understand its mistake.

dvNull

Re:Why I hate the software industry

[geekoid]

Software failure can be just as catostrophic as a brige/building failure.
What id the software the engineer usues is faulty, and the bridge collpases?
what if vital signs our misreported to a doctor?
etc, etc, etc ...
Speaking of engineers, here come custodial engineering now.

Whiplash lawsuit?

[TNT_JR]

"Lost 100GB"? Sounds like a new backup strategy is in order? ;)

the REAL code, and how this did happen

[Anonymous Coward]

In the installer is a small shell script to remove any old copies of iTunes. It contained the following line of code:

rm -rf $2Applications/iTunes.app 2
where "$2" is the name of the drive iTunes is being installed on.

The problem is, since the pathname is not in quotes, if the drive name has a space, and there are other drives named similarly then the installer will delete the similarly named drive (for instance if your drives are: "Disk", "Disk 1", and Disk 2" and you install on "Disk 1" then the command will become "rm -rf Disk 1/Applications/iTunes.app 2

The new updated version of the installer replaced that line of code with:

rm -rf "$2Applications/iTunes.app" 2
so things should work fine now.

Re:the REAL code, and how this did happen

[crayz]

Yeah, and you needed to install on that specific drive too. I have two drives: "DVD 1" and "DVD 2" and they are fine, because I was installing in the app folder on my OS X drive(just called "Mac OS X")

So this sucks, but it's really a very specific bug that would've affected only a few users. I can see how Apple would've missed it

Re:the REAL code, and how this did happen

[twenex]

The real code checks for the existance of the file/dir first. So.... you would lose your "Disk" drive if you installed on "Disk 1". So, I would guess that anyone losing their entire system is installing on something called "/ Drive" or something similar to that, so that "/" is the component before the space.

Re:the REAL code, and how this did happen

[MaxVlast]

Crap!

My drives used to be "Vyacheslav 9", "Vyacheslav X", etc., I just renamed them to remove the spaces (to make shell scripts work better, actually.) I'd have lost all of my data had I not done so. Wow.

Re:the REAL code, and how this did happen

[timsuth]

> rm -rf $2Applications/iTunes.app

Yes, you should always quote variables in shell scripts, but given that iTunes.app is a single file, there is no point to the -r (recursive remove, for removing everything below a directory) option to rm.

Had it only been "rm -f $2...", the damage would have been minimal. (rm -f will not remove directories, hence the only risk would have been if we had e.g. "/Volume1/some directory/iTunes.app" and there was a file named "/Volume1/some".)

Renumeration Remuneration

[eyewell]

Renumeration has been addressed. It is aparently no longer 2.0, but 2.0.1.

However, this does not address any Remuneration!

not mine

[IRNI]

my drive is partitioned. the installer went fine. my drive isn't messed up. nothing missing. must suck for whoever it did happen to who doesn't have a backup.

No problem

[alfredo]

It installed with no problem, works well. I didn't like the fact that I had to log in as administrator, and thought that odd. Kudos to them for fixing it so fast. It is not like them to release defective software. Sorry for all those who lost data.

I am glad they are not like another software maker who would insist that it is a feature, not a bug.

The equalizer actually works!

This is how I fixed it....

[capedgirardeau]

This just happend to me, well a day ago it did. This is how I totally fixed the problem: In iTunes I put visuals on and set to 'Full Screen', the volume to max, hit the play button and sparked one up. What was the problem again ?

0 bytes?

[krog]

*no one* uses Macs? then i hope you don't like:

• digital video
  
• high-end publishing
  
• computer art
  
• electronic and digital music
  
• desktop publishing
  
• CGI
  
• Talking Moose
  

Re:0 bytes?

[MaxVlast]

The talking moose owned yoor world in 1992.

And what if....

[Guillaume Ross]

iTunes was made by Microsoft? It'd be a riot here..

Re:Learn to use exception handling you caveman

[AndroidCat]

You people who code in C don't realize that reading your code is almost like reading tape from a tape reader.

Actually, we read tape from the tape punch, but other than that, what's your point?? :^)

Worked for me.

[prwood]

I have a Pismo PowerBook with MacOS X 10.1, and I downloaded iTunes 2 immediately after it was released. My hard drive has two partitions, one for MacOS 9.2.1, and one for MacOS X 10.1. I also already had a previously installed copy of iTunes on both drives. I ran the iTunes installer, and everything worked fine. It didn't wipe out any data, and I am quite enjoying the new iTunes 2. I

Gee, I guess I was just lucky?

Aplle lies

[frost22]

This problem only came to light today, and they have a fix out the same day.

No. Apple lies:
- they have not even admitted failure. Instead they engage in Microsoftian doublespeak talking about an "issue". Instead, they should write: "due to a defect in our software installer ... "
- they don't take responsibility. No compensation for people who sufferered sever data losses.

Posting a patch isn't always enough.

You just don't wipe people's hard drives. Never. Having an installer that is even capable of such is a sign of faulty design. They are to blame.

f.

Re:Aplle lies

[stripes]

Having an installer that is even capable of such is a sign of faulty design.

How do you make an installer that can remove the old version of a program, and yet have zero chance that it never removes the wrong thing?

(yes, it was a bad bug, very bad, but I'm not sure how you design a drool proof installer that has zero chance of deleting the wrong thing -- I know how to do a not drool proof one though. Put iTunes2 on a disk image along with a README that says "delete the old one before installing the new one").

Re:Aplle lies

[haruharaharu]

ow do you make an installer that can remove the old version of a program, and yet have zero chance that it never removes the wrong thing

Easy. require apps to declare their contents to the system and provide an interface to manage these apps and their versions. Then require that apps use that interface exclusively to manage them.

It's really hard to wipe an entire drive with uninstallComponent("iTunes", VERSION_ALL);

Re:Aplle lies

[jrockway]

I'll tell you... Debian GNU/Linux (or GNU/Hurd) would never have this problem :-D

Re:Aplle lies

[MaxVlast]

My understanding is that's how it does work. (Each package installs a receipt with a bill of materials.)

The problem could still have occurred if the BOM path was wrong ( Drive 1/iTunes instead of "Drive 1/iTunes") Just in a slightly different place.

Re:Aplle lies

[stripes]

Easy. require apps to declare their contents to the system and provide an interface to manage these apps and their versions. Then require that apps use that interface exclusively to manage them.

Already does, you put them in a direcory named Foo.app where Foo is the name of the application, then users can move it around without things getting out of sync...

It's really hard to wipe an entire drive with uninstallComponent("iTunes", VERSION_ALL);

That depends on how "uninstallComponent" is implmented. If it is written in a shell like language it can be pretty easy...

Wrong, mind you, but not so hard to do wrong. I would guess some Linux and other generic Unix installers have problems with spaces in file names... I know for sure they have problems with display numbers other then zero, and this mistake is about as basic.

Re:Aplle lies

[jcr]

How do you make an installer that can remove the old version of a program, and yet have zero chance that it never removes the wrong thing?

The obvious way is to rsync the existing app's directory with the new version, instead of delete-and-replace.

-jcr

Re:Aplle lies

[stripes]

The obvious way is to rsync the existing app's directory with the new version, instead of delete-and-replace.

That still deletes files, merely saying "resync" does not magically fix that. I liked the "move the old one to the Trash" suggestion better.

Re:Aplle lies

[jcr]

Check the man page on rsync. When you rsync a directory with another directory, you're not running a risk of deleting anything outside that directory. (Like, say, /*)

-jcr

Re:Aplle lies

[stripes]

Check the man page on rsync. When you rsync a directory with another directory, you're not running a risk of deleting anything outside that directory.

Yes, but rsync uses unlink (or it did in 1995, which is the last time I looked at it's source). It might manage to destroy everything, like if you accidentally specify the target wrong (great...my whole disk has been replaced with a single copy of iTunes and nothing else).

I'm not saying nothing is better then what Apple uses (an installer that runs a shell script), but that nothing that aims to do the same thing will have zero chance of deleting the wrong thing. In passing I'll also mention that it isn't obviously worse then many other Unix installers (you know, the ones that just run random shell scripts!), but Apple has the misfortune of having a user base that tends to use spaces (and punctuation) in their file names...

Re:Aplle lies

[frost22]

- they don't take responsibility. No compensation for people who sufferered sever data losses.
Wow what kind of moron is iTunes on a server? Doesn't that eat a lot of CPU cylces.

You misinterpreted my typo. I was talking of severe data loss. It really doesnt matter if the affected machine is a server or not.

f.

Re:Aplle lies

[rew]

How do you make an installer that can remove the old version of a program, and yet have zero chance that it never removes the wrong thing?

Well, you give it the "list of files" of the old installations, and make it delete all files from those old installations that do not contain any user-data. (e.g. leave out the "initial database" that is distributed with the program).

In effect you have:
for (tp = file_list; tp;tp=tp->next)
unlink (tp->fname);

See the missing check if the unlink works? That's intentional. We're not interested if the remove of the file worked. If for example the user decided to write-protect some of the files (actually directories) we don't want them deleted.

Roger.

Re:Aplle lies

[stripes]

Well, you give it the "list of files" of the old installations, and make it delete all files from those old installations that do not contain any user-data. (e.g. leave out the "initial database" that is distributed with the program).

That still doesn't have zero chance. Your linked list might be damaged by another part of the program, either overwriting the link part, or a name part. You can complain that that would be due to a bug, but so is the shell script problem.

Zero is a hard goal to reach. Impossable even given the chance of DRAM errors.

You can get a little closer if you try this:
for (tp = file_list; cnt = N; tp;tp=tp->next) { assert(cnt--); unlink(tp->fname); }

Of corse we both left off finding the package install location (i.e. translating Foo.app/MacOS/Foo into "/Volume/Name of Volume/Applications/Foo.app/MacOS/Foo"), which is where the script went wrong, so we have a few more chances for error...

There is also a problems with the resource fork, in OSX it has been replaced. Resource Bar in program Foo is stored as Foo.app/Resources/Bar, so knowing all the files that Foo has requires knowing all the resources it creates, which may be hard to know ahead of time.

It also can be a problem because one may not know all the files in the iTunes package, it is clearly known for past versions, but not future versions.

I won't mention dumb-ass users replacing iTunes files with their own :-)

See the missing check if the unlink works? That's intentional. We're not interested if the remove of the file worked. If for example the user decided to write-protect some of the files (actually directories) we don't want them deleted.

I see it, it might be a bad idea if the file is a kext, or some random incompatable shared lib or kernel module. It is still a valid design choice though. It would be better if you told the user which files you wanted to delete but didn't because the user had protected them.

I still like to solution of moving the things to the trash. It also doesn't have zero chance of not deleting things, but it probbably won't. It definitly doesn't have no chance of moving the wrong stuff to the trash, but if it does it is not so hard to move them out, definitly simpler then undoing an rm.

Re:Aplle lies

[bzcpcfj]

"How do you make an installer that can remove the old version of a program, and yet have zero chance that it never removes the wrong thing? "

You could take the approach Acrobat Reader 4.x did. If it detects something that looks like an older version of Acrobat Reader (pre-4.x), it pops up a box telling the user to uninstall it.

Ahh... Imagine if this were a Microsoft product.

[sheldon]

There would be great shouts of anger. Oh those bastards, they do this because they do not want you dual booting!

It's because they are a monopoly! They are purposefully targetting Linux partitions.

Ohhhh!!! AARRRGGHHH! Grand Conspiracy!

Yet here in these responses I see the simple realization that some idiot made a mistake writing the installation script.

Are you sure?

Are you sure it isn't Apple trying to subtly encourage people to use Microsoft's Media Player instead? :)

Why using scripting anyway?

[SCHecklerX]

If you'd use a scripting language that allows system-independent calls, shit like this would be avoided. Instead of shelling out to another command (rm in this case) You would call "unlink" with the PROPER parameters in the first place!

Oh well.

So old, it's a classic joke

[McD]

From the classic old joke list "How To Shoot Yourself In The Foot" in various programming languages and computing environments, here's the entry for Unix:

% ls
foot.c foot.h foot.o toe.c toe.o
% rm * .o
rm: .o: No such file or directory
% ls
%

Same bug. Welcome to the world, Apple. :-)

Peace,
-McD

Re:Ahh... Imagine if this were a Microsoft product

[Special J]

If it were a Microsoft product then it wouldn't be a bug.

It would be a "feature".

Bias and stories?

[SilentChris]

I know we usually have the anti-MS, pro-everything else bias around here, but I was wondering: does anyone have any stories of Microsoft software doing anything similar? I know the Linux installs are pretty adament during the partitioning process ("Watch out! Look out!"), but I can't think of anything by MS doing the same thing.

Can Mac really be the anti-Christ? :)

100BG! Ouch

[dbretton]

I personally lost more than 100GB of data.
....
....
Somewhere, in a little corner of the basement of a house, someone is installing their new iTunes...

{blip, squeek}.. Oh man, this is sooo cool!
{HD Grrrrinnd!!}
huh? What thee.. !!

NooooOOooooOOOOO!!!! My PORN!!!
Oh my God!

Later that day, at a Starbucks, we see a man, trembling as he sips is triple MochaBucka Latte-chino...

Brtney... GONE!
Pam....GONE!
Margolis....GONE!
That chick doing the horse...GONE!

My life is over....

Did Apple send out email warnings?

[lsommerer]

I noticed on the download page that Apple asks for email addresses. Does anyone know if they emailed warnings to people who had downloaded the effected version? Seems sort of crappy if they didn't do that.

Re:Did Apple send out email warnings?

[Alan Partridge]

totally optional information

Half-Life did the same thing ..

[Idimmu Xul]

This isnt the first time this has happened..

Although not quite as disasterous as formating the entire hdd, originally when installing half-life if the install dir was changed from say C;\Sierra\Half-life to D:\games\half-life then when half-life was uninstalled it would take the whole of d:\games with it :(

distintly not fun :/

When did the..

[GISboy]

BOFH start writing shell scripts for Apple?

I just checked the register.co.uk and he still seems to be employed there?

Almost sound like a BOFH gag:
User: I need more space in XXXX account.
BOFH: Ok, I just freed up 8M of space.
User: Great, now I have 16M.
BOFH: No, now you have 8M free.
User: Arrrr......!
BOFH: (click).

Do not turn to the Dark Side of the Force...but, hey, the Aqua side ain't so bad!

Re:Ahh... Imagine if this were a Microsoft product

[phillymjs]

Yet here in these responses I see the simple realization that some idiot made a mistake writing the installation script. Are you sure?

Yes, because on the Mac, the presence of multiple partitions != dual-booting user. OS X will reside quite happily on the same partition as OS 9.x, and you can choose between them with the Startup Disk control panel.

The only advantage I found to putting OS X on a different partition is, you can select which operating system you want to boot in every time you power on the machine, by holding down the Option key until presented with the menu.

~Philly

Exactly.

[mindstrm]

This is not to say that you can't negate yourself of responsibilty... but that would usually require some kind of negotiation, and contracts signed and notarized by lawyers or something.

ie: Where it can be very clearly demonstrated that both parties thoroughly understood, in detail, the terms of the contract.

Just curious.

[mindstrm]

But what kind of drive/storage device did you have that had 100GB on a partition?

The shell script not the problem...

[A_Non_Moose]

You have to realize that iTunes has an equalizer now.

After the installer formats your HD, you can record the high pitched scream you emit into an mp3 and then change the pitch to all bass.

So, now Apple just "equalized" itself with all other unicies.

To Apple I say, "I feel your pain" but you need to "strategize" some more.

Ow, crossing OS, platform and political lines... for shame! for shame!

(on a side note, modding me down as overrated because of my +2 bonus makes about as much sense as hating people for being intelligent...Oh, wait, that is what happens to "us" nerds all the time... I just answered my own question, never mind...so, being different is ok, as long as you are different like everyone else? Heh, makes sense...NOT!)

Yes I'm an esoteric, tenacious, longwinded SOB.

I'll never need therapy as long as I can post to slashdot.

Orbb: "keep talking, I'm reloading"

Re:The shell script not the problem...

[Rendus]

(on a side note, modding me down as overrated because of my +2 bonus makes about as much sense as hating people for being intelligent...Oh, wait, that is what happens to "us" nerds all the time... I just answered my own question, never mind...so, being different is ok, as long as you are different like everyone else? Heh, makes sense...NOT!)

You forget the purpose of the +1 bonus. It's not to be left on all the time, but only used on comments that deserve a +1 off the bat. At the very least, this post doesn't qualify. Therefore, the comment IS overrated, and deserving of a -1 moderation because of it.

The +1 bonus is NOT a "Hey, I'm better than everyone!". It's a slight bonus to being a well-behaved member of Slashdot, which are few and far between. There's an expectation that you use the +1 accordingly.

Re:The shell script not the problem...

[A_Non_Moose]

Ah, yes.

You do make a very good point.

I do however counter with this and this as an example.

On the first link, pay attention to the moderation and the reply of "kvigor" saying I was off topic, then the second link about why I disagreed with him/her.

Like this response it is on topic to both your statement and the subject of this particular news piece.

I think with the person's loss of a 100G of data we all knew it was going to happen because there are too many variables, people are fallable and maintaining a dual OS strategy is doing more harm than good.

And to answer your question, Rendus, I've pointed out both gently and rudely the "death by a thousand cuts" some moderators seem to be inflicting on some users...dunno why, hence the second link...modded as offtopic when it was not.
(shrug)

I don't know why, and in some respects I've ceased to care, after all as someone said "Karma ain't your dick size". If this is true, then why are some moderators such as kvigor such "pricks of the litter" (yeah, an ad-hominim attack meant as humor...fullfilling yet in poor taste)
And why kvigor was not modded as offtopic, I'll never know.

The question is, perhaps was I abusing the +1 bonus? Perhaps.
Are moderators such as kvigor abusing thier privelage? Perhaps.
Should a post be modded down because it is not understood? Hell, no!
Should Calculus be no longer taught because people don't understand it, save for the very few?

Same logic applies, does it not?

I suppose I can sum up this entier thread and its replies as:
"Bummer dude. Didja backup? Better you than me."

This thread reminds me of the first time I moderated, I saw a thread that was so incidiary, so inflammatory and so way on/off base that I bought it hook, line and sinker 3/4 of the way thru untile I realized it was satire!

And even then I did not mod anyone down for the simple reason that people on /. are normally good, intelligent and accepting individuals.
A few out there are proving me wrong over and over.

To quote Shrek "They Judge me before they even know me".

Won't be the first time, won't be the last, but I thought at least I'd get away from that kind garbage here.

and for those of you keeping score...up 5 points and down 6 in about what 3 or 4 hours...

I find it funny in a sad sort of way.

Moose

Don't we agree to the license after the damage...

[melvin22]

is already done? I was discussing the whole iTunes2 issue with a friend over breakfast, and he said that we only agree to the license after the product is installed, and the damage is already done. I haven't had a chance to install iTunes2 yet, because I'm not on my computer so I don't know about that. But could the post-install agreement to the EULA give people grounds to fight Apple (sue them, or whatever)?

Re:The power Mac

[Alan Partridge]

Of course, the "Classic" MacOS was so thoughtfully laid out that you didn't actually NEED installers at all. Us MacOS users feel like we've turned around and are now rushing back down the road we've been travelling along these years - straight into the path of the oncoming Microsoft juggernaut. iTunes X 2.01 is a marginal improvement, now crossfades tracks rather than clumsily cutting and has a few UI tweaks on top.

Apple. So easy to use....

[CritterNYC]

iTunes.pkg *click-click*

"Ah... Apple... So easy to use, now wonder it's number on.."

*quack* *quack*

"what?... huh?... oh, son OF A BIT#$%!"

Have EULAs been tested in court?

[Greyfox]

There are a lot of questions that arise with EULAs. I seem to recall a click-through EULA on a web page being rejected by a court. Maybe it'd be a good time for a legal-system test.

Even if the court says the data on the drive should have been backed up, you should be compensated for time spent restoring that data, reinstalling the OS, and generally getting the computer back in shape. Not to mention the woeful negligence factor as a simple test should have uncovered this problem prior to the product being released. An application install should not wipe your hard drive out.

As usual, IANAL (But I play one on TV)

Possible Fix...

[DragonPup]

Andrew Welch of Ambrosia Software posted a method that MIGHT work on recovering the files here. Basically sometimes the installer, according to Andrew, just messes with file permissions and visability, not actually deleting them.

I didn't test this because iTunes didn't mess up my 5 partitions, thankfully.

-Henry

Uncle!

[Voline]

Ok, Ok. I'll back up my hard drive. Today. I swear.

Re:iTunes 2.0 problem the exception, not the rule.

[jchristopher]

Before OS X, partitioned hard drives on Apple boxes were all but nonexistent.

Actually, classic MacOS always had partitions too, they were just hidden from the user.

Re:Don't we agree to the license after the damage.

[Apotsy]

If I remember correctly, there are in fact two license agreements. One when you start the install, and another when you launch the software for the first time.

Hard to remember though, since I always just click "Agree" and forget about it right away...

this is not just limited to iTunes!!

[warrior]

I installed the latest quicktime player software, and it also deleted all the data it could. The only thing left was an (almost) empty win98 directory. Perhaps apple should look at ALL of their install programs!!!

Mike

Re:this is not just limited to iTunes!!

[MacBoy]

Highly doubtfull. What plattform? What version of the quicktime installer? Where are all the other reports of this problem? Why spread FUD?

In my experience, the Quicktime installer is and has been an efficient and effective download-on-demand installer, and it doesn't require users to reboot, like almost everything else in the windows world. There have never been reports of data loss resulting from it.

/. said something bad about the iPod? hells yeah.

[Geek Dash Boy]

And there was a /. story about the iPod - and it had nothing bad to say about Apple

Go back and read it again. There's a negative comment by Taco in the initial post.

Slashdot is inherently anti-Apple due to the attitude of editors which is quickly imitated by the troll hordes and flamebaiters. 'cuz everyone knows the easiest way to generate wind is to get a bunch of Mac zealots in one place and say one bad thing about Apple or the Mac OS.

For the record, I'm a long-time Apple customer that got tired of the OS wars a long time ago. I'd much rather be coding in my very nice Mac OS X setup, thank you very much.

Once Again Apple is Ahead of the Curve

[Anonymous Coward]

I don't understand what benefit formatting or deleting data has, but I am SURE it has something to do with great usability! I predict within the year, MOST software installers will also reformat/delete your hard drive if you dont have enough space. Apple is just that good.

Re:Wait a second...

[kvigor]

I attempted to mod this as offtopic; slashdot had a tizzy and decided I'd chosen 'underrated' (which I can assure you I didn't). Fortunately, posting this content-free message should at least undo the moderation damage.

Stepwise lamented the problems of pkg's before

[Ghoser777]

Part 1 and

Part 2.

In summary: Unix + lazy installers = bad

F-bacher

I bequeath to you...

[A_Non_Moose]

a dual-boot to the head...(whap!)...OW!

ok, this is a slight tangent I have to point out, but then again I'd just be NIT-picking.

Give a man a fish, and if he makes a crack pipe out of it, well, you've got serious problems to deal with.

Yes, I'm a tenacious, longwinded, esoteric SOB who needs therapy...why else do you think I'm posting here?

Survey says...

[doorbot.com]

Interestingly, there is some logic beyond this argument as well.

Assume for a moment that you're a hard working, productive worker building your product or otherwise producing more income to the company than they pay you. You are a "good" employee, and the company would prefer to keep you. Now, another employee who was hired along side you is promoted to your manager. But you reflect back at all the asinine questions that person asked, and how perfectly obvious they made the fact that they should be working at McDonalds and not making $100,000 a year bossing you around (let alone in this industry).

Why did they get promoted? Is it because they're buddies with the boss? It is because everyone is in some maniacal conspiracy against you?

No. It's because of two things. Either they get promoted, or they get fired. Sometimes a company will chose to promote rather than fire an employee. But more likely, the simple fact is that you are productive in your position. If you were to be promoted to manager, your job would radically change and the company would not have the benefit of your skills working towards completion of the product.

That's why most management is "stupid" and "doesn't get it." Because they were workers who were stupid and don't get it. Granted, there are some members of management who are good at what they do -- these are the few that won't get fired when management has its shakedown. And not everyone is promoted to management... but if they were hired on, does that necessarily make them good at their job?

Think about it, if management were that great, it'd cost a lot more.

Re:Who clicked?

[QuickFox]

I didn't click on anything, so that means that I can sue them, right?

You know, you might be able to settle out of court, if you're fortunate enough to deal with a mature and responsible virus writer.

windoze

[h4b1t]

see, if you had been running windows xp then all that would have happened was the files would all be converted to a secure media format by the windows media format converter wizard and then everytime you tried to play one it would contact the riaa and fine you.

Make Apple read this...

[A_Non_Moose]

It will be punishment enough.

(this will help some people to follow along, for I forgot (thank god) about some of these languages...maybe that is why 'code == speech'..heh what a connection"...anywho...)

Ok time to make you C the error of your ways, before looking at the ++ side.

We are not going to dole out useless pearls of wisdom because everyone should be aware by now software is not made on an assemblerly line...damn, that LISP is coming back, sorry.

There is a certian Basic Motif that creates this Paradox that creates a Revelation: Programmers are human and machines execute mistakes with alarming speed.

So, go FORTH and remember it will be a cold day in Cobol when you get on that FORTRAN and have to make Smalltalk or some Hypertalk over a cup of Java and hope and Cray you won't be NeXT.

And if you have Access to a Window seat, you'll see that the Sun will rise again.

MOOving forward, Apple, remember the old Adage that "IBM is sometimes and acronym for It's Better Manually or I've Been Misled".

Yes, I'm making and ASP out of myself right now, but I had to Scheme a little bit to be this SASsy.

All your XBase are belong to US!

'Moose out (of his mind I'm sure some of you are saying).

Stupid -- this is a non-issue

[Lurkingrue]

THE HORROR, THE HORROR! Several, perhaps even DOZENS of Mac users may have been severely inconvenienced by a BUG! I can't begin to express the proper outrage! Stop the presses! Confiscate all Macintoshes! Obviously, they're too unstable and dangerous to use. Obviously, we should lead off with a Slashdot Mac-bashing story! If ONLY all these poor Mac users had used Linux instead, the tragedy that would have been averted! Oh, the humanity!

Please.

This topic has gotta be a huge joke, and certainly doesn't warrant such an outlandishly over-blown story. While a very bad bug, this installer problem will likely have affected only a few people, probably in the three-digit range. Here's why:

1) If you weren't one of the early-adopters, you'd never have heard of this -- Apple pulled and replaced the defective installer in less than 18 hours.

2) If you followed the instructions in the original "Read Me" file, and deleted all pre-existing copies of iTunes, you wouldn't have this problem.

3) If you weren't using multiple partitions on your drive -- which most OS X users aren't -- you'd not have this problem.

4) Finally, if you did have multiple drive partitions, but didn't include a space in the name, you also wouldn't have encountered this problem.

I won't even go into the fact that the primary partition is left untouched by the installer-bug.

So, in summary, yeah -- a really bad bug. But, no, it didn't really affect many OS X users.

Re:Stupid -- this is a non-issue

[Lurkingrue]

All the more reason not to be broadcasting anti-Apple FUD.

How to do it

[RAVasquez]

Rather than having the script delete iTunes.app, have it move it to the appropriate .Trashes folder. Nothing gets deleted, and if there's a problem, the old version is sitting in the Trash can.

Re:How to do it

[stripes]

Rather than having the script delete iTunes.app, have it move it to the appropriate .Trashes folder.

Ok, that's a good idea. There are a few problems with it, but I think they are smaller the the problems with real removal.

• I think iTunes (the old one) is probbably owned by root, so I'm not sure it would go to the current users trash, or work well there unless chown'ed. If it gets chown'ed just dragging it back where it was might not be a good idea (if nothing else it means an attack on user Foo can overwrite itunes and if someone elses runs it sneak in...; not a big deal since user Foo had to be an admin to install iTunes2 anyway). However untangling that kind of mess is better then dealing with half (or all) your files being rm'ed
• There may not be enough space to install unless the old iTunes is deleted first, which would frustrate people.
• More crap in my trash can? Have you seen how full the thing is already?
• There is no easy way to invoke the Finders "move to trash" from a shell script. It would be a fine idea to fix this problem.
• From a UI point of view I'm not sure people would expect upgrading some software to leave the old version on the trash. Of corse two years after we start doing it that way everyone will expect it, and think it would be dump not too...

Very good idea. Someone tell Apple :-)

P.S. it is owned by root: drwxrwxr-x 3 root admin 58 Nov 3 17:30 /Applications/iTunes.app/.

Apple need some UNIX people?

[Mongoose]

This is clearly an issue of impromper trianing on unices... I mean what are we going to do when Microsoft releases it's BSD distrobution to combat apples? hahaha

Kernel modules

[hearingaid]

Disclaimer: I don't run OS X, and I've never configured a Linux kernel.

However, I have configured FreeBSD kernels. And in FreeBSD, some kernel modules are externally loadable, but they always need hooks into the compiled kernel (and it's generally better to put the kernel modules you always need directly into the /kernel file, that way you don't need to do so much disk access)... And, so far as I know (I haven't installed any FreeBSD past 4.0), device drivers are not hookable in any version of FreeBSD. IIRC OS X & FreeBSD share a lot of kernel traits in common, this I would expect to be one of them.

Arghh!

[Lars T.]

Stupid "preview". Stupid, stupid, stupid.

Users don't follow directions (RTFM)

[Head]

I'm not defending Apple but the instructions for the install clearly state "If you have previously installed iTunes for Mac OS X, delete the old version before installing the new one."

Then the code that checks for the existence would fail and the bug wouldn't rear it's ugly head. However, we all know that most users don't RTFM.

Of course, if they really meant what they said in the instructions, the check for existence should have asked the user to remove the old version instead of doing it automatically.

Misleading statements

[FastT]

This is a chronic problem that most commercial software houses have.

While technically true, the connotation of your statement and the statements that follow are misleading. Sure, the state of testing is poor in commercial software, but it's even worse in open source software. At least commercial companies provide some minimal testing with each release, actually have people assigned and paid to do testing, and have someone to call when you have a problem. Compare this with the schlock you get with most open source software, which was at most tested on nothing other than the author's machine, with no code review, no test cases, and no accountability.

Of course, there are some open source exceptions, just as there are in the commercial world. But remember that 99.9% of open source software is NOT Apache or Linux, and the people developing the rest of this software are frequently more enthusiastic than skilled, and definitely have no resources dedicated to testing. In fact, many open source developers are also commercial developers who prefer to work on open source because they can work alone and it doesn't come with all the "crap" like code reviews.

So, let's lay blame where it belongs--on software development in general, the lack and/or cost of resources, and the general disregard for software quality as something important. None of these things are limited to commercial software development.

I *knew* dial-up would pay off!

[Powercntrl]

All my friends brag about how they can download things so fast - I still haven't bought into the broadband hype. Now I'm even more glad I didn't.

See, I have this huge queue of MP3s waiting to download and I was going to download iTunes 2.0 as soon as it finished... Okay, so it looks like the Xbox will be out before it finishes, but thanks to the slowness of my connection, a bugfixed version of iTunes has already been released, while people with broadband got their hard drives wiped.

...and you wonder why broadband companies are hurtin'. Give the programmers time to fix their bugs - use dialup! ;)

/. flamers suck

[liq-bin]

There really could have been some decent discussions that followed this, and a few of us tried, but the signal to noise ratio on /. has gotten pretty bad. At the very least, if you don't have something constructive to say, keep your comments out of my thread if you can't keep them to yourself entirely.

Apple's bug in the installer, could be blamed on any number of issues, but I think the real underlying issue is more fundamental and philosophical than technical. Yeah, so they screwed up a shell script, and it was kinda obscure and probably outsourced, but more importantly, why was the installer trying to delete ANYTHING? Why would an installer delete anything? it's an INSTALLER!

If something needs to be removed (which should be rare if all of your depondencies include version numbers and such) then the offending item or folder should be moved to the trash. This is after all the purpose of the trash can.

Another problem - why is the iTunes installer installing with root authority? I only use it to play mp3's. I don't need it mucking with my system. I'd use something else if there was something else worth using. SoundJam still wins over iTunes but the Apple Corporate Machine ate that product and had it removed from the market. It's an app. if I need the firewire interaction / automation I'll ask for it. I don't think it should have installed with root authority, there should have been a non authorized install or a drag install.

So the issue we should really be concerned with is not how Apple was removing files, but why!?! And this question is not specific to Apple, it is a concert on all platforms, including yours. This shoud have been an opportunity to discuss, question, and learn. Not flame. And if you do flame, your only justification should be that you were actually affected by the bug.

Re:/. flamers suck

[liq-bin]

also, why doesn't rm have better limiting functionality for scripting use? Or if it does why is it not used? Imagine this command
rm -r --limit 25M

if rm sees the potential deletion set to exceed 25Meg it returns without deleting anything. I know that I would often use this, or maybe default it to something so that I didn't accidentally delete WAY more than I intended to.

There is a post that asks if Apple needs some UNIX people. I think alternatively that UNIX could also use some Apple people.

your point?

[lsommerer]

I don't see that it particularly matters that the information is optional. My question was whether or not they attempted to use that information to inform people about the potential problem that could result from using the installer.

Presumably, some people provided the optional information, and perhaps some part of that information was not bogus, but included a real email address.

This says a lot about QA departments

[Eusebo]

Ironically one of the easiest ways to cut costs in the short term becomes very expensive in the long term. My experience seems to be consistent with that of others: The QA department is staffed with bottom of the barrel technologists who don't understand what it is they are doing and aren't paid much to do it.

Hopefully many companies (not just Apple) will learn a lesson from this little blunder. The QA process should, in many respects, be larger than the development process. Many companies have the notion that good developers alone can produce good products. IMHO, most good products have good developers and great QA folks behind them.

Completely Unacceptable

[PbHead]

I find this to be completely unacceptable, and here's why:

1.) Communicability: Code like this should email itself to other users BEFORE it wipes the drives.

2.) Transparency: Item's of this nature should do their best to hide themselves from the user.

Apple has succeeded in making either the most worthless 'Virus' I have ever seen, or possibly the best due to a lack of common viral standards.

It would have been better if they gave the user an option to E-Mail an "Upgrade Now" to their friends before eating the data. The best option of course would be to go through the users address book and E-Mail the "Uprgade Notification" to everyone in the background.
Go back to the Drawing Board Apple, if your going to wipe some drives, at least do it with some gusto.

Re:The cure.. in 4 easy steps!

[MacBoy]

Sell all your Mac hardware and build yourself a (much faster) PC with commodity components

You are either misinformed or merely ignorant if you believe that any x86 based PC is faster than a G4 based Mac. Mhz means virtually nothing, but some people will never learn, I guess.

There is only one step necessary to avoid having been affected by this bug. RTFM. If users had done as instructed and removed previous versions of iTunes from their system before installation (a very good idea indeed, before installing any piece of software) then they would have been perfectly OK.

not really...

[Anonymous Koward]

you implicitly (I hope I got that right :) have given your son priviledge to install software on your behalf if he 1) has an account capable of installation of programs on the system and 2) have even just plainly allowed him access at all (i.e. did YOU have a disclaimer, that your underage son couldn't have legally signed anyway, as a requirement BEFORE being allowed access through the system in the first place?).

I think you get the idea..you're responsible for the USE and SECURITY of your own internet connection and/or system.

The NASA model is not commercially viable

[Ars-Fartsica]

Sorry, but if you try to push out software using the NASA model, you will be late to market and ultaimtely out of business.

Rapid deployment and subsequent revisioning is the secret behind many of the largest software companies in the world.

Ultimately you have to sell a product and pay your employees, and the NASA model is never going to do that.

Re:The NASA model is not commercially viable

[CKW]

But is the currenty "commercially viable" marketplace the best way to determine which software is best for mankind?

I look to my left and see Microsoft. And I must say, NO. Just because we currently *use* the capitalist marketplace model for a lot of things, does not mean that it is the best. It's there because of historical reasons. Personal rights reasons. And utility.

The general base of lusers and pointy haired managers that determine who wins in the "commercial marketplace" are too damn stupid to pick the proper combination of power, performance, price, and stability, that is important. That's why we spent so many years with DOS, Win3.x and even 9x, while things like AmigaOS (1985) lingered and died.

Doomed by the mediocrity of the masses. What fun it's been.

Fairness in reporting about Apple

[CaptainDogCow]

Hey, what if Slashdot posted well informed and unbiased info about Apple. It didn't even mention the fact that this bug only exists in the MacOS X installer. If this had been a post about a Microsoft bug and they hadn't distinguished between a Win 9x problem and a Win 2k / XP I know everyone would be up in arms. It was an unfortunate mistake apple made in releasing this installer. However lets be fair in the future. Apple has a hard enough time competing against MicroSerf complacency.

Alternative headline:

[GISboy]

Apple bites user. Says It'll do it again until s/he changes his/her (i)Tune...

Oh, wait! Isn't that how this happened in the first place?

Or

User loses data, Apple gets bit. (Just look at the Aqua apple on the front page...hehe)

It's not a meritocracy...

[The Infamous Grimace]

...it's a mediocracy.

(tig)

Possible fix?

[mr100percent]

Anyone having permission problems, or thinking they lost a hard drive: Logout, then login as ROOT. Hopefully you can now see all contents of your hard drive again. Go to your hard drive(s), right click and select Show Info. Go to Privileges in the drop down. Change the selection EVERYONE to READ & WRITE. Click Apply to All Enclosed Folders. Wait while it copies the privileges. Now logout and login as yourself. See if that resurected your data

See if that's it. I heard of it on Macnn.com forums. You may have a chance...

Whoa...

[BryanHughes]

I don't know whether to laugh or cry. Who the hell could write code that bad? Just one more reason to stay out of the OS X BSD arena and stick with a true winner, FreeBSD.

Re:Ahh... Imagine if this were a Microsoft product

[spitzak]

The bug only destroys partitions that are mounted. Unless the Linux partition can be mounted (can OS/X do this yet?) it would be untouched.

If I understand the bug right, if you have Itunes installed on a disk named "foo bar" (with a space in it) *and* you have a disk named "foo" (ie it matches the part before the space) the "foo" disk is wiped with the Unix rm -rf command.

Re:Geez...

[Darth Turbogeek]

It's a pity that the poster you were commenting to is actually 100% correct. If an employee is godd at their job, promoting to management is a stupid idea. Your taking them away from what they are valuable to. Thence, if I had a long serving Sys Admin, I'ld ather find some way to give a benefit to him for that, rather than the pseudo reward of a promotion to mamangement.

I personally have no other way to advance up the corporate tree. If I want to go up, it's management, yet that is something I suck at (Been there, done that, hated the t-shirt). I am, on the other hand good at most IT related jobs. Thence, I try to make myself too valuable to be shifted upwards. Unfortunantly, that marks me as "management material" in the eyes of the MBA's.

Good at Job, so promote him!

No thanks. I'd rather be in the computer room fixing servers than sitting at some f******* meeting. How about you pay for my ADSL / travel / car instead?

Besides, in the time of layoffs, who is the last person you want out the door - the manager who no one is really sure knows what they manage or the sys admin who knows the network absolutly backwards and just keeps everything running?

Re:how can microsoft top that?

[Glytch]

You apparently lack experience with NT.

iTunes.app IS a folder

[Anonymous Coward]

Yes, you should always quote variables in shell scripts, but given that iTunes.app is a single file, there is no point to the -r (recursive remove, for removing everything below a directory) option to rm.

But iTunes.app IS a folder, under MacOSX a .app is a folder with some extra flags, a .plist, and the actual executable code inside (among other things).

Re:iTunes.app IS a folder

[timsuth]

>But iTunes.app IS a folder, under MacOSX a .app

Ah, I stand corrected. Thank you.

NetManage beat Apple to the punch...

[thesteveco]

I believe it was version 4.1, round about '95, that had a beautiful "feature" in its mail application.

If you'd already deleted the contents of a users' mailbox at the file system, and then went to remove the user in the application the software would delete every file (but not subdir) in the C:\ root dir, after which it pruned ever file AND subdir one directory deep... lead to some fun tech support calls!

What a disaster... talk about poor QA!

a MS flaw

[Tyler-Durden255]

I believe there have been serious problems with the installer on OSX circa 1999, this is just the first one to hit the big time.

http://www.stepwise.com/Articles/Technical/Packa ge s/InstallerWoes.html

You weren't paying attention.

[jcr]

Oh, and I'd love to hear how making the package containing the burning framework a separate item would solve this problem. That would simply move the buggy pkg file to another place.

The problem wasn't the framework, it was a bug in the installer script. NeXTSTEP, and the first version of Mac OSX Server had an installer app, based on tar.gz files, that never exhibited this kind of problem.

Why Apple abandoned tar for pax, we're all still wondering. You can find a detailed explanation of exactly how Apple botched the installer system in Mac OS X on stepwise.com. Search the site for "pax".

-jcr

Re:You weren't paying attention.

[Apotsy]

The installer script was in the .pkg file, hence the words "buggy pkg file" in my above post.

You started this thread by saying iTunes shouldn't need an installer, but several people rightly pointed out why one is necessary. You then implied that putting the system-level frameworks iTunes depends on into a separate installer would somehow have prevented the problem, which it wouldn't.

You finally got something right by saying it was a bug in the install script, but you still don't seem to grasp the facts that 1) The install script is necessary and 2) Making it a separate install from iTunes would not prevent bugs in said install script.

It seems you're the one who wasn't paying attention.

minimized their collateral damage

[Pope]

You do realize that the phrase "collateral damage" is militaryspeak for killing civilians. I think you've been watching the news too much recently.

Re:The bug Spaces?

[SimCash]

that wouldn't have any problems with spaces

Gee, and feel so retro when I insist that "My working directory" should be spelled "my_working_directory".

IF the hard drive is partitioned...

[Magius_AR]

It seems Apple's new iTunes 2 installer deletes the contents of users' hard drives if the drives have been partitioned

As opposed to a hard drive with data that ISN'T partitioned? How's that work?

Ha! I'll show them.

C:\> fdisk
...
Delete primary partition? (y/n) y
Primary partition deleted.

Ah, all partitions gone. Finally my data is safe!

Magius_AR

research before opening your mounth

[axelbaker]

You are dead wrong. The EULA comes up while you are mounting the disk image. Before you even see the installer.

Power Tools can kill!

[dublin]

Ah, I see, you live by the "tough man" approach to programming. The fact is, people keep shooting themselves in the foot with this.

Some truths just stubbornly remain the same: Unix and its derivatives are great because they are power tools - but always remember, "Power tools can kill!" (If they couldn't, they wouldn't be power tools, anymore, now would they? For a great insight to this important aspect of the Unix philosophy, do a net search for Neal Stephenson's "In the beginning, there was the command line", and his quite apt comparison of Unix to the amazingly powerful Milwaukee HoleHawg, which is, as the saying of aviation goes, "terribly unforgiving of any carelessness, incapacity, or neglect."

Why is this on Slashdot?

[TheInternet]

Am I the only one that the fact this gets an entire topic to itself is a little weird. Maybe in a "quickies" or something...

- Scott

Even better

[mr100percent]

Call Apple, they'll reimburse you for the cost of norton utilities or a disk recovery.