No, there is no iPod virus

Thursday, 2007-04-05; 22:27:00



Articles like this one over at ArsTechnica piss me off.

So. Much.

What's the impression you get from a title that says "Lab claims first 'proof of concept' iPod virus"? That there's a demonstrated and effective 'proof of concept' virus for the iPod, right? That, if someone malicious were to get their hands on this 'proof of concept', they could potentially cause headaches for iPod users, right? That this "virus" probably matches up with the typical definition of "virus", right?

Only you'd be wrong.

Let's go through the facts: first, this "virus" does not work on stock iPods: you have to have iPodLinux installed. So, not only do you have to install an entirely different operating system on the iPod (the only purpose for which would be shits and giggles), but iPodLinux does not work on all iPods. (Correction: there is partial support for iPodLinux on iPods other than 1-3G, except the 2G nano and any iPod shuffle since they don't have a screen.) That means that the virus only works on 1st, 2nd, or 3rd generation iPods. Not the latest 5G iPod videos, not the 2G iPod nano, not the 2G iPod shuffle, not even 1G iPod shuffles or nanos or even the discontinued iPod mini. So basically, if you bought your iPod in the past in the past 2.5 years, you're safe from this "virus".

But wait! It gets better! You have to manually copy the virus onto the hard drive of your iPod before it can do anything! What does that mean? There's no vector for self-propagation. Once on an iPod it can affect the entire iPod by infecting all .elf files on the disk, but it's completely harmless to other iPods.

Oh, right, and then you have to manually launch the virus anyway.

Then Ms. Cheng over at ArsTechnica reports on this "virus" while maintaining a straight face. There's no bullshit-calling on Kaspersky Lab for creating something that doesn't work on 99.9999% of all iPods. There's no bullshit-calling for calling it an "iPod virus" when, for all intents and purposes, it doesn't infect iPods. If that were true, it would affect the operating system that is installed by default on the iPod. When you encounter a virus on Microsoft Windows, do you call it a "Dell virus" or an "Acer virus" or an "Alienware virus"? If you do, you're just dumb. It's called a "Windows virus" for a reason. These articles could reasonably called sensational even if they called it the first "iPodLinux" virus, just for having the word "iPod" in the description; this is, at best, a "Linux virus that just so happens to work on iPods if Linux is installed on them, which happens to be exactly sixteen out of the hundred million iPods sold".

But that wouldn't be very attention-getting, would it?

Now, see, I could care less if the lame-ass Kaspersky Lab issued a press release about it. But news sites picked up on it, because whenever something that Apple makes is infected by a virus, it's instant news. CNET and Help Net Security are two websites that published articles on the "virus". No I won't link to them because they don't deserve the hits for these lame-ass articles. And I really couldn't care less about them anyway because I expect it from sites like this.

But it's sites like ArsTechnica that piss me off when they print it. As I said above, there was no bullshit-calling at all on Kaspersky Lab. Kaspersky Lab is just trolling for hits, just as Maynor and Ellch were when they released "details" about a "hack" that "affected" "stock" "wireless cards" in MacBooks. (It is of no consequence that they demonstrated an actual flaw a year later, one which Apple already fixed. They trolled for hits, and they damaged their reputation. And thanks to many of the Mac weblogs out there, they got called out, and they got a bucketload of shit dumped on them.) Ms. Cheng seems reasonably intelligent, from the previous reviews and articles she's written. And ArsTechnica on the whole seems to be composed of writers that are usually more well-versed in technology than sites like CNET. But she can't see through this smokescreen?

I mean, seriously. I can write a virus that's just as bad as Kaspersky Labs'. I'll even show you how to do it. Open TextEdit. Type "#!/bin/bash". Press return. Type "rm -rf ~" Save the file as "super-cool-virus.command" onto your iPod's hard disk. Open Terminal. Type in "chmod +x ", drag the .command file you just created to the Terminal window, press return. Double-click the .command file on your iPod. OH MY GOD YOU JUST DELETED ALL THE FILES IN YOUR HOME FOLDER.

Are you shitting me?

NEWS FLASH: Journalism is not re-hashing press releases. That's not even blogging. That's called "being a shill".



[Full disclosure before continuing: about a month back, when ArsTechnica was looking for more people to write for their "Infinite Loop" journal, I applied for one of the positions. They were looking for people who could commit to 40 articles a month; I included a letter of intent, proposing to do 30 articles a month by covering some of the discussions in the Mac weblog community that aren't usually on the radar of Infinite Loop. I was not hired.]

ArsTechnica's "Infinite Loop" journal has seriously declined in quality in just the past few months. This started before they announced they were looking for more writers; Infinite Loop had already reached critical mass in covering most newsworthy Mac happenings, in my opinion. Some of the articles were on things that most people really didn't care about, and weren't really interesting. That was part of the impetus behind including a letter of intent with my application.

After hiring the new writers, though, the quality went off a cliff and is in the process of free-falling. It seems to me that Infinite Loop is still trying to only report on "major" news, which means that with the addition of three new writers, they're really scraping the bottom of the barrel. One new writer in particular, Iljitsch van Beijnum, seems to write articles on the most inane issues, and also seems to lack the knowledge to actually competently report on them. For example, he spends a complete article reporting on how the BBC reported that some guy, Kevin Finisterre, says that Mac users are lax on security. Supposedly he's some authority on the attitudes of Mac users just because he was the instigator of the Month of Apple Bugs project. (Don't get me started on that one: a full 7 of the 31 bugs affected third-party applications, and the last bug was vaporware.) This is just like the Kaspersky Lab situation: why the hell do we care what he thinks about the "attitudes" of Mac users with regards to security?

He also wrote about "The iPhoto-disk space conspiracy", which basically amounts to, "OH MY GOD iPhoto saves the originals of your photos so even if you edit your photos and save your edits you can always revert to the original! Oh, yeah, and iPhoto supports a ton of languages which explains why it takes up so much disk space!" And he includes such gems of statements as, "We're supposed to think that's so the application can restore your photo to its original state after making unfortunate edits, but we all know the real reason, don't we?" and "There you have it, ladies and gentlemen of the jury. The facts speak for themselves. Or could it be that Apple's software engineers are too busy working on Leopard's secret new features to save a few hundred megabytes here and a gigabyte there? No, that can't be it..."

Yes, Iljitsch, the facts speak from themselves. You're reporting on a non-issue.

And Iljitsch's crowning achievement in reporting is one in which he demonstrates that he doesn't know what the hell he's talking about with regards to Objective-C 2.0 garbage collection and Core Animation. The article has been updated, but in a nutshell, he thought that "when a program renders some complex graphic in multiple stages, it makes sense for the system to hold on to the intermediate images even if those aren't needed right now" and that somehow garbage collection did that. It does nothing of the sort: all it does is periodically check whether a certain class instance residing in memory has anything pointing to it, and if not, it frees up the memory that was being used. So, as a Cocoa programmer, you no longer have to worry about retain counts.

What's funny is that in the updated article, Iljitsch says that "he overlooked something"; he did nothing of the sort -- he didn't research the topic, he didn't understand it, and he tried to report on it thinking that he would be able to pass his supposed "knowledge" by the Mac community. Even John Siracusa took the time to leave a comment pointing out the gross inaccuracies in the article.

I don't necessarily blame Iljitsch for his crappy articles, even though it would be nice if he realized that he seems to be in over his head. But with so many writers on Infinite Loop, it's apparent that they have too many writers and too little material about which to write, which is why they're reporting on non-issues and sensational press releases about fictional functioning viruses.

And that's a shame, 'cause ArsTechnica is usually one of the sites with a higher standard of writing.


Technological Supernova   Rants   Older   Newer   Post a Comment