Warm bodies are still smarter than silicon (When “the cloud” delivers a thunderstorm, part 2)

Recently, I posted about Dylan M and his sudden unexplained loss of his Google account. The aftermath of the story is given in a follow up to the previous article on Consumerist. While it is nice to see a happy ending, the truth as to why Dylan had his account locked in the first place is yet another cautionary tale about trusting cloud-based services.

Quoting Dylan as quoted by Consumerist:

I am a former art student and for the past year I have made my living as an artist. Three years ago I had been preparing a compilation of images to participate in an art show entitled “The Evolution of Sex” featuring a set of images, not my own, which I felt depicted the increasing violence and growing absurdity of pornography over the past 2000 years.

The image that they considered a violation of the Terms of Service is not among them and was more explicit, but it was created by the same photographer as the overtly suggestive last image, whose work is apparently well known and contentious for the obvious reasons of skirting legal boundaries.

Translation: this photog’s work pushes the line of what’s considered kiddie porn. (It would seem that Dylan’s account was flagged simply because he had a picture from a photographer known to test the limits, caught by an automated scan. I’ll get back to this point at the end.)

The only thing that is aggravating is that in the same folder they flagged, which was also titled “The Evolution of Sex,” are images of well known ancient Pompeii fertility statues, pre-historic examples including the Venus of Willendorf, a page from a French anti-pornography series from the 1800’s, one version of a common and well known advertisement that has been snuck into phonebooks nationwide since the 1950’s that is subversively pornographic (check your phonebook, or Snopes, it’s still very common in the UK), the cover from an issue of Rapeman, an infamous Japanese comic book about a superhero who rapes the wives of his enemies as retribution and can also be hired by corporations to rape the wives of thieving employees, and a picture of a vending machine on a street in Japan which claims to sell used young girls underwear.

Google employs an automated system to scan user storage for violations of their ToS and in the process erroneously flagged one of the images in the folder as child pornography… I am not angry at Google about this, as some might suggest… Google was unable to speak with me about it for legal reasons and it was Vic Gundotra who fast-tracked the appeal process once he learned of the situation through Twitter and personally investigated. When I asked him what would have happened had he not intervened he said the case would have gone through the regular appeals process and may have taken weeks to be sorted out.

Translation: We’re Google, we know what’s child porn better than you do, and even if we occasionally seem to be wrong, you can do without your Google account for a few weeks, right?

I can understand Google policing their servers for child pornography. However, I get the impression this was not reviewed by a human. It’s obvious that Dylan is not a pedophile, but an artist. It infringes upon Dylan’s free speech and free expression for Google to be “trigger happy” and assume that one picture from this photographer had to be kiddie porn based completely on that photog’s reputation.

So in some ways it’s worse than I thought. Put an image Google decides not to like on your Picasa account, and one could wind up losing one’s entire Google account, not just Picasa. I think that’s a bit too heavy handed and serves to underscore the need to make backups off of the cloud. There is something about plugging in a USB flash drive, copying data to it, taking it back out, and actually touching the physical medium one’s data is stored on. Maybe it’s just me, I don’t know.

Google definitely needs to find better ways to handle situations like this. I think just locking an account with no explanation is inexcusable. We should not have to do what Dylan did, should the same thing happen to us. At the very least, I will probably never use Picasa after reading about this, and will back up my Flickr and other photo service accounts on a more regular basis from now on.

When “the cloud” delivers a thunderstorm

(NOTE: in the time it took me to get this post ready for publication, Dylan did get his access restored. I will be following up with the conclusion of the story, which raises more points for discussion than I wanted to add to this post.)

I have been cautious with regard to the new wave of “cloud-based” services. I keep backups of all my data on my own media. And I’m sure some of you laugh at me as old-fashioned. Well, wait until you read this story.

A recent post to Consumerist tells the story of Dylan M. (he is identified only by first name in the article; the last initial is on his Twitter account) and the sudden deactivation of his Google account.

Dylan was a happy user of Google’s services for the last seven years, until 2011 July 15 when he found his Google account was deactivated. Dylan has lost “approximately 7 years of correspondence, over 4,800 photographs and videos, my Google Voice messages, over 500 articles saved to my Google Reader account for scholarship purposes… all of my bookmarks… over 200 contacts… my Docs account… my Calendar access… [which includes] not only my own personal calendar of doctor’s appointments, meetings, and various other dates, but I have also lost collaborative calendars, of which I was the creator and of which several man hours were put into creating… saved maps and travel history… my website, a [B]logger account for which I purchased the domain through Google and designed myself” according to his Twitlonger post.

Dylan goes on to write he has been a loyal fan of Google, encouraging the company he works for to use Google Business Apps and purchase storage with Picasa, and encourgaing his friends and family to open a Google or Gmail account and use Google’s Chrome browser. He also goes on to slam Google for behaving in such an abusive, monopolistic fashion.

Note that Dylan went to Twitter to air his greivances? Remember that at one point rumor was that Google was about to buy Twitter? This incident is a prime example of why such an acquisition would have been a disaster for the computing public. Thankfully, it didn’t happen.

So, how do we protect ourselves from what happened to Dylan? Back up your data to a storage medium you physically control, whatever that may be. Keep multiple copies of things that are truly important. Back up everything as though Google’s (or Microsoft’s, Apple’s, etc.) datacenters will lose everything for everyone (or at least everything for you) sometime in the next week.

At the very least, it is a bad idea to trust one company (such as Google in Dylan’s case) with everything. I’m not sure if one can, for example, run both the Delicious and Google Bookmark plugins without fear of conflict. I know that a Gmail account can be accessed via IMAP and backed up using tools such as archivemail for Ubuntu. I don’t use Gmail for truly important email, but if I did, I would back it up with archivemail --copy --all and the appropriate URL and other switches.

If I am ever a significant adopter of cloud-based services at all, it will be a relatively late adoption. With the ubiquity of USB flash drives which can hold upwards of 4 gigabytes being easily affordable (some, such as this one made by LaCie even resemble a door/car key and can easily be carried on one’s existing keyring), I see no reason to put important data “on a server somewhere” which can go down when I least expect it. I have known of exactly two USB flash drives to fail during the timeframe I have used the technology; one (mine) was because I used it like a small hard drive for an Ubuntu install (it actually lasted for almost a year, though it did corrupt quite a bit of data during that time), and the other (my mom’s) failed due to a defective USB hub, apparently melting something plastic on the connector to the point where it won’t even insert into a USB port. My first USB flash drive, a 32 MB Memorex model (which at the latest probably dates from 2005), still works and has been used for everything from moving small quantities of documents to a boot medium as recently as this year.

That said, I still recommend optical discs such as CD-R, DVD-R, BD-R (recordable Blu-Ray), etc for long-term archival of data, particuarly data that should definitely not be changed after it has been written such as legal documents.

As noted above, Dylan did get his Google account back. However, the circumstances under which it was shut down deserve a rant of their own.

[To be continued…]

A strike against the “music industry”

A recent osnews.com story chronicles Google’s troubles with its music store project. From the article:

Sadly, however, rumour has it the negotiations with the music industry have been so frustrating, Google is contemplating abandoning the entire project altogether. This has led some to wonder – why doesn’t Google, or a consortium of technology companies, just buy the music industry outright?

The talks with the music industry to set all this up have apparently been so frustrating for Google that they are now contemplating shutting the entire thing down. Wayne Rosso, who once ran Grokster, claims that sources familiar with the matter stated that Google is “disgusted” with the music labels, and is considering options ranging from just following Amazon’s lead and not seek the labels’ consent, to just shelving the entire project.

If you’re a regular reader, you’ll know I’m not Google’s biggest fan. I’ll admit they have done some good, but they’ve also done their share of things that are at least dubious if not outright evil. A music store project of the sort Google is planning, however, definitely falls in the former category, and it’s a shame that the music cartel is stonewalling it. A sticking point of the negotiations is that Google wants the service to be free of charge for the first 500 tracks; the music cartel (Warner Music Group in particular) wants a $30 per year fee.

I’m not sure what the “music industry” (who I well henceforth refer to as the music cartel) is trying to prove by stonewalling Google; it only makes them look like the fools they are. Today, artists can sign up with Bandcamp, Magnatune, CD Baby, among others. Dozens of companies exist to press “short runs” of CDs, or the artist can burn copies to CD-R. The music cartel we grew up with, where artists relied on record companies to get copies of their music in the hands of the public, has changed radically in the Internet era. In fact, physical copies are now closer to becoming the exception rather than the rule. (In a way I’m not a fan of this, as I prefer a physical CD to a digital download in most cases, but that’s another rant for another day.)

Remeber, the same music cartel that charges $20 for a CD, and gives the artist that actually made the music around $1 on it, is the same music cartel trying to stonewall Google’s music store. I’m not surprised. It took the music cartel (specifically the RIAA) most of a decade to figure out that suing their customers for alleged copyright infringement was a lousy business plan. That hasn’t stopped them from trying the same shenanigans in other countries; France has a draconian “three strikes” law (the HADOPI law) with regards to unauthorized copying (though it’s possible this is more the MPAA’s doing).

Google appears to be following Amazon’s lead, which is to offer the service first, then clear up any legal problems later. It’s hard to blame them; Google epitomizes “filthy rich corporation” at this point in time. But it’s also just a bit frightening to me that Google can do this. It’s as if we have The Rules The Rest Of Us Must Play By on on side, and The Rules Google Gets To Play By on the other. Could any of you out there just open your wallet if you were facing a lawsuit by the RIAA?

Google prefers serving its own dog food at the users’ expense

Even though this is over a month old, this is a story I really wanted to discuss.

Techcrunch covered a story in 2010 December about Google and its increasing tendency to serve up links to its own properties in search results, which is particularly noticeable in regards to searches for local businesses (such as the example search for “ny chiropractor”).

And then Google responds with a post entitled “Local Search: It’s all about the best answers for users.” What nerve! The title of the TechCrunch article sums it up rather nicely: “Google, These Aren’t Really The Best Answers For Users. They Are The Best Answers For You.”

Honestly, if this is the future of web search, then it’s time for someone to set up a company with the same role as Google, but structured as a non-profit. It’s obvious to me that with self-favoritism such as this, the desire to make a profit off of its own properties is getting in the way of Google’s nominal mission of access to information. Remember “Don’t be evil?” Profit at the expense of delivering the best user experience qualifies as evil in my book.

The companies behind sites like Citysearch, Yelp, TripAdvisor, etc are outraged at this. And rightfully so. Google is dangerously close to acting like a monopoly–and we know what the law says about monopolies. In fact, I would not be surprised, were this to continue, if it resulted in a Department of Justice investigation.

If one happened, it wouldn’t come a moment too soon.

At your own risk, indeed

Mashable recently reported on a Google Maps user that was injured while following walking directions from the site, and later sued Google and both the driver of the vehicle that hit her.

Lauren Rosenberg used her Blackberry to get this set of directions in Park City, UT which include an 0.6 mile stretch on Deer Valley Drive, also known as highway 224. The Blackberry’s Google Maps application apparently does not give the standard warning about walking directions being in beta, which is the focus of Lauren’s lawsuit against Google.

I would imagine the Blackberry application also does not show the choice of two to four different routes, as the third choice as of the time I ran the directions uses Park Avenue for most of the distance and only requires one to walk 279 feet (or just over 1/20 of a mile) down the apparently pedestrian-hostile Deer Valley Drive. There is no route that completely avoids Deer Valley Drive (every street in the area which intersects Deer Valley Drive does not go through to the other side, probably due to geography). The Google Street View pictures show a speed limit of 35 mph for this stretch of the road, probably about the only pedestrian-friendly thing about it.

I’m not sure the lawsuit against Google is entirely deserved; the directions are about the best that can be managed, though Google was sloppy to not include the warning in the Blackberry app. There is no question that the driver who hit Lauren deserves to be sued, of course. However, I am wondering why the government responsible for maintaining the roads isn’t being sued as well, as the roads could be made somewhat more pedestrian friendly than they are. I understand the limits of geography and our motor-vehicle-centric society (and thus the assumption that “everyone has a car” in a small town like this). However, cars do break down, and sometimes the only way to get to the auto parts store is on foot. Sometimes civil engineers forget this simple fact of life.