Both ends of this are wrong, at least in the short to medium term: our data's not that accessible, and most things you call the help desk for are not Googlable (if that's a word). Things like JPL's internal policies and procedures, for example -- we have an internal, Google-based search engine, but it's not able to find everything by a long shot.

Also, as it happens, our help desk is very good -- even if it is run by Lockheed Martin -- and it would actually be a shame to see it go away. This might change someday, but right now, humans are irreplaceable.

Here in North Dakota, we only have three seasons:

• Winter
• Flood
• Construction

Winter lasts 7-8 months, of course. Flood and Construction can overlap. If there's no snow on the ground, much of the state feels a deep sense of uneasiness arising from chronological uncertainty.

These LEGO figurines look awesome! My hat's off to the JUNO team!

We sent (flattened) LEGO figurines aboard the MER landers, too (not on the rovers proper). Their names were Biff Starling and Sandy Moondust, if I recall correctly. See my blog for a terrific color picture of Spirit's LEGO figurine before she drove away.

The MER landing system was pretty damn complicated as well -- you shouldn't think otherwise just because it worked perfectly, twice. (And the late add of the DIMES system, which used images taken during descent to make decisions about rocket firing, made things even more complex but probably saved Spirit.)

I know some of the engineers working on the MSL landing system, and they're well aware that it all needs to work properly. They're also terrifically smart, the kind of people who can make this work if anyone can.

Do I think it's a guaranteed success? Of course not: guaranteed successes don't happen in this business. But I also recognize that it's not more complex than it needs to be, to solve this very hard problem.

... we'll have sufficient bandwidth that video shot from a mobile device can be uploaded straight to the web, with only a brief "buffering" stop on the actual filming device. Then they can confiscate the device as much as they like, but the video will be beyond their grasp due to the technical difficulty of 1) figuring out where it went, 2) getting the host to take it down, and 3) doing so before the original filmer (or friends) can spread copies of it all over everywhere.

Shortly after that, some bright lad will suggest jamming devices to disrupt the transmission, which will pose all kinds of problems for them, such as disrupting their own signals. So then they may try short range hand-held EMP devices, which will work great right up until they fry somebody's pacemaker. Meanwhile, people will busily be miniaturizing the technology even further, so that observers could be filming the cops' activity without any obvious sign of it.

And eventually they'll give up and conclude that they'll just have to put up with being filmed by whoever happens to be standing around.

Ah, technology.

(Yeesh, let's try that again.)

Here's a remix of that xkcd strip that the MER team liked better than the original: http://twitpic.com/52etk7.

Goodbye, Spirit, love, and well done.

What should simply happen is that universities should publish their own journals, online, using the simple, cheap web distribution methods.

Interestingly The University of Michigan is doing exactly that. They combined their university press with their library, and shifted the press focus to digital instead of print. As their site says, "The press mission is to use the best emerging digital technology to disseminate such information as freely and widely as possible while preserving the integrity of published scholarship." They also do some print-on-demand stuff for people who want paper copies.

I hope to see this approach adopted by a great many more universities. It cuts profiteering parasitic publishers out of the loop, and simultaneously reinvigorates the university library by expanding its mission to include publication and dissemination of new research in addition to the more traditional roles of archiving existing materials.

If you haven't already, check your university library's holdings for those. They have likely already paid a great deal of money to crappy parasitic publishers for access to databases full of journals, and you have to check through the library web site. Your library's holdings usually won't show up in Google results, again because the crappy parasitic publishers don't let Google index the library's licensed databases. Pro tip: look for a "journal title" search or something similar, and search for the title of the publication first, then narrow down to the specific issue/volume containing the article you want.

Gimme a break; I was twelve, and had not yet heard of open standards. I just used the software that came on the computer. Now I'm a web services librarian. I write software too, and I sing the open standards gospel daily.

ASCII is great, though I'd actually prefer UTF, thank you, on the grounds that diacritics actually do matter, not to mention the ability to encode things in Cyrillic, Korean, or Scandinavian runes. Though even UTF has its limits. Let me know when you work out a way to store NTSC format video encoded in some damn proprietary codec as text, okay? Or, for that matter, video games, which are literary and artistic works worthy of preservation.

The simple fact is, computers are inherently more complex than older information storage methods. The information they store cannot be read directly by a human. Unless you can hold a hard drive to your head and sense the magnetic charges directly, the information must be interpreted by software first. That simple, undeniable requirement adds several layers of complexity to any attempt at long-term preservation of digital data. For ample demonstration, just go read Keeping Stuff, a delightful essay by a comp sci professor at Grinnell in which he discusses his attempts to preserve his own undergraduate work from the early '70s.

Oh, and you can wag your finger at me some more as soon as you've worked out an open standards solution to the fact that basically every non-geek does all their work in proprietary programs that spit out crappy proprietary files, and then expect them to last forever.

As a librarian, this makes my head hurt.

I guarantee you, those books are sitting on a shelf in a library someplace. Probably within a thousand miles of you. And we have this lovely thing called "interlibrary loan," an arrangement under which you can walk into your local library and borrow those books from the library that has them, either for free or for a small processing fee, depending on how badly the library's budget has fared in recent cuts. We saved those books for you. That's what we do. Please come borrow them.

As for the future, well, digital copies are actually a LOT harder to preserve long term. I myself have files that I can no longer open, because I no longer have a copy of the word processor "Sprint" running on MS-DOS 5.0. They're less than twenty years old, and are essentially unusable.

By contrast, I once held and read a hand-written breviary from fourteenth century Italy, a good six and half centuries old and still usable. If we could find a way to archive digital information which would guarantee its usability a mere century from now, I'd rest a lot more easily.