Sorry for lack of posts, podcasts, etc….getting back into the swing of things after vacation and such has been a little tough. Plus, I have been working on backend code stuff, and I end up tired and unwilling to spend the time at the end of the night to work up a good post. That being said, I have noticed that having a unpublished flag for entries would help give me time to edit, prep, etc. Yes, that’s right…I don’t even have a non-public attribute on these posts. As soon as they’re in the db, they’re available! If I edit after first commit, it’s live in front of Google and everyone!

I’ve been dallying around with hacking a flag onto the posts table, but that felt kinda icky. Then I just finally realized today that Django has an excellent permissions system! Duh. I may poke it with a stick. It might be overkill…we’ll see. If nothing else, it’ll give me some experience with it.

I have also been experimenting with Twitter and tagged del.icio.us posts for lightweight updates of present state (twitter) and notable urls that I don’t have much else to say about (del.icio.us “toblog” tag). I’m thinking of actually importing that data into a blog post, though, rather than just pointing to it; I’ll check those feeds in the early morning hours, get any new data, and create a blogpost with it via a script. That way, I can save posting for slightly meatier thoughts.

Just a quick note to sock away my code snippet, and hopefully help others. When working on Django development, the SQLite backend is awesome; it’s simple, self-contained, and easy to carry about (I carry apps on my USB fob). But be careful with datatypes; datetimes in particular (at least for me).

SQLite (v2) stores everything as text under the covers, even though is accepts typed column declaration (it keeps track of the datatype you declare, but doesn’t do anything with it really. v3 adds some typing possibilites, but I don’t think Django leverages this. Consider everything a string.). So you can insert an invalid datetime fragment via Django code (this is most common when you’re calculating it yourself, rather than using the form widgets), and you may not realize anything is awry until you try to pull data back out. You’ll start to get unpack errors and such, as the Django SQLite datetime parsing code fails. After a bit of munging about, I tracked down out the incantation to take a 9-tuple time struct and pop it properly into a SQLite datetimefield:

if item.has_key('date'):
    foo = item.get('date_parsed') # date_parsed is a 9-tuple 
    pc_date = datetime.datetime(*foo[:6]) # this generates the correct input type

Hopefully I can save someone a bit of pain. *grin*

Holy…crap. I knew about Universal Feed Parser, but it slipped my mind, and spent two days reimplementing about 20% of what it does (the easy 20%, too). Do yourself a favor…if you’re hacking on feeds (RSS or Atom) in Python, just go get it and import feedparser. You’ll be glad you did.

We’re back! Our London holiday was great…it’s an awesome city. So much to do, so much to see. It’s fairly rare that we pull a repeat, but London definitely falls into that category. Perhaps as a springboard to the countryside…we spend a day in Sussex, walking some trails, and it was beautiful.

Pics on the way. (I’ll update that link specifically to the vacation photos as soon as they’re all organized.)

GNUCITIZEN included a link today to a Discovery Channel video over on Google Video: 2057: The City. GNUCITIZEN’s comments are worth reviewing: he’s a security researcher, and understands how powerful and fragile our increasingly computer-mediated reality will become. The video is your typical Discovery Channel; half future vision, half present-day documentary on precursor tech. But not bad, though the gung-ho surveillance stuff gives me the same queasy stomach feeling that it always does. Time to re-read The Transparent Society!

UPDATE: Ugh…it got cheesier as I went along, but still, I’ve seen worse. Apparently there’s an entire 2057 series of shows; City, Body, World, etc.

Holeee frack.

Any Singularitarian out there knows that brain simulations are a big indicator of major steps towards the Big Discontinuity; once human beings are running on silicon, at 100X bio-speed, something weird and wonderful should result post-haste.

So brain simulations are always interesting, and some IBM researchers have taken a functional model of half a mouse brain (8,000,000 neurons, 6,300 synapses per neuron), and run it on a BlueGene/L supercomputer. 1 second of in-model time took 10 seconds of computer time to crunch…performance, of course, will only improve.

Jamais Cascio has some interesting thoughts on the matter…for example, what happens when a mouse brain can be simulated at FASTER than real-time? 1000X faster? Does something emerge? Fascinating to consider.

FORA.tv is pretty interesting…they are focusing on recording and distributing the audio and video of discussion in forums like C-SPAN, the EFF, the Hoover Instituion, and the Cato Institute (just to name a few). Included are tools to build RSS feeds based on your interests, discussion boards, etc. If I can get a good podcasting feed set up, I might become an avid user; I’m kind of a wonk when it comes to these things.

UPDATE: URLs are a bit of a pain to share; they should make it easier. A couple of example videos: Neil Gaiman at Cody’s Books, Oct 2 2006, and Kevin Phillips discusses American Theocracy: The Peril and Politics of Radical Religion, Oil, and Borrowed Money in the 21st Century at the Commonwealth Club of California on April 18, 2007.

I had NO idea DIY rapid prototypers had gotten this far: RepRap — RepRap is short for Replicating Rapid-prototyper. It is a practical self-copying 3D printer…The RepRap build cost will be less than $400 US for the bought-in materials, all of which have been selected to be as widely available everywhere in the world as possible. We hope to announce self-replication in 2008.

Yowza. We are close. This I have to try.

I’ve long been a fan of the Skepticality podcast, and I cheered when they became the official podcast of Skeptic magazine. They just keep getting better; episode 49 is a must listen.

Dr. Shermer talks to author and Stanford emeritus Professor Philip Zimbardo on the release of his new book, The Lucifer Effect which ponders the question, what causes some people to commit acts of evil while others become unexpected heroes.

Anyone with an interest in humanism, the nature of evil, and how good people are convinced to do bad things should definitely grab this episode. Dr. Zimbardo (conductor of the infamous Stanford prison experiment) provides scientific theories that affirm and reinforce intuitions I have about human nature; I’ve put The Lucifer Effect on my booklist!

Here is the blip.tv page, and direct MP3 download link for the March 30th episode.

I talk about the first chapter of Yochai Benkler’s Wealth of Networks. Summarizing the arguments out loud should help me to better integrate the points into my worldview, and might give you a decent overview of Benkler’s work. (Though I don’t do it justice…do yourself a wonderful service and read it yourself!)

I haven’t re-tracked down the vlog survey link yet…I’ll add it when I find it!