James Tauber

journeyman of some

blog > 2007 >

James Tauber's Blog 2007/08

Quisition Updates

My flashcard site, Quisition has now been moved over to WebFaction and I took the opportunity to clean up some code and make some incremental feature enhancements.

You can read more about them on the Quisition site. One new feature is the ability to reset long overdue cards. So if you tried out Quisition a while ago and would like to come back without a huge backlog of old cards, this feature is for you!

I've also implemented an alpha version of the testing parts of the site specifically styled for the iPhone. If you're interested in trying it out, email me and I'll tell you how to access it.

by jtauber : Created on Aug. 27, 2007 : Last modified Aug. 27, 2007 : Categories quisition iphone : 6 comments (permalink)

Speech Accent Archive

An archive of (at the time of writing) 795 people from around the world, reading the same passage, many entries including a narrow phonetic transcription:

http://accent.gmu.edu/

Oddly enough, I found out about it watching an interview of the lovely Zooey Deschanel where she cites it as the latest website she bookmarked.

by jtauber : Created on Aug. 25, 2007 : Last modified Aug. 25, 2007 : Categories linguistics : 1 comment (permalink)

Stack-Type Vectors, Part II

Previously, we introduced the stack-type vector and showed the identity vector, how inverses are formed and how these stacks are scaled. We haven't yet show how they are added.

Consider the two stack-type vectors marked (a) and (b) in the diagram below:

The geometric way stacks are added is to place them on top of each other and draw lines through each point of intersection, as we have done at (c). The resulting lines, show by themselves in (d) are the resultant addition. (Remember the number of lines drawn doesn't matter, it's how dense they are that determines the "magnitude" of the vector)

With a bit of experimentation, you can see that this definition of addition (in the limit as the lines become closer and closer to parallel) fulfills the requirement that v + v = 2v. In fact, you can establish that all the axioms for linear spaces hold true.

So we have so far seen three different types of linear spaces: one type made up of tuples of numbers, one type made up of arrow-type vectors and one type made up of stack-type vectors. The latter two have the special characteristic that they are geometric: they have some meaning in the context of a manifold.

Because physics models phenomena in the physical world (which can be viewed as a manifold) the vectors used in physics are of this geometric type. Displacement, velocity and acceleration are examples of measurements that can be models as arrow-type vectors. But what about stack-type vectors? We'll get to their use in physics in the next Poincaré Project entry.

by jtauber : Created on Aug. 23, 2007 : Last modified Aug. 23, 2007 : Categories poincare_project : 0 comments (permalink)

Trailing Slashes

The biggest mistake I made in Leonardo was making "foo" and "foo/" mean the same thing. I don't mean directing one to the other with a 301, I mean returning the same content at two different URLs. If you add the fact I listened on both jtauber.com and www.jtauber.com, it meant that every resource had 4 distinct URLs.

That's now fixed: www.jtauber.com redirects to jtauber.com and "foo" is redirected to "foo/".

Unfortunately, most of the del.icio.us bookmark counts I include at the bottom of each page are now effectively reset to 0. While 164 people had bookmarked "poincare_project", none had bookmarked "poincare_project/". Even though the former now redirects to the latter, I've effectively split the vote between past and future bookmarking of that page.

As far as I know, del.icio.us never heeds 301s and updates its database accordingly. Google Reader doesn't seem to either, judging by the continual checking of "/atom/full" in addition to "/atom/full/".

by jtauber : Created on Aug. 23, 2007 : Last modified Aug. 23, 2007 : Categories django leonardo this_site : 12 comments (permalink)

A Very Short Introduction

Via Greg Mankiw, I just found out about Oxford's Very Short Introductions series. Greg has high praise for Mathematics: A Very Short Introduction. When I looked it up on Amazon, I found numerous other books in the series recommended to me: Particle Physics, Cosmology, Music, Economics, Theology, The Brain. There were about 20 that immediately sounded interesting.

As I explored more I found more and more books in the series. Oxford's Very Short Introductions site indicates over 150! Damn, almost every single one of them looks interesting.

As well as individually, they sell them in boxed sets like the "Basics Box" (Philosophy, Maths, History, Politics, Psychology), the "Brain Box" (Evolution, Consciousness, Intelligence, Cosmology, Quantum Theory) and the "Thought Box" (Hegel, Marx, Nietzsche, Schopenhauer, Kierkegaard).

As is generally the case with these sorts of series, I'm guessing the quality varies a lot. Still, it's hard for me to resist buying a large proportion of them right now :-)

by jtauber : Created on Aug. 20, 2007 : Last modified Aug. 20, 2007 : Categories books : 2 comments (permalink)

jtauber.com Switches to Django

I've spent the last couple of days doing a crude port of Leonardo over to Django including a converter from the Leonardo File System.

I apologize in advance for any problems with the new system.

Functionally, it should already be better than the old one. In particular, categories are handled in a better way which, amongst other things, means that you can navigate through all the pages in a particular category with a previous and next link (see the categories in this post for an example) and category pages list entries with a last modification date and number of comments. Also there is a path of links at the top of any page that isn't at the top level.

I still need to clean up some of the content in light of changes like this so excuse the construction work for the next few days.

Lots more new features planned though, which is one of the reasons I did the port in the first place.

by jtauber : Created on Aug. 19, 2007 : Last modified Aug. 19, 2007 : Categories django announcements leonardo this_site : 0 comments (permalink)

Thoughts on the Social Graph, DataLibre and Aggregation versus Hosting

The recent activity stemming from Brad Fitzpatrick's Thoughts on the Social Graph reminds me of something I was thinking a lot about back in 2004. See especially Mark Atwood's post Five billion social network sites, each about one person on the Social Network Portability list.

The idea I had three years ago was the notion of people hosting their own data and the social networking sites merely being aggregators.

Actually, the way it occurred to me first was in the context of DOAP and the next Advogato but I then generalized the idea in Aggregation Versus Hosting.

Steve Mallett started a mailing list and a website about this same idea, calling it DataLibre. Unfortunately the site no longer exists but here is an announcement of the original site: http://www.oreillynet.com/ruby/blog/2004/09/datalibre_update.html

Steve really drove the vision but I posted numerous thoughts on the idea. Because my blog didn't have categories at the time and I haven't gone back and categorized my old stuff here's a list of my relevant posts:

Brad's article (which is a lot more concrete than anything I ever said) seems to be rejuvenating discussion about this sort of thing. See, for example, the post More on social network portability on Plaxo's blog.

by jtauber : Created on Aug. 18, 2007 : Last modified Aug. 18, 2007 : 0 comments (permalink)

Blog Stats

I've posted here every day for the last 10 days straight. I wondered if I'd had such a writing streak before, so I write a quick program that generated a bunch of stats:

(not counting this post...)

  • I've blogged 721 entries
  • I've never missed a month
  • Worst months were August 2006 and April 2007 both with only three posts the entire month
  • Best month was March 2005 with 47 posts
  • Most number of posts on a single day was six on 14th March 2005 (pi day!)
  • Of 477 blogging days, 303 had one post, 118 two posts, 45 three posts and 11 four or more.

and...

by jtauber : Created on Aug. 18, 2007 : Last modified Aug. 18, 2007 : Categories this_site : 0 comments (permalink)

List-Unsubscribe

I tend to swing back and forth with my mailing list subscription habits. The two endpoints of the pendulum's path are:

  • I rarely read this list, but a have a rule that puts it in its own folder so it's always there if I ever do want to read it
  • I rarely read this list and it's cluttering my mailer with 10,000 unread messages so I'll unsubscribe

I'm currently on a swing to the second one.

In one case, I wasn't immediately sure how to unsubscribe from the list, so I looked at the raw headers for the List-Unsubscribe field (described in RFC2369). Sure enough, it was there.

It would be really nice if more mail apps supported this. Okay, it would be really nice if Mail.app supported this :-)

by jtauber : Created on Aug. 18, 2007 : Last modified Aug. 18, 2007 : 3 comments (permalink)

A Lot Many People Talk Like This

I work with a number of people from Maharashtra and I've noticed they all say "a lot many" where I would say "a lot of". I wonder if this is based on a similar construction in Marathi or Hindi or if there's some other reason (analogy with "a good many", for example).

A Google search for "a lot many" (with quotes) shows numerous examples (if you ignore the cases where punctuation separates "a lot" and "many" — I wish you could tell Google not to count those cases).

Anyone have any insight?

UPDATE: I sat down with one of my colleagues and I understand now. In some dialects of Marathi (especially around Pune, for example), the "proper" way to express a large quantity is not just khup but khup sare (खुप सारे). In some regions in Maharashtra, it might be fine to just say khup but the use of both words together is preferred by those that consider the Pune dialect (and those similar) the more "pure". A literal translation to English would thus be something like "a lot many".

by jtauber : Created on Aug. 17, 2007 : Last modified Aug. 17, 2007 : Categories linguistic_observations : 3 comments (permalink)

PGR2 Maps

A few years ago, my dad won an XBox as a door prize at some event and, for his next birthday, we got him Project Gotham Racing 2 (PGR2). I played it a fair bit too and absolutely loved it.

When I was in Florence at the end of last year, I confess I actually recognized some of the roads from PGR2. I went looking online to see if someone had mapped out all the courses on Google maps but couldn't find anything.

There are some screenshots on Flickr from Bizarre Creations showing the course models. For example at http://flickr.com/photos/bizarrecreations/304326348/ but nothing to really tie the course to the real place.

This evening I found http://www.moondookie.org/pgr2tracks.html which has 12 of the courses drawn over satellite images. It would be nice to have a more complete set, though, and in machine readable form.

Actually, what would be really nice is having the actual models (like the one depicted in the first link). I guess it's too much to hope for that :-) Google's 3D Warehouse is probably best bet of having something anywhere close.

by jtauber : Created on Aug. 16, 2007 : Last modified Aug. 16, 2007 : 0 comments (permalink)

Potter Predictions Now Has Comments

I've added comments to Potter Predictions.

So if you've read the 7th book, head over to http://potterpredictions.com/ and you can comment on each prediction (including arguing if you think I got the "answer" wrong).

by jtauber : Created on Aug. 15, 2007 : Last modified Aug. 15, 2007 : Categories potter_predictions : 0 comments (permalink)

Logo Design For Cats or Dogs

I'd like a logo for the upcoming relaunch of Cats or Dogs. I pretty much know what I'd like: the words "Cats or Dogs" with a cartoon cat on the left and a cartoon dog on the right, both leaning on their respective word with a smug look on their face as if to say "of course you're going to pick me".

In return, you'll get a credit (with a link to your site) on the footer of every page.

Email me if you're interested.

by jtauber : Created on Aug. 15, 2007 : Last modified Aug. 15, 2007 : Categories cats_or_dogs : 0 comments (permalink)

Two Google engEDU Videos

I really enjoy the engEDU videos (aka Tech Talks) that Google makes available.

Two interesting ones I recently saw:

both in some senses about the same topic of measurement.

Ironically it took me a while to get used to Pat Naughtlin's broad Australian accent. I wonder if that counts as cultural cringe.

by jtauber : Created on Aug. 15, 2007 : Last modified Aug. 15, 2007 : 1 comment (permalink)

Stack-Type Vectors, Part I

Part of the Poincaré Project...

So I've promised a couple of times to introduce stack-type vectors. This is a type of geometric vector that works a little differently than arrow-type vectors. They still meet the requirements of a linear space but their behaviour on a manifold, in particular with regard to a coordinate system on that manifold is a bit different. We'll get into that difference in subsequent posts (as well as talk about where these stack-type vectors show up in physics).

Here is what a stack-type vector might look like in three-dimensions:

Hopefully you can see why I call these stack-type vectors. From now on, though, we'll mostly draw two-dimensional stack-type vectors, but you should still get the same idea.

Note that the number of lines drawn and how long they are are not important (any more than the thickness of the line and shape of the arrow are important in an arrow-type vector). Instead it's the direction the lines are stacked in and how tightly they are packed that characterizes the stack-type vector.

But for a set of these to form a linear space, we need to say how to add them and how to scale them.

Here's how this vector scales (by 2x and by -1x):

Notice that scaling by 2x means the lines are packed twice as closely.

In part II, I'll show how addition of these vectors is defined.

by jtauber : Created on Aug. 14, 2007 : Last modified Aug. 14, 2007 : Categories poincare_project : 0 comments (permalink)

Blogging By Day of the Week

A colleague of mine observed in his reading trends on Google Reader that he reads more blog entries on a Thursday than any other day. His hypothesis was that by Thursday he's bored of office work and other routine stuff and so is more likely to choose to read blogs rather than undertaking other activity.

I checked my own reading trends and, sure enough, I read, on average almost twice as many entries on Thursday than the second most common day. Saturday is the day I read the least number of entries. The thing is, unless I'm travelling or really swamped at work, I generally clear my queue of blogs to read daily. Which means the Thursday thing could be more indicative of when entries are posted than when I take the time to read them.

My colleague points out, his argument could equally apply to posting as well as reading.

So I did a quick three-line Python script to output which day of the week each blog post of mine was made on and a sort | uniq -c dance gave me the following results:

  • Monday 67
  • Tuesday 106
  • Wednesday 108
  • Thursday 93
  • Friday 115
  • Saturday 125
  • Sunday 93

So my lowest reading day is my highest posting day and my highest reading day is my equal second lowest posting day.

I guess I don't blog when other people do :-)

by jtauber : Created on Aug. 14, 2007 : Last modified Aug. 14, 2007 : Categories blogging : 3 comments (permalink)

49

I'd heard stories of people getting huge AT&T bills (size, not cost) for their iPhone because AT&T lists every single data transfer session as a line item. Sure enough, when I got my bill today, the envelope was thick and inside was a bill 49 pages long.

by jtauber : Created on Aug. 13, 2007 : Last modified Aug. 13, 2007 : Categories iphone : 1 comment (permalink)

Math Doesn't Suck

Speaking of Numbers...

A few years ago I was on a flight from Boston to LA when I saw Danica McKellar (Winnie Cooper from The Wonder Years) sitting in business class. I later looked up IMDb to see what she'd been up to and discovered she'd graduated summa cum laude from UCLA with a BS in mathematics.

Furthermore while there, she published a paper that got her a finite Erdős number, making her one of the few people that has both a finite Erdős number (hers is 4) and finite Bacon number (just 2 in her case). (See Wikipedia's Erdos-Bacon number article for a list of other people with both.)

For a while, Danica's had a feature on her website where she answers maths questions posed by fans.

Then, earlier this month, she published a book entitled Math Doesn't Suck : How to Survive Middle-School Math Without Losing Your Mind or Breaking a Nail. I'm clearly not the target audience but if I had a daughter, she could do a lot worse than have Danica McKellar as a role-model.

Way to go Danica!

by jtauber : Created on Aug. 12, 2007 : Last modified Aug. 12, 2007 : Categories mathematics : 1 comment (permalink)

Installing iWork '08

I just got back to Boston from San Francisco (had a great dinner and chat with Graham Glass last night) and my copy of iWork '08 had arrived.

I'm looking forward to trying out Numbers as I've been eagerly anticipating it for close to two years (I was disappointed '06 didn't have it).

From what I've seen online, I suspect Numbers won't be suitable for some people's high-end number-crunching but for the everyday spreadsheet work I do, it looks like a dramatically better approach than the mainstream spreadsheets of the past.

It will be interesting to see what the "more word processor than page layout" additions to Pages are like as well.

by jtauber : Created on Aug. 12, 2007 : Last modified Aug. 12, 2007 : Categories apple : 2 comments (permalink)

A New and Improved Cats or Dogs Coming Soon

Cats or Dogs turned out to be a great success. It was fun writing and I got a lot of great feedback.

I always meant to go back and add more polish and a bunch more features.

I'm finally starting to do that (reusing some of Potter Predictions too) and, rather than continuing to live off quisition.com, decided to get a dedicated domain: cats-or-dogs.com

Only thing there at the moment is the ability to subscribe to get an email when the site launches. I'll add an Atom feed too, for people who would prefer that means of notification. Both will also help me to gauge interest.

by jtauber : Created on Aug. 11, 2007 : Last modified Aug. 11, 2007 : 2 comments (permalink)

We Like Our Own Stickers Better

Great answer to a dumb question...

audio

by jtauber : Created on Aug. 9, 2007 : Last modified Aug. 9, 2007 : Categories apple : 0 comments (permalink)

Predictive Text

I always used to think it funny on my old Sony Ericsson that the T9 predictive text proposed the name of one of my sisters whenever I was trying to type the name of the other, right up until the last character.

When I try to type my own name on my iPhone, it proposes one of my heroes instead...

If that's predictive text, I guess I have a bright future :-)

by jtauber : Created on Aug. 9, 2007 : Last modified Aug. 9, 2007 : Categories iphone : 1 comment (permalink)

In San Francisco

I'm in San Francisco from today until Sunday.

If anyone is interested in catching up Friday evening or Saturday, let me know — we'll see if we can get a Geek Dinner going.

by jtauber : Created on Aug. 9, 2007 : Last modified Aug. 9, 2007 : 0 comments (permalink)

Everything is Miscellaneous

I've mentioned before that I've long held a fascination with taxonomies, classifications systems and the like.

I've also been a long-time advocate of hierarchy being derivative (via ordering of facets) rather than primitive, and of faceted tagging (back in 2005 I called it tag the tags but see Google Code Project Hosting for a great example of how simple tag structure achieves this).

So I'm having a great time (half way through) with David Weinberger's Everything is Miscellaneous which is a wonderful book about these kinds of topics, not just in the computer age but going back to topics like the alphabetization, library catalogs in Alexandria, the Dewey Decimal System, the Periodic Table (and alternatives) and the Linnaean Taxonomy.

The book doesn't fall into the books that changed my mind category but more the I never thought of putting it that way before category.

by jtauber : Created on Aug. 9, 2007 : Last modified Aug. 9, 2007 : Categories books : 0 comments (permalink)

Linear Spaces: Duck Typing and Tuples

Part of the (very slow moving) Poincaré Project.

We've introduced the general concept of a linear space and shown how it applies to the specific notion of arrow-type vectors.

Recall that a (real) linear space is a set of objects that can be added and scaled. What 'add' means and what 'scale' means are arbitrary as long as the following rules are followed:

  • the space must form a commutative group under addition, in other words:

    • for any two elements u and v, u + v is also an element of the space

    • there is an element 0 called the identity element such that for every element v, v + 0 = v

    • for every element v there is a corresponding inverse (written -v) such that v + -v = 0

    • the addition is associative, i.e u + v + w = (u + v) + w = u + (v + w)

    • the addition is commutative, i.e u + v = v + u

  • for any real number a and any element v, av is also an element of the space (we say v has been scaled by a)

  • for any two real numbers a and b and any element v, (a + b)v = av + bv

  • for any real number a and elements u and v, a(u + v) = au + av

  • for any real numbers a and b and element v, a (bv) = (ab)v

  • for any element v, 1v = v

(other properties fall out naturally from the above requirements. For example, it's easy to show that -v must be v scaled by -1, that v + v = 2v and so on. Because these follow from the rules above, they are true of all linear spaces.)

Note that as long as these rules (or axioms) are followed, you have a (real) linear space. It doesn't matter how the elements are actually "implemented". The arrow-type vectors we saw earlier are just one type of linear space.

Programmers might find it useful to think of this as like duck-typing. If it looks like a duck, walks like a duck and quacks like a duck: it is a duck. If you'll pardon the paraphrase, if it adds like a linear space and scales like a linear space: it is a linear space. We never need to peer into the internals of the elements themselves.

Another "implementation" of the linear space idea which is probably familiar to many of you is the n-tuple. Arbitrarily picking n to be 3, the set of 3-tuples form a linear space if we define addition

(m, n, o) + (r, q, r) = (m + r, n + q, o + r)

and scaling

a(m, n, o) = (am, an, ao)

which means, of course, that whatever m thru r are, they must themselves be able to be added and able to be scaled by a real. An obvious choice would be that m thru r are themselves real numbers. Then we have a linear space often referred to as R3.

It is important to note that a tuple-based linear space is not the same as a linear space based on arrow-type vectors. While it may be possible to map arrow-type vectors to tuples (even in a way that is obvious and intuitive), they are not interchangeable and an infinite number of possible mappings exist between an arrow-type vector based linear space and a tuple based one.

It isn't possible to point to some arrow-type vector and say, oh, that's (3, 2, -7) without having first defined how to map the two spaces (which is an extra bit of structure).

One interesting characteristic of the arrow-type vector which the tuple in and of itself lacks, is its geometric nature. An arrow-type vector can be tied to a manifold and mean something in relation to that manifold and indeed this is what makes it a geometric type of linear space. (Remember in a previous post I used the phrase "travel in a particular direction at a particular rate" — this interpretation of an arrow-type vector only makes sense in the context of a manifold)

Next post in the Poincaré Project, we'll look at another type of linear space that is geometric (the one I promised I'd talk about last time) — the stack-type vector.

by jtauber : Created on Aug. 9, 2007 : Last modified Aug. 9, 2007 : Categories poincare_project : 4 comments (permalink)