Following Bob, Joe and Ryan, here are my Browser / OS stats for the last month (rounded to nearest tenth of a percent):
| Firefox | 54.6% |
| Internet Explorer | 24.7% |
| Safari | 9.9% |
| Mozilla | 6.5% |
| Opera | 2.5% |
| Windows | 60.3% |
| Linux | 20.5% |
| Macintosh | 18.3% |
(iPhone at 0.2%)
For comparison, here are the figures for the same period a year ago:
| Firefox | 44.8% |
| Internet Explorer | 37.1% |
| Safari | 9.6% |
| Opera | 3.4% |
| Mozilla | 2.21% |
| Windows | 69.4% |
| Macintosh | 16.9% |
| Linux | 12.9% |
by : Created on Jan. 30, 2008 : Last modified Jan. 30, 2008 : (permalink)
For a while now, Marcie Lascher has been asking me to write a blog entry about her. She doesn't care what I say, she just wants to have a blog mention her so she feels more Web 2.0 enabled. But while I'm at it: Marcie, I think you are awesome!
by : Created on Jan. 22, 2008 : Last modified Jan. 22, 2008 : (permalink)
Lots of good stuff of great interest to me.
No wow moments during the video given rumours and spoilers beforehand, although I did like the way he announced the studios behind iTunes movie rentals, announcing the mini-majors before the majors and the latter with the line "and by the way...these six too".
I've already bought external drives for use with Time Machine and already have an Airport Extreme, so Time Capsule came a little too late for me. It is a pain unplugging the drive from my laptop all the time, though. And if I get an Air (see below) I'd need a Time Capsule anyway; unless they release an update that allows any external drive to connect to an Airport Extreme.
I've already updated my iPhone. Ability to put bookmarks on the home page is nice (currently, I have Facebook and Quisition). I've tried out the locator and it works great in my apartment. I'll try it more tonight when I'm out on the road.
Updated iTunes too and rented a movie. Worked beautifully. I will definitely be watching a lot of movies this way now.
I don't have a TV here in the US (watch everything on my computer) so Apple TV doesn't interest me at the moment. Steve's admission of failure on the first release was refreshing and it's nice that Take 2 is a free software upgrade for existing Apple TV users.
I only heard Steve say "Boom!" once (around the 45m mark). Coincidentally, it was shortly after the Flickr demo blew up.
The Air is very appealing to me. I used to have a 12" PowerBook from 2004 when I was traveling between Australia and the US a lot. My current laptop (a 17" MacBook Pro I bought mid-2006) dates from a time when I was living for months at a time in a hotel and it was basically my primary machine. Now I've settled into an apartment, I have a Mac Pro as my primary machine. The times I do travel, the 17" is just a little too big.
So the Air is a nice option as a travel laptop. The battery life is appealing. I haven't heard anything about whether the SSD increases the battery life even more. The HDD is slow but the SSD is a lot more expensive. I haven't decided yet. I'm at least going to wait a couple of weeks to see one in person and hear initial feedback.
by : Created on Jan. 16, 2008 : Last modified Jan. 16, 2008 : (permalink)
Yet again I can't watch the Macworld Keynote Address Quicktime because of a "Bad Request" error. Has happened every year and I have to wait a day or two. What's a fanboy to do?
UPDATE: Working now.
by : Created on Jan. 15, 2008 : Last modified Jan. 15, 2008 : (permalink)
Last year my goal for Quisition was to hit 1,000 users.
My goal this year is much more ambitious. It is to have 1,000 active users. To put that in perspective, with current rates of engagement (prior to the redesign, anyway) that would mean attracting 50,000-100,000 users.
That might be possible with $100k of AdWord spending, but I don't want to do that.
So the primary sub-goal is to raise the engagement rate from 1-2% to 5%-10%. More on that over the next few months (bottom line: improving functionality to make Quisition a more effective—and hence compelling—learning tool)
That still means I need to attract 10,000-20,000 users. Or 10-20x what I had at the end of last year. That's still a sizable increase in advertising costs, all other things being equal. Too rich for my blood.
I can reduce the CPC (cost-per-click) with better ads in cheaper placements. But the more important metric is the cost-per-conversion—how much does it cost to get a new user. Lower CPC doesn't matter if it doesn't result in a new user. Also, I can't think of a way of tracking where the long-term users came from so the thing I need to watch out for is that the cheapest conversions may not be the ones that last.
A more general notion, that applies regardless of whether the traffic source is an ad or, say, this blog, is how many people that visit the site actually sign up for an account. Before the redesign, that number was 5.6%. The redesign seems to be working already because that number is now 8.2%.
Just for the record, last year I was getting a 0.24% CTR (click-thru-rate) on my ads and 6% conversion rate. So far this year, I've increased the CTR to 0.3% and, since the redesign am getting a conversion rate of 10% from ad click-thrus.
So, to achieve 10,000-20,000 users, I need to:
where the latter involves me trying to:
The product of all those factors needs to be in the 10-20 range.
Of course, that all assumes that AdWords continues to be the biggest source of leads. I probably need to pursue other forms of promotion, but I really don't have any experience with that other than this blog.
by : Created on Jan. 14, 2008 : Last modified Jan. 14, 2008 : (permalink)
I don't think I've mentioned it here before but next week, I'm one of the keynote speakers at the BibleTech 2008 conference in Seattle. While I've given talks a number of times about my Greek linguistics research, this will be the first time that I'll get to talk about how I've used technology in that research.
I plan to give a history of the MorphGNT project and the various sub-projects I've worked on over the last fifteen years, covering the evolution of data models, text encoding, tool sets and more. I then want to talk about the opportunities that lie ahead and where I hope the work will go in the future, particularly given my collaboration with Ulrik Sandborg-Petersen.
by : Created on Jan. 14, 2008 : Last modified Jan. 14, 2008 : (permalink)
I decided it was time for a redesign of the Quisition site. I started with the logo and things just took off from there.
I wanted to make the logo more modern and "lickable" and I found a really easy way to do it.
I started with the old logo, flat black text, Gil Sans Bold with a bit of kerning. Then I simply overlaid on top of the text an ellipse about as high as the text and maybe 50% wider, shifted up so that the bottom of the ellipse was roughly in the middle of the text. I then put a gradient fill on the ellipse: all white but varying in opacity from 15% at the top to 60% at the bottom. I then moved the blend midpoint about a third of the way up from the bottom.
I did the same to the green question mark, and voila!
I'm really happy with the result. I used the same technique for the shiny tabs and sign up button on the new site too.
BTW, I used OmniGraffle. It's not really designed for this sort of thing but did the job nicely.
by : Created on Jan. 13, 2008 : Last modified Jan. 13, 2008 : (permalink)
Back when I was editing Alibi Phone Network in 2004, I found that cuts that seemed fine to me when I watched them alone would make me cringe when I watched them with a friend. I didn't even need to get feedback from the other viewer; I just got embarrassed at certain points and immediately wanted to start making excuses. I dubbed this approach to testing the "cringe test".
I discovered a couple of days ago that the test is useful for websites too. I was showing a friend Quisition for the first time, and without him saying a word I started thinking to myself "oh no, it's not clear where he has to click" or "it's not obvious what that means; I should have explained that better".
by : Created on Jan. 11, 2008 : Last modified Jan. 11, 2008 : (permalink)
by : Created on Jan. 10, 2008 : Last modified Jan. 11, 2008 : (permalink)
At the start of the year, I launched Quisition, my flashcard site. By the end of the year, I had reached my goal of 1,000 users. I've already added a bunch of new features in the new year with lots more to come.
One Friday evening in February, I conceived of, implemented and launched Cats or Dogs. It rapidly took off in the Python community and by PyCon two weeks later, I was being stopped in the hallways and asked if I was the "Cats or Dogs" guy.
Also at PyCon, I chaired a session, participated in the panel on web frameworks and volunteered to be the organizational admin for the Python Software Foundation's participation in the Google Summer of Code. As well as administering, I ended up mentoring two projects over the summer.
In April, I started playing Lord of the Rings Online. The last few months I've probably online averaged no more than an hour per week, but at the peak I was playing a lot. Although I never got around to blogging it, my main character reached level 50 (the current limit).
In May, I got an apartment in the US and said goodbye to long-term hotel stay.
One of my songs, You, Me, Us We had some success, coming in the top 20 in a songwriting content and later being named top song of 2007 by ElectroQueer.
In June, I bought an iPhone. I take it for granted now.
In July, I conceived, implemented and launched PotterPredictions which had a reasonable amount of success in its naturally short life (I still wish I'd thought of it earlier).
In terms of open source, it was a fairly slow year although I did make a lot of progress on django-atompub and pretty much finished support for the syndication format RFC.
I started work on yet another django-based website called habitualist that will hopefully have a closed beta in the next month or so.
In November, I bought a Mini Cooper S and resumed pre-production work on my first feature film.
Finally in December, just before leaving to the UK and Spain for the holidays, I received both an OLPC XO laptop and a 23andMe DNA test kit. More about those later!
They were the 2007 highlights from my blog. A number of other exciting things happened that didn't get blogged about at the time. mValent had a great year and an amazing last quarter; Ulrik Sanborg-Petersen and I made slow but significant progress on a number of MorphGNT projects, the fruits of which should be available soon; Facebook became, for me, both a great way of reconnecting with old friends and a promising platform for development. I made some progress on my PhD but not as much as I would have liked. I need to change that in 2008.
So all-in-all, it was a pretty awesome year. I'm expecting 2008 to be even better. Watch this space!
by : Created on Jan. 9, 2008 : Last modified Jan. 9, 2008 : (permalink)