Simon Willison’s Weblog

Blogmarks in May, 2008

Filters: Type: blogmark × Year: 2008 × Month: May ×


Python + Hadoop = Flying Circus Elephant. Last.fm have released Dumbo, a Python module that lets you easily write Hadoop map/reduce tasks using Python and generators. # 31st May 2008, 2:14 pm

Obscure bugs revisited: IE, HTTPS and plugins. Filed for future reference: IE breaks mysteriously if you serve it up plugin content (e.g. Flash) over HTTPS with a no-cache header—it deletes the file from cache before the plugin software gets a chance to open it. # 30th May 2008, 9:54 am

Twitter, or Architecture Will Not Save You. Kellan is not an armchair architect. He also doesn’t mention Rails once. Well worth reading. # 29th May 2008, 1:16 am

Google Gears renamed “Gears”. “We want to make it clear that Gears isn’t just a Google thing. We see Gears as a way for everyone to get involved with upgrading the web platform.” Support for Firefox 3 and Safari is being added and Opera are integrating Gears with both their desktop and mobile browsers. # 29th May 2008, 12:38 am

Google Earth in a browser (sort of), Scriptable, a quick peek and poke. Dan Catt on Google’s new browser plugin version of Google Earth... which conveniently exposes a JavaScript API to the browser in the form of the “ge” object, which can then be poked at interactively using Firebug. # 28th May 2008, 11:13 pm

Using Memcache with Google App Engine. Brad Fitzpatrick’s 20% time project. # 28th May 2008, 11:11 pm

OpenID phishing demo (via) A demonstration of the OpenID man-in-the-middle phishing attack. idproxy.net OpenIDs are immune to this particular variant due to the landing page not asking for your password (the phishing site could still provide their own redesigned landing page and hope users don’t notice though). # 28th May 2008, 8:09 am

Google AJAX Libraries API (via) Google are hosting copies of jQuery, Prototype, mooTools and Dojo on their CDN, with a promise to permanently host different versions and an optional JavaScript API to dynamically load the most recent version of a library. I wish they’d stop capitalising Ajax though. # 27th May 2008, 5:56 pm

Tracking Christmas Cheer with Google Charts. Brian Suda’s Google Charts tutorial on 24 ways has proved invaluable for figuring out how to handle grid lines and axis labels, both of which are pretty unintuitive (and not hugely helped by the official documentation). # 26th May 2008, 9:43 pm

QUnit. The jQuery unit testing framework is now documented and supported as a separate project. # 26th May 2008, 5:31 pm

Twitter / MarsPhoenix. NASA’s Mars Phoenix lander, due to land on the planet today, has a Twitter account. Bio: “I dig Mars!”. # 25th May 2008, 7:41 pm

Debugging Django, a slidecast. I used SlideShare’s slidecast tool for the first time to synchronize audio of my Django London User Group talk with the slides. The talk included several live demos which aren’t represented in the slides so it’s a bit piecemeal in places. # 25th May 2008, 2:47 pm

Easy way to reset your sleep cycle: Stop eating (via) New research shows that you can quickly reset your sleep cycle by not eating for 12-16 hours and then using breakfast to flip in to another time zone. I get clobbered by jet lag when I fly from the US to Europe; this could be really useful. # 25th May 2008, 2:11 pm

LastGraph 3. Andrew Godwin’s last.fm profile visualisation tool, now in its third incarnation. # 25th May 2008, 2:05 pm

Walk, Don’t Run (via) A retrospective look at Grim Fandango (possibly my favourite game of all time) and the fan community that are keeping it alive, nearly a decade after it was first released. # 25th May 2008, 2:04 pm

Richard Feynman and The Connection Machine. Too much great stuff in here to attempt to summarise. # 25th May 2008, 2:01 pm

modswgi: Debugging Techniques. mod_wsgi is excellent software, and the documentation is equally superb. I used these instructions recently to run the Python debugger inside a running instance of Apache, which helped my track down some import errors that weren’t occurring with Django’s development server. # 25th May 2008, 1:34 pm

On the spot. Did you know Jupiter just grew a third spot? Apparently the spots are storms, and the largest has been raging for several centuries. # 24th May 2008, 6:25 pm

Search Engine Optimization Through Hoax News. Devious new black-hat SEO technique: invent a news story that’s pure link-bait. The recent “13 year old steals dad’s credit card to buy hookers” story was a hoax: it was a pure play for PageRank. # 22nd May 2008, 6:09 pm

On-board vs. Off-board Comet. Useful distinction. On-board comet runs on the same server as the rest of your application; Off-board comet is served from a separate server (generally a subdomain) and a separate stack. If you want to stick with PHP, Rails or Django for the rest of your site off-board comet looks like the way to go. # 22nd May 2008, 5:02 pm

AOP aspect of JavaScript with Dojo. Fantastic post—concisely explains Aspect Oriented Programming, then shows how Dojo’s dojox.lang.aspect brings AOP to JavaScript, including some really useful built-in aspects for logging, profiling and more. Aspects are like Python decorators on steroids. # 18th May 2008, 10:45 am

GeoNames Commercial Webservices. Wikinear has been loading slowly recently, so I’ve signed up for GeoNames very reasonably priced commercial plan which provides access to better servers at their end. This might speed things up to the point that I can reliably run the site on Google AppEngine, which times out aggressively if an external HTTP request takes too long. # 18th May 2008, 10:32 am

Dopplr place googlemaps, with and without Yahoo Geo API bounding box adjustment. Dopplr uses Geonames for most geo information, but is now mixing in bounding box data from the Yahoo! Geo web service to improve the default zoom level for their maps. The JSON callback API means no server-side code is required on Dopplr’s end. # 17th May 2008, 11:35 pm

Firebug Command Line API. Another thing I didn’t know about Firebug: you can set a breakpoint at the start of a function with “debug(fn)” and log all calls to it with “monitor(fn)”. # 16th May 2008, 12:14 pm

Using Git as a versioned data store in Python. gitshelve supports the same interface as Python’s built-in shelve module but stores things to a versioned Git repository instead of just a pickled dictionary. I’ve been casually wondering what a Git-powered CMS would look like. # 15th May 2008, 3:25 pm

Cubescape. Beautiful isometric cube building tool by Cameron Adams, written in JavaScript and jQuery. # 15th May 2008, 8:40 am

Crossdomain.xml Invites Cross-site Mayhem. A useful reminder that crossdomain.xml files should be treated with extreme caution. Allowing access from * makes it impossible to protect your site against CSRF attacks, and even allowing from a “circle of trust” of domains can be fatal if just one of those domains has an XSS hole. # 15th May 2008, 8:06 am

Engineering @ Facebook: Facebook Chat. The new Facebook Chat uses Comet (long polling with a hidden iframe) against a custom web / chat server written in Erlang, designed to handle a launch to all 70 million users at once. It was tested using a “dark launch” period where live pages simulated chat request traffic without showing any visible UI. # 15th May 2008, 7:55 am

goog/useragent/iphoto.js. The Goog library includes code to detect the user’s installed version of iPhoto, based on reverse engineering the Mac.com Gallery RSS feeds. This has Mark Pilgrim written all over it. # 14th May 2008, 9:21 pm

Doctype: /trunk/goog. Google’s newly released JavaScript library (pure JavaScript, so more along the lines of YUI and jQuery than GWT). I haven’t found the documentation for it yet, but the code is extremely well commented. UPDATE: The documentation is spread throughout Doctype. # 14th May 2008, 9:12 pm