Do yourself a favor and upgrade from whatever steaming pile of garbage you're using to browse the Web. Because that's what Firefox 3 makes the competition look like: steaming piles of garbage.

Mozilla Firefox lets you assign "keywords" (or "shortcuts") to bookmarks by entering a word or phrase in the bookmark's "keywords" field. The advantage to setting a keyword for a bookmark is that you can load the page more quickly by typing the keyword in the location bar than by finding the bookmark in your bookmarks menu or toolbar.

However, if you assign the same keyword to multiple bookmarks (for example, assigning "tech" to both Slashdot and TechMeme), Firefox will only open one of your bookmarks when you type the keyword in the location bar. This mis-behavior is documented in bug #237180, which was submitted in March of 2004. Some argue that you shouldn't be able to assign the same keyword to multiple URLs, and others (like me), argue that you should be able to, and Firefox should open all bookmarks that have been assigned a given keyword. After all, isn't that what tabs in a browser are for?

So, if, like me, you would like the ability to quickly open a set of bookmarks by typing a single word (think "stats" for a set of pages showing various statistics about your website traffic, or "news" to pop open your three favorite news sites), vote for bug #237180, and leave a comment indicating how you'd like the multiple keyword problem to be resolved.

More information about this poster is available at John Slater's blog, Into the Fuzz.

This is bad news for Twitter. Up until now, they've been the market leader in the "Popular websites that are unavailable" category, but with Amazon.com as competition, they'll have a hard time keeping that position.

Caveat: This whole thing was thrown together without any regard for coding or design conventions. Also, I can't guarantee that you won't be banned from Digg for using any of this information - I was, a few times. Proceed with caution.

First, the data structures. The Top Diggers lists are supported by two database tables: `digg` and `digg_users`. `digg` holds the set of front-page stories and their submitters, while `digg_users` stores aggregated information about each user in the system.

CREATE TABLE `digg` (
  `user` varchar(128) NOT NULL default '',
  `submission` varchar(255) NOT NULL default '',
           `date` datetime NOT NULL default '0000-00-00 00:00:00',
  PRIMARY KEY  (`submission`),
  KEY `user` (`user`)
);

CREATE TABLE `digg_users` (
  `user` varchar(128) NOT NULL default '',
  `frontpage` int(11) NOT NULL default '0',
  `dugg` int(11) NOT NULL default '0',
  `submitted` int(11) NOT NULL default '0',
  `profileViews` int(11) NOT NULL default '0',
  `frontpagestatic` int(11) NOT NULL default '0',
  `frontpagetotal` int(11) NOT NULL default '0',
  `submittedstatic` int(11) NOT NULL default '0',
  `image` varchar(128) NOT NULL default '',
  UNIQUE KEY `username` (`user`)
);

Next, some data to get you started. Here is a SQL dump of the two tables described above, including information on about 13000 Digg users (for `digg_users`), as well as the last 4 stories on the Digg homepage at the time of the last update (for `digg`). Import this data into your database in preparation for the next step.

The next step: data retrieval. The main work of updating the top 100 list is done by a Python script:

import digg

# Grab the last X pages of popular stories
digg.update_news(100)

# Update the top X profiles
digg.update_profiles(110)

Of course, this code makes no sense without the digg.* methods, downloadable here. This script also requires the excellent BeautifulSoup Python HTML parser. You will have to modify digg.py to change the database connection parameters.

For those who don't care to read through the code, it achieves two main objectives: Find out who submitted any frontpage stories since the last update (it stops when it hits a story already in the `digg` table), and using that information, determine the new top 100 users and update their profile information.

The last step is data presentation. The information in the database tables needs to be transformed into a readable HTML file. I'll leave this step as an exercise for the reader, but to get you started, this SQL query will get you the data you want in an easy-to-read format:

SELECT
   user `Username`,
   frontpagetotal `Frontpage Stories`,
   submitted `Stories Submitted`,
   dugg `Stories Dugg`,
   profileViews `Profile Views`
FROM digg_users
WHERE
   frontpagetotal <= submitted
ORDER BY
   `Frontpage Stories` DESC,
   `Stories Submitted` ASC,
   `Stories Dugg` DESC
LIMIT 100

So to sum up, if you want to manage your own "Top 100 Diggers" list, take the following steps:

1. Import the dump of digg data linked above.
2. Set up and run your scripts
3. Create a readable version of the data.

Have fun, and beware the Digg ban-hammer.

If this happens to you, here is some information that you could send us to help us out:

1. Your Operating System and Firefox versions. Example: Mac OSX 10.4.11, Firefox 3.0RC1.
2. Whether you're logged into Mahalo when it happens.
3. prefs.js. This file is found in your Firefox profile directory, and it contains any changes you've made to Firefox's default settings. It doesn't contain any especially private or personally identifiable information. (Where is my Firefox profile directory?)
4. If you can get Mahalo to crash Firefox in a clean profile (one without any personal information like passwords or an extensive browsing history), then sending us a ZIP of your entire profile directory would be extremely helpful. If the crashing happens in your regular, every-day browsing profile, please don't send this to us if there's any chance that it includes information that you don't want anyone else to know, like usernames, passwords, or browsing history.

So if you'd like to help us fix this problem, send as much of this information as you can to finke@mahalo.com. Just remember: We're all in this together. We don't want Firefox to crash any more than you do.

"You should be expert in massively scalable architectures, how MySQL and Linux interact, how MySQL and memcache interact, sharding, replication (including multiple master replication) and how to tune MySQL based on various schemas for maximum performance and availability. You are a HANDS ON implementor, a get-it-done kind of developer. The right person is a self starter with the 'general get it factor.' You work well with a team of like-minded engineers, and have a genuine desire for excellence."

See Jason's blog for instructions on applying for this Systems Engineer job in Los Angeles. Tell 'em Finke sent you.

If the Wii and Mario Kart is more your style, check out my brother's blog. He updates regularly with new insights into the Wii and Wii games.

If you like to read about food, or if eating is one of your hobbies, check out Fast Food Critic. Just try not to read the nutrition information posted with each review.

And for no reason at all, here's our dog Pedro trying on his Halloween costume a few years back. You can see the Scape-o-lantern on the coffee table behind him.

I've updated YouTube Comment Snob to work with YouTube's latest redesign, as well as to add compatibility for Firefox 3. It's available for download from the Mozilla Add-ons sandbox until it is approved by an AMO editor for mass consumption.