Simon Willison’s Weblog

Subscribe

Entries

Filters: Sorted by date

Credibility and CSS

James Buckley links to a new report on How people evaluate a web site’s credibility. His comments:

[... 324 words]

Yahoo to one day go Google

Jeremy Zawodny drops a so-subtle-that-I-probably-misread-it hint that the Yahoo search page might some day follow AllTheWeb in Google’s footsteps.

HTTP status codes

Craig Saila has a minor rant about HTTP error codes. Did you know that a 410 should be served instead of a 404 when a resource has been deliberately, permanently removed? I didn’t.

BCSS

Here’s the reason I’ve been blogging at a relatively low frequency lately: BCSS—the Bath University Computer Science Society. The site is still under heavy development (there’s a surprise) but there’s enough information there now for it to be worth linking to. At the moment the site is valid HTML 4.01 Strict but it works as XHTML as well (try appending ?xhtml=1 to the URL of any page on the site) thanks to an ultra flexible page template class (outlined here) and a few simple tricks to convert XHTML in to HTML beforethe page is displayed.

[... 386 words]

Mozilla for bloggers

Matthew Haughey (freshly redesigned) has published a Mozilla advocacy article explaining why Mozilla (and variants) are excellent tools for bloggers. Spot on.

Sitepoint redesigns

I don’t know how I missed it, but SitePoint have redesigned in funky valid structural XHTML and CSS. I quite like the new look (not so keen on the new logo though) and the navigation is definitely a huge improvement—instead of the previous confusing arangement of several “sister” sites they now divide content in to “Articles” and “Forums”, which makes a lot more sense. They have some clever DHTML enhancements as well, such as a collapsible left hand navigation panel. The right hand navigation that scrolls up and down (jerkily) with the window is quite distracting unfortunately.

[... 155 words]

The importance of titles

Gordon Weakliem reminds us that the most important RSS element is <title>. I’m painfully reminded of this each and every time I add a new entry—I have well over 800 entries now, and I’ve promised myself that next time I perform a major upgrade on this blog I’m going to go through and manually add titles to every single one. The task should be made slightly easier by the camelCase permalinks, which I can convert in to “suggested” titles to help the task along. It’s still not going to be much fun though.

Creative commons query

Aaron Swartz has been talking to Google about indexing Creative Commons licensed works:

[... 120 words]

Dependencies suck

Sigh. I guess I’ll stick with the console version.

[... 37 words]

An interview with Cory

An interview with Cory Doctorow, via Leonard. Provides some great background insight in to the world described in Down and Out, along with Cory’s thoughts on such topics as the recording industry and the Disney corporation.

Vector search engines

Building a Vector Space Search Engine in Perl by Maciej Cegłowski:

[... 307 words]

Problems in Nirvana

Simon Phipp is Chief Technology Evangelist at Sun, and recently participated as a speaker on the “.NET Nirvana” Geek Cruise. His blog entries covering the event make interesting reading. To cut a long story short, after a couple of presentations on Java web services and open source coftware the Microsoft contingent at the event requested that he be barred from attending an evening Q & A. Simon makes the following observation:

[... 228 words]

Blogging and journalism

I’ve been pretty much ignoring the whole “Blogging vs Journalism” thing but recently I’ve begun to understand what the big fuss is about. One of the most popular arguments put forth by journalists concerned by competition from blogs is that the information contained therein isn’t as reliable thanks to a lack of an editor to check facts. Rubbish. I can’t remember the last time I read a technology article in the main stream press about something I have more than a passing interest in that didn’t have at least a few errors. Some of the blogs I read on the other hand are written by subject matter experts—these people are not being paid to knock out 750 vaguely relevant words on a breaking story, they are voluntarily providing their insights because they are heavily involved with the topic at hand.

[... 308 words]

Quasar usability

I had my first ever Quasar (otherwise known as laser tag) experience last night with the Bath University Computer Science Society (BCSS). Due to various misunderstandings our booking failed to materialise, but Quasar were very nice and said we could play against another group at 7.30pm. They failed to mention that the other group had an average age of about 9. Suffice to say it was an entertaining match, which we won by a very small margin (little kids spend all day running about; lazy students don’t).

[... 424 words]

PHP5 Preview

A great preview describing the features we have to look forward to in PHP 5. Proper object reference handling is going to be sweet.

Doing forms justice

Dan Loda: Doing forms justice. A demonstration of how labels, accesskeys, fieldsets and CSS can make a form that’s usable, accessible and looks great. He even uses an optgroup in a select box, an element I’ve never seen demonstrated before.

Pingback redux

I think I’ve worked out a way of implementing Pingback (or a Pingback-like system) without any need for XML-RPC, <link> elements or custom HTTP headers.

[... 663 words]

Browser detection reconsidered

Leonard Lin on The Folly of Depending on CSS Parsing Bugs:

[... 333 words]

Mail models

Matthew Thomas offers a fascinating example of usability gone horribly wrong (the explanation is provided here). Incidentally, while showing the above to my girlfriend she uncovered an interesting usability issue on Matthew’s site itself— “is this a collaborative weblog then?” she asked, having spotted the Posted by mpt on 2/24/03; 3:51:24 AM text below the entry. Clicking on the name link provided no extra information, bringing up a screen with no useful content at all. I imagine this is an issue with the CMS powering the blog, but it does neatly demonstrate how some CMS features can detract from the unserstandibility of a site.

Safe HTML checker

I’ve finally enabled a subset of HTML in my comments. In doing so, I had several requirements that needed to be fulfilled:

[... 227 words]

Slow professional suicide

Al Sparber makes perfect sense in article from June last year:

[... 180 words]

SSH public key authentication

I’ve been having fun with SSH lately. Did you know you can set up SSH so you can log in to servers without having to provide a password? It’s called “public key authentication” and is apparently more secure than using a normal password. You generate a public/private key using a program called ssh-keygen, and store a copy of the public key on the server(s) you wish to authenticate with. When you attempt to log in, the server sends you a message encrypted with your public key—your machine decrypts it and sends back the original message, proving your identity.

[... 194 words]

Python for Java programmers

Python and Jelly: Scripting Power for Java and XML incorporates an excellent introduction to Python and Jython for Java programmers, with a whole bunch of comparative code samples and comprehensive coverage of differences between the two languages.

More Python advocacy

Intelligent Enterprise: For all of you unfamiliar with Python, get ready for the “next big thing.”

Watch out for Javascript in referrals

Here’s a good reminder why you should always encode < and > as HTML entities when displaying content from an untrusted (i.e external) source: Kasia in a nutshell was hit by a false referrer containing javascript deliberately aimed at hijacking the page the referrer was displayed on:

[... 76 words]

Get a better browser!

Via Scott, this oh-so-true quote from a Microsoft “next-generation technology” consultant:

[... 377 words]

Calendars and crawlers

Douglas Bowman has been having some amusing problems with robots and his calendar. The calendar, visible on every page of the site, automatically adds a “next month” and “previous month” link to allow surfers to browser through the archive in both directions. Unfortunately, Doug ommitted the logic to stop showing a “previous month” link when there were no earlier entries. An enterprising crawler started following the links, and didn’t stop until it had reached 1542!

[... 113 words]

DNS mess

As the recent lack of updates demonstrates, I’ve been getting stuck in to a pretty time consuming new project. It should have launched several days ago but I made a right royal hash of the DNS settings—hopefully everything will be working fine in about 24 hours time.

[... 119 words]

SQL slammer analysed

Robert Graham’s analysis of SQL Slammer cleared up quite a few things I had been wondering about the worm. It confirms that the majority of the infections were caused not by SQL Server (as reported widely by the press) but by the embedded MSDE component, which is far less likely to be patched (or firewalled off from the public internet) than SQL Server.

[... 128 words]

Eric Meyer’s colour blender

Eric Meyer’s Color Blender is an incredibly useful tool for picking colours for a web site. Give it two different hexadecimal colour codes and it will calculuate and display between 1 and 10 “midpoint” colours. It’s fun to play with and great for tracking down that elusive perfect shade of green...