Solitude 

At Solitude, we like our open source software. There are all kinds of interesting ideas out there that people are giving away for free. The problem is the websites are almost always shit. Now, I’m not knocking the visual look; it probably doesn’t matter that the website for most programs isn’t attractive.

The main issue, illustrated by almost every open source project I’ve ever seen, is that they assume the user knows what the project is about. Wrong. I usually find out about these things through a vague link. Putting the news section immediately on the front page is a bad thing. The first thing I want to know is exactly what the project is about, not that version 0.65c has just been released with new enhancement X.

Three guidelines:

1. Have a section on your website explaining in plain English, no marketing jargon and as little technical gibberish as is necessary (which should itself be explained), exactly what the project is for, what it does and why anyone should care.
2. Begin the front page with a careful, one-line abstract of the explanation page followed by a “read more” link. If you can’t construct such an abstract, you’ve already got a problem with the structure of the project.
3. Clear, single-click downloads. Note that Sourceforge pages put most people off.

These guidelines will improve your site greatly and are not hard. We’d name names of projects who don’t live up to expectations, but I’m sure that it would take far too long.

Yesterday, I read about Tabtastic (via Jack) and was a little impressed by how simple it is. Follow a simple naming convention and data structure on your page and you can have a javascript-implemented tab pane effect on your site. It looks good, and the html is reasonably semantic (if that is your thing.)

The problem, however, is with usability. When you click on a tab, you change the anchor position of the page (the url fragment beginning “#”). This kicks off whatever logic is required to ensure the correct information is showing for that page. It also scrolls the tab bar off the screen, completely breaking the illusion of having a tab pane. A change of tabs should only change the content, not the window position.

A small matter, but one that immediately hinders the usability of the system.

A question: if you had a tool which did what you wanted, had a decent interface, some nice features but only worked 95% of the time, would you keep it? That 5% is probably not that much unless it happens to be a tool that you use frequently and for long periods, then the 5% really starts to matter.

Macromedia Dreamweaver is that tool. For years now, I’ve been using it (in code viw) to build websites, write PHP and manage web based projects. In the last year or two though, the problems have been mounting up:

• Huge amount of time to create a new site folder and associated data. It should not take minutes to open up a dialog. A minor nuisance, since new sites are relatively infrequent.
• Partial function colouring in PHP. Some standard functions get the function call colouring (blue), others don’t. More than once this has made me second guess the function name.
• Making .htaccess files incredibly difficult to open. It’s just a text file, it should not be that hard.
• Munging data. Because of some angle bracket handling code that has gone horribly wrong, Dreamweaver occassionally changes my data without telling me. What do I mean? Removing the “less than” character from a conditional, for example, or taking chunks of what it believes to be XML away. Unforgivable.

In those same years, I’ve learned to be far less tolerant of problems with tools because I can generally be certain that the problem is not with me. I’ve also become fairly proficient with emacs, and my set-up suits me down the ground. I’m constantly impressed by new extensions tht I find, the commands have become second nature and the minimal interface works. Goodbye Dreamweaver, you should’ve worked 5% harder.

While checking for information on bus timetables yesterday, Arriva (the main operator in my area) astounded me with some piss poor design.

The left hand-side of their page features a form that allows you to input a town and route number to find information on it. Great. Except that upon submitting the form it doesn’t just send your information to a page in which results will be returned, or redirect you to the appropriate information. Oh no, that would be far too simple.

Instead, it executes a javascript function using the long since deprecated javascript: protocol (you do have JS turned on for something as simple as this, right?), pops-up a small window which does nothing but cause a redirect in your original window and then closes, and then takes you to a results page. Javascript and pop-ups for a redirect? What? Don’t get me started on the crufty URLs either.

My final year project is now complete (expect more on that at some stage), celebrating has been done thrice (just another two more to go), pictures were taken (of the CompSoc Easter celebrations at Cafe West), sleep has been caught up on, work has begun on the first project of the holidays, and it feels like a huge burden has been lifted. I’m not the kind of person to get stressed or panic, but it feels very good to no longer have such a constant and heavy workload.

Enough of this self indulgent tripe. Real posts begin again this week.

A quick message to say that Solitude will largely disappear for the next two weeks or so. Coincidentally, that’s how long I’ve got to finish my final year project (something about efficient plagiarism detection) which I’m sure will get a post or so here in the future.

The odd post may appear and random links certainly will continue.

Some new toys will appear when I return, as well as documentation and source for the recently updated ActoRSS, which provides RSS feeds for most actors, producers and directors (it is now more accurate in selecting the correct actor).

Back in a few.

Five films in this months film fight, a little bit more interesting than last month.

Steven Soderbergh impressed a lot of people when he remade rat pack classic, Ocean’s Eleven. For some people’s tastes the script was over-polished, a hundred writers making every line zing until the words practically lit up Clooney and Pitt’s teeth. It did, however, feature an entertaining (if somewhat silly) heist. In true Hollywood fashion, a sequel was inevitable.

Ocean’s Twelve. It lacks the polish of it’s predecessor (a good film), but is also missing a plot, reasonable nudges by the director, and crucially a heist. You see, although the team desperately need to get millions of dollars, you never actually see them commit a robbery. A few glimpses are shown retrospectively but it’s not enough to satisfy. That the key heist was pulled off in such an incongruous fashion to what one would hope for (it’s not a real spoiler to say it was a simple bag switch) is disappointing. If you can’t be satisfied with looking at the mugs of Hollywood A-listers as they show off what is essentially their summer vacation mess around, don’t bother.

Another remake next, Assault On Precinct 13. Almost entirely predictable, following the standard spot the stiff rules to the letter, only verging off where a modern film must. The cast put in reasonable performances and the isolated in the city motif works well to create an urban tension. Not bad, just not original; perhaps what should be expected from a safe remake.

We’re now several days after Million Dollar Baby took home the lion’s share of this year’s Oscars. It deserved every one of them. Although the character played by Clint Eastwood is an awkward stereotype to begin with, his good natured mean streak wins out in the end. The story itself is gripping, punching you through every blow that Hilary Swank’s character (perfectly realised, I might add) had to go through to get respect, and the struggle beyond that. Fantastic film.

A documentary to get a little change. The premise of The Yes Men is solid: activists set up a satirical website for the WTO and were then mistaken as representatives. They then proceed to give atrocious talks which represent WTO policy as it really is, as opposed to how it is stated by the organisation. The pranks are amusing, and the response is gob smacking, but the film is let down in several ways.

Firstly, the Yes Men themselves come across as prima-donnas. They seem to be in it for the attention more than the issues at hand. In fact, there is one scene fairly far into the film when they are speaking to another activist and he educates them on WTO policy; the very same thing they’ve been apparently blindly protesting. The editing is also lacklustre. While their lateness for one conference (forgetting about time zones) is an amusing aside, the film is filled with, well, filler. This might have made an amusing 45 minute TV show, but there is clearly not enough material to warrant an 80 minute cinema release.

Finally, live action anime from Japan, Casshern. Now, you might be asking how a live action animated film would work. The answer: human actors in appropriately ridiculous costumes, in front of huge CGI sets. Strangely the aesthetic almost works. Almost. It’s let down by an emptiness and some poor colouring choices. Sure, the world is supposed to be a dystopian future, but a few more people (drones) in the huge streets would make it fit in with perceptions better.

Sadly, the film itself is dreadful. Truly awful. I like anime, but this is of the low quality, silly, ill-planned variety. Stupid characters, overly long monologues about vague rubbish, an overarching philosophy that has only a few scraps of meat to it; bad. The timing is off (way too long), the directing is shameful, the characterisation is woeful. After seeing it, you’ll be dumbfounded by how it could go so wrong after an almost promising first half hour. Bad.

So, without a doubt, this months winner is Million Dollar Baby. The Academy finally got something right.