Alex Muller

Here’s hoping Tweetie for Mac fills the void.

I’d known this for ages, but only just consciously realised that there was a way to get rid of those tweets that make me sigh painfully to myself. These regex expressions are what I’ve put into Twitterrific’s built in tweet filter.

New blog post.*http

I hate these notifications of blog updates on Twitter. This is exactly what RSS is for, and it works just fine. If I might like your blog, put a link to it somewhere (ooh, how about that section on Twitter that’s handily labelled More Info URL?) and I’ll check it out in my own time and decide if I want to subscribe to it.

^@.*http.*twitterfall

I absolutely love Twitterfall as a product and use it whenever there’s breaking news or events, but following the two developers means a whole load of replies to people I’ve never met advertising their service. Using this is an easy way to keep up with the interesting stuff they post, while getting rid of the links I don’t want to see.

twizmaster and [mM]ac[hH]eist

These two annoy me. The former is an account that’s used as a Twitter quiz service which people reply to with their answers (and reply to all too much). The latter? Search it using search.twitter.com and see for yourself.

And how to implement them…?

The easiest way is to use Secrets.prefpane and add them to the Filter Tweettext option (separated using the pipe character, as they’re regular expressions.

A few days ago on Twitter, I wrote this:

Hashtags are "a community-driven convention for adding additional context and metadata to your tweets." To see what I mean about normal words being tagged, go to search.twitter.com, pick a random word and put a hash sign in front of it.

Updates like this annoy me, because there’s absolutely no need to tag normal words in a tweet. For every person on Twitter who’s versed in all things #, there are many more who haven’t had the trend explained to them. This means there are people tweeting both "#episode" and "episode", so we’ve got two terms that need to be searched for to find tweets relating to TV.

I’ve seen some nice Twitter guidelines in the last couple of weeks (some on corporate usage spring to mind), so I thought I’d have a go at some for hashtagging…

When should I tag?

There are a couple of things in this category:

• When at an event, or a gathering of people – or trying to keep track of an event. Future of Web Apps and The Pirate Bay’s trial are two that have worked quite well in the past.
• When Twitter’s search might not give proper results for a phrase. #safari4 was one from yesterday, as I suppose “Safari 4″ could give some strange results.

And when should I not?

Essentially…

• When the search function on Twitter will give you exactly the same result. I cannot see the point in #plurk or #youtube. I suppose #lost (for the TV show) does make some sense, as it’s trying to differentiate from people just using the word "lost".

I’m not too sure where I stand on how Twitter is evolving and growing at the moment. Guess if I was forced to opine, it would be that they’re not developing, improving or evolving fast enough. I know they’re trying to keep the service simple, but the lack of a few would-be-amazing features (groups, anyone?) makes it seem stale and unloved.

This was prompted by me thinking up ways that they could make the site more useful, primarily in terms of adding more user-generated information while still keeping the original, simple tweet structure. How about a community wiki-esque (oooh, but limited to 140 characters – I just thought of that) system for explaining what the hashtags actually mean. I searched for #pop this morning (it’s a course I’m doing) and it came up with a whole load of results about burgers. I’d love to have been able to hover over #pop in the tweet and see a 140 character explanation of what this place was.

Spam’s an issue, clearly, which is why the system would be community moderated. Once a site has the giant user base that Twitter does, I reckon they could leave a lot up to the users.

Finally, a really, really basic screenshot of the kind of thing I’m thinking about:

Spent half an hour this evening trying to think on what the deal with iTunes’ new Genius feature is… nobody’s totally clear on how it works, so I thought I’d half-heartedly check into whether it works better for people with a more mainstream music taste or not. Here’s the results from me pulling a few dozen random opinions from Twitter and checking them against Last.fm profiles. The x-axis is an arbitrary score I assigned based on how the Tweet sounded, and the y-axis is an “eclectic” score given by some service. I’m also doing something fun with tags, but don’t know where that’s going yet.

As you can see, it’s sort-of kind-of maybe conclusive, but nothing that’s going to wow anyone. The people satisfied with Genius tend to have a less eclectic score (more mainstream taste), if that makes sense.

If you want lend me a hand, and can be bothered, give iTunes Genius an overall "satisfaction score" of between 0 and 1. I’ll compare it against your Last.fm music profile (and please leave a link if I don’t know where to find it). This should really be automated, but I don’t have the time, motivation or know how to put it together right now. Cheers!

I’ll post something more substantial when I have something more substantial to work off. Shockingly.