Author: Jonathan Sanderson

Via the Guardian’s Jemima Kiss on Twitter: Morris: A Life With Bells On

Fantastic.

The French Navy have retired their former flagship, aircraft carrier Clemenceau. It’s heading to Teesside to be broken up, and the steel will be… recycled. Yes, they’re recycling an aircraft carrier.

Not clear what they’re going to do with the 700 tonnes of asbestos, though.

Slideshow at the BBC

Oh, wait – this is a blog, not Twitter. I can embed things rather than merely link to them. I remember now.

Not my picture, I hasten to add. Wish it was, though.

I should probably retweet this, but until I integrate the Action Streams stuff with the main page, not so many of you will see it. Oh, actually – more people follow me on Twitter than read this blog, come to think. Ah, well, too late to back out now.

Via Matt Locke:

genius http://www.flickr.com/photos/hano1/3246320809/ (via @danhon)

Brilliant real-time snow report mashup. Gosh. Here’s mine.

Incidentally, they’re saying that South-East England has had the heaviest snow it’s seen in 18 years. Funnily enough, I was working in London, the winter of 1991. We had a couple of inches, and all the transport packed in. So I shoved my walking boots and woolly hat on, and walked in to the lab: Kilburn to Piccadilly. Central London was almost completely shut down. Oxford and Regent Streets were empty, barely anyone bothered.

The lab was shut, so I spent the day hopping between the few shops that were open, looking for a dinner suit in the January sales. I’m still caught out by how badly London copes with a little snow; I’ve never thought of a couple of inches as being anything much. Ah, well.

Alom’s Why is Science Important? project continues apace, with new postings most days and the film in full production. This contribution, however, is worthy of particular mention:

Maya Hawes replies.

She’s 12. Her article is brilliant. Spread this one far and wide, folks, it deserves to be read.

Tweet of the week already:

“Earl Grey is the final Ceylon.”

Loudmouthman.

I was a late adopter of Twitter, joining just a little under a year ago, and tweeting mostly about making tea. I’ve written about Twitter here rather a lot in the year since, either referencing it directly when discussing OhMyScience or my favourite STEM engagement project of 2008, the Mars Phoenix Lander twitter stream, or simply reposting things I learned from people I follow.

One convention on the service is to tag tweets about specific events or subjects with ‘hashtags‘, short unique strings prefixed by #. For example, for the BIG Event this year we’ll likely use #BIG09. They’re only a convention (at least for the moment), but they do at least allow people to define ad-hoc groups around events or shared experiences.

Looking for that group, however, is not straightforward – the tools are lagging the convention somewhat. However, as an aide-memoir (and a post for posterity, for some indeterminate time in the future when this all seems so primitive. Perhaps three months from now), here are some services that help:

• Search.twitter.com does a decent job of finding hashtags, and updates the page to alert you to new hashtagged posts.
• Hashtags.org attempts to catalogue hashtags, but seems more about trending topics than ‘finding everything under a given tag.’ Twitter search also collates trending topics, but from a straight keyword frequency plot, rather than by hashtags.
• Twitter Groups looks interesting – using hashtags to define groups by interest rather than event. Anyone using it?
• Twemes is another attempt to pick out what’s hot and trending in the Twittersphere.
• Twitterfall is something that’s been asked for regularly – a constantly-running search that updates as a column flowing down your screen, so you don’t have to refresh.

Ultimately, I worry that the ‘trending topics’ type analyses can’t conceivably scale, as Twitter grows. I can sort-of believe Twitter Search being able to sample a representative proportion of tweets, but surely anything else is going to be flaky, with access to only a tiny fraction of ‘the firehose’? In which case, they’re simply showing what lots of people are talking about. The whole point, surely, is to spot fast-growing minority topics?

Twitterfall is nicely done, but I fear it’s not necessarily that useful in practice. If I’m following, say, #omc09 (the Oxford Media Convention), I’m more likely to want to see everything tweeted there, not miss stuff if I blink. But there may be situations when I want auto-refresh, I suppose. Nice to know it’s there.

Finally, there are activity aggregators, Friendfeed being the principle player here. Many people swear by it, but I’ve never quite got my head around it.

An interesting development is the next version of blogging/CMS software Movable Type Pro – Motion is a roll-your-own activity aggregator, based off the system driving the Action Stream page here. I’ve yet to play with it, but look forward to doing so – I should also play around with Action Streams stuff, and integrate it with the main flow of the blog somehow.

There’s something rather appealing about being able to pull content from around the web – people talking about your product or project, for example – and catalogue it in real time, on your own branded page. It’s a dangerous concept, but a powerful one.

I’m (finally) posting new films to SciCast. Want an example? Try The Immiscible Love Story. Genius.

Insight of the day:

Don’t waste time trying to influence decision-makers. Just make decisions.

Apple’s Compressor, which ships as part of Final Cut Studio, comes with a distributed processing/rendering system. In principle, one can turn networked Macs into processing nodes, and push compression jobs across the network across all the nodes. This is of interest to me because:

1. I have ~300 short films to compress, to several formats each.
2. I have four MacBooks used for workshops as well as my MacBook Pro; each sports dual-core 2GHz-class CPUs.
3. I have a wired gigabit network.

Unfortunately, I’ve never managed to get Compressor/QMaster to, you know, work. I’d follow the instructions, define nodes, add them to a cluster, and … nothing. No jobs pushed, no error messages, nothing. Reading around the boards, Compressor clusters are notoriously finicky, and setting them up is something of a black art. Nothing I tried ever worked, and a year ago I gave up entirely.

But today, I read (and did) this page of set-up instructions. Now, I have four CPUs munging through an H.264 compression batch at ludicrous speed, with three more MacBooks back home in Glasgow to add to the cluster.

Wheee-hah!