Following some comments and feedback today on my live blogging from ILI08, I thought I’d post a few more thoughts on this.
I found using Twitter a pretty good way of posting to the live blog. When I lost wifi access on my laptop and switched to using my iPhone for a bit, there was absolutely no problem – I could just keep blogging.
Rather than using my normal twitter account to do the blog I setup a new account that was meant to be event specific. This allowed me to keep the live blog tweets discrete from my personal tweet, and meant I didn’t overwhelm my followers with tweets (except those who chose to follow the new account). However, I did find that Twitter expects a unique email address for each account – and the idea of setting up a new twitter account, and a new email for each event I want to blog is not appealing. So I suspect that I just need a twitter ‘live blog’ account alongside my personal one.
One final thing was that some of my existing followers sent comments to my normal twitter account, so I retweeted to my live blogging account – clearly there is scope for some confusion here, and another reason for not having more twitter account than absolutely necessary.
One of the other issues I wanted to tackle was differentiating between reporting what the speaker was saying, and my own comments on it. I had an idea that by using my personal account alongside my live blog account, I could differentiate between these two things. However, this felt a bit artificial, and I think risks losing the ‘voice’ from the blogging – I’m not sure I would use this device again.
One of the commenters on the UK Web Focus post that started me off thinking about Twitter for live blogging pointed at a service http://livetwitting.com/. I think this used in conjunction with a dedicated ‘live blogging’ Twitter account could well be a great solution – I’ll try to remember to give it a go next time I’m at an event – I especially like the way it supports annotating the blog with session and speaker names (and Q&A bits). The other thing is that since you are doing it via Twitter, even if livetwitting doesn’t work so well you’ve still got the twitter stream.
One of the other things I liked about the idea of using twitter was that it would be possible to manipulate the output, and this was true to a certain extent. My preferred way of extracting the liveblog was using the Twitter search API – I used a search for all tweets from the ostephensili08 account, and all tweets referencing the account – the syntax is extremely simple, and you can output results as atom or json. However, one issue is that you can output a maximum of 100 tweets at a time, and there doesn’t seem to be a way of knowing how many tweets in total have matched your search result – so when pulling these results together I have had to manually work out I have 4 pages of results.
I pulled together these 4 pages of results into a single RSS feed/JSON file of results using Yahoo Pipes. However, in some cases (see below) using the twitter search results in their raw atom format.
Chris Keene left a comment in my last post suggesting the use of FriendFeed – I need to have a look at this and see how it works. Chris also shows how Dipity can be used to display the twitter stream – so thanks to him I’ve setup an account and used the Twitter search api to bring in all tweets from @ostephensili08 and any replies sent to this account (which is mostly me talking to myself) – there seems to be a problem with Dipity consuming my merged results set via Pipes, so I’ve just used the raw atom feeds from the search api, giving dipity 4 URLs as RSS sources.
The Timeline is perhaps the obvious way of outputting the results – but, I found the map display very interesting as well, as although I only had a few tweets with places in them, I actually found it interesting to have these picked out and see what the context was.
If anyone has any other visualisation suggestions, or ways of displaying the output, leave a comment
Link to my dipity account dipity / ostephens