With regard to Laconi.ca, it's open source so in theory you could actually modify it to somehow feed into Twitter for people who want the information there while it also being a completely separate system that you could run entirely as you see fit without the need for Twitter's backbone. That way Twitter becomes just one way for people to access Twaffik and it's still there if Twitter goes down.

Anyway, just a couple of comments in response to the others.

Twitter seemed the natural option as that's where the crowd is and we didn't want to go creating a 'new crowd' using laconi.ca – but having said that I'm ignorant to the capabilities of laconi.ca, so maybe there is an angle to investigate.

But, the underpinnings of Twaffik are that it's an experiment. We're not a startup in it for financial gain, which maybe people expect us to be, but actually running this as action research as part of the Institute for Social Media. It came out of frustration of getting stuck in Manchester with little information about the travel infrastructure…and because people leave work at all kinds of different times of the day, I think it's valid to the messages pinging throughout the day.

We're interested in what it takes to get something like this going, how it grows and what utility people get out of it. It's an innovative approach to traffic news and a different model to commuterfeed, which I must admit I hadn't heard of when we started on Twaffik. Even more than that, we want to see if we can turn something like twitter into a service that has appeal to the mainstream – that's a big ask because twitter clearly haven't done it themselves as yet.

The hashtag option is interesting and feasible, but I'm starting to experience hashtag overload and there's no opt out – if you are following someone (or more likely a number of people) who are using a particular hashtag then you're likely to get bombarded – unfollowing them all isn't an option – is there another way? With the follow model you can simply unfollow if it becomes a pest and just look at the twaffik.info website.

Whether that little red bus becomes a pest we'll have to see and as it's action research we'll adapt if it does, based on the feedback we get.

Similarly, in that vein, we're really interested in the suggestions from the Manchester Twitterati…maybe we'll open this up for further development to willing volunteers. We'll see.

It also highlights something that Alan has been talking about for a while. There's a wrinkle in Twitters model somewhere here that sits on a fine line between being able to filter out the noise and retain the good stuff…but I'm not sure if they are mutually exclusive facets of the service – as JNathan alludes to.

As for the future, we have plans in the pipeline for extending the experiment in Manchester and beyond, but for right now hopefully we can get a few more manchester twitter users on board.

-pc.

I guess they are halfway there with their device notifications system, which can be set on a per follow basis. That effectively makes a two group system, though sadly now useless in the UK without IM or SMS.

Also, we are aware of the potential problem of contributors' Twitter streams being overrun by lots of big red buses, but is this more a problem with the way Twitter works? There's a similar problem when people you follow use #hashtags to live blog an event.

It's also worth making clear that Twaffik is an experiment at this stage. It's part of the work of the Institute for Social Media's research work – see http://i4sm.org for more.

Thanks, Alan H.

The problem with the current implementation is that I probably don't want to know about every traffic situation in Manchester and there is no way of automated filtering of tweets received. And as you point out as the service scales its usefulness in appearing in the regular Twitter timeline diminishes.

Perhaps long term we could get some location aware goodness in here. i.e. a Twinkle-esque show me all @twaffik replies within a 5 mile radius