But which is better? Twitter or Tumblr
Only one way to find out… FIGHT. Actually, I was a bit dismayed to see Harry Hill mention the user-friendliness of Macs on TV Burp at the weekend. Et tu Harry?
One of the most appealing things about Twitter is the basic lack of rules – just type some stuff – but the elegance of using some of the features for cool results. If you’re not convinced, think of the smart #uksnow mashup which is a lot better (more relevant) than the the new BBC Weather site.
Structured Microblogging
Tumblr is a bit different in that it provides what is essentially a regular blogging platform but with structured types of post (Text, Video, Audio, Quote, Chat etc.). I like it a lot, but it suffers from limited clients, mashups and lack of conformance to Blogging APIs (so I can’t use Tumblr from Live Writer) meaning that I have to use esoteric clients or mechanisms (Tumblweed, or the odd Firefox extension) to work with it.
So Tumblr has its issues, but then so does Twitter as it becomes increasingly hard to filter the flotsam from the jetsam (which is more useful, Flotsam or Jetsam..? You get the idea with that…) This will only be compounded as the number of users grows.
Flotsam and Jetsam
So what you get with Twitter is an expression of “Now”: this is what is happening RIGHT NOW. If you’re tuned in, and a good processor, then there’s a lot of value. There’s not a lot of help to do that though (Tweetdeck helps a little). RIGHT NOW on the internet is an increasingly small amount of time. (Whether Twitter is a cause or an effect of this is a debatable point).
In principle, with a little more structure, then I could “dial down” the noise that I don’t want (#mutterings, #nonsense) and filter for (#codesnippet, #giantnewsevent). You could use hashtags to fake that up, but it’s an honour system of course.
Twitter As A Search Engine
Similarly, I can’t quite use Twitter as a search engine. Go on, give it a try: replace your browser search engine or home page with http://search.twitter.com and see if you can survive a couple of days (or compare the value of the search). Maybe you can… (Oh and then switch (back) to www.live.com (Well, I can try…))
Now, it’s a shame that this doesn’t work, because the short messaging of Twitter tends to force the use of keywords (metadata!) and less of the chat (see the length of this post already?). But a bunch of intent is lost because of the URL shortening services and the lack of parsing of those links when considering search. So tweeting: “i love this: <some service>/shorturl” is not much help in a search scenario.
It’s also a shame because an answer that works RIGHT NOW is an important part of search (I’m thinking techy, as that’s what I am). I find, as I’m sure everyone does, that wading through competing advice owing to “differing versions” (or similar) is a real pain with any search engine. This might be OK for a developer as you may just need the general gist of a solution in order to path find, but it isn’t good enough for many other disciplines.
I find that I have enough connections now that I don’t need a search engine – I use Live and Google to retrieve information that I know is already there rather than to actually find anything. The stuff I need just seems to arrive. So can human-organised micro-blogging replace search engines? Do I care about the (now gigantic) long tail that search provides?
What’s Happening?
The combined effect of the users of Twitter reminds me a lot more of “The Machine Is Using Us” than the actual examples used in that video. I wonder what Prof. Wesch thinks? (An aside, this awesome World Simulation with Twitter video is worth the watch if you haven’t seen it.) But it is weakened as the information organisation itself is too limited to appeal beyond immediate observation. (Do you go on holiday for 2 weeks and then read all the Tweets you missed?)
What would be nice?
A couple of things from me:
- Enhanced structure/semantics for messaging. This way I can follow more people and get more of what I need. There’s an opportunity here for a successful client (like Tweetdeck) to do this using an informal mechanism (styling a tweet using a hashtag (or some other notation – like ! notation)). You could even charge brands to skin incoming tweets from a brand… I expect some kind of nod when you implement this and make some money out of it ;)
- Improved metadata. The tiny urls are all well and good, but it’s devaluing the benefits of the overall organisation of information. Similarly, there are other attributes - mood is a good one – that aren’t represented well.
- Triggers and Workflow. I loved that #uksnow mashup so much that I’d be happy to contribute to a crowd sourced weather “now-cast” every day. But I need to be prompted to do that. Building in a structured way of achieving this kind of information assimilation would be cool. Again, this could be achieved by a client only with some smarts.
This stuff would really provide an engine: a command line for the web as my friend Jeremy said.
So I didn’t answer the opening question: which is better? I don’t think it matters, but this is a compelling space for evolution of how we interact, and the value in that.
I find it interesting that Twitter is teaching the world to pseudo-code, and organise, even though they don’t realise it a lot of the time. Let’s hope they give us all stock options for our efforts!
