Archives for August 2007

Stop this modern type tagging right now!

There have been some interesting reactions to posts by Paolo and I recently about the state of tagging in 2007.

Marc Canter says we're moaning which I really didn't think we were. However my old guvnor, Mike, described my post as a whinge so I so I guess it must be true :-)

Marc then goes on to lambast Technorati somewhat. I think del.icio.us and Technorati are to be commended for doing what we could not, i.e. make tagging commonplace. I'd rather have unsophisticated tagging than no tagging at all. As Paolo notes, when we did the first K-Collector demo at Blogtalk we got significant push-back to say that people wouldn't tag things themselves.

Among those who gave us food for thought was David Weinberger who also mentions in his neatly titled Tagging like it was 2002 Paolo's efforts with N??va100 (which, although based on K-Collector, I was not involved in).

Feedback from people like David was valuable because it pushed us to work harder on making tagging a simple process with recommended tags, auto-discovery, smart UI shortcuts, integration with blogging platforms, and so on. I think we did a good job given our resources. But, ultimately, we felt we had little widespread impact.

These days everyone tags and the idea that the bulk of people won't seems faintly ridiculous. That was pretty hard to predict back in 2002/3.

Phil Pearson, who I never met but did talk to from time to time, reflects that tagging is alive and well in services like Flickr. However I would argue that the impact of tagging on such services is narrow and that innovation has been largely absent. More importantly Phil remembers that the genesis of both K-Collector and his own Topic Exchange service lay in the role of tags (and in particular tag sets) as channels.

There was a lot more to do with the channel metaphor that Paolo and I had talked about but never got around to implementing. In particular the role of expertise and reputation in creating personalized channels.

Thomas Vander Wal whom, alas, I also have never met also reflects on stale tagging. Although he aludes to more innovation going on in the enterprise space I can't say I found his examples terribly compelling. I mean, we had tag-stemming, faceted classifcation, related tags, and tag sets all working in K-Collector in 2003.

It seems to me that the last four years have largely been about re-inventing and re-implementing the UI for adding tags to things. Ajax has helped but making it easier to tag has lead to new problems.

As David Sifry notes Technorati has done a marvellous job in making tags uber-prevalent. However I was glad that Stephen Downes relinked to his talk about community blogging from Northern Voice in 2005:

Now one way, a very popular way, of trying to fix meaning to a blog post, is through tagging. Tagging has been the rage. I'm also anti-tagging. Why am I anti-tagging? Well, take a post, any post, and ask yourself, what would a graph of all the possible tags for this post look like? You are going to get a power law. So you have a post - somebody's written something about the Prime Minister - and so, you know, you have 'Martin', very popular, that would be a very commonly used tag, 'tax break', that might be a commonly used tag, 'my goldfish', maybe once, by somebody who didn't get the concept of 'prime minister'. You're going to get a power law curve of tags.

But the thing is, if you do it that way, then the meaning of the post becomes whatever tags are sitting there in the big spike. Right? So the post becomes, it means, that tag. But that tag contains only a part of the meaning of the post. It's a very narrow, one-dimensional look at something that might be a lot more complex.

Stephen is yet another guy who I've had some interesting discussions with but never met. I think he highlights one of the root problems. As tags have become more popular the spikes he describes have gotten spikier but the tools to extract meaning from the totality of tag information and relations just hasn't improved. That's the trough of disillusionment right there.

What is lacking is tools to get at the meaning. The genesis of my work in tagging first in liveTopics (oh so many years ago now) and then in K-Collector was in my reflections on David Snowdens work on complexity and sense-making. To me tagging was a necessary step on the way to building tools to help us make sense of our world.

I saw tools like K-Collector as being the first rungs of a ladder leading towards more sophisticated sense-making tools that have just never materialized. It's as if the lesson was that tagging, in and of itself, was important and not what tagging enables you to do.

So I stand by my original post. I prefer to think of it less as a whine about the state of tagging and more of a sense of puzzlement coupled with excitement that there are so many opportunities remaining to do good things in this space!

30/08/2007 13:33 by Matt Mower | Permalink | comments:
More about:

Moving on

Just a quick note to say that I resigned from Cominded this week to pursue another opportunity about which I'll probably have more to say later. However I wanted to say that I'm grateful to Paul for the opportunities he's given me and in particular for the chance to work with such a nice group of people. I wish the team every success for the future.

I'm now taking a few days off and, simultaneously, grappling with going back to a G4 powerbook after a MacBookPro. It's still a nice machine but it does feel like a step into the past ;-)

26/08/2007 10:20 by Matt Mower | Permalink | comments:
More about:

To the trough of disillusionment we go!

I read a post on Paul Walks blog some time back: Tagging: Are we in the Trough of Disillusionment? where Paul is enquiring as to the state of tagging.

I started tagging on this blog back in June of 2002 and, with Evectors, I co-created K-Collector which was all about tagging. I think I know a little bit about tagging and the value of tagging. From what I can see we are nowhere near the "Slope of Enlightenment".

I have been surprised, disappointed, and excited that, despite the widespread adoption of tagging across many applications, the state of the art in tagging seems firmly wedged in 2003. Surprised because there seemed, despite the expectations of many that nobody would tag things, to be a momentum building in the use of tagging. Dissappointed because I expected to be using applications that really used tagging to do some interesting things. Excited because it means the field is still open.

I think the uses of tagging definitely forms part of my personal research agenda. Although I can't claim to have done anything innovative in the field since 2004 I have not stopped thinking about it and have ideas on the drawing board that will hopefully see the light in 2008.

Q: How can we develop tagging to fundamentally change how we think about and explore ideas?

Tagging in 2007 seems to have advanced no further than a means by which one or more users of a site (or application) can group content around a loose framework of concepts. If you are lucky those concepts are may express relationships but often they do not. In K-Collector we expanded tagging to include relationships as well as a prototype faceted classification along the lines of Who, What, Where.

What new developments in the use of tagging and classification systems will change the way people manage, transform, filter, and use the kinds of information they are already working with? What sort of applications will benefit most from advances in tagging?

23/08/2007 15:21 by Matt Mower | Permalink | comments:

Tunnelling BitTorrent over SSH

I'm making a note in case I hit this particular wall that you can tunnel BitTorrent over SSH. This came up when I read a post about Comcast breaking BitTorrent uising a technology called SandVine.

The usual ISP argument goes like this "If everyone uses the bandwidth they've paid for our network will collapse." This argument is horse shit. What the ISP is saying is that they are massively overselling their bandwidth and if people used what they thought they were paying for the ISP couldn't cope. It's like when I turn up to the airport to find that my flight is "overbooked." How do you sell more tickets than you have seats? Willfully, that's how.

I would counter that ISP's should have their feet put to the fire and sell what they can provide at prices that are profitable. That gets rid of "fair usage" at a stroke. It's not about some arbitrary measure of "fairness" but about what I have paid for and can legally use. Of course the ISP's and Telcos will fight this bitterly because it would put them on a level playing field where they could no longer operate a cartel and would actually have to compete leading to a reduction in price and improvement in service.

Our problem as consumers is that telcos & ISP's have lobbied so successfully to corrupt the law and strange all methods of competition that as consumers we are forced to put up with them if we want an IP dialtone.

23/08/2007 12:02 by Matt Mower | Permalink | comments:

Me and my research

I read Brad Neubergs post Creating a Personal Research Agenda yesterday and it lead me to actually reading the paper of Richard Hammings talk You and Your Research at last. The thrust of Hammings talk was about "doing great work" and why so many people who have everything going for them don't seem to amount to anything great (although many amount to something good).

Hamming sums things up this way:

I claim that some of the reasons why so many people who have greatness within their grasp don't succeed are: they don't work on important problems, they don't become emotionally involved, they don't try and change what is difficult to some other situation which is easily done but still important, and they keep giving themselves alibis when they dont. They keep saying it is a matter of luck. I've told you how easy it is; furthermore I've told you how to reform. Therefore, go forth and become great scientists!

Now I'm not a research scientist. I'm not paid by anyone to do research. But this smacks a little of an alibi. I think in this blog back in 2002-3 I was getting close to at least defining some important problems to work on. Then I went off the boil.

Reading and re-reading Hammings paper has refocused me on getting involved again. At the moment my interests are still very eclectic but there are themes:

I am reading about factor analysis. What interests me about factor analysis is that it is a tool that allows you to discover structure. I am also interested in tools like clustering which also allow you to discover structure but, in many cases, they seem to require guidance. For example when using a clustering algorithmn like K-means you decide how many clusters you want and, often, the order in which you feed things to be clustered affects the outcome. Factor analysis is a statistical technique closely related to pychology. I'm not sure how widely used it is outside psychology. It's an area I would like to explore.

The structure of information and of relations between information has always been of interest to me. K-Collector was, at heart, a way of exploring the structure of related information through the evolution of a taxonomy. It was a primitive tool by the standards of what will come but I think it was an interesting one. I'm still very interested in tagging and tagging systems. I'm especially interested in faceted classification because it exposes the underlying structure of the taxonomy in what I think are useful ways.

I'm interested in what we we do with information, how we structure it and how we make use of it, how we filter it, combine it, and transform it. Information has never been so malleable as it is right now and this trend will only increase. I think links are fascinating and a good starting point for this problem. I am also interested in how we see information, how we visualize it.

Now this is just a set of interests. What it is not is a list of important problems, great problems. I think this is where I went wrong, I lost sight of the problems. It's hard to be emotionally committed to an interest, it's too nebulous. But a really thorny problem that holds out great reward is something you can commit yourself to and, in many cases, permits no other response.

But I am also aware that it's very hard to focus like this. Hamming himself did not try to work 100% on great problems but set aside Friday afternoons for it. He figured that 10% of his time, compounded over a long period, would be significant. I wonder if this is behind Googles 20% "personal time" practice although I am not clear that people are encouraged to work on "great" problems with that time.

I need to reformulate my interests into a set of problems and work out whether solving them amounts to a hill of beans.

23/08/2007 10:28 by Matt Mower | Permalink | comments:

The only good reason...

...not to be reading Alexander Coburn's blog is that you are already utterly depressed by the cesspit that is American politics. From his piece today on Karl Rove's departure we can see that, though Rove may cut an ugly figure, Coburn doesn't paint him as the svengali he is usually portrayed to be:

The prime task of a political counselor is to keep his patron's polling numbers high, and enhance his political clout. Rove leaves his employer with ratings in the low thirties, with almost zero political capital in the bank.
[...]
Rove's last turn as arch demon of choice was as the alleged instigator of the firing of several federal prosecutors. Federal prosecutors serve at the pleasure of the president, so their arrivals and departures are always part of everyday political logrolling, as practiced by every presidency in historical memory. As with many of Rove's political duties, this was handled with matchless incompetence and so the Democrats have had a field day.

His conclusion:

The administration will plod on, with the grand vizier still ensconced and unlikely to depart until the full term is served. You can bet vice president Cheney will be at Bush's elbow right to the end, making sure all the names on the pardons list are spelled right.

The best reason I can think of to impeach Bush and Cheney is that, out of office, presumably they wouldn't be able to pardon anyone. Still something to play for in the lame duck year then...

21/08/2007 15:46 by Matt Mower | Permalink | comments:
More about:

Could Ron Paul be for real?

Okay is there something about the state of Illinois that would make it a whacky outlier? Because I read on Steve Dekorte's weblog that Ron Paul just came third in a straw poll of republican candidates polling nearly a fifth of the votes and only 1% less than the 2nd place man (although still over 20% behind front-ronner Mitt Romney).

Knowing as little as I do about the mechanics of presidential nominations I am bound to wonder: Is this significant?

18/08/2007 15:00 by Matt Mower | Permalink | comments:
More about:

Not getting Google lovin?

As long-time readers will know back in 2005 this blog was mysteriously demoted in the Google index for a number of search phrases that it was normally ranked #1 for (e.g. "Matt Mower"). It didn't disappear altogether, you could still find it if you were prepeared to navigate to page 150 of the search results or thereabouts. I was never able to figure out what caused, my pages were being indexed by Google and none of the SEO optimizations I subsequently did seemed to make any difference. It reappeared, as mysteriously as it had disappeared, in the #1 spot about a year later.

I just read an article about Google Proxy Hacking that suggests how it might have be done. I've certainly been made aware of proxy sites that replicate content from this blog as original material. I hope this outting might raise the profile of this problem and get it addressed.

17/08/2007 13:05 by Matt Mower | Permalink | comments:
More about:

Dock the Blue Danube

Here is a link to a movie demo'ing Dave Dribin's entry at C4[1]'s Iron Coder competition. It's called The Bouncer and it's unlikely premise is an app that can bounce other applications Dock icons. How could this be cool you ask? Watch the movie and, if necessary, skip forward to half way through and watch from there. It's a beautiful thing.

17/08/2007 12:03 by Matt Mower | Permalink | comments:
More about:

Only freedom has compassion

I'm watching Ron Paul speaking at Google back in July. I've just listened to his defence of the value of the US constitution and doing things in a constitutional manner and now his views on free trade and contracts. The man is thoughtful, eloquent, articulate and seems to be principled.

It's very hard then not to contrast Ron Paul with the smirking, vindictive, half-wit that a majority of Americans have twice elected twice to head their government. You might not agree with everything Paul says but he seems to have a more logical basis to his views than most candidates. I've heard it argued that his financial ideas are crazy but he's given a robust defence of sound money and free markets (and by that we mean a market free of the kind of government intervention that creates protected monopolies such as we have now). I find it hard to believe that what he's suggesting could be so bad but, worst case, could he actually do more harm than the shrub? Give the man a chance.

A while back someone told me that there was no way Ron Paul would ever become president. I've been heartened that his campaign has, so far, seemed much more alive and relevant than I might have imagined. I just pray that America begins to wake up to the McPolitics they are being served by the traditional company men from the "left" and "right".

Almost at the end now. Although I've been following Ron Paul for some time this is the longest time I've seen him speak and answer questions. His message of freedom is impressive. If I were a US citizen I would be proud to vote for him and hopeful about him representing me.

16/08/2007 22:23 by Matt Mower | Permalink | comments:

Everything Nu is old again

Just read about Tim Burks' (the guy behind RubyObjC) latest brainchild, Nu:

Nu also is the name I chose for a project that relates a different C to a different lambda. Nu is a new programming language that binds the expressive power of Lisp to the pervasiveness and machine-level efficiency of C by building on the power and flexibility of Objective-C.

Another goal of Nu was to make it easier to build and reuse software components. Consistently with that, Nu was built from many preexisting components and many well-developed ideas. There's not much that's new about Nu except for its particular combination of recycled parts.

Nu is object-oriented and functional. It is written in Lisp-like S-expressions but conforms to no preexisting Lisp standard. Instead, Nu was designed to follow Yukihiro Matsumoto's "Principle of Least Surprise" while staying tightly integrated with C via the Objective-C runtime and object model.

Reading Tim's outline it sounds great. As a Ruby programmer first I find Objective-C generally okay to work with but often cumbersome. Not being able to pass blocks around, for example, is a clear step backwards in usability and leads to less-expressive code. However the problems that Tim outlines with the various bridges have always, when push comes to shove, made them unattractive to me and I resign myself to sticking with Objective-C.

Because Nu is a language layered on top of Objective-C it offers all the benefits with none of the downsides. I'm not afraid of Lispyness and think this might be the first Lispy language that I might learn with an expectation of using. Chatting with Marko Karppinen in #macsb:

[09:04] nutrimatt: hrmm... Nu looks very interesting
[09:04] nutrimatt: did Tim demo it at C4?
[09:05] Markonen: he did demo it
[09:06] nutrimatt: Markonen: any thoughts?
[09:06] Markonen: looks awesome

According to Marko, Tim is looking to get together a team to support Nu before going public with it. I hope Apple might take an interest in this project. Their continued use of Objective-C shows that they know when they're on to a good thing, and their increasing support for scripting languages like Ruby and Python show that they're still looking for competetive advantages in app building.

Nu, if it lives up to expectations, might be a win-win for Apple.

16/08/2007 09:21 by Matt Mower | Permalink | comments:

Lazy function definition in Javascript

Came across a neat technique in Javascript from Peter Michaux's blog:

var foo = function() {
    var t = new Date();
    foo = function() {
        return t;
    };
    return foo();
};

the idea is that the first time you call foo the function creates an object (in this case a Date instance but it could be something considerably more expensive) and then redefines itself to be a new function that returns that object. His application (which you can read in more detail in his post) is optimizing browser detection/dependent code.

16/08/2007 08:52 by Matt Mower | Permalink | comments:

First impressions of Numbers

I'm not exactly a spreadsheet maven but I do use Excel occasionally and, usually, have a hard time with it. It works but it's not the easiest application to get along with. Since I will probably upgrade to iWork '08 for the new Keynote I thought it would be a good idea to try Numbers as well.

I installed the 30-day trial (not sure why they seem intent on hiding that link, I couldn't find it anywhere on the iWork site) and fired Numbers up. I started a new sheet based on the budget template. Immediately you see how Numbers shines over Excel. Instead of a big grid of cells you have a nicely laid out set of tables that present the structural content of your spreadsheet in a way that is both friendly and easy to manipulate.

The formulas in the various tables are all interlinked and you can add new rows to the tables just by dragging them (or using the add above/below context menu items). The pop-ups (for selecting months in this case) work nicely. In fact the whole thing works really nicely and I was very quickly able to build a nice budget spreadsheet like I never quite managed to in Excel.

I've heard that Numbers performs poorly on even moderately sized data sets that Excel has no problem with. It's a 1.0 product so I guess I don't find that too surprising and expect it will improve. In terms of usability though I think Numbers kicks Excel's ass. If it weren't for the Ubiquity of Office I could see a lot of people using Numbers. It's a pity for them that they probably won't.

12/08/2007 12:51 by Matt Mower | Permalink | comments:
More about:

Slow going on the app front

I feel like I may have reached the high-water mark of underachieving: I have a bunch of updates to do on Diffly and Transference that I'm not getting around to, a new Mac app in the works that I'm bogged down on, and a Rails app that I should have launched a month or more ago.

Maybe I need my chakras realigned or something.

12/08/2007 11:50 by Matt Mower | Permalink | comments:

Web&Walk + N73: A mixed but generally good bag

On Friday I picked up a Nokia N73 on T-Mobiles Flext 25 + Web&Walk tariff. That's the cheapest 12 month contract (I've not yet come across a cell-phone company I'd willingly sign an 18 month contract with) they do and, although it's a little more expensive than my orange rate, it includes all my mobile calls not just same-operator calls, weekend calls, and 1GB of mobile data. That seems pretty value good for £32.50.

I'm still trying to decide whether handset insurance is a good idea. Normally I'd go for it for the peace of mind but given that the replacement value of the phone appears to be about £150 paying between £7 and £9 per-month for a year seems like a pretty poor deal. What's not clear is how easy it is to get a replacement handset from T-Mobile without buying another contract. Will they sell me one. Probably I'd better check.

Having broadband internet on my phone is proving to be great (Woo!) with some caveats (Hiss!). Connections are slow. I remember this was the case with GPRS and doesn't seem to have improved in the last 5 years; However once the data starts flowing it's about as fast as my 1Mbps ADSL at home which is more than good enough for my purposes. What I don't like though is the frequency with which I seem to get gateway/connection error messages. Sometimes retrying gets me through, sometimes I have to leave it and try again later. That's frustrating. I seem to be getting a good 3G signal if I am reading the phone right. Is this contention on T-Mobile resources? Anyone else on T-mobile have any suggestions?

The built-in web-browser for the T-mobile edition of the N73 is pretty good. My colleague Alan Bradburne tells me it's based on Webkit which is pretty neat. I also downloaded Opera-mini at his suggestion and that looks good too. $30 is not a lot to pay for a marked improvement in browsing speed. I am testing out a couple of instant messenging solutions: IM+ and Fring. I don't care for either interface that much although IM+ would edge it if I could ever make it connect to AIM. As it is I seem to be stuck with Google Talk and Skype, neither of which are preferred IM protocols for me. Anyone who knows the trick for making IM+ connect to AIM I'd be grateful if you'd share it.

I've got both the Google Mail and Google Maps for S60 installed. Mail's not bad although it doesn't seem to support the concept of draft messages which is a pity because it's the first idea that presented itself for note taking. Maps seems less good but I've had connection errors most of the times I've used it in anger which may have coloured my impressions.

Other than AIM instant messaging my primary need now is a good note-taking facility. Ideally a way of recording voice notes since typing on the N73 keyboard isn't exactly a pleasure. I suppose I could record video but I'd like something a little more convenient for note-taking on the fly.

In use the N73 is good. The camera is great and the gallery makes it easy to upload photos to Flickr. The Nokia Media Transfer software for the Mac works nice and the iPhoto and iTunes integration both did their job. The N73 music player is adequate and the sound through the headphones is okay for casual use. I'm pretty pleased with the handset and feel my decision not to get the N95 was the right one.

At the moment I am waiting for my number to be ported across to Orange. Since I bought the phone Friday I am hoping that will happen some time tomorrow and I can get my Spinvox voicemail back and stop carrying two phones. Why this takes 5+ days is beyond me. Orange are claiming that the delay is T-mobile. Maybe. But I don't trust Orange at all. I will be (at least momentarily) happy to get away from them.

All that said I am left with two gripes. One small and one not so small. On the small side my experience is that real-life battery life for the N73 is about 48hrs when you factor in bluetooth, data, and some MP3 playing. I guess that's not bad but I would have been a lot happier if it could get to 72hrs so that I could comfortably go a weekend without needing my charger.

However the real kicker was when I asked about the price of roaming data and was told it was £7.50 PER MEGABYTE. £7.50 per megabyte? WTF? Pro-rata that makes the 1GB allowance I get in my contract (for £5) cost £7,500!! I can only assume that is a money trap for the unwary and it sucks because having mobile data while traveling would be extremely useful.

In terms of the N73 & T-Mobile combo I'd have to say I'm very pleased and I'd give it 3.5/5 which would rise to 4.5/5 if I didn't get any more gateway/connection related messages while browsing.

08/08/2007 14:19 by Matt Mower | Permalink | comments: