blogging california england firefox glassfish google hacking j2ee java openid opensource roller skype soccer sun sunray thewaronliquid travel treo ubuntu vaio voip web2 work yahoo
May
1

Here's an idea.

I'm a fan of delicious and tagging in general but unfortunately a lot of content I'm interesting in doesn't find me while I'm surfing the web - I actually still read books, magazines, billboards, watch TV and listen to the radio. The following happens to me a lot - I see an interesting book review; make a mental note to buy it then forget about it - ditto for tunes, films and virtually any product you can imagine. I'm sure I'm not the only person with the short term memory problem. But the people that really have the problem are the companies that are trying to sell those products - every time the above scenario plays out (and I'm sure it does a million times a day) - a potential lead turns cold and a sale is lost or at least delayed.

Imagine a service where all you had to do was remember a short word or alpha-numeric identifier and you could use mail, SMS, the web, the phone to register the things you're interested in. The identifiers could be categorized (ie. radio, print, tv, billboard, etc.) to keep the namespaces smaller and therefore restrict the size of the identifiers easier to remember; avoid errors, etc.. Those identifiers then appear in you delicious queue of tagged items so you can re-categorize them or research them further.

For example - I see a film poster on a  billboard - I remember the Identifier ZZG238 for as long as it takes to send an SMS mesage - and I add the context / categories "billboard" "film". When I get home - I connect to the tagging service and see the identifier has been found and resolved and it gives me further options - buy, search, queue, etc. Very simple.  There's margin for error - sometimes I'll get the identifier wrong but the context should allow for error correction or allow me to resolve the ambiguity - ie. ZZ?238 - did you mean song or hairspray ? The SMS route can also provide additional context - ie. when and where.

I did see a demo from (HP I think) that used a camera phone to scan a kind of barcode - but that limits you to thing you can see and a fairly close for long enough to take a picture - so it's pretty limited. 

Or I could carry a notebook and pen with me at all times :) 

What do you think - is anyone working on this ?  I'd happily be a beta customer :)

 

Jan
8

OK, I just tidied up the tag cloud code a bit; there's now some scaling to ensure the 'intensity' is sensible. The template code relies on some styles which set the font size and opacity.

    <div class="tagcloud">
       #set ( $maxtags = 25)

       #set($mytags = $model.weblog.getPopularTags(-1, $maxtags))

       #set($maxtagcount = 0)
       #set($mintagcount = 99999)

       ## number of styles s1-s10
       #set($steps = 10)

       ## determine min and max tag count so we can scale the fonts and opacity
       #foreach ($tag in $mytags)
          #if($tag.count > $maxtagcount)
             #set($maxtagcount=$tag.count)
          #end 
          #if($tag.count < $mintagcount)
             #set($mintagcount=$tag.count)
          #end 
       #end

       ## generate the link for each tag assigning style accordinly
       ## the styles : s1-s10 increase in font size and opacity   
       #foreach ($tag in $mytags)
          #set ($intensity=(${tag.count}-${mintagcount})*${steps}/(${maxtagcount}-${mintagcount}))
          <a class="tag s${intensity}"  href="$url.tag($tag.name)" 
             title="$tag.count entries">$tag.name</a>
       #end
    </div>
     
Relevant fragments from the stylesheet follow, first I set the abolsute position of the tag cloud then  reduce the line-height for effect. Then I define s1-s10, increasing the font size and opacity proportionally. The opacity requires different hacks for different browsers and I still don't think I have all the IE versions covered. Sigh.
   div.tagcloud
   {
      line-height:18px;
      position: absolute; 
      text-align: right; 
      top: 90px; 
      right: 20px;   
      left: 600px; 
   }

   .s1 {
     font-size:18px;
     filter:alpha(opacity=20); 
     filter:progid:DXImageTransform.Microsoft.Alpha(opacity=20);
     opacity:.2;
   }

Nov
6

Finally - b.s.c now has support for tags. It even does delicious style completion in the built-in editors! I've been waiting for tags for a loooong time. At the weekend I mucked about with the macros for displaying tags - when I get time - I'm going to see if I can produce some tag eye-candy. I also think I'm going to remove the categories - I think they're a bit redundant. That of course means the Dojo fisheye category chooser has to go :(

One problem I now have is that I can't post from Performancing as it doesn't support Roller's tags. Wouldn't it be nice if there we're one universal publishing protocol for the web. Wouldn't it be nice if I could choose any tool for publishing to any repository - too much to ask ? Oh wait.

Nov
6

I just had a quick IM chat with Joel Selvadurai. He's just launched Messagr.com which overlays Skype's user directory with some interesting social networking meta-data. Essentially you can tag your Skype account with your own interests so people can find you and contact you easily and in real-time.

I'm suprised LinkedIn haven't made the Skype connection - Oh wait - imagine all those recruiters IMing and calling you at the wrong time !

If there we're a coolness rating for web startups, Messagr.com would be right up there - it merges web and IMS, has a social networking aspect, search, tagging, etc.

One word of caution with sites that use identity federation (I realize that's not quite the right term). Think about it - you give your user-names for all the accounts the service needs to access (in messagr's case - only Skype) and you protect that with a password. What are the chances that the password is the same as your password for the external services - pretty good I would imagine. Be careful what information you give up !

Find it

Subscribe

Contact Me

My status

follow pixelfodder at http://twitter.com

Links

The Aquarium (from the source)

Images

sharps. Get yours at flagrantdisregard.com/flickr