Scouta #6 Web App in Australia

Our social media recommender playground scouta.com has come in at number 6 in the top 60 web applications in Australia as named by Read/Write Web.

Google Developer Day

Today I’m heading into London to join in the fun at the 2007 Google Developer Day.

Pretty much all the day is going to be focussed on getting the most out of Google’s APIs. I’m most familiar with youtube’s APIs at this point, but I’m pretty intrigued with what is going on in the mapping and mobile areas.

The rhythm of IM is a powerful thing

A year or so ago I proposed that we needed a timeline graph for Skype online-ness.

Well, I’m pleased to see that somebody has done something similar for Twitter.

Sean Voisen has made TwitterViz which does exactly what I was talking about — but for Twitter, not Skype. I want to know more about the rhythms of life of my online community, and this will do it.

I mainly think of this kind of tool in a business environment, because I’m always working with people on the other side of the globe, and understanding their rhythms really helps.

But, for me, I want this averaged over days or weeks. I want to help my brain get a clue of when they are about, based on the last month or few months with the exceptions thrown out.

Nice once Sean. I aim to catch up with you for a chat and see where this might be going. When are we both online, I wonder?

How people and the phone work together

I did a really useful days sales training yesterday with Robin Nyman of Unique Selling Pointers.  Thanks Robin.

I think I learned a lot there, just how to turn sales and partnering calls into a process that even I can do.  And he did some live call examples, showing us how to get to speak to the person you want to in an organisation. Behind it all there’s lots of psychology I guess.  People really want to help other people, and big organisations either purposely or unwittingly make it really hard to find the right person to talk to.

It all feels a bit like social engineering, where we use people and psychology to defeat security systems, however this is done with honesty and integrity, not to cheat and steal.

I’ll be at Hack Day

I just got the word that I’m accepted into the BBC/Yahoo Hack Day. Which sounds like a phenomenal event.
So, now to cook up something interesting to do. Hmm. Something with Pipes and BBC content and a recommendation engine or two perhaps… I wonder.
[Update: I've added the Python language category to the unofficial hack-day wiki to see what sort of group forms there.]

Web application response time monitoring

I’ve been out there looking for a simple and affordable online service that will give me web app performance monitoring for Scouta. Ideally, I’m looking for something that give more that simply working out if the server is up or down. There are lots of services that do that. That’s a bit simplistic for a web service where we have multiple servers running, and where response time to our visitors and members is pretty important in the user experience.

What I’m really looking for is something that works on response time across a bunch of requests, and when a modified average goes above some response time limit, I want to get an email and an SMS.

So, today I’m brewing up something in python that will test one or more URLs and will SMS and email me when this modified average goes above some limit I set.

Delivering emails is easy once we’ve detected a response time issue. For SMS, I’ve signed up with TextAnywhere to get SMS delivery from a http web service interface.

Media Object Metadata

One of the things I’m working on at the moment is to model a reasonably complete set of metadata associated with an online media object. This is to give us in Scouta the best internal representation of media objects that we can get, to help with recommendation generation, searching and content delivery and embedding.

The source for all this metadata? Well, we start with RSS 2.0, then add on the Media RSS extensions and the creative commons licensing extension.

Then add in the Atom Syndication Format. Then add in the data returned from who knows how many online media site APIs. Also ID3 tag information from MP3s and other embedded tags from other file formats.

There’s a bit to work through here, as you can see. I’ll try and post a summary here when I get something reasonably complete.

The Digg Revolt: turning point for DRM

It is starting to look like the Digg Revolt will be the next step in the turning tide against Digital Rights Management. And what a day for people power. But it does raise some interesting questions that will be played out over the coming weeks.

It all started when Digg responded to a DMCA takedown notice issued by the MPAA and AACS Licensing Authority, the body responsible for managing the DVD-HD rights managment system. Digg administrators deleted the offending posts, a fairly normal operation, in line with Digg’s terms and conditions. All ok so far.

[Read more]

The second act

In the first act, I spent most of my time working in the city of London for an investment bank, which meant my time here was rather limited.

Well, that all changes as of today, 1 May, when the second act begins. There’s a bit to decide on just which direction to take, but it will involve working on new and different web applications, especially our growing Scouta.com, and expanding into some new consulting and project directions.

More on that as it evolves. For now, it is vision and planning time. I’m going to start with a few mind maps and take it from there.