Archive for November, 2005

One Number: What about SMS?

Interesting posts over on BP’s Weblog about his experiences getting one-number services to work.  His biggest problem seems to be routing SMS traffic.  Not possible using the VoIP one-number service he implemented on Asterisk.  He also writes about using the Plazes API to determine location for call routing, which is something I’ve been thinking about for a while. 

2005-11-28 11:58 am | 1 Comment »

Tags:

Prioritizing Podcasts Low: How?

Regarding how Rogers might prioritize podcast traffic lower than email or web, a reader wrote to me this afternoon with this:

I work for a small service provider and talked to the four largest traffic monitoring solutions: Sandvine, Packeteer, Cachelogic, and Allot Communications.  I can’t recall who said it (I think it’s either Sandvine or Cachelogic), but one of them said that N.A. cable operators are controlling their peak traffic (and costs) by selective rejecting certain P2P requests at certain times.  So rather than blocking it out outright, they are automating the selective rejection at certain times.  Which is different again than apply rate policies.

I presume Rogers is doing something similar with podcasts.

Not blocking, but degrading.

2005-11-27 6:08 pm | 1 Comment »

Tags:

Torcamp Rocked!

Torcamp, Toronto’s single-day Barcamp event, was held yesterday.  It rocked! 

This was the first time I’d ever participated in an Open Spaces workshop, and I have to say that it worked very well.  The basic premise behind Open Spaces is that the floor is open for anyone to speak on any topic.  You simply sign up to lead a session on the "grid", and then lead it.  If you wish to participate, then participate.  If not, go to a different session.

The first session I participated in was led by Reg Braithwaite.  I’m paraphrasing, but my takeaway was that software should "just work".  He used numerous examples to illustrate situations in which a little more contextual information could be used to produce a more relevant result for the user.  For instance, Reg pointed out that Yahoo consistently presents advertisements for dating services to him, when his Yahoo profile says that he’s married.  Naturally, this theme resonates with me.  iotum’s technology focus is the use of context to filter communication requests on the basis of relevance. 

Following that, David Janes gave a short presentation on microformats.  Very interesting, but too short.  This could have used a lot more discussion.  The basic idea is to embed structured data for applications like calendars, contacts, social networks, tags, lists, outlines, reviews, votes and so on, into web pages, feeds and streams.  Applications can then consume these.  David showed Greasemonkey consuming calendar and contact information from a web page, which presumably could then be inserted into Outlook, or some other contact management solution.  The real power, I think, is in formats like the hReview format.  For instance, I write reasonably often about wine.  You could use the hReview format to construct an aggregator of wine reviews from multiple sites.

Patrick Dinnen hosted a session on Wireless Toronto.  Not of interest to me, living in Ottawa, but there was some interesting discussion about how to create public / private hotspots.  Apparently a firmware upgrade to the Linksys WRT54G routers is available which will allow public access to the internet, and (via a WEP key) private access to the rest of your network.  This could be a real benefit to owners of coffee shops and restaurants who want to provide internet access to their customers.

Leila Boujnane and Paul Bloore showed us idée’s Espion visual search engine.  This was a very impressive demo.  What they’ve built is a tool for searching millions of images, based on image content.  Think of it as Google for pictures.  In the same way as Google revolutionized web page search by returning relevant pages, Espion will revolutionize image search by returning relevant images.  There was a spirited discussion around whether it really "worked" when one person asked if it could differentiate between a dog and a cat, or a dachsund and a bassett hound.  There’s no image recognition involved.  Rather there is an algorithm for calculating the signature of an image based on colour, shapes, luminosity and so on.  Very impressive!

There was a very stimulating session a little later talking about the role of channels in media distribution.  Or, at least that’s how I internalized it.  The organizer Mark Kuzniki is consulting to the government on interactive media, but the session devolved quickly into a discussion of business models as they relate to the media industry.  Consensus was: we don’t want more government intervention in cultural products. It would be better for the entertainment industry to choose new channels themselves.

The second to last session I attended was Ken Schafer’s "Making money".  An excellent, and stimulating conversation around the ways to monetize product and content.  Ad-funded, affiliate, and other models were discussed, including Google’s Click-to-call. 

And the last session of the day was a discussion of how to keep the momentum going.  Albert Lai has offered up his offices to run a demo event in December.  That would rock.  I am personally interested in seeing if we could build a Barcamp event for Ottawa. 

Check out Kate’s writeup here.  Technorati tag Torcamp may also provide other points of view. And session notes can be found on the wiki. 

Many thanks to David Crow for organizing this.  It was great fun to spend the weekend hanging out with bright passionate people.  I met a lot of new people, and hopefully made some new friendships.

| 7 Comments »

Tags: ,

Rogers Says “We’re not blocking Podcasts”

Mark Evans made a call to Rogers to investigate the rumour that Rogers is blocking podcasts.  They claim no, although they say they are prioritizing email and web higher.  How?  Packet inspection?  Prioritizing ports?

| 2 Comments »

Tags: , ,

Skype: Trouble in Paradise

Following my November 22 post on a Skype reorganization being in the works, Andy Abramson posted some more on what he’s been hearing, and then Om Malik and Mark Evans jumped on the story.  My take:  I’ve been through plenty of big company reorgs, and this is probably a lot like one of those.  It can take as long as six months to figure out where all the second order bits are, and what everyone’s new job is.  Unfortunately, with Google breathing down their necks, the Skypers are short on time.

2005-11-26 12:15 pm | No Comments »

Tags: