Tuesday, 17 October 2006

Thinking of the 35th Time Before the 1st

FeedSeeder was deployed on its first publicly visible server this morning. While it took much longer than I expected (better part of my non-client time for a week) much of that time was setting up the infrastructure; prepping the box, loading the project into Subversion, setting up the deployment recipes, etc.

While it’s not rock solid yet and I maybe could have deployed without the infrastructure in place – once, the promise of easier, faster deployments down the road was enough to halt actual development for a little while.

That’s all for now. Off to the morning’s coffee.

Saturday, 7 October 2006

Tuesday, 26 September 2006

Saturday, 23 September 2006

More FeedSeeder Demos – Organized and Otherwise

As I’ve mentioned before, I’ll be at Portable Media Expo a this coming weekend. If you see me, feel free to introduce yourself and ask for a FeedSeeder Demo. I’m looking for comments and suggestions.

Also, as soon as I get back from PME, I’ll be doing a more formal presentation as part of the University of Minnesota’s Institute for New Media Studies’ Emerging Digerati series.

UPDATE: Sounds like Tim from the Winecast, Frosty from ITRadio, and Johnee Bee from Mostly Trivial will also be representing PodcastMN.

Tuesday, 19 September 2006

Tags Less For Me and You, More For Us

In working with FeedSeeder, and interesting transformation has happened.

I actually see value in tagging text.

For video, audio, and pictues – text tagging is a no- brainer. There’s really no better way to get the media categorized than by asking the community to do it. Text for me was a much tougher sell. I kept asking – aren’t the tags already in the text? Can’t text parsers pull them out and ‘tag’ appropriately?

I still think in 95% of the cases the answer is yes. In those cases, tagging isn’t really useful – in-fact, it’s annoyingly redundant.

What I’ve discovered is a great desire to tag not for me (I’ll just post it on my blog) or for everyone in the word (that’s what Google is for) but rather tag for a small, defined, active group, in a folksonomy that only makes sense to them. This translation, re-organization, and re-contextualization of all information (whether image, audio, or text) leveraging words (and meanings) unique to a specific group – this is where I see the real value in tagging.

As an added benefit, it reduces the synonym problem (‘movies’ vs. ‘film’ vs. ‘cinema’).

Wednesday, 13 September 2006

Wanted: More Publishers and Better Filters

David Newberger and I were talking about information filtering models and much of the thinking behind Feedseeder came though as did a couple much larger issues. The conversation was so good, I thought you’d enjoy it as well.

David: I was talking with a few people today and each one mentioned Information Overload in the seperate conversations. I knew that is was gaining more of a following but I didn’t think this many people were getting overwhelmed.

Garrick: I believe it. I think there are two tools that havent caught up w/ the information tidalwave yet; filtering and integration.

David: Yep. Sites like Digg and Memeorandum are not working it tells me also.

Garrick: Nope.

David: There is also the whole age factor in this as well.

Garrick: Seems to me there’s a fairly simple question we’re not getting a good answer to: ‘What should I be paying attention to right now’

David: Yep. I think if someone can effectivly answer that then they are going to be the next Google but the approach is something totally different then Google.

Garrick: It is. I believe it requires intimate knowledge of a persons social groups, and personal/professional goals.

David: I was thinking about a contract I heard about. The DoD is willing fork over $20M to figure out what the media is focused on at home and abroad……this seems like something that would be a jumping off point for the question you posed.

Garrick: Huh. I don’t think there is a focus. At least not in our house. Either TiVo or Netflx playing in the background while Jen and I are on our laptops.

David: Not in ours either but when aggregated over the global scope of media I think that there are some stories will have more attention then others. They would like to know the direction of the stories and such.

Garrick: Ah.

David: It seems to be an attempt at the wiretapping execpt this time they are looking at the public information and not the private information. Back to the question of “What should I be paying attention to right now?”. It seems to me that this is a deeply personal questions. It is a question that would require a lot of knowladge in anthropology, sociology, and tech field.

Garrick: Yes. for sure. and marketing needs to be left out of it.

David: Yep. I wonder if it is even possiable to write an algorithm that takes into account things like the social questions.

Garrick: Maybe not to answer the best thing, but it would be helpful to cut out 80% of the noise.

David: True. But the static that is out there is so large that you would have to figure out some of the values you can qunatify for some social issues.

Garrick: Oh sure.

David: I would even argue that cutting out 90-95% of the static would be needed for it to prove useful.

Garrick: Hmmm. Maybe I need to think about this differently. As J Wynia has said – we all know what spam looks like.

David: Think about it, there are what, 40-60 million blogs in the US or is it around the world. And at least 10-20 thousand newspapers in the US.

Garrick: Yes. 90-95% of them don’t talk about things relevant to me at any given time.

David: Plus, you have to take into account the podcasts and vlogs now.

Garrick: Same there.

David: The one that are relevant to me change as the issues change this is something that needs a lot of consideration.

Garrick: Exactly. The publication as a whole is far less important than any individual article.

David: Exactly.

Garrick: Google is our best tool for finding relevant articles. but, it searches the entire world.

David: And there is the splog problem with Google.

Garrick: Mostly likely, I don’t care what the entire world thinks.

David: But I think that narrowing it to the point of keeping it US Centric means it is not doing its job.

Garrick: Oh.

David: in the case of say a war my viewpoint might be one way because of how my government is spinning it and then you have the rest of the world.

Garrick: I didn’t me to restrict it to geographic boundaries. but rather communities-of-influence boundaries.

David: Oh, got it

Garrick: For example, I don’t care what gem transportation companies say about ‘ruby on rails’. I do care about what you say about it.

David: Good point. so now we need a reputation system that is based on a persons status in the subject matter.

Garrick: Oh. I dunno.

David: But we also must allow for bleed from the edges as well.

Garrick: I think we need a filtering system that can pick up patterns from people we declared we care about.

David: Could that skew the persons view to one direction and filter out opposing views? and would that be a good thing if that happened?

Garrick: Only if you’ve declared you don’t care about what the opposing view’s think. And your ‘friends’ have as well. And their ‘friends’ have as well.

David: now that seems like it would be fun to see, it allows for the bleed, reputation, and to a degree subject matter now to filter out the static that is left.

Garrick: So, in the most extreme cases, that’s how blogs work today.

David: I can agree with that.

Garrick: My ‘friends’ write about stuff that’s important/interesting to them. if I find it important/interesting, I write a post on it and link to them. You read it and continue.

David: But with only 40-60 million people blogging right now we have a problem of scale. There are 3.3 or so billion people and we are only getting the view of 40-60 million.

Garrick: Yep. Sounds like we’re doing a good job of cutting out the noise already.

David: True but what is causing it and how do we start to engage some of that untapped well.

Garrick: Find an angle that makes sense for an individual, and make it easy.

David: One could argue that we are only using something like 1% of the collective brain.

Garrick: Exactly. Once the other 99% gets online, we’ll have the same problem we have now….but 99x worse.

David: So, we need to put the dam and locks in place now. Can you imagine 100 or 150 million people blogging on a consistant basis? Think of the mental power that would be. Think of the different ways people think and see things.

Garrick: Exactly.

David: So, my goal would be get 100-200 million people blogging on a constant basis on there passions of interest sure some might like to talk about dogs, fish, girlfriends, and such but other will talk about philosophy, RoR, Computer Science and other areas.

Garrick: They’re all the same. that’s why we need to talk about all of them. Here’s an example – the Google results for garrickvanburen.com include: “keith ellison”, “ical wordpress”, “linksys router setup mac”, “punch pizza in mn”, “house season premier fox”. I see that as a good thing.

David: Yeah it is diverse.

Garrick: Any 1 topic to me is flat. like a bad stereotype.

David: True.

Garrick: It’s combinations of topics and perspectives that make people and their views interesting. I read Doc Searls for everything he writes – though, I’m not into his photography. I read Mark Cuban for everything he writes, though, I’m not into basketball. Those topics round them out in my head – and make them real.

David: That is a good example.

Garrick: Arguably – I’m exposing myself to opposing ideas and concepts by letting Doc’s photography and Mark’s basketball through my filters.

Tuesday, 5 September 2006

Demo FeedSeeder – Check


(image thanks to PFHyper)

The FeedSeeder demo went pretty well tonight at MinneDemo. I’m pleased. Thanks to everyone there for letting me share it with you. Thanks to Luke & Dan pulling us together.

If you weren’t able to catch it, I’ve added the most recent Quicktime screencast to Feedseeder.com (~60mb, ~7min).

If you’ve already seen it or want to see more, or talk about some of the concepts behind it, you know how to reach me.

I Don’t Care What Everyone Thinks

“…when it comes to choosing what I read and what is relevant to me, I’m only interested in opinions from people i know and trust.” – Shel Israel

A while back, I re-grouped all the feeds in my aggregator – notably the ‘Everyone’ group contains authors that said something interesting once…but that I neither know personally or consider consistently insightful. A clunky hack to be sure. All the other groups; my feeds, ‘Friends & Family’, ‘Minnesota Blogs’, and of course ‘Must Reads’, are all a higher priority than ‘Everyone’.

I’m in full agreement with Shel. The aggregation tools we have today make it easy to read what the entire world finds interesting. Still, there isn’t an easy way to glean the conversations, emerging and otherwise, within the comparatively small group of people I trust. It’s really a simple question:

What are the people I trust talking about?

This is one of the problems I’m hoping FeedSeeder will help with, at least for me.

Something’s in the air – Ross Mayfield calls it MeMeme.

I found Ross’ post just after saving my Intro presentation for tonight’s MinneDemo.

Thursday, 31 August 2006

More Hay Makes Better Needles

I’m starting to load my 320 feeds and various groupings into FeedSeeder, the development version. Even with a good chunk of that to go, I can see how it’s changing my reading habits for the better.

The interesting bits are unhiding themselves.

This, before it’s even ready for public release and just nearly in time for next week’s MinneDemo. Cool. I’m excited. About browser-based software. For those that have heard me rant about the uselessness of it before – I might just have to change my tune.

Still, probably won’t move my calendaring, email, or to-do lists – but there’s a better-than-decent chance I’ll move my feed reading.

Wednesday, 30 August 2006

The Benefits of Screencasting in Software Development

Yes, I’ve been too busy working on the FeedSeeder project to write up too much about it. That said, I’ve been privately documenting it’s development with screencasts.

At the end of each development day, I start at the beginning of the app, load up iShowU and walk through all the functionality as if I’m presenting it to someone new.

Sometimes things still don’t work – or don’t work as expected (‘Huh?’). Sometimes they do (‘Yeah!’).

Either way, I’ve found the benefits huge;

  • I see what’s broken, and have a record of how it happened
  • I get practice walking through the app
  • I have a list of things to work on tomorrow
  • I can share it’s status with the small group of very excited people – and guarantee their experience with this early-stage, in-development app.

Also, yes, I’m still on track to present it at MinneDemo next Tuesday and it should be in real good shape in time for the Podcast and Portable Media Expo.

Finally, Dave Winer‘s current work on the River of News has helped frame quite a bit of thinking around this app.