Announcement: APML Open Source Libraries in C#

In yet another milestone for the APML Workgroup and the APML format, we have published the first open source libraries for loading and manipulating APML files.

From the site:

APML will allow users to export and use their own personal Attention Profile in much the same way that OPML allows them to export their reading lists from Feed Readers.

The idea is to boil down all forms of Attention Data – including Browser History, OPML, Attention.XML, Email etc – to a portable file format containing a description of ranked user interests.

These libraries are a result of months of R&D and iteration by the Faraday Media development team and we donate them to the community in the spirit of open collaboration and mutual benefit. They are released under the extremely liberal Apache 2.0 license.

We encourage anyone who would like to support or modify the libraries to get in touch so we can help in any way we can.

My thanks to Ashley our CTO and Mike and Paul the two ninja programmers who have been involved with the library and the APML workgroup who's input has helped to create the 0.6 spec.

We look forward to seeing what new and interesting projects get created with this resource.

The repository can be found on Google Code.

I'm falling in love with my APML file

When you are intimately involved with developing a piece of software you grow to become unattached to your application configuration. At any given time the next experimental build might blow up and kill your settings or you might have to delete your config files to see what a 'fresh install' might look like for a new user.

Over the last year I have destroyed many an installation of Touchstone - each time thinking nothing of it... Deltree *.* (don't you remember DOS?)

Lately though, I have noticed a change. As the app has switched from a Swiss cheese set of features into a complete product my APML file - the file that contains my Personal Attention Profile - has started to become precious to me.

I can no longer just delete it and start again. It has grown to identify me. It produces content results that I like. I want to protect and nurture it. In fact now I back it up and carefully ensure that I never let it die in the process of trying the next experimental build from the dev team.

My APML file is becoming "the digital representation of my physical self" (Morpheus - Matrix 1).

It is obvious that APML is going to become something quite personal for people, and I would like to publicly recommit ourselves to protecting and nurturing it with all our might.

We are the user's ally in the fight against information overload and the search for great, personally relevant content. And APML is our BFG.

How does Touchstone work? Find out here...

For the life of me I can't work out why Paul, our new (for about 4.5 months ago) uber-programmer, didn't get a proper welcome to the team on this blog. Ash and I have always welcomed new team members etc. I will blame myself entirely.

Paul, over those last 4.5 months, has done an amazing job. You will notice a huge difference in the beta in the areas of performance overall and the ticker usability specifically. You need to email to thank him :)

The reason I am pointing all this out, though, is because Paul has started a series of great posts about how Touchstone works. Or at least how we have solved seemingly simple problems that actually amount to a complex web of systems and subsystems to make an application work 'as you'd expect'.

So.. Check out the Touchstone Developers blog and specifically you can start with these few:

Authenticated Feeds (A big feature request from the testers!)

Indexing Firefox History (Firefox is our friend)

Proxy Support (a HUGE feature request from our testers)

More to come so be sure to subscribe to the feed.

Thanks Paul!

Best of Breed Future

As Richard mentions on Read/Write Web today - John Milan wrote an excellent two-part article for R/WW earlier in the week, about the future of software. However - I have not yet had a chance to read it. It was too dense and long and I have not had the time to dedicate.

But Richard, being the excellent site editor and blogger that he is, recognized the possible problem and posted a summary today.

He summarizes it like this:

In Part 1 John argued that data should become open and accessible, just like the code in Open Source software. Code is often re-written and re-factored, but systems only work if they agree on the data.

Part 2 contends that people will demand more access to their data and more integration with their apps. This will result in the single minded, all-encompassing applications of today dying off - in favor of multi-celled, specialized solutions. So the future will be combinations of best of breed technology, rather than monolithic software.

John's conclusion was thus:

"And what trait will the eventual winners in this brave new world share? The solutions that can hone their data requirements, move results from system to system, use the best form factor for the job and still keep it on a human level."

Richard compares this to the new Firefox and I think he's right. But I would also like to compare it to Touchstone (surprise, surprise).

With Touchstone's input and output/hub and spoke model, it effectively moves data from one system to another and at each point making a decision about the best form factor. The example we use most is 'the more important the info the bigger the alert'.

This is probably not what John meant - but I think it still holds true.

Thanks to RSS as the universal syndication format and Microformats as growing standards, users can pick and choose the best apps to use together. We hope that Touchstone will be the best notification platform in that mix.

Via Touchstone

Note: I will start to say 'Via Touchstone' on posts when my post is based on something that I found from Touchstone. I have found that Touchstone has started to become one of my main information sources as it evolves into a complete solution and more and more of my posts are based on info it alerts me to.

Paying attention to broken windows

As we have alluded to over the last few weeks, we are now working on what will become the first Beta build of Touchstone. So as we start to tie up loose ends, work on optimization and generally focus in on the little details it's great to come across posts like solving big problems by paying attention to the small things from Matt of 37Signals.

It's true that by polishing off what is already there and making it painless, easy and even glossy - the application truly feels more stable, pleasurable and intuitive.