Older news

2006: Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec

7 Feb 2006: Added to the list of books in Further Info.

20 Feb 2006: Started development of a new product called Associate. It will allow users to manage their files without thinking of paths or filenames. Break free of the hierarchy! (see here for more details)

Older blog entries

2006: Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec

Jan 2006: Attended the TableTop 2006 Workshop where new interface ideas pertaining to horizontal display surfaces were presented. It was interesting to see how user interfaces can be viewed when basic assumptions are changed (orientation, hand input, etc.). I found it exciting to see basic interaction between the user and the computer re-visited!

5 Feb 2006: Read about a new idea in network stacks, presented at the Australia Linux conference. Unfortunately this year I didn't attend, but the slides from that talk look very interesting ( pdf format). Van Jacobson and others have revisited basic assumptions about the network stack and have been able to present amazing speed increases. This revisiting basic assumptions is exactly what I found exciting at TableTop.

20 Feb 2006: I've been narrowing down some research mentioned on the Products page to start creating a new product! To better represent that on this website, there is now a section on the product page devoted to development. So Products is really research, development AND products. That's what you get for living in the real world.

The new development that I am starting is all about the computer keeping track of your files, and better, more useable, integration with applications to support that. All of the techniques listed under "Current Research" on the Products page will be used, but the main emphasis is about transparent access to your information. See this part of the Products page for more info.

22 Mar 2006: Listened to Herb Sutter's talk that he gave to people at PARC (currently it can be accessed from Herb's website). Herb is selling the idea that concurrency is the next big programming revolution. Before hearing it I thought he was overstating the case, but he brought up some pretty compelling points. Two things in particular:

• Chip manufacturers are producing more hyperthreaded and multi-core microprocessors, hence more programs (including client applications) will need to use multiple threads and writable, shared memory.
• I was using the ACE library when problems with the Double Checked Locking idiom first emerged. Shared memory locking is hard.

Well worth a listen.

23 Mar 2006: On track (mainly) to finishing a functional prototype of a key concept in Associate. Over time I have found that usability testing is so important -- especially when the concepts are unusual. For web sites, or process flow, paper prototyping can be enough, but to test a new user interface paradigm requires a working user interface. So I've been using Python and pygtk which I have found excellent for the task. My main struggle has been to leave functionality OUT of the prototype so that I can test the usability of the concept early in the development cycle. That's not a bad problem to have.