Thursday, December 26, 2013

In memoriam, John Knight, May 3 1936 - Dec 22, 2013

I'm sad to say that my father, John Knight, passed away on Sunday evening. There will be a service Monday the 30th and I'm appending the text of and a link to the obituary. He was a wonderful father and a much-loved professor, and will be very much missed by all of us.


John Knight passed away peacefully at home in Ottawa on December 22, 2013. He is survived by his wife Norma (nee Evans); brother Bill (Eileen); three sons, Alan of Seattle (Kirsten Carlson), Douglas of New Westminster (Ruth Silverman), and Scott of Toronto (Natalie Fong-Yee); two grandchildren, Marisa and Charlotte; and several nieces and nephews. Predeceased by his mother Mary, father Rixford, and sister Margaret (the late Robert Diehl).

John was a Professor of Electrical Engineering at Carleton University since 1967 and remained there until well after retirement. Thousands of engineering students will remember him fondly as an inspiring and entertaining lecturer who was always willing to help.

A memorial celebrating John's life will be held Dec. 30 at 2:00 p.m. at Riverside United Church (3191 Riverside Dr, Ottawa). In lieu of flowers, donations to Stride Wheelchairs Plus or the charity of your choice.

Tuesday, June 18, 2013

STIC 2013 - video and slides for my talk

I've recovered a bit now from the STIC 2013 conference. The video for my talk "Modern Web Programming with Dart" is now available. I give a very quick overview of Dart and Web Components as approaches to improving web programming. You can also download the slides, which are a bit hard to read in the video.

Other presentations are being added at Jim's site. The most recent is Dan Ingall's keynote talk. On the loose theme of "My Favorite Things" Dan talked a selection of things he's been involved in over the years, starting with some quite interesting history about very early Smalltalk versions and how much they differed from Smalltalk-80, and ranging up through various Squeak projects and a short bit on the Lively Kernel. It includes running versions of some remarkably old software doing some remarkable things. This was part of the historical theme of the conference on the 30th anniversary of Smalltalk-80 being released to the world.

Sunday, March 17, 2013

Words of wisdom from Joachim Tuchel. Glorp expects your model to be internally consistent. I've had a number of questions from people who expect, when working with databases that it's enough to modify one side and then the database refreshing will take care of the other side. If you have a bi-directional relationship, and the two sides don't match, Glorp will complain when you try and write. Unfortunately, it complains when it gets down to the level of writing rows and sees that there are two different sources trying to write different values to the same field, which isn't always the most helpful in figuring out the original problem.