Craig Johnson's Radio Weblog

Great article on Korean crisis - with implications

Wed, 22 Jan 2003 15:59:17 GMT

Via Rafe Colburn at rc3.org. Seymour Hersch in the New Yorker provides a lengthy overview on the current Korean crisis. Now that the nuclear cat is out of the bag, policy is likely to be very dodgey until a method of deterring the new stateless threats is devised. I am pessimistic that will be via pre-emptive strikes, as our intelligence in many areas is so bad.

I wish I had said that

Wed, 22 Jan 2003 01:45:21 GMT

Paul Ford at Ftrain writes eloquently on staying the course.

"And yet I want to live there, and so I'm going to keep trying to get to the place, to the big rock candy democracy, for as long as I can."

Often the most important fights are fights we could walk away from. Saying no is a powerful act. Go, read.

A modest suggestion for killing the Marc Holzman appointment at CSU

Tue, 21 Jan 2003 20:23:32 GMT

Given that neither Mr. Holzman nor Gov. Owens seem to have the common sense to put the "Holzman for President of CSU" thing to sleep, I have informed all members of the CSU Board of Governors, Governor Owens, and Marc Holzman himself that I will cease any contributions to CSU if he is appointed President. I urge all CSU alumni to join me.

Let's just say no to an amateur running CSU!!

The drought continues - YMMV

Wed, 08 Jan 2003 18:10:05 GMT

This last weekend we traveled west to Breckenridge via US 24 and CO 9 from Colorado Springs. The whole trip to Hoosier Pass was brown, dry and it is clear that the whole South Park area is very dry. Once we crossed over the pass, we were immediately into the snow which was at least a couple of feet deep in shade. I would guess that it is mostly left over from the early season big snow that come through the northern and central mountains, but which had minimal impact on us.

So if you want to find snow, best to stay north of Hoosier Pass. That would be between Alma and Breckenridge on this map.

Nice writeup on the complexity problems in networking

Wed, 08 Jan 2003 17:31:49 GMT

Via Tomalak's Realm, a link to the latest Bob Frankston article called "Bad Coupling". This is related to the "leaky abstraction" discussion that took place recently on Joel Spolsky's "Joel on Software" blog.

There seem to be two entirely different ways that abstractions can leak. The first is what I will call original sin:) In this case the development team seeks to make an abstraction that simplifies the development effort, but inadvertantly does not deal with some error or boundary condition well and suddenly, the underlying software reaches out and bites. This kind of leaky abstraction simplifies software development but can lead to nasty failures.

The second way an abstraction can leak is in the ways described in the Frankston article. Well meaning engineers stretch the existing infrastructure abstractions in ways that work, but in a very different dimension than the original abstraction. In these cases, the leaky abstraction complicates software configuration while enabling capabilities that were not envisioned or accepted by the original abstraction's developers.

If you are developing software, the first kind of abstraction will be your biggest concern (unless you are one of those who gets to write the configuration wizards:)). If you are a purchaser or user, the second kind of leaky abstraction will intrude in unpleasant ways. When this second type of leak becomes embedded in the infrastructure, the whole community pays with configuration taxes for a very long time!

Mass Layoff tracking killed again by Bush 2 (like father, like son:))

Mon, 06 Jan 2003 18:05:32 GMT

Ok, so why is it the Bushes felt compelled to kill the messenger on mass layoffs? Must file in the "no news is good news" style of market capitalism:) I am often amazed at how thrifty the GOP can be:). Via Scott at the FuzzyBlog, go read it all from the SF Gate site.

Soviet Design - wish I had said that!

Fri, 03 Jan 2003 21:57:25 GMT

Peter Van Dijck in Ease uses the term "Sovjet Design (Soviet Design)". I love that term!!

There is an additional point. Process-centeredness is based on a mechanistic view of the world that holds, at its worst, that the skills of the individuals involved in the work doesn't matter so long as "they follow the process". It is frequently the result of over-internalizing the lessons learned from past problems (otherwise known as fighting the last war). Excellence in any endeavor is rarely the result of unskilled or moderately skilled people looking hard in the rear-view mirror.

TDD and emergent architecture

Fri, 03 Jan 2003 21:01:33 GMT

I have posted a short expansion on an excellent TDD for web apps paper by Edward Hieatt and Robert Mee over on the Agile Development side. Great experience report on TDD and emergent architecture.

Why small code increments are good if tinkering is expected

Tue, 31 Dec 2002 17:05:46 GMT

Stefano Mazzochi (via Sam Ruby) states the case eloquently for small increments of code in an open source project.

"The hardest and best the code is, the more harm it creates to the
community; this is because people will rather use the software rather
than extend it. Normally, if more than one blackboxware submission is
donated, the community will ask for a complete refactoring. (see
Xerces2)"

This is a great insight. Every time this has happened on a project that I am working, the incentive to stay away from that code has been high. When a large chunk of code drops in (I would argue even if it is relatively easy to understand) it creates a large barrier to understanding and thus to extension. Now this may be a virtue if you don't want extension:), but if the expectation is that things will be changing over time then incrementalism is a good friend.

"The good old Software Engineering practices they teach you in college
are bullshit: making architecture decisions without continous
reversibility is expensive because design constraints change too much.
Those who want to apply hardware engineering practices miserably fail.
Open source is here to prove that such a "messy" way to do code is
actually the only one that works and scales."

This is another key insight. Architecture as an emergent property rather than a static set of constraints or characteristics. This is particularly important where adaptation is more important than optimization (see Jim Highsmith here). I like that! Architecture should follow design whenever possible to allow the system to evolve rapidly to where it wants to go. Build it often, build it to run, make sure you can always go back to what ran previously.

Back from the holidays

Tue, 31 Dec 2002 16:51:01 GMT

The holidays are always a trip for those of us who live in the Great Plains of the USA. Given some remarkably nice weather (and we would prefer a bit-o-winter given the drought), we zipped up to Sturgis, South Dakota for Christmas, visited the in-laws and almost got to see Gail's brother get married. The most interesting thing was seeing the public information spots on TV relating the the winter emergency kits everyone should carry when traveling in winter. Food, something to melt snow into water, something to keep you warm, all based on a fact that you may need to live in your car for a few days until help arrives to pull you out. Weather is the primary force on life in the Great Plains.

Some thoughts on LibraryLookup

Fri, 20 Dec 2002 18:33:58 GMT

I have posted some thoughts on the interesting LibraryLookup, Jon Udell's recent experiment, here. REST, postmodern computing, agility, and the power of blogs ... I think it could be a movie!!