Opinion

What I learned on my summer vacation

Kent Sharkey — Tue, 15 Jul 2008 23:52:42 GMT

There are (at least) two truisms when you work for yourself:

You either have too much work, or almost none.

and

You can take a vacation anytime you want, as long as you're willing to spend just as much time apologizing to customers and struggling to catch up, and you're willing to not be paid for that time.

With those two in mind, we[1] decided to take a week off and head over to Yoho National Park. I decided to document this trip using this 21st century equivalent of the old elementary school teacher's favourite, "What I did on my summer vacation."

We have been to Yoho in the past [2], but this time we were finally going to see the Walcott Quarry site: "The world's most significant fossil find of all time. Ever!"[3] of the Burgess Shale. We've been wanting to see these for ages, as – all joking aside – I do think they are incredibly important in understanding how our world (and us) developed. They represent some of the first important multi-cellular organisms on the planet, they are a documentation of early ecosystems in action, and they provide yet more insight into our evolution. In addition, they show examples of all of the major phyla[4] that currently exist (and a few more).

The walk itself was a fairly brutal (for this aging keyboard jockey) 5-1/2 hour trip up, and 3 hours down. It's about a 760m (2500 feet) elevation gain, and – this is significant as I'm acclimated to sea level – that is from a starting point about 1243m (4000 feet) above sea level. There were 15 of us, plus the guide and his assistant. Most of the attendees were university students of various sciences, but there were also a couple of engineers and a lawyer. As with any guided hike, there were the stops to discuss topics related to the area, as well as plant and animal (hoary marmots) identification. The highlight, of course, was the quarry itself.

Opabinia (from Wikipedia)
Anomalocaris (from Smithsonian)Ninety-nine years ago, Charles Walcott was up there to find examples of what the railroad workers had called, "stone bugs". He (or his wife) saw them while riding over the Burgess Pass (thus the name Burgess Shale, despite the fact that they are on Mount Field). This led to multiple expeditions and digging to extract and identify some of the strangest creatures who ever lived. The shale samples represent creatures from the Cambrian (about 500 million years ago), well before we (or the dinosaurs) cluttered the place up. They are believed to have lived on a ledge in a shallow sea, and were swept down by a mud slide into the depths, where they were preserved. In between now and then, they were lifted to their current location in the midst of the Rockies through the wonder of plate tectonics. The animals (and plants) show great diversity in overall body shape: from the sponge-like Vauxia, through various insect-like trilobites and Canadapsis, to the WTF?! of Hallucigenia and Opabinia. Alas, they didn't actually have examples of those last two[5], and only the "headless shrimp" forelimbs of Anomalocaris (those things dangling at the front of the creature in the picture to the right), but we did see great examples of the others. Yes, I'm a biology geek, but touching a piece of stone containing more stone that was a replica of something that once lived was magical. Photos and more to follow as I get the stuff off of the camera.

And now for the obligatory (and gratuitous) link to technology.

During periods of hypoxia, it struck me that the evolution demonstrated at the Burgess Shale was paralleled in the evolution of programming languages. After a slow start involving various machine languages (bacteria), we had an enormous explosion of creativity and experimentation, where multiple language forms and styles came into being. Back in the Dark Ages while I was in University, my friends [6] studied Fortran, APL, COBOL, Algol, Spitbol, Snobol and many, many others. These were just a sampling of the varied and creative languages of the Seventies (for some reason, I don't recall them taking C, Lisp or Smalltalk). Some of these languages continue (at least in spirit) to this day, others mutated into 'something else', while others died out – either completely, or essentially – when was the last time you saw a job posting requiring Pike expertise? After some time, more 'niches' emerged: GUIs, networks, the Web. As each niche emerged, there was a burst of development and experimentation, just as we see when organisms are introduced into new ecosystems[7]. The initial Mac development language of choice was Pascal, and of course Windows was home to MFC [8] and Visual Basic. The first networks really drove the power and joy of C (and its many offspring: C is truly the arthropod phylum of programming languages), although it was fairly dominant before. The Web of course gave us HTML, CSS, JavaScript, VBScript and more. We're now seeing the increasing usefulness and importance of scripting languages, such as Ruby, Python and PHP. Along the wayside we have Rebol, Eiffel, MUMPS, Prolog and many others.

I believe we are now beginning the colonization of new niches, in particular multi-core, parallel development. We've already seen some movement in this area in the form of Map-reduce, Intel's concurrency checker, Microsoft's Parallel extensions to .NET, Erlang, and others. I suspect that we will see much more in the coming years, as various attempts are made (and failed) to create both easy, transparent ways of coding for parallel systems, as well as fine-tuned "know what you're doing" methods.

We can learn much about the future of our favourite languages from what happened to the Cambrian creatures. In particular:

Pikaia: notice any family resemblance?
(from Berkeley's Evolution site)You can't really predict future success from current success. Looking at the Cambrian creatures, many would say that Anomalocaris – the most dominant predator – might rule the future. Or perhaps the trilobites, the most common body form of their time. Few – if any – would have picked a little slug-like creature that happened to have a novel structure: a backbone [9]: Pikaia.
Stuff happens, and bad stuff can easily happen to successful creatures (and programming languages). You may be highly successful, but pile a mound of dirt on you, and you'll still be eliminated (see Delphi), leaving only fossils and legacy code.

[1] By 'we' I of course mean that M decided, planned, scheduled and executed the trip. I was pulled away from my keyboards to navigate and 'enjoy myself.' Bless her soul.

[2] On the previous visit many years ago, I was accosted by a squirrel. While walking through the campground, a squirrel ran up to us, and ran up my pant leg (about up to pocket level). After we matched gaze for a few moments, he finally decided I wasn't a pine, and continued his trek across the campground.

[3] A running gag of the guide's.

[4] Phylum (plural phyla) is a major grouping of organisms within a Kingdom. You can think of a phylum as a "body plan". There are many phyla, but the major ones are: sponges, molluscs, jelly fish (and anemones), worms, insects, stars and chordates (the phylum that includes us). The broad categories that are used to identify groupings of organisms are kingdom, phylum, class, order, family, genus, species (if you don't remember – or have ever learned – the mnemonic, it's, "King Phillip came over from Greece, singing").

[5] These have been removed by miscreants and hoodlums from the Smithsonian and ROM gangs. Sadly, they also didn't they have any sample of Pikaia.

[6] Alas, I didn't get a degree in Computer Science, much to my chagrin.

[7] See Darwin's Finches, cichlid evolution in Lake Victoria and Lake Malawi, and others.

[8] Yes, I know MFC is not a programming language, but I see APIs that are as huge and "development model changing" as MFC as different enough from their core language as significant to track as real programming languages.

[9] OK, technically, Pikaia didn't have a backbone, but a proto-notochord. However, this did distinguish it from the other animals of the Cambrian period, and it did eventually evolve into our backbone.

Silverlight inspiration

Kent Sharkey — Wed, 02 Apr 2008 18:41:33 GMT

While these pages are in Flash, they can/should also be inspirational for those working in Silverlight. Probably even for WinForms and ASP.NET developers.

As the good Jon Galloway pointed out the other day, Silverlight developers should really think of what they are doing from the user perspective, not from the developer "oh, shiny!" perspective. Don't just create something that could be done in HTML with a little JavaScript. Doing this just creates a site that is less functional, less searchable and overall less worthwhile. (yes, I'm looking at you Download center beta)

There was a backlash against Flash when people added it to their sites without consideration. Hopefully Silverlight can avoid a similar backlash, but only if developers avoid making the same mistakes.

It takes a village

Kent Sharkey — Tue, 01 Apr 2008 00:04:55 GMT

Or at least an effort.

I went down to the regional district office last week for one of their planning meetings. The masses had gathered to determine a vital matter for the three communities of the valley: Should there be a new gas station?

While the matter certainly wasn't anything of global proportions, it did bring one characteristic of the citizens here to light: Just how long have you been living here? One speaker after the other started their proposal with, "I've been here n years..." I imagine the message is supposed to add weight to their opinion.

You see and hear this all the time in dev circles as well: want ads ask for "Java developers with 30 years experience." Developers make blog posts and statements that begin with, "I've been using .NET since n" as though it really matters. I use myself as an example. I wrote my first .NET code a few months before it was shown at PDC, before it was even called NGWS (pronounced Nigwiss). However the set of .NET developers more skilled than myself is roughly equal to the set of .NET developers in the world. It's not how long you've been doing something, but the degree to which you're pushed yourself at really learning it. Don't get me wrong: I have pushed myself, but as my life is not measured in kloc, I don't consider my skill equivalent to someone who has to keep metal in motion.

So, how can you learn something so that it actually means something? Well, the first (and obvious) method is to actually use it, but to use it in a way that you're actually learning something. Stepping through samples from articles and Web casts is fine, but you need to break things, change values, make it fizz. Scott Hanselman's series on reading code is quite useful, especially if you try to "guess whodunnit" before-hand (predict how the code should work, what would happen if you changed it in some way, how you can use it in your own applications, etc.) Another great resource is Larry O'Brien's series on learning a language. Run through them every six months or you. You'll be surprised at how different your resulting code might be as new techniques and larnin' enters your brain and fingers.

While it may seem counter-intuitive, teaching is an amazing way of learning -- I learned more about programming^[1] by teaching it to others than I did by actually writing code. Different developers have different (sometimes way different) ideas about what is good code, and by discussing it with them, you grow (even if you were right in the first place). It also helps solidify your own arguments by forcing you to actually think them through and explain them in ways that others can understand. Plan a lunch workshop series at work, do a talk at your local User Group, or even sign up to help at the local continuing education facility.

Learning and mastery are not states you get to by accreting years of service.
You learn by doing, so go do something.

(For those interested, they still haven't decided if they'll allow the new gas station.)
[1] Don't even get me started on what I learned administering and teaching Microsoft Mail for years

'Some other issues'

Kent Sharkey — Wed, 20 Feb 2008 04:46:33 GMT

Don't get me wrong: I really like my Macbook Pro. Almost to unhealthy levels. I truly believe that it is the fastest laptop I've owned, while still being one of the lightest and the easiest to carry around.

However, Apple support boggles. Here's the *full* text of an update that came out today:

MacBook, MacBook Pro Keyboard Firmware Update 1.0
This MacBook and MacBook Pro firmware update addresses an issue where the first key press may be ignored if the computer has been sitting idle. It also addresses some other issues.

The update package will install an updater application into the Applications/Utilities folder and will launch it automatically. Please follow the instructions in the updater application to complete the update process.

For more information about this update, please see About the MacBook, MacBook Pro Keyboard Firmware Update 1.0

"It also addresses some other issues". Can you give me a hint? Oh, and if you do click the link, you get exactly the same text, plus the list of machines affected. Still no hint. Compare to a recent hotfix from "the other guy":

Security Update for Windows XP (KB946026)

A security issue has been identified in the mrxdav.sys driver that could allow an attacker to compromise your Windows-based system and gain control over it. You can help protect your computer by installing this update from Microsoft. After you install this item, you may have to restart your computer.

Look at that: an actual explanation of what it's trying to fix.

I'll leave it to the conspiracy theorists to figure out what would happen if Microsoft published fixes that "fix stuff -- really you don't need to know."

10 goto 20: 20 goto 10

Kent Sharkey — Wed, 16 Jan 2008 02:34:07 GMT

I know I'm just a caveman, but at the moment this world confuses me.

M forgot her Hotmail password. So, I decide to go through the wizard to help her get it back. They will kindly mail out a new password. To her Hotmail account. Who came up with that one?

Bearable lightness of bandwidth

Kent Sharkey — Sun, 13 Jan 2008 00:15:55 GMT

I decided to work off-site on Friday -- partly (I admit) to get away from the usual distractions at work (Miss Simon, and the three cats), but also to get off a network for the day.

The network gives me a "convenient" distraction most days: waiting for a Web project to start up? Check a few blogs. Need to grab a few graphics? Well, I'll just go to this Web site.... Ooo, shiny, I should look at this article. Finish a task? Well, I'll just check the Auction House to see how things are selling. Etc, etc. Just disconnecting for a little while was invaluable. I was better able to focus on the job(s) I needed to get done, and just do them. Granted, I was still working on three documents, but it was much more conducive to focus than usual. I really think I need to do it more often, perhaps declare a "Day without network" once a week or so.

It also reminded me of that great Microsoft tool -- "The Off-site". Every once and a while, a group will go off and work "somewhere". It may be at some hotel or other facility (although these are very rare these days, I imagine), at another building on campus (or another campus), or even just in a meeting room for a day. They serve two purposes that I can tell. The first is the above: they tend to be disconnected affairs, where you are supposed to focus on some task. The second is that it breaks up the normal order of things: by being in a different location, you don't fall back on normal behaviours. You don't end up going to the same cafeteria/coffee bar, you don't have all the niceties of your normal work environment, you're forced to talk with people in your group you don't normally talk with. In short, a reset. A fresh seed in your randomizer.

So, my recommendation. Once a month or so, just work somewhere else for a day (assuming you're not a call centre drone or other job where you can't do that), preferably without a network connection. You'll be (hopefully) surprised at how much you can actually get done, and how it may recharge your energies and mind going forward.

Review: The Canon

Kent Sharkey — Thu, 03 Jan 2008 20:51:13 GMT

I just finished The Canon last night, and ... whoa. Great book.

The premise of The Canon is simple: most people have very limited knowledge or interest in science. That frightens and saddens me (and the author, Natalie Angier). To solve it, she has written a book to provide the basics of science and scientific thought. It's not anything that you wouldn't get from a decent high school or college education in the sciences, but many people don't have even that. However, her explanations are phenomenal -- she helped me understand a few concepts that I never did get throughout university. I'm certain that specialists in the various fields my quibble at her "simplifications", but it's a good basic grounding in the current knowledge in a wide variety of fields.

The book is divided into nine chapters:

Thinking scientifically - one of my personal pet peeves. Too many people can't seem to do this, and fall prey to pseudoscience and quackery.
Probabilities - I've met so many people, including some of the smartest people I've met that can't grasp randomness and odds. I guess that's one of the reasons lotteries and casinos do so well.
Calibration
Physics
Chemistry
Evolutionary Biology - my biggest pet peeve, covered much more lightly (and approachably) than The Ancestor's Tale
Molecular Biology
Geology
Astronomy

The writing. Oh, the writing. I have heard the term "effervescent" used to describe writing in the past, but never really understood it. Until now. The writing in this book dances - it makes me feel like I should never sully paper or phosphors with my drek ever again. Ms. Angier covers complex topics with light, love and beautiful prose. (Although I have to admit a few times I felt it did reach the "overly precious" level)

As far as I'm concerned, this book should be required reading for all university and/or high school students. And their parents.

The version goes up before the features go in?

Kent Sharkey — Mon, 10 Dec 2007 06:22:24 GMT

I am agog. Exchange 2007 Outlook Web Access doesn't support rules? Wasn't that in the first version of OWA years ago?

Look at all those Xes!

Kent Sharkey — Mon, 03 Dec 2007 23:32:45 GMT

Congratulations to Chris, Dave, Jeff and everyone else over at The Code Project for what must have been a fairly herculean effort - rebuilding CP for ASP.NET while a few million people continued to hammer on it.

CP is still one of the best resources for just about any stripe of developer, be they native C++ wonks, .NET scribblers or even Java folk. Now, it joins the ranks of the most battered and heavily used ASP.NET sites -- hopefully to be a showcase for many weeks to come.

Late to the party (again): Version Control for all your files

Kent Sharkey — Tue, 30 Oct 2007 18:55:47 GMT

Like all good developers, I keep my source in version control (you are a good developer, aren't you?). However, I started to expand its use lately, keeping my home directory in version control (in my case Subversion). Yes, I know, that's nothing new, lots of people do it. I now understand why.

I tend to work on one of three machines, and I've always had the problem of keeping my work in sync on the machines. I've dabbled with Google Docs, SyncToy, RoboCopy and other solutions, but none of them have seemed all that satisfying and/or cross platform enough. Then I realized that I was already using something that would help: subversion (with TortoiseSVN on my PC machines, and SCPlugin on my Mac). I shove requirement docs, presentations, blog posts in progress, notes, everything in. Then, from any machine, it's just a SVN Update away to have my machine synced up. One place to backup, and every machine is happy, and I can get my files from anywhere.

I may never need to go back to a "pre-teach" version of "Upgrading to .NET: Data Access I", but having it available is a nice bonus that other sync solutions wouldn't provide.