Coding Sanity

Went to DevDays 2006 yesterday, it's the first DevDays I've ever attended (not through lack of trying, scary how something always comes up). I unfortunately couldn't stay past lunch, but below are some of my thoughts.

Welcome and KeyNote

I thought some of the demos they showed here, especially of the Avalon stuff, was very cool. However, it seemed to be a bit like an Amway conference or something, with way too much forced jolliness. Maybe it's just 'cause I'm a curmudgeon, but I would have preferred less jollity and fun and more information content.

The demo of Windows XP in IsiZulu was pretty cool, and I liked the presenter, but I sort of wonder about why it was shown here. Developers worldwide tend to code in English. Don't get me wrong, I think an IsiZulu Windows is a great idea, as are the Afrikaans, Sesotho, Afrikaans and IsiXhosa versions they announced. The biggest value I can see for these is in schools, where students can get to grips with computers in their mother tongues. That said, it did seem a bit strange to push this for so long in a developer conference where pretty much every single one of the developers would be writing code in English, no matter what their home language. Perhaps the greatest value to developers was showing them what a good localisation push looks like, something we can all learn from.

Hacking Revealed

I went on the Security Track, which was perhaps a mistake. I'm interested in security and wanted to learn more, but on reflection, DevDays would not be the place for this. These sessions were definately targeted at people with perhaps a more limited security knowledge. That said, I really enjoyed this session. Peter Willmot was an entertaining speaker, and I did actually come out with a very good suggestion on hardening Web Servers (don't run your web sites off the C: drive). I wish this session had been much longer and more in-depth, and he hadn't been forced to rush it.

Privacy, Security, and Compliance

This speaker (Ayal Rosenberg) didn't seem as quite as passionate and interested as Peter Willmot, but that would be a hard act to follow. Once again, the session was way too rushed, with the speaker flipping through slides like a madman trying to stick to the schedule. Whilst the information was generally pertinent and valuable, I didn't gain much from it, but I'm sure others being introduced to this information for the first time would have. During his talk I picked up two complete inaccuracies, the one was so egregious, that I must have misheard, so I won't delve into that. The other was a complete urban legend about the supposed computer expertise of the much-overrated hacker Kevin Mitnick. No, he could not duplicate a modem tone with his voice. His skill was in social engineering, a topic not much in evidence in the sessions I went to, but probably the most insidious and dangerous attack, bypassing all your systems and concentrating on the weakest link, people.

Overview

I was really upset I couldn't make the Protect your Applications session, if anyone has any info on this, I'd really appreciate it.

So all in all, I was a touch disappointed in DevDays, but this was probably due to going in with too-high expectations. It was professionally run and had interesting and informative speakers. I really liked the Share Your Pain demo ;-D

I just wish that the speakers weren't so rushed though, and this would in fact be my only real gripe with DevDays. I don't know if that was the fault of the organisers for shuffling the schedules at the last minute, or of the speakers for not timing their presentations beforehand.

I've noticed that neither of my last two articles on sadeveloper have been given any rankings. I went and investigated, and found that no-one has bothered to give any rankings or comments for any articles (there are 4) in the last two months. I don't rate articles because it might be seen as a conflict of interest (you know, dragging another article down to make mine look better). But not everyone in sadeveloper writes articles.

I don't know about the others, but I personally enjoy having feedback. I've heard some complaints in dotnet.org about article quality on sadev. Go have a look at the sadeveloper main page and look at the most recent articles and the most popular. If you don't like anything you see there, rank it down! If you'd like to see more of one kind of article, rank it up! The thing is that when you're thinking of writing an article for a site like sadev, you go and look and see what people like.

The comments you receive on your article influence your future ones. If someone says "I'd prefer more in-depth detail", then you'll keep that in mind when writing your next article. The point is that these rating systems are a sort of feedback look that the readers can use to train the writers into writing what the readers want.

I was chatting the other day to a friend of mine, one of the very few South African developers who I regard as better than myself - yeah, that ego's quite a thing huh? Anyway he was concerned that he was beginning to price himself out of the market. The main source of his concern came from such trends as outsourcing to India, automated tools and so on. I've been thinking about it, and while I don't agree that he should have to worry, the fact of the matter is that I think he does have to worry.

I don't know how many of you have read Steve McConnell's book Code Complete, but he has the results of an interesting study done some years back. Basically they took a whole bunch of developers with approximately the same experience and found up a 20 times productivity difference between them. That's right, some of the developers were 20 times more productive than others, with the same level of experience. Things like automated generation just allow the savvy developer to amplify their productivity.

So this is why I feel that he (and I) shouldn't have to worry. We're both passionate about development, and love what we do, we both keep current with new technologies, read development blogs, buy development books, and subscribe to development magazines. We code in our spare time for goodness sakes! Is there any doubt that we'd be better than some guy just in it as a standard career option? In fact one of the worries that both of us have is that our careers are pushing us towards more management kind of levels, and we would prefer to stay developers.

But the fact of the matter is that we do have to worry. The reason is quite simple, most managers (in South Africa at least) think about development teams primarily as a cost centre. Their concern is to rein in costs, not to boost productivity. Look at how leery many managers are about buying tools such as code generation or ORM tools. It's a short-sighted mentality, but a common one. Such a manager would be very unlikely to hire a very expensive senior developer, since they would not fully appreciate the benefits that such a hire could provide. I'm not just talking experience and productivity, I'm also talking mentoring, which I personally consider a major part of my job.

So what are our options? Do you guys think that it's possible to have a never-ending upwards career in development? Do we have to move to management in order to justify increases, despite the fact that we don't want to? Perhaps we could do a kind of Thinktecture thing where we get very senior devs to assist project teams in skilling up their staff both on the job and off it? Or is the only option that allows us to stay in development and earn bigger bucks to start our own product-focused company?

I don't know if the South African market is ready for a Thinktecture kind of beast. I doubt many teams would be happy with outside experts being brought in to skill them up, I think it'd raise too many hackles. The only long-term option as it seems to me is to start a product shop, but maybe I'm wrong. You tell me.