Blogging isn't my natural medium and my LiveJournal withered on the vine a couple of years ago. But I keep being told I need a blog if people are going to believe I'm actually doing something worthwhile with my time so I've been debating for a while now the thorny issue of an inaugural post.

It should really have been something related to my Go research but ironically I find it very difficult to write about code. Sure I have three decades of clever tricks locked in my head and the anecdotes flow easily enough over beer or when I'm holding court in front of a projector, but the art of committing them to text generally eludes me - though I plan to do something about that in future posts.

I guess this mental block is why speaking at conferences has in recent years become so important to me. The immediacy of interacting with a live audience somehow loosens my tongue allowing me to pass my often hard-won and eclectic knowledge to other developers, whether they be hardcore Rubyists or professional programmers in general.

I'm the first to admit that my presentation style is somewhat unorthodox. It lacks the practiced ease we've all come to expect from marketeers and is more akin to an unmanned fire hose, liable to veer in any direction at a moment's notice. Several critics have picked me up on this over the years and I don't dispute their complaints. But my thinking is that if you're not making mistakes, you're not learning, and in true hacker style I'm teaching myself on the job. That means missteps aplenty along the way.

However I feel many of these critics have missed two key points about conferences.

The most obvious is that a conference is transient and fleeting: a group of geeks gather for three or four days, drink a little too much, exchange war stories, and go their separate ways. Some of the sessions they attend during this period will leave an impression but aside from keynotes most will slip by without remark. This means that how well a session adheres to the traditional script may have some effect on the buzz created around it in situ, but once the punters head home the chances are that precious little will remain with them.

My response to the transience problem is to cram as much useful code into my slides as possible - even if that means having too much code for the audience to study in detail - and getting them online in a timely fashion. On average about a month of research and development will go into a fresh talk, and I'll then rework this a few times to hone it for different audiences creating a virtual artefact that can be studied independently. Thanks to sites like Slideshare it's possible for this code to reach thousands when they need it rather than the few dozen who are normally able to attend a session.

I've discovered that conference organisers seem to be much more amenable to this approach than many audiences and I think this relates to the key defining characteristic of conferences: their ability to nurture and support community. Having a few known mavericks on the bill mixes things up and creates a sense of a vibrant community growing in many different directions, whilst speakers who produce significant content with a lifetime measurable in years rather than weeks are a subtle but significant boost to community credentials.

Now community can certainly exist without physical meetups, and indeed for the many years I was a member of ruby-talk before attending my first Ruby-related conference I already felt part of a welcoming and active community of developers. But the rise of the unconference in recent years is ample proof that geeks really enjoy getting together and sharing their passions, independent of their desire to listen to a talking head.

In essence the most important track at any event is the hallway track, and the value of that depends very much on having a good mix of attendees who are passionate about their interests as well as a reasonable number who have deep knowledge of the conference's core subject matter. Because conferences in a sense are the rare occasions when our virtual communities become a tangible thing and should therefore be a place where developers of all skill levels can both learn something new and feel they're teaching something valuable as well.

My friend Greg Brown briefly alluded to this in a recent blog post inspired by my ongoing efforts to crowd source funding for my RubyConf outing in September so he's in large part responsible for my deciding to tackle this topic today. If you don't know about his Ruby Mendicant University efforts you should check them out.

Returning to the subject of conferences, it certainly is a strange world if you're a speaker. Competition amongst organisers is cut-throat and very few conferences can cover travel costs and/or accommodation. Some go even further and will only cover admission for one speaker per talk, or not cover admission at all - and for some supposedly professional conferences this can add hundreds or thousands of pounds to the cost of speaking, effectively making it the preserve of the made men and women of the corporate vendor space. Personally I think this is greatly to the detriment of attendees as it excludes the vast majority of startup and open-source developers, i.e. people doing interesting work.

Such a negative economic framework creates a terrible impasse for those of us working on negligible budgets. Should we submit proposals that we believe are worthy of a much sought-after slot, knowing we may have to pull out if we can't raise the funds needed to attend? Or should we ignore conferences outside our local geographic area?

My answer to these questions has always been to submit to every conference I think will be a good outlet for my current work, trusting to providence and the generosity of others to make attendance feasible. Sometimes this has worked well, others I've either taken a modest financial hit or else pulled out with sufficient alacrity that the organisers haven't been left in the lurch.

I take this stance because I firmly believe a conference proposal should stand or fall purely on its individual merits, otherwise we cede our communities to those with their eyes on the dollar signs: instead of valuable technical content the average conference would become a parade of marketing schlock, a terrible blend of cupidity and gullibility.

If you see things similarly what can you do to help?

1. If you're a potential speaker submit that proposal. This counts double if you're a woman, because the gender imbalance on most conference rosters is dismal.
2. If you're an attendee, keep an eye out for speakers who need your help. For those of us struggling to raise an airfare or cover an hotel bill even a £5 contribution is a big boost to both our morale and the practicality of our attending.
3. Conference organisers, consider setting up a voluntary speaker subsistence fund that attendees and sponsors can contribute to. Give cool subsistence sponsor or community supporter badges to those who contribute.

Because speaking as one of the many independent g[uy|irl]s in our community, we need all the help we can get.