Workshops and conferences have a specific and important place within science, and academic life more widely; they are good places to present early work, to get feedback rapidly. They are good places to find out about other peoples work, and to benefit from the random interchange of ideas with people who you do not already know. For instance, I often choose to go to the talks that I think will be most interesting, but then stay for the next or the whole session, just for serendipity. Mostly, I end up reading or writing, but sometimes, these lucky talks turn out to be most interesting of all.
There is a problem with workshops though; basically, they require a significant degree of committement. They can be expensive, both in terms of travel, accommodation and, of course, time. Now, some scientists enjoy the travel; I do not mind, and it is nice to see parts of the world you might not otherwise but increasingly I dislike being away from home. But there is a larger scientific problem; part of the point is to meet people you do not know, but given the overheads of attendence, the range of people likely to go a conference is quite small.
Of course, there have been attempts to work around this in the past. Virtual Conferences are a nice idea; do away with the travel, and just have the presentations. This, of course, limits the opportunity for serendipity; you will not bump into people at dinner or in the hotel lobby in a virtual conference. And, worse, they do not solve the time zone problem. Virtual Conferences for the British often mean sitting up bleary eyed in front of a machine in the wee small hours, listening to some fresh-faced, recently awoken Californian talk. This does not inspire collegiality. Likewise, the degree of committment required limits the scope.
The NearCon idea is an attempt to make a journal feel more like a conference. Ironically, this is how scientific journals arose. Instead of sending letters to many individual scientists, you would write to a journal which we broadcast it. But, again, the enormous expense of publication (whether in page charges, OA fees or just simple manuscript preparation time) works against this process. However, using a platform like Knowledgeblog , which is inherently cheap, we can work around these problems.
The process would work like this: scientists would be invited to submit, short 2 or 3 page articles, probably using EasyChair. A small PC would peer-review the articles. Then, on a specific date (advertising in advance), the papers would be begin to be published onto a kblog. However, rather than publishing everything at once, papers would be released incrementally, over about a 1 week period, spread throughout the timezones of the authors. At the same time, a mailing list would be created with all authors and any “attendees”. On publication of each paper, an email with links to the paper and content would be sent to the mailing list, providing a natural seed for questions, and discussion. The mailing list would operate like normal: there would be no expectation that authors would reply instantly, and it is likely that discussions on two or three papers would happen at once, over a period of days. However, because papers would be released slowly, say 2 a day, attendees would be less likely to suffer from the information overload that getting 10 papers at once would cause. After a short period (say 1 week after the last paper), the mailing list would be archived, so as to prevent zombie posting and general spam issues.
Clearly a NearCon would not have the same opportunity for serendipity that a conference or workshop has. But it might potentially attract people who would not come to a conference. For instance, at meetings like SePublica, amoung attendees, authors and PC members, most of the scientists come with some technology to push (myself included!). There was no one writing papers as a “normal” scientist just saying “this is what I would like to do”. A NearCon might allow this. Likewise, Science Online London  has very few scientists and a lot more publishers.
I have several ideas already for NearCon topics. First, and perhaps, most obviously would be the “Future of Scientific Publishing”; the NearCon format would allow a number of types of papers that authors would not otherwise submit. For instance, requirements papers written by scientists who want to do something, or have tried and failed. For example, I presented my own struggles getting courier font to mark up Ontology terms in a paper . Another friend described writing a paper in S/Rweave, but eventually having to turn the whole thing into a Word Doc. Yet, another described their desire for a data journal. These papers would be valuable, but a scientist whose only involvement in publishing is as a user would not go to a conference on the publication process.
A second idea it to have a Bioinformatics Core Facility NearCon. The special case here is that many of those working in core facilities have a service role and often do not have conference budgets. We have already shown with our Bioinformatics kblog  that short articles on new techniques or experiences with new tools are very popular. At Newcastle, we now have a local Bioinformatics Special Interest Group ; a NearCon would help to supplement this, breaking down the geographical barriers this sort of group faces. Finally, short papers giving a short statistical overview of a Core Facility workload (50% microarray, 30% Next Gen, 20% data integration, for example) would be easy to write and when combined, fantastically useful for tool developers, planning and so forth Supported by a kblog, with citable articles, backed by a set of web archives, this form of paper, would move from being transient grey literature, to become a valuable historical record.