First, there are some things I don't blog about because they're currently confidential information that IBM hasn't chosen to make public yet. That's what an internal blog is for.
Second, there's personal material that doesn't have much to do with work. That would be me discussing my hobbies and other parts of my fascinating social life. That's what a personal blog is for.
The third option, obviously, is the professional blog, where you discuss public information pertaining to a professional topic publicly. My blog that you're reading now is a professional blog about J2EE development. Other examples are political blogs, news blogs--commercial sources of information.
So the three-blog approach that Bob uses makes sense. But I think there are still issues around what belongs on a professional blog. There are some broad guidelines, like don't make your employer look bad. But then there's a range of what I consider being on-topic vs. off-topic. I find that my blog postings fall into one of there categories:
- original content -- Postings with new information that isn't available elsewhere, or new summaries or analysis of previously existing information.
- references -- Links to existing content that don't add much to the content, but make the reader aware of the content and where to find it, and enough information about the content for the reader to decide whether to pursue it.
- tangential discussions -- Commentary about topics that are not the purpose of the blog but that are interesting to the blogger and hopefully interesting to the readers.
Original content tends to be on-topic. References are usually on-topic too, but not as valuable as original content because the referenced material would still be available and discoverable whether or not the blogger pointed it out. Many blogs do little more than simply point to other material, and many of those are more news sites than blogs (or are, indeed, news sites, with all the personal flavor of a TV newscaster). References are still of value, though, helping readers find material of value that the might otherwise not know about. Tangential discussions are the least valuable, but sometimes the most interesting. They're usually quite off-topic, but fun.
For tangential discussions, I think the important consideration is not so much that the blogger thinks the tangent is interesting, but that they think it will be interesting to their readers. When I go onto a tangent (graphing statistics, swarming downloads, Simpsons quotes), I try to ask myself, "Will the J2EE developers who hopefully are reading my blog find this interesting? Or, is there something about this that makes it interesting to J2EE developers?" For example, I've done a couple of entries on happiness. Why? Not for the J2EE content! But I find this topic interesting and useful, which is why I read up on it. I blog about it because I find a lot of J2EE developers, and technical/engineering people in general, often struggle with being unhappy even though we generally have a lot to be happy about. So I would say that discussions of happiness are in many ways as important for this readership as discussions about J2EE (although a topic I myself have much less expertise about). After all, wouldn't you rather be a happy coder and work with happy coders?
So I have and continue to wrestle with the mix of material on this blog. Obviously original content is good, but there's only so much of it that I can produce, in terms of inspiration and in terms of time to document it. References are frankly a lot easier to blog and hopefully still valuable. Tangents add color and keep things interesting, if not for the readers than at least for me (so that I'm motivated to continue doing original content and references!).
I often wonder what readers want. What little surveying I've been able to do, they like the original content and want more. But that's hard to do. Given that the amount of original content is fixed (bounded by inspiration, time and commitments, confidentiality, etc.), what other options are there? Is it better for a blog to just have original content, even though that's less material total? Or better to have more material, references and tangents in addition to original content, because that stuff is valuable too? I think all of us bloggers are trying to figure that out.