Bill Seitz notes that Instapaper can’t grab the meatier plaintext from Markdown files on GitHub.

I’ve noticed that GitHub doesn’t seem to play nice with other services at all. It breaks the WordPress bookmarklet I use (to write this post, for example).

No idea why. Is it a bug or a deliberate thing?

Bill Seitz has launched his short e-book : Hack Your Life with a Private Wiki Notebook on the Amazon Kindle.

There’s a fairly complimentary review by me on the Amazon site, which should be read in light of the fact that I’ve been watching this book grow over the last few months and sending comments on the drafts back to Bill. So I have some personal involvement here.

I stand by the message, though. That this is an important book for someone to write and has a wealth of interesting ideas and references. It’s well worth a couple of dollars and an hour to read if you aren’t already immersed in Bill’s world.

However, as I pointed out some time ago, wiki is surprisingly bad for authoring longer texts that structure an argument for a reader. And, to an extent, the style of this book betrays its origins as a set of notes on wiki-pages. Both in the way that it’s fragmented – forcing the reader to navigate links rather than follow a narrative in the text – and in the way you see history sedimented. What makes wiki bad for authoring structure is that, at least traditionally, it offers little support for revising / refactoring. What often happens on individual wiki pages is that you write something substantial when you launch a new page, and then in the flow of updating, just paste extra links, ideas, thoughts at the bottom.

That soon leaves you with longish pages which are effectively little rivers, except upside-down with newest stuff at the bottom and the oldest at the top. Which can be fascinating for archaeologists who want to reconstruct what you thought at different times in the past, but doesn’t make them particularly useful as “book chapters”. Pages are both too narrowly specific (they only talk about one thing) And too full of historical cruft. Yet it’s very tempting to see them as the basis of such chapters.

And there’s nothing much to help you transform / refactor standard wiki-pages into approriate segments of a longer text.

As an aside, the Smallest Federated Wiki’s refactoring turns out to be highly flawed in that it enforces a rigid idea of paragraph in the structure of the document. In practice, when writing and rewriting you often want to grab and move different sized chunks of text. Sometimes single paragraphs, but sometimes just a couple of sentences from one paragraph to another. And often a larger chunk of three or four paragraphs. SFW makes single paragraph movement between pages or re-ordering very easy, but at the cost of making single sentence or multiple-paragraph re-ordering far more difficult. There’s no way to select / cut / copy / paste chunks of text with arbitrary boundaries the way that an single “edit” box for the whole page would allow.

So Bill’s is a nice book. And it’s a worthwhile read. But it hasn’t escaped the curse of being wiki-written. There are still moments when you encounter that sedimented history and think “if you were writing this from scratch, that wouldn’t be there”.

Still, I think it’s going to be interesting to see how successful this wiki-style is in an e-book format and for an e-book readership. One advantage of e-book readers is that at least they support hyperlinks, so the contents in the book can be surfed and meandered through. OTOH, is that the pattern of reading that people expect or want when sitting down with a Kindle? It’s in interesting experiment.

Over on my main blog you may have seen that I’m musing about my online presence again. Increasingly fed up with Facebook I’ve now taken the plunge to remove myself entirely. (I haven’t, as of writing, deleted my account only because I need to extract some more writings before I do.)

I’m also increasingly concerned about my dependence on Google for so much of my online life.

One man who has few such qualms is Bill Seitz, who has consistently stuck to his home-brewed WikiLog concept over the last 10+ years. I’ve criticised the idea of WikiLog before – with one of my high-falutin conceptual arguments – but actually I’ve had to admit that Seitz is right and I’m wrong. The virtues of combining wiki and weblog functionality in your own software (which means very easy, high-density linking between both types of entry, and consistency of managing the address, full ownership etc.) outweigh any qualms about the difference of addressing philosophies.

Now Seitz has gone back to adding functionality to his wiki : the WikiGraphBrowser adds dynamic visualisation that shows the links between pages, creating an instant “TouchGraph” style mind-map. I’m excited, partly because of the software he’s producing, but partly because here’s another smart person investing in wiki’s future.