Simplified Blogging: Part 1 - Creating an Entry
Posted on October 6th, 2005
Hot on the heels of my previous post, I want to work up a definition of what a simple blogging system could look like. The way I see it, blog entries are fundamentally documents. But the usual interfaces that are involved in creating blog entries aren't the same ones that you'd normally turn to when you wanted to create a document. Why?
When I think about creating a document, the type of application that first comes to mind is a word processing program like Microsoft Word or Open Office. The last thing that comes to mind is a from on a web page, no matter how much you dress up said form to look and act like like a full-blown word processor. It's more than just having a toolbar with a lot of buttons to click on. I can do all sorts of things on a Word document that I can't do from a web-based interface: save incrementally, select blocks of text reliably, spellcheck against a customized dictionary, or undo and redo like a madman.
Maybe someday soon web-based interfaces will be able to do all these things and then some. Still, a desktop-based word processor has a pretty strong head start. And they're abundant in the here-and-now. I need something that works with today's tools.
So it would be awfully nice if I could start a blog post by firing up a word processor instead of a browser window. But of course, that's tricky. The file formats! The inevitably horrible HTML that will come of attempting to convert them! The screaming babies!
Blosxom dodged that bullet by sticking with text files. The simplicity and universality of text files is well established (craphound.com, 43folders.com, oreillynet.com). But going the text file route means you have to give up the comparative wealth and splendor of a word processor's interface. For example, the 3 links in the past sentence were hand typed from the "href=" all the way to the "</a>". For me that's no big deal, but I'd rather hit Ctrl-K in Word. Even though HTML is relatively simple to learn, it's not part of the primary mission. The primary mission is writing a captivating blog entry, not captivating HTML.
So how do you get from your Word processor to HTML with as few casualties as possible. I think AbiWord might be the way to go. It can convert a number of file formats to HTML via the command line, which is just the thing I'm looking to do. And it does a decent job of it too. I've also looked at wvware, but it's more Microsoft Word-centric. AbiWord uses it internally anyway. From some preliminary comparisons between the two, AbiWord leaves behind more structural clues.
That's the cincher of course-- structure. Who's to say whether 14 point bold text is a top level title or a subtitle? I'll need as many clues as I can get, because this conversion process will by definition be imperfect ("lossy", as they say). I think that's okay though. My blog entries at least aren't exactly spectacles of layout daring-do, they're just a bunch of consecutive paragraphs. Sometimes with an image. Maybe with a blockquote. It would be unrealistic to expect a conversion tool to take a tricked out Word document with headers and footers and master pages and all that kind of crap and expect the end result to magically keep everything intact. But I think there's plenty of middle ground between that extreme on one hand and text files on the other. It's a blog entry after all, not a magazine layout.