My Dream Machine

I know what I want. It’s pretty ambitious:

I want a database-centric personal publishing platform.

The front end is sort of like a word processor. But it is really a database. It has little tools that allow one to quickly assign fields. In fact it suggests fields, tags, etc as you write text, sort of like auto-complete. This intelligence could come from the Freebase commons, or even from personal use. When text is finished it lives in this database. But you can also drag in images, in whatever shape you get them. But as much as possible an author writes sentences and text and works in images in a editorial way.

Then I can extract this data in many forms. My first extract would be to send it to a blog for posting. Perhaps I highlight the group of text and images i want to send. The blog has a style sheet which styles the output. I check the result and look and I get to manually override styles before I publish.

The same information is also sent out in a an email with a different set of styles and parameters. Or it may have some extra bits of info taken from the database.

I may then take a whole bunch of postings and want to reformat it for a more permanent web page, with a different look, structure, function.

And I will definitely want to extract the same data into a page-makeup layout program and make a long-form “book” out of the same stuff with different formatting, style, etc and make a PDF or print it.

When I change something in the root database, the effect of the change needs to show up instantly (say after refreshing) in the derived format. Even  better would be able to make the change, like fixing a typo, in any formatted edition and have it really change the database so the same change shows up everywhere. So if I hyperlink a term once in the database that hyperlink travels wherever that linked term goes, whether to the web or a PDF.

As far as I can tell this publishing system does not exist. Abobe has many parts, including the ancient word processor Framemaker which has metadata inputs, and RoboHelp, but no unified database centric platform. There are systems for producing catalogs, such as QuarkXPress Server, and  In essence they have the tools for formatting info all different ways, but no unified platform. And there are a lot of databases with horrible output control, not to mention primitive input. Folks working on smart input, aren’t thinking about beautiful, or multiple, ways to output.

And while the temptation is to make an “enterprise” level program, some superduper thing that a big publisher would use, I think this is the future of personal publishing, and I want a personal version.

Whose going to write this?

And if someone is or has already written this, point me to them quick!

—