Tinderbox isn't yet groupware  10/21/04

Tinderbox is designed as a single-user application, and the awkwardness of trying to make it work as groupware is extremely frustrating considering the potential of the completely imaginary Tinderbox Server (which I will call Tinderboxen). I am amazed at many aspects of Tinderbox (and will be posting on them), but this frustrated me-- because I want it to be better still, not because it is broken.

Tinderbox would need a lot of work to facilitate sharing and groupwork. There's no client/server architecture available for it such as you find with FileMaker. There's no way to work with Tinderbox data except through Tinderbox, such as the clients-and-webform model of LiveJournal that makes groupwork across networks or platforms possible. There would need to be more integrated wizards that would help get users to where they wanted to go in the UI or data

There's also the more abstract issue of how to provide users other than the creator with easy entries into a Tinderbox file. With data that can be seen geographically, in outline form, as html, as text, or as a hierarchical chart, how do you communicate to a new user the structure of your particular file? There are rudimentary techniques for this such as "readme" notes and adornments, but these are cumbersome for trying to communicate an entire way of thinking-- and Tinderbox is flexible enough to accomodate extremely divergent tasks and ways of thinking.

At the same time, Tinderbox is almost there, and that proximity is agonizing. It stores data as xml, so the data structures are clean and accessible without proprietary closed-source markup. Other applications could work with the data. Tinderbox is great for seamlessly accomodating both live and static (archived) work, and has features built for workflow management. Recent versions of Tinderbox accomodate interactive wizards that could be the solution to the issue of paradigm-introduction. If I could share it, it would rock! (harder, that is)

An example of why it's so frustrating comes from a conversation with my girlfriend the other day. Her company is considering a production site for the show she's working on. Tinderbox could be great for this-- it would be part production blog, part static site, with FAQs and write-ins and enough other stuff that the organization and simplicity of Tinderbox would be an asset. The various staff members that might contribute wouldn't have to learn HTML or the structure of the site to provide content. The content would be abstracted and separated from the visual style (which the studio would likely want to have a say in) so that the design and copywriting for the project could proceed in parallel. It would accomodate funky HTML for those who want to use it, but would make the coding transparent to those who don't.

But to do this, they would have to get copies of Tinderbox for everyone who might contribute, or tie the project down to a single editing station. Tinderbox isn't very expensive as software goes, but it's really expensive if you might not use it. And then even if they got 15 copies they'd still have to pull a time-share on the data. If they could serve Tinderbox and have keyed licenses or the like, it would be the perfect tool. But, for now, sharing is a dealbreaker.