Dan rants: More musings on web manglement

This is an extension of my musings on the next generation of Newwwsboy .

After Burning Man , my next priority is revamping my web pages. Part of that is colocating a server so I can have some database accesses on-line, but some of that involves cleaning up the text and form of the pages themselves. And being the lazy programmer that I am, rather than spending a few hours going through my pages and making all the updates directly in HTML, I'm going to spend a couple of days writing a content mangle... errr... management system to deal with common problems.

For instance, I have a couple of terms that, when they're outside a link, should probably always link to something. Any time I use a trademarked term I should probably put a tm after it, any time I reference a common company name I should link to their home page, things like that.

I should look for entries in my text that appear to be the names of people and look in some of the usual sources to see if they are. Similarly for things that might be book titles. Why not automatically check those names against the databases of Amazon or Border's , for instance?

And once I find a name I should keep it in a local database cache, both so that I can override it and so that I don't keep hitting remote sites for every little question.

I should automatically insert meta description tags so that I don't end up with my boilerplate template stuff in all the search engines.

I should be able to keep glossaries around, and the glossaries should be able to be heierarchical; a word may have a different meaning in my whitewater pages than in my graphics notes .

I should be able to relocate trees quickly and easily, so that I can keep and edit local copies of the SHS-SF web pages, for instance, without having to duplicate the top part of the tree on my local servers.

I've heard rumblings from various people that such a product has reasonable commercial application too, and I was wondering what other products people have seen that try to solve these same issues. I've seen the advertising stuff about Frontier , but my impression is that you're tied into their outline database form rather than being able to pull data from where people are already commonly using it.

I'm also interested in what sort of tools people are using to edit XML, because manually inserting tags for indexing and glossary management doesn't seem like a rational option for your average doc writer.


Friday, August 20th, 1999 danlyke@flutterby.com