Alright, got some time on my hands right now, so... For the glossary on klf.de I created a database of possible entries, i.e. 3 a.m. Eternal, It's Grim Up North or whatever you think should have a place in the wiki. Next, I set up a second database with synonyms, abbrevations and alternate spellings for the entries in database #1 - 3AM, 3 AM Eternal or similar. Whenever a text entry gets indexed, the glossary reads all entries in database #2 and matches them against the text. If there is a hit, the occurance will be stored in database #3 :-) Not where it is, or how many hits exist for one entry, just the info that there IS one. These are stored along with the index of the text we just searched through. Whenever a visitor accesses a page that has been indexed, the glossary reads all previously found hits from database #3 and converts them into links (like you can see in the entry for 808 State or Doctor Who). The advantage is that the system doesn't need to match every possible entry against this text (which would take quite a long time) but only those that have been found earlier when the administrator (me) indexed all entries. So all you have to do is to come up with a new entry and its alternate spellings (see above) and the respective text, and the glossary will place links to this new page whereever the entry is mentioned. As you can see in the FAQ section, this isn't even limited to the glossary itself, but can be applied to whatever section you want. So, basically it is a Wiki-like structure. It needs some small additions now, but it provides a very good platform for an extensive KLF database. ..::// Thomas Touzimsky A best man's speech should be like a mini-skirt; short enough to be interesting, but long enough to cover the bare essentials. (Unknown)