In principal it should be fairly simple to write a script to search all entries. Use the fact that all threads have the same URL except for the 5 digit number at the end to generate a list of all possible threads, just replace the numbers with 00001 to 99999. In practice you could limit it much more. As Chandler mentioned most of these won't be hits so if you wanted to get all the content you could write some sort of script to try and load each page and only save the ones that aren't empty. Once that is done most computers have plenty of horsepower to search for a word or phrase. You could also filter various posters, pro or con, in your search.
No I haven't written one. No I don't plan to. We revisit topics on a pretty regular basis, so it isn't really that necessary. If I ever do, I'll let you know or send it to Founder to impliment if he sees fit to do so.
Shadowcat: Are you talking about attempts using Google domain search? That's what I mean by spotty results. I think it's going to be hit or miss no matter what tricks you try, because I don't think Google indexes the contents of this site very regularly or thoroughly. If it's not all getting indexed, it's not going to show up in search results. As for why TUSCL content is not indexed like some other sites are, that's beyond my level of geekdom.
Try a Google domain search. I think that's what the Founder recommended in answer to an earier question. Select Google Advanced Search and enter [view link] in the domain field. The results will be spotty. It's hard to detect any pattern as to what kind of search terms return the best hits. You can also try a regular Google search and get similar results if your terms are specific enough. Also, be aware there was a period when domain searches weren't working at all with TUSCL, so I don't know if the site's registration lapses sometimes or what.
Messages don't appear to get purged, at least not for years. I've seen results that go back over two years. As you may have noticed, topics drop off the list after one week without any replies. However, the topics remain online. If you bookmark a topic, you can browse to it after it has dropped off the list. What I've done is save the topic list as a web archive file every week or so when I remember to. Then, I can use that to go back and view old topics.
Also, you can type in old topic numbers in your browser's location box, e.g., try replacing the 93846 that follows "DiscussID=" at the end of this topic's URL with a lower number. Not every number has a corresponding topic. Only about 1 in 8 or 1 in 10 produce anything. I'm not sure if there's a formula for the numbers or if it's random. I've meant to ask the Founder about that, or if there's some way to pull up one, long list of old topics.
3 comments