Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search


From: brett:codeconscious at: 10-Dec-2000 21:32

I had a glance at the NQL site. NQL is interesting for its Neural net functionality and other gizmos, but it looks to be handicapped by an over-enthusiastic marketing push. I've my doubts about the "simplicity" of manipulating a stack as well (I never quite learnt Forth, nor used the HP financial calculators - so that might explain it :) ) The pricing and positioning of NQL as being a new standard for computing (akin to SQL) is probably to boost an overvalued share price. All up, me thinks Rebol has a much more solid future than this one, on the other hand, we all have seen that the best technology does not always succeed.
> I've read that web mining is going to be huge in a couple of > years, and the more competition, the better each player will > have to be.
Before Rebol I was quite taken by WebL. A research project at Digital Research which was investigating languagues suitable for automatic information gathering from the Web. WebL is quite straightforward and very specific. I thought it succeeded quite well at specific tasks. It is not however supported and for a while, disappeared altogether. I now use Rebol to solve the same problems even if it is a little harder at the moment than WebL. My plan, given a spare huge amount of hours, is to develop a capability similar to what WebL has but using Rebol scripts. Unless of course, someone else will kindly do it and give me the results. Somethings though, I'll not be able to do without changes to Rebol. [RT if you're listening, WebL allows parallel evaluation. Can we have reduce evaluate the items of a block in parallel too?! This could be very handy smoothing out the waits for multiple network requests, or taking advantage of SMP machines.] The thing is, I don't think web mining will be all it's cracked up to be. The main reason being that, the original provider of content may change the structure of the content at any time. Unless that content conforms to some specific standard, you'll never be sure that what you web-mining bot brings back will be useful in any way. If it does conform to a standard, then it most likely won't be the web, it'll be XML or whatever. In other words another protocol.
> RT have that capability there, but it's rather > underdeveloped at present. Awaiting another dialect I guess > - make datamining as easy as VID! > > NQL's ODBC interface looks much easier. And both still lack > IMAP support. > > However, I think one of Rebol's main advantages over NQL is > it's user base, and online support from the RT staff :-).
Indeed. It will be interesting to see if NQL is still around after the DOT.Crash has finished working itself out of the stock market. Brett.