on 5/4/00 5:43 AM, M. Uli Kusterer at [EMAIL PROTECTED] wrote:
> Adrian,
>
> could you probably do a few things in it:
>
> 1) Could you add some docs? I have no idea what's happening, why I get
> "false" and "true" answered when clicking "parse sentence". It'd be nice to
> have a liitle text field that just describes the different options and what
> they do. Also, some labels on the fields would help a great deal already.
Adrian: Yeah, they're coming. I had hoped to get them completed by now, but
the move to SourceForge filled my weekend instead. Work will continue soon.
> 2) It'd be an added nicety if you made the stack smaller. I have an 800x600
> screen and need to use the scroll box (!), and there's so much whitespace
> and all the fields are scrolling anyway.
Adrian: Oh yeah. I'll fix that. I have a 17" screen, I don't mind wasting
space while developing. I had always intended to clean up the interface.
It should wind up being 640x480 compatible.
> 3) Could you prepare a small blurb essay about these networks? I've heard
> about the general concept since my sister is studying computational
> linguistics at Uni but I really don't know more about it than that I could
> say that something is a network when you sketch it on a sheet of paper. I'm
> not talking about an in-depth description, just an attempt to provide a
> rough description of the concepts behind it and what its benefits are. I'm
> sure Anthony would like this, too, in case the design of his NuParser would
> benefit from that.
Adrian: I'll do this as well, I'll need it for my final report on the
project, it would be good to get it out of the way now.
> It's great to see that you put so much work into ATNParser and I'll try to
> give it a closer look by the end of the week. BTW, does it handle
> ambiguities? I had the impression you declared "studies" as a verb, whereas
> it could be a plural noun in English pretty well, too.
Adrian: The network that's currently stored in it is purely for testing,
don't pay too much attention to it. The network will be set up so that it
can handle ambiguity, though the HyperTalk syntax eliminates it pretty well.
> -- M. Uli Kusterer
Adrian: Expect to hear more from me about this soon.
Adrian Sutton
PS: Use Netscape to get an account at www.sourceforge.net now!!!!
**************************************************************
Email: [EMAIL PROTECTED] Ph: 3714 4649
Never put off until tomorrow what you can do the day after tomorrow.
-- Mark Twain.
**************************************************************