On 06/17/2012 09:22 AM, Kent A. Reed wrote:
Ditto, altho I don't have the technical skills to do such. Step 1 (parse) and step 2 (canonical) should not be too difficult. From there the curve steepens considerably. Tool paths with multiple geometry tools and cutter comp is non-trivial. ( probably the understatement of the year).On 6/17/2012 10:19 AM, Matt Shaver wrote:On Sat, 16 Jun 2012 10:28:32 -0400 "Kent A. Reed" <kentallanreed@xxxxxxxxx> wrote:I'm hoping to get over to the standards library at NIST to read these as well as old versions of ANSI X3.37. NIST used to be an open campus but with security the way it is now, I have to get an invitation from a staff member.Did you hear back from Fred? Thanks, MattFunny you should ask, Matt. Fred responded with a suggestion I consider coming over to do some interesting things in his lab. (NIST has always taken advantage---and I mean it in a good way---of retirees who feel the need to continue technical work.)To their credit, Fred and Tom Kramer have looked around, but neither has come up with any documentation I don't already have. It was a long shot at best: their research was oriented toward machine control, not design automation.In reading the APT360 programmers manual, I've been struck again and again by how easy it would be create a Python-based parser of the APT geometry statements. It would be similar to a parser I wrote nearly 20 years ago when we were testing our understanding of the STEP geometry model. The trouble is, the geometry is just the tip of the iceberg. I think I'd have to do too much work to make a useful tool for the APT language (which is about machining, after all) as opposed to the geometry.Regards, Kent
Dave
