From LOJBAN@CUVMB.CC.COLUMBIA.EDU Sat Mar 6 22:59:10 2010 Return-Path: Delivered-To: veion@XIRON.PC.HELSINKI.FI Received: (qmail 14115 invoked from network); 13 Nov 1996 17:13:21 -0000 Received: from SEGATE.SUNET.SE (192.36.125.6) by xiron.pc.helsinki.fi with SMTP; 13 Nov 1996 17:13:21 -0000 Received: from listmail.sunet.se by SEGATE.SUNET.SE (LSMTP for OpenVMS v1.1a) with SMTP id <8.CEFACDCF@SEGATE.SUNET.SE>; Wed, 13 Nov 1996 18:13:18 +0100 Date: Wed, 13 Nov 1996 09:12:20 -0800 Reply-To: jimc@MATH.UCLA.EDU Sender: Lojban list From: Jim Carter Subject: Re: place switching cmavo... To: LOJBAN@cuvmb.cc.columbia.edu Content-Length: 1112 Lines: 20 Message-ID: On 1996-11-12 "Trevor C. Hill" writes: > ... It seems to me that because the places are, for the most part, > ordered by frequency of use, these words should simply take a certain > place (x2 for se) and put it at the front, leaving the other places in > their original order... (because you usually want to fill x1 and so need "fa" as well as "se".) I also have noticed that effect. If it were five years ago I would have (and I think I did) argued to have conversion operators just yank a place rather than swap places. But we now have the inertia of a fair amount of text written to the original TLI behavior of conversions, plus a baseline, so my support for the idea has to be virtual, not actual. -gua!spi conversion operators do not swap places; they leave remaining places in the original order. James F. Carter Voice 310 825 2897 FAX 310 206 6673 UCLA-Mathnet; 6115 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA 90095-1555 Internet: jimc@math.ucla.edu (finger for PGP key) UUCP:...!{ucsd,ames,ncar,gatech,purdue,rutgers,decvax,uunet}!math.ucla.edu!jimc