From meissner@lynx.cs.usfca.edu Thu Feb 1 03:48:46 1996 Received: from lynx.cs.usfca.edu (lynx.cs.usfca.edu [138.202.160.1]) by dkuug.dk (8.6.12/8.6.12) with SMTP id DAA04049 for ; Thu, 1 Feb 1996 03:48:42 +0100 Received: by lynx.cs.usfca.edu (AIX 3.2/UCB 5.64/4.03) id AA31078; Wed, 31 Jan 1996 18:43:01 -0800 Date: Wed, 31 Jan 1996 18:43:01 -0800 From: meissner@lynx.cs.usfca.edu (Loren P. Meissner) Message-Id: <9602010243.AA31078@lynx.cs.usfca.edu> To: Craig.Dedo@mixcom.com, jwagener@ionet.net Subject: Re: (SC22WG5.1021) (x3j3.1996-41) Re: fodder for meeting Cc: , List@mixcom.com, Mailing@mixcom.com, WG5@mixcom.com HOLD EVERYTHING. Don't go off half cocked, at least until you can reconstruct the discussion when the requirements were put into F90. I think I was there -- wasn't there a meeting somewhere near Dallas in mid 1990? Joan Brixius as I recall argued strongly for the present situation. Something along the following lines: When I put a RECL integer in READ, I can control the representation because it is an integer that I create in my program. But the integer that comes back from INQUIRE is generated by the system so we need to force it to be DEFAULT so it won't foul me up in case my program is not expecting a weird KIND. I am pretty sure that was the gist of the argument, and everybody seemed quite convinced at the time. Maybe somebody knows how to get hold of Joan Brixius; I am sure she would remember. --Loren Meissner