From jwagener@ionet.net Tue Jan 23 17:55:19 1996 Received: from ion3.ionet.net (ion3.ionet.net [204.96.200.8]) by dkuug.dk (8.6.12/8.6.12) with ESMTP id RAA23282 for ; Tue, 23 Jan 1996 17:55:09 +0100 Received: from hwmac4 (tsip45.ionet.net [206.28.164.54]) by ion3.ionet.net (8.6.12/8.6.12) with SMTP id KAA10610; Tue, 23 Jan 1996 10:52:56 -0600 X-Mailer: InterCon TCP/Connect II 2.1.2 MIME-Version: 1.0 Message-Id: <9601231055.AA47110@hwmac4> Date: Tue, 23 Jan 1996 10:55:47 -0600 From: "Jerrold L. Wagener" To: camiz@mat.uniroma1.it Cc: sc22wg5@dkuug.dk Subject: nonadvancing I/O Content-Type: Multipart/Mixed;boundary=part_AD2A6F3000041EBA00000002 --part_AD2A6F3000041EBA00000002 Content-Type: Text/Plain; charset=US-ASCII Content-Disposition: Inline Prof. Sergio CAMIZ Department of Mathematics "Guido Castelnuovo" University "La Sapienza" Piazzale Aldo Moro, 2 I-00185 ROMA ITALY Tel. (+39+6) 4470.0988 FAX (+39+6) 445.4642 / 4470.1007 E-mail CAMIZ @ MAT.UNIROMA1.IT Dear Dr. Wagener, I follow since years Fortran Forum and the progress of definition of new Fortran standards. I use Fortran since ever for both short and long programs, mainly oriented towards calculus and statistical computations: I am planning to move towards Fortran 90 in the next future, although I still work in strict Fortran 77. I have no particular comments to do on the Committee work, but I wish to outline the interest of a particular format control edit descriptor, presently available in both Microsoft Fortran 77 and Vax Fortran 77. It is either a backslash (\) or a dollar ($): its effects is to avoid that an end-of-record mark be sent to the output, so that next I/O statement will continue on the same line. Its typical use is to prompt to the screen a particular input during a dialogue, such as the following: Do you want to save the output (Enter y/n)? --> _ that would be written and used as such: Character yn Print '(a$)', 'Do you want to save the output (Enter y/n)? --> ' Read *, yn If (yn .eq. 'y') then .................. Else .................. End if It is evident that it is but an aesthetic fact, neither other use of it I found up to now..., maybe in the future it will be overcome by dialogue boxes, windows, buttons, etc. Nevertheless, it is presently very useful for building dialogues and comfortable in its use. I wonder why nobody considered to include it in the new Fortran specifications. Yours sincerely, Sergio CAMIZ Professor ============================= Dear Prof. Camiz: This functionality was discussed during the development of Fortran 90, and a general form of it is included in the Fortran 90 standard - see section 9.4.1 of the Fortran 90 standard. This is the ADVANCE= option in the WRITE (or READ) statement, and when ADVANCE=NO is specified there is no advancement to the next record (line) at the end of the WRITE (ADVANCE=YES is the default and provides the standard F77 functionality). In the event that your intent is that the standards committee consider your less verbose form for input prompts, I am copying this note to WG5 for consideration as a Fortran standards requirement. If you would like to see that happen you should send an appropriate note to sc22wg5@duug.dk or to the WG5 convenor, Miles Ellis (miles.ellis@etrc.ox.ac.uk). Thank you for your note; please let me know if I have not responded adequately to it. Best regards. Jerry --part_AD2A6F3000041EBA00000002 Content-Type: Text/Plain; charset=US-ASCII Content-Disposition: Inline =========================================== Jerrold L. Wagener jwagener@ionet.net 6 E 5th, Suite 308 918-592-3023 Tulsa, OK 74103 918-592-3103 --part_AD2A6F3000041EBA00000002--