From Mok-Kong.Shen@lrz.lrz-muenchen.d400.de Wed Oct 6 11:28:36 1993 Received: from cd1.lrz-muenchen.de by dkuug.dk with SMTP id AA25265 (5.65c8/IDA-1.4.4j for ); Wed, 6 Oct 1993 11:28:36 +0100 Received: by cd1.lrz-muenchen.de; Wed, 6 Oct 93 11:29:20 +0100 X400-Trace: de*d400*lrz-muenchen; arrival 06 Oct 93 09:28:48 Z action Relayed X400-Internal-Trace: MTALRZCD1; arrival 06 Oct 93 11:29:08 +0100 action Relayed X400-Internal-Trace: MTALRZVEE; arrival 06 Oct 93 09:28:48 Z action Relayed P1-Message-Id: de*d400*lrz-muenchen; 931006111946886-MTALRZVEE Ua-Content-Id: 931006111946886- Original-Encoded-Information-Types: IA5-Text Message-Id: <931006111946886-MTALRZVEE*Mok-Kong.Shen@lrz.lrz-muenchen.d400.de> Date: 06 Oct 93 09:28:48 Z From: Mok-Kong.Shen@lrz-muenchen.de To: SC22WG5@dkuug.dk Subject: Comments on Proposals by BSI and Schonfelder X-Charset: ASCII X-Char-Esc: 29 Dear WG5 members, BSI Fortran Panel has done a magnificent job in putting forward very convincing arguments for the obsolecense of a number of outdated, bad and redundant features. In fact I don't see there could be any room for counter arguments. I agree with Aurelio Pollicini that all features listed in the BSI proposal should be declared obsolecent in 9X. I consider it to be the responsiblity of WG5 to help to eradicate these chronical diseases (fortunately already steadily diminishing like smallpox did in the past) of the Fortran community as soon as possible. If a good conversion tool can be soon developed, ideally to be validated by the institutions currently validating Fortran compilers, and put in the public domain, then I think the issue of obsolecense will be entirely free of objections from the public. I don't know exactly whether the BSI proposal has first to be treated by X3J3 before being considered by WG5. If there is no procedural difficulty, I suggest that WG5 immediately conduct an informal letter ballot on the obsolecense issue. A fast action is further desirable in view of a recent (in my opinion justified) critique in the present mailing list that the progress of Fortran standardization processing has been too slow. I like also to suggest that discussions on the paper by Lawrie Schonfelder on parametrized derived types be carried out as soon as possible, if there are differing and better ideas from other members, and that an informal letter ballot be conducted subsequently on that issue. Best regards, Mok-Kong Shen