From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-0.8 required=5.0 tests=BAYES_00,INVALID_DATE autolearn=no autolearn_force=no version=3.4.4 Path: utzoo!utgpu!watserv1!watmath!att!cbnewsl!arny From: arny@cbnewsl.att.com (arny.b.engelson) Newsgroups: comp.lang.ada Subject: more silly portablity problems Message-ID: <1990May25.171457.26472@cbnewsl.att.com> Date: 25 May 90 17:14:57 GMT Organization: AT&T Bell Laboratories List-Id: Aside from the Ada portability concerns I'm used to dealing with, such as Integer and Float type differences, rep specs, etc., I ran into an unexpected one the other day. I received something close to 20_000 lines of Ada source code from someone, which I believe had been compiled under Alsys. When I attempted to compile it under DEC's compiler, I discovered (to my disappointment) that the compiler limits the input line length to 120 characters and the source code contained many lines (hundreds I think) over this limit. The compiler simply ignored everything beyond column 120 (it did print an error message). I don't mention this to fault the compiler or the source code (although personally I find line lengths > 80 columns to be most annoying during those times I am stuck with an 80 column terminal or printer). This just seemed to be an interesting portability concern. Fits in with the maximum length of identifier names problem which can be much more cumbersome to fix (like after you globally shorten the name and find out it is now equal to some other name). The situation was easy to fix (my choice was to write a quick program that splits the line at the last token prior to column 120), but still something we should be aware of when writing code, pretty printers not withstanding. P.S. In case anyone cares, the code was CADRE's Ada binding to their Teamwork/ Access routines. (CADRE Teamwork is a CASE tool, for those unfamiliar with it) -- Arny Engelson att!wayback!arny