jon@astro.ucla.edu (Jonathan Eisenhamer) (05/06/89)
To All Spots, A basic question: Is it possible to use dynamic memory with fortran programs compiled with SunOS 3.5 f77(1)? There doesn't appear to be fortran versions of alloc/free and no mechanism in fortran to allow the use of pointers. Now for some catching up on old stuff: A while ago, I asked about unqualified accounting errors without any explanation using Transcript 2.0. There were two errors, accounting error 2 and accounting error 3. No one seems to know about the error 2, but error 3 indicates that the start and stop pages returned from the Laserwriter were inconsistent (i.e. start page # greate then end page #, etc.). No suggestions about what would be causing this were given though. Also, a while ago, I asked about an apparent bug in the system(3f) fortran call. The problem is that, under 3.5 f77(1), a call to system(3f) would result in all opened files being set to the end-of-file. One person (lost the address) responded that this is a known bug but did not know whether there was a fix for it. Thank you for your time, Jonathan Eisenhamer jon@astro.ucla.edu jon@uclastro.bitnet bonnie::jon (span 5.708)
zjat02@uunet.uu.net (Jon A. Tankersley) (05/18/89)
In article <8904202125.AA01698@mira.astro.ucla.edu> jon@astro.ucla.edu (Jonathan Eisenhamer) writes: >The problem is that, under 3.5 f77(1), a call to system(3f) would >result in all opened files being set to the end-of-file.... There is a patch for system() for 3.5 f77. We were losing fd's after system calls. This was before they increased the number of possible active fd's... We'd still have lost them though with the bug... -tank- #include <std/disclaimer.h> /* nobody knows the trouble I .... */ tank@apctrc.trc.amoco.com ..!uunet!apctrc!tank