|
From: | John E. Malmberg |
Subject: | Re: VAX/VMS 7.3 and build from copy of master. |
Date: | Fri, 28 Feb 2014 19:45:27 -0600 |
User-agent: | Mozilla/5.0 (Windows NT 6.2; WOW64; rv:17.0) Gecko/20130215 Thunderbird/17.0.3 |
On 2/26/2014 1:57 PM, h.becker wrote:
On 02/26/2014 08:07 AM, John E. Malmberg wrote:I have removed it, because we need to move on.Yes. I think beautifying the build environment wasn't really the goal.It is an issue on search lists, in some cases if the sys$disk:[] is missing, the file may not get found due to the stickyness of DCL file handling. I may be putting in a few more places than absolutely needed just as a defensive measure.There weren't sys$disk:[]s in all the places. In prepare_vms.com I prefixed the parsed template file with one. It looked strange to search with that prefix but then to have a result without. That broke the pipe command, which I changed again to fit the 255 maximum of a DCL command token (for versions prior to V8.2 of VMS).
The sys$disk:[] is needed only in the output, or the current version of EDT will attempt to put the file on the same volume as the input file.
In makefile.com, having objdir := sys$disk:[] but no such prefix in the source file list and in the cc command seems to be inconsistent. I'm still curious to see a problem without this prefix in the make build environment.
In makefile.com they are not needed for the C compiler output.
I think it is sloppy to have error messages for expected conditions.Unfortunately VMS has no equivalent of "rm -f", delete/nolog isn't it. But one can redirect the output of make.
Which is worse because it suppresses all messages, which means legitimate warnings can get missed.
We can go with this patch so that we can move on. Regards, -John
[Prev in Thread] | Current Thread | [Next in Thread] |