peter@baylor.UUCP (Peter da Silva) (02/22/86)
Has anybody with this "make" actually used it? As near as I can tell it requires a seperate makefile for each target you want to make, and seems functionally equivalent to a batch language with a "if older file.obj file.asm" type of construct added. -- -- Peter da Silva -- UUCP: ...!shell!{baylor,graffiti}!peter; MCI: PDASILVA; CIS: 70216,1076
cramer@kontron.UUCP (Clayton Cramer) (02/25/86)
> Has anybody with this "make" actually used it? As near as I can tell it > requires a seperate makefile for each target you want to make, and seems > functionally equivalent to a batch language with a "if older file.obj file.asm" > type of construct added. > -- > -- Peter da Silva > -- UUCP: ...!shell!{baylor,graffiti}!peter; MCI: PDASILVA; CIS: 70216,1076 We're using it. It does NOT require a separate makefile for each target you want to make. It is very similar to the UNIX make. Major differences: 1. With UNIX make, the make file is sorted first so that if an executable is dependent on the .o files, and the .o files are dependent on some source files, the .o will be made and then the executable will be made, even if the executable dependency is before the .o dependencies. The Microsoft MAKE needs the dependencies in the order of production. 2. With UNIX make, I have yet to create a line too long -- and some of my dependencies are 250 or more characters to a line. Microsoft MAKE is limited to 132 (?) characters to a line, so you have to use continuation characters (back slashes) to tell it to continue on the next line. This isn't really so painful, except that we have a script on our VAX that generates the make file for UNIX make, and we can't easily modify the script to generate the makefile for Microsoft MAKE.