From: Malcolm Dunnett

> I am using ghostscript 8.54 on my Alpha servers, using a kit that Mark
> Berryman build back in 2006.
>
> I have tried downloading the latest version, but the build fails with
> MMK spitting out a "syntax error" from one of the .mak files.
>
> Has anyone built a newer version of ghostscript for VMS? If so, any
> hints on how to get past this error (and on how well it works)


Assuming that your "the latest version" is 8.63, then it looks as if
no one could have built it on VMS from _that_ kit. You can clear the
worst of the initial complaints with changes like these:

ALP $ gdiff OPENVMS.MMK_ORIG OPENVMS.MMK
278c278
< DEVICE_DEVS21=
---
> ### DEVICE_DEVS21=

419c419
< .suffixes : .c .obj .exe
---
> ### .suffixes : .c .obj .exe



ALP $ gdiff DEVS.MAK_ORIG DEVS.MAK
422c422
< $(GLOBJ)lvga256.so: $(lvga256_)
---
> $(GLOBJ)lvga256.so : $(lvga256_)

426c426
< $(GLOBJ)vgalib.so: $(vgalib_)
---
> $(GLOBJ)vgalib.so : $(vgalib_)

530c530
< $(GLOBJ)X11.so: $(x11alt_) $(x11_)
---
> $(GLOBJ)X11.so : $(x11alt_) $(x11_)

685c685
< $(RINKJ_OBJ)evenbetter-rll.$(OBJ): $(RINKJ_SRC)evenbetter-rll.c
---
> $(RINKJ_OBJ)evenbetter-rll.$(OBJ) : $(RINKJ_SRC)evenbetter-rll.c

688c688
< $(RINKJ_OBJ)rinkj-byte-stream.$(OBJ): $(RINKJ_SRC)rinkj-byte-stream.c
---
> $(RINKJ_OBJ)rinkj-byte-stream.$(OBJ) : $(RINKJ_SRC)rinkj-byte-stream.c

691c691
< $(RINKJ_OBJ)rinkj-device.$(OBJ): $(RINKJ_SRC)rinkj-device.c
---
> $(RINKJ_OBJ)rinkj-device.$(OBJ) : $(RINKJ_SRC)rinkj-device.c

694c694
< $(RINKJ_OBJ)rinkj-config.$(OBJ): $(RINKJ_SRC)rinkj-config.c
---
> $(RINKJ_OBJ)rinkj-config.$(OBJ) : $(RINKJ_SRC)rinkj-config.c

697c697
< $(RINKJ_OBJ)rinkj-dither.$(OBJ): $(RINKJ_SRC)rinkj-dither.c
---
> $(RINKJ_OBJ)rinkj-dither.$(OBJ) : $(RINKJ_SRC)rinkj-dither.c

700c700
< $(RINKJ_OBJ)rinkj-epson870.$(OBJ): $(RINKJ_SRC)rinkj-epson870.c
---
> $(RINKJ_OBJ)rinkj-epson870.$(OBJ) : $(RINKJ_SRC)rinkj-epson870.c

703c703
< $(RINKJ_OBJ)rinkj-screen-eb.$(OBJ): $(RINKJ_SRC)rinkj-screen-eb.c
---
> $(RINKJ_OBJ)rinkj-screen-eb.$(OBJ) : $(RINKJ_SRC)rinkj-screen-eb.c



ALP $ gdiff LIB.MAK_ORIG LIB.MAK
1222c1222
< $(GLOBJ)strmio.$(OBJ): $(GLSRC)strmio.c $(AK) \
---
> $(GLOBJ)strmio.$(OBJ) : $(GLSRC)strmio.c $(AK) \



ALP $ gdiff OPENVMS.MMK_ORIG OPENVMS.MMK
278c278
< DEVICE_DEVS21=
---
> ### DEVICE_DEVS21=

419c419
< .suffixes : .c .obj .exe
---
> ### .suffixes : .c .obj .exe



And then it fails much more cleanly.

Judging from all the non-space-separated colons, it's pretty clear
that no one tests this stuff on VMS. At least not using MMK or MMS. At
a glance, I'd guess that some actual work would be needed to make it go.

------------------------------------------------------------------------

Steven M. Schweda sms@antinode-info
382 South Warwick Street (+1) 651-699-9818
Saint Paul MN 55105-2547