On Wed, Dec 5, 2018 at 9:48 AM Paul Winalski <paul.winalski@gmail.com>
wrote:

> Another curious compiler at DEC was VAX Ultrix Fortran.  DEC had
> gotten a lot of push-back from the research community, who wanted to
> use VAX and Ultrix but considered the f77 compiler inadequate and
> wanted to use VAX Fortran, which only ran on VAX/VMS.  There was a
> rush project to port VAX Fortran to Ultrix.  It was decided that the
> quickest way to get a quality compiler to market was to have the VAX
> Fortran compiler continue to emit VMS-format object files, and to
> modify the VAX/VMS linker to accept a.out object files and to emit
> a.out images.  Four of us worked on the linker port.  Two of us from
> the VAX/VMS languages team did the linker mods and two engineers from
> the Ultrix group wrote code to translate VMS debug information to Unix
> stabs.  The resulting linker was called lk.  The VAX Fortran RTL was
> also ported, and since we had a way to produce Unix executables from
> VMS object files, it meant we didn't have to rewrite the RTL, which
> was mainly in BLISS but also had modules in Fortran, VAX assembler,
> and Pascal.
>

Nice work! We ran VMS on our MicroVAX rather than Ultrix or 4.[23]BSD
because of the FORTRAN compiler being so much better on VMS and our group
needing it for its hydrological simulations.

Warner