Re: Is there another ada compiler

[Andreas ZEURCHER <ZUERCHER_Andreas@outlook.com>]

On Monday, August 3, 2020 at 10:18:39 AM UTC-5, nobody in particular wrote:
> We should note, GNAT / Adacore were created on the backs of American 
> taxpayers via a grant to New York University. Unfortunately, the 
> taxpayers got the shaft and a profitable business was born to continue 
> the fun.

Well, I am not usually in the habit of saying nice things about GNAT, but let us compare FSF's GCC GNAT with FSF's GCC CHILL.  Ada and CHILL are fierce competitor languages:  one from NATO military and the other from ITU-T telecom, where Ada trended a little more toward Wirth family of languages as inspiration whereas CHILL trended a little more toward PL/I as inspiration.  Both languages had a 2-decade mandate to be utilized in their respective industrial sectors, but each's mandate had evaporated by the latter half of the 1990s.

Ada had AdaCore arise through several mergers as the for-profit support company for open-source software, analogous to Cygnus Solutions during the 1990s, and its acquirer RedHat until this day.  CHILL had a different business model entirely.  CHILL compilers were produced by the telecom companies that were self-mandated to use CHILL.  If Ada had that business model, Raytheon would have authored its own compiler, Lockheed-Martin would have authored its own compiler, Boeing would have authored its own compiler, Airbus would have authored its own compiler, and so forth.  Eventually the telecom companies in Europe fatigued of the effort needed to write a compiler for an evolving language standard (ITU-T Z.200 and ISO 9496), so 2 of them (Alcatel or Siemens, IIRC) outsourced their internal compiler development to Per Bothner, who eventually landed at Cygnus Solutions, after University of Wisconsin at Madison (years after Randy).  Eventually, Cygnus Solutions convinced FSF to allow their CHILL compiler into GCC.

Shortly after FSF GCC admitted CHILL into its compiler suite, RedHat bought Cygnus Solutions and nearly all of the European telecom companies were finalizing the financially painful governmental reform where PTTs (postal-telephone-telegraph agencies of governments) were divesting their relationship with the equipment manufacturers—much like AT&T divested WesternElectric/Lucent and Bell Canada no longer had Northern Telecom as favorite-son supplier during much the same 1990s time period.  Long story short, when FSF pleaded for someone anyone to update GCC CHILL to GCC 3.X internals, no one stepped forward to fund the effort with money, and most especially no one donated source code as in-kind support.  GCC CHILL as donorware ended as of GCC 2.95.

Whatever or however one might critique FSF GNAT versus AdaCore GNAT Pro differences or delays or never achieving perfect congruence among any pairwise matching of any of their releases, GNAT's viability to continue maintenance & evolution is far better that CHILL's donorware-based approach that failed miserably under the same FSF GCC umbrella during the same time period.  So matters could be far far worse than they are.