In article <2025Oct24.171132@mips.complang.tuwien.ac.at>,   
   Anton Ertl  wrote:   
   >Buzz McCool  writes:   
   >>Anton, do any of your undergraduates need a senior project that could   
   >help out with a Debian release?   
   >   
   >There is a release as a .deb package.  And you can install it with   
   >apt, once you have added net2o.de to your apt sources (see the section   
   >about "Debian >= Bullseye" on https://gforth.org/).  As for having it   
      
   So if you are into gforth, you can install it. Big deal.   
   The real goal should be that if you are in a graphic Debian package   
   manager, gforth should show up, if you search for forth.   
      
   >as an official Debian package, one would have to become a Debian   
   >developer in order to do that, and, from what I read, that's a rather   
   >long process.   
      
   I have tried to get ciforth in the Debian distribution. You don't   
   have to become a developer, "merely" you have to find a developer   
   that sponsors the package. The mere fact that gforth was in Debian,   
   means that gforth has a sponsor.   
   I had a sponsor for ciforth, but   
     - he had no affinity to Forth   
     - he didn't understand the quasi FIG-forth philosophy behind ciforth   
       (enable everyone to build a Forth from 1 "one" assembler file).   
     - he concentrated on the insane rules to automatically build new   
        package, cause me to waste enormous time to work that on github   
     - I could not respect him as a software builder. He had one project   
       on github. There was no specification/comment about functions,   
       files. There was not even a readme, where you could guess what   
       that package was for, supposedly.   
      
   You could generate the assembler file in different assembler formats.   
   For Debian this meant that the mechanism to do this, was part of the   
   source, negating the genius of FIG-forth that you need minimal   
   understanding to modify. In fact this was far superior to mere "open source".   
   That is to say, my experience was on the whole negative.   
      
   >But if someone volunteers, they can start with Bernd's .deb and then   
   >maim it according to the requirements of Debian (the current official   
   >Debian Gforth package does not include the manual and disables   
   >native-code generation (for the current development gforth-fast that   
   >causes a slowdown by a factor 4-11).  So I am not so sure that having   
   >no official Debian gforth package is a real loss.   
      
   You are short sighted. It is not about having the latest and greatest   
   gforth in the Debian distribution. It is about having gforth available   
   to the Debian world. Preferably a modern one.   
   A compiler is a tool. Modifications should be made cautiously with   
   an eye on upward compatibility.   
   Hell, if you could convince a Debian developer to sponsor the package   
   that is a real win. At the least gforth 0.7.3 remains available.   
      
   Now the sponsor of gforth has left the package orphaned. That means   
   without action, gforth will disappear altogether from the Debian distribution.   
      
   >Concerning the idea that undergraduate students help out, I would need   
   >some that are interested in whatever I want help for, and my   
   >experience is that most of the time, advising the student consumes a   
   >lot of time, and the results are disappointing (not necessarily the   
   >student's fault); however, sometimes the results are great.  In any   
   >case, unpaid student work is good for proof-of-concepts, not for   
   >something that needs long-term involvement like an official Debian   
   >package.   
      
   Let us assume there is a modern gforth in Debian. The priority is that   
   compilations work from a Debian release to the next. If there are no   
   changes that is not much work.   
   It is much work if you treat gforth as an academic exercise to   
   introduce new concepts. Even then, a student could introduce a   
   certain aspect (say recognizers) from a version to the next.   
      
   Exposure to the unwashed Debian users, will generate (a lot of)   
   bug^H^H^Hdefects reports. There is no reason that sometimes a student   
   could take up one deficiency to lift, and get study points for it.   
      
   >   
   >- anton   
   --   
   The Chinese government is satisfied with its military superiority over USA.   
   The next 5 year plan has as primary goal to advance life expectancy   
   over 80 years, like Western Europe.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)