From: arne@vajhoej.dk   
      
   On 3/26/2025 8:54 AM, gcalliet wrote:   
   > Le 26/03/2025 à 13:02, Arne Vajhøj a écrit :   
   >> On 3/26/2025 1:09 AM, David Meyer wrote:   
   >>> Is there anything in the VSI licensing that would prevent a community of   
   >>> VMS and Rust (for example) fans from developing a VMS port of a Rust   
   >>> compiler and releasing the compiler as open source?   
   >>   
   >> No.   
   >>   
   >> VMS users can write or port all the compilers they want to. And   
   >> they have done so in the past: old versions of GCC C and C++ ran on   
   >> VMS VAX and VMS Alpha, old versions of Gnat Ada ran on VMS Alpha   
   >> and VMS Itanium.   
   >>   
   >> The reason it is not happening is not license restrictions, but   
   >> lack of interest (willing to do work type of interest - not   
   >> it would be nice if somebody else did the work interest) in   
   >> the VMS community.   
   >>   
   >> The specific discussion was about the LLVM compiler backend,   
   >> that VSI use for their compilers. If VSI made that available   
   >> (it is open source) then it would be easier for people to   
   >> write or port new compilers using LLVM as backend.   
    >   
   > It's the most important point.   
      
   Note: this is pure speculation - I have no idea about   
   what VSI actually thinks about the topic.   
      
   It is obvious that VSI will need to upgrade LLVM several   
   times within the next 10-20-30 years. It is obvious (to me   
   at least) that it does not make sense for VSI to re-port aka   
   re-apply changes every time they take a new cut - it makes   
   sense to get the VMS changes into the upstream product to   
   make LLVM upgrades much easier. So when Intel & AMD   
   release AVX-8192 in 2034, then LLVM add support for that,   
   VSI grab the latest LLVM and the VMS compilers support it.   
   It has to happen.   
      
   But things does not happen by magic - they happen due to   
   hard work. It takes effort to setup a usable public   
   Github repo (usuable = no dependency on VSI internal build   
   environment) and it takes effort to get open source changes   
   into an upstream project. And VSI compiler team is not a 1000   
   man team. So they prioritize.   
      
   I suspect that if projects to write/port native compilers   
   based on LLVM (for languages not supported by VSI) to VMS   
   started to materialize, then that could help push the   
   VMS LLVM release up the priority list.   
      
   Arne   
      
   --- SoupGate-DOS v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)