RTEMS Source Builder | Next (Bleeding Edge) Version Numbers are Annoying (#92)

[Chris Johns (@chris)]

Chris Johns commented: https://gitlab.rtems.org/rtems/tools/rtems-source-builder/-/issues/92#note_120007


!127 discusses the change from the release buildsets to a common named directory of configuration files. I stated on discord my objection:

> **opticron:** even if we deprecate it over 2 major versions and provide stubs with warnings (and solutions) in the interim?
>
> > **kiwichris**: Yes this is my current view. I am not seeing any advantage other than a localised name changing. Any wrapping script will know the RTEMS version being built already so that will be something like the deployment fragment I posted in !127. You will need equivalent git churn on release of RTEMS I have been publicly active for a number of years now saying the the 5 to 6 BSP buildset change was the last major interface facing RSB churn after I was received strong feedback on the breakage I caused and this is a bigger change

The MR !127 conversation https://gitlab.rtems.org/rtems/tools/rtems-source-builder/-/merge_requests/127#note_119902 repeated what I said in discord and added a requirement for any change to happen:

> It breaks backwards compatibility of tools used to build and integrate RTEMS and for me that is the top priority. To have me agree and approve any related MR it has to build the K26 stack with `rtems-deployment` `HEAD` plus build the K26 with the 6.1 RSB so show backward compatibility.

-- 
View it on GitLab: https://gitlab.rtems.org/rtems/tools/rtems-source-builder/-/issues/92#note_120007
You're receiving this email because of your account on gitlab.rtems.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/bugs/attachments/20250215/4027ff46/attachment-0001.htm>