Build failure in amber-demo branch
brian.goetz at oracle.com
Sat Oct 6 14:59:11 UTC 2018
Actually, the intent was far less ambitious than that — it was a “trial-fit” mostly for purposes of, well, doing some demos, and seeing how they fit together. But, the three features feeding into it are moderately stable, so it will probably be useful for more than just this for a while. If the merge overhead becomes significant, we may take it down.
More broadly, we haven’t yet enumerated a policy for when a branch is stable enough that we should try merging it with other feature branches. There’s definitely desire for a “all the amber features that are good enough to try out” branch.
For the record, here’s a catalog of the current branches:
condy-folding: enhanced constant folding (JEP 303) and intrinsics for indy/ldc.
jep-334: Constants API (dependency of condy-folding), in the process of final review for PTT
patterns: experimental branch for pattern matching
patterns-stage-1: curated branch with the subset of pattern matching features to be launched first
enhanced-enums: dormant, associated with Enhanced Enums JEP, which ran into a fatal collision with generics
stats-before-this: experimental branch, dormant
lambda-leftovers: experimental branch associated with Lambda Leftovers JEP, dormant
datum: prototype for records, relatively stable
concise-method-declarations: recent JEP for concise method declarations
The “demo” branch derives from patterns, datum, and concise-method-declarations
> On Oct 6, 2018, at 5:39 AM, Aleksey Shipilev <shade at redhat.com> wrote:
> Thanks for creating amber-demo branch! I assume that is the branch that merges all development work
> in one convenient place? I set up builders to produce binaries for that branch:
> Current branch fails to build though. Part of it is fixed in upstream jdk/jdk, but there are
> amber-specific failures, it seems:
> error: variable dc is already defined in method blockStatement()
> Comment dc;
More information about the amber-dev