# [isabelle] questions about PIDE architecture



Then I made a counter-movement away from JVM-based project management tools, just with minimalistic shell scripts around Isabelle tools. The latter is briefly documented in the "system" manual as follows:

section {* Scala compiler \label{sec:tool-scalac} *}

Makarius




I see, you have found a way to replace IDEs' "run configurations", and I had the opportunity so see that used on your laptop efficiently such that I never could follow what is going on ;-))

But may I ask for "notes on PIDE architecture" ?


In PIDE communication between Java (i.e. Isabelle/jEdit) and Isabelle/Scala seems straight forward, because both run in the same JRE.
But how is communication between Isabelle/Scala and Isabelle/ML possible?


Over the years watching the advent of *.scala in ~~/src/Pure/ I imagined some magic about the parallelism between the *.scala files and the *.ML files. But in

https://github.com/larsrh/libisabelle/tree/master/pide-core/src/main/scala


there are some directories from ~~/src/Pure/, and all of the copied directories contain exact copies of all the *.scala files and while dropping all *.ML files.

So, where comes communication between Isabelle/Scala and Isabelle/ML from?

Walther



This archive was generated by a fusion of Pipermail (Mailman edition) and MHonArc.