Hi,
from time to time I am playing with the approach to move powerbuilder to an open source platform. I concluded that there is no technical reason not to build an open source compiler (maybe a vm).
I see there some "phases".
1) Build a working commandline tool to build PowerBuilder applications.
In the first step I would not build a new compiler from the scratch but extend PBOrca to use Powerbuilder Orca functions. Orcascript is a dead end. I tried this but it is not "big" enough.
My "model" are the tools TypeScript (tsc) or Cordova. It should be possible to write PowerBuilder applications from Notepad and build them with the command line tool. From this point it doesn't matter what IDE is used to build apps.
2) Alternativ IDE
I build a small Visual Studio extension to provide a new Project type to hold *.sru, *.sra, ... yes it is a bit like PB.NET but keep in mind: this is Classic! And yes! My approach does not include working in PBLs, but in plain text files. PBLs are only to compile and build.
When pressing "build" the cli tool of point 1 is called.
3) Leave PowerBuilder 3rdparty code
This is the hardest part but to this the project is already autonomic. The PowerBuilder IDE is not longer needed.
From this it is necessary to build an own compiler (Wowowowo, are you kidding my?). My idea is to use microsoft technology. The .NET compiler opened to the open source community (Roslyn) and there is no show stopper to compile PowerScript to MSIL and to .NET nytecode (Do not compile to C#, like SAP does, it's bull ... it's not a good idea). This is not just a compiler but a whole framework in .NET. That means the PowerBuilder objects should have a "pendant" in .NET. (DataWindow, SQLCA, Pipes, ...).
4) We are free!
Insert new language specifications to PowerScript (We will get interfaces! We will get a foreach keyword!!!).Build new DataWindow source (REST, ODATA). Build predefined DataWindows like plugins to get data from specific datasource (AmazonDW, GoogleDW). It will come with NuGET.
Conclusion
But all these ideas are stopping on one point: Who will use it? Who will contribute? My answers are not optimistic.
best regards
Benjamin