Thank you both for your answers. It seems I posted with my regular email, rather than the alias under which I am registered on the list, so I did not receive notice of your updates on the topic - hence my delay. I also see that I left out some important info re version, etc. OS: FreeBSD_11-Current_amd64. Wine: 1.7.11,1. >> if you know that the application is 100% Java, why bother with WINE at all? If you know its not, then using the Windows JRE is almost certainly necessary. Yes excellent point. The app is not 100% java, it installs in windows, and has java component, along with dotnet and probably other stuff. For the dotnet bit I could try to run it through native mono, but the "other stuff" section of the code would probably break the app. >>Don't forget that there are half-way houses too: 100% Java client applications that talk to local Windows-only servers, e.g. a DBMS with no Linux port. In that case you might run the server under WINE but use the Linux JRE for the client. This is most likely the case and the direction that I should further test. How can your suggestion be implemented: "Run the app under wine, but link to native JRE for java calls coming from the wineserver"? > A lot of tools (eg winetricks and playonlinux) work by creating separate wine prefixes for indivdual apps. It is indeed a common use case, though I haven't throughly tested what happens when you run them concurrently and spawn multiple wineservers. That observation is what led me to ask the question. Multiple / concurrent sessions are not a problem, you just get multiple wineserver pid's showing up. I'll just look through the winetricks code or ask their mail list re this question. Your answers did help with confirming for me that I had indeed perceived the winetricks approach to the problem correctly. Thanks Again.
