[email protected] said: > [explanation of librumpkern_sys_foo vs. rumprun-posix]
Thank you for the detailed explanation. > I seem to recall that there's a mail in the list archives discussing the > relative merits of syscall compat vs. rumprun, but I couldn't find a > link with a few minutes of searching. I hunted through the GMANE archive back to its beginning and couldn't find anything related either... Would be interested in understanding more about how this works, will put on my list of code to read :-) > Anyway, getting back to your proposal, I think I would prefer to guide > new users just towards rumprun, where things just magically work on any > platform. Less explaining => better. If learning about the type > incompat becomes necessary, it can be an expert topic. Makes sense? IMO they (librumpkern_sys_foo, rumprun-*) serve different use cases. If I understand you correctly the rumprun-* model serves when you want to run unmodified applications using the entire rump kernel stack. In the case of eg. fs-utils and also the project I've been working on where you want to interact both natively with the host filesystem and a rump kernel filesystem, rumprun-* does not really fit...? Martin ------------------------------------------------------------------------------ _______________________________________________ rumpkernel-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/rumpkernel-users
