I didn't see anything in the build 3036 announcement and ensuing discussion about benefits for Linux VMs. My primary VM is Ubuntu Edgy, and although I occasionally run a Windows XP VM as well, I don't need Coherency, Bootcamp support, or any of the other goodies in this release. So is there any benefit to Linux VM users, or should I sit this one out?
While I use Parallels for many guest OS' I have found the ability to have multiple nics very useful in linux.
There was a lot of improvements unrelated to guest OS. Check the list please - http://forum.parallels.com/thread5997.html .
Real Linux improvements We read this list of course but we would like real features like a X11 Video driver or copy-paste working with Linux. In OSS projects its not polite to ask for features but as I payd 70 dollars and propably have to pay again soon for the next release: Please listen to the second largest userbase. Another solution would be publishing some specs so we can we can create stuff ourselves. but I reckon that's not an option.
That may just be the best option if Parallels can handle not letting the fear of competitors get to them. In this way they can free themselves the burden of officially supporting the tools and yet still get the benefit of having happy technical users. As a side benefit, this lets Parallels become the platform of choice for all of the first mover/technical types out there. If I knew that Parallels gave the specs to the community and VMWare didn't for instance, guess which one I'd be using and also guess which one everyone that I support (family, friends, business...) would be running as a result. A sort of geek halo effect. For example, my main VM environment is FreeBSD 6.1 (which 3036 just broke BTW). Since there are no tools for that right now I came up with a hack that sort of works where I use the NoMachine client to connect to the VM so I can do things like cut&paste. But I'd love to have a tools package sort of like in Windows. If it was open then this is a possibility - if it's binary only then it'll probably only work in a few Linux distros and supporting this could also slow the pace of the full product. My suggestion is to put out a basic open source Linux tools package (even if not fully feature complete) and give enough info to let the community take it from there.