Web-based Virtual Machine Management - Discussion 1

V0.2

- What actual things could be in ‘mozilla branch’? o For instance: I don’t really understand how CVS/tinderbox works; with SVN it’s simple: r3201; r3202; r3203.. i.e. linear o With CVS, how does it work? Can I actually specify a time (12:36 on 11/12?) Would that even be useful?

- Toolchain still unclear: assume skip for now

- What is the next step for operating systems? o Should I be setting up (i.e. installing manually) several ‘base VMs’? o What role does my web-interface even play? (It seems rather over-engineered to have a bunch of ad-hoc VMs, and the only purpose of the web-interface is to download a Mozilla source tree to the /opt dir of that VM.) o What role does LVM play? Is the intention for ‘COW’ to be that many people can be tinkering on the “same” VM at once, i.e. Fedora 7 with thunderbird trunk + Fedora 7 with firefox trunk on two different VMs?

- In the long run o It seems like a lot of the effort is in cloning virt-manager. How can I not do this?