From a5d7b72899d5959a3dc305d1ee39c5cb64966fbc Mon Sep 17 00:00:00 2001 From: Daniel Kahn Gillmor Date: Mon, 6 Jun 2011 19:20:25 -0400 Subject: updated TODO with various not-unreasonable ideas --- TODO | 31 +++++++++++++++++++++++++------ 1 file changed, 25 insertions(+), 6 deletions(-) diff --git a/TODO b/TODO index 3601eaf..f10242a 100644 --- a/TODO +++ b/TODO @@ -1,18 +1,37 @@ TODO for KVM-Manager -------------------- -(much) better documentation + * (much) better documentation -file re-organization/restructuring + * file re-organization/restructuring -debian packaging? + * debian packaging? -more flexibility with networking (currently only supports + * more flexibility with networking (currently only supports tun/tap-style bridging) -/usr/local/sbin is hard-coded in places (e.g. in the run and finish + * /usr/local/sbin is hard-coded in places (e.g. in the run and finish scripts); it probably shouldn't be. -screen logging is still sub-par, and could potentially fill the host + * screen logging is still sub-par, and could potentially fill the host disks if the kernel spew is immense -- rotate? filter through svlogd somehow? same problem as cereal. + + * how do we send break signals over the serial line in the new setup? + + * we aren't currently supervising the console screen well. if the + user kills it, it will be gone. + + * can we separate the privileges for the monitor from the privs for + the console? + + + * can we start up the console screen first, listening on a UNIX + socket, and then launch kvm with the console as a client instead + (so that it works the other way around)? + + * can we get our logs to be timestamped? + + * di-maker fails if you give a full path to the iso. + + * allow handing a kernel + initramfs to the guest as a special case + of booting -- cgit v1.2.3