You may already know this, but just in case you don't . . .you need to
understand that the decision to track testing instead of stable or
unstable is a serious one, with potentially serious hazards. Many
people have the concept in their heads that testing is a fairly
robust distribution that's less risky than unstable -- something like
"more current than stable, but less dangerous than unstable." That
isn't true, because that isn't the purpose of testing. When things
break in testing, they can often stay that way for an extended period
of time -- viz. two occasions last year when first GNOME, and then
KDE, were uninstallable out of testing, and broken for those tracking
testing closely, for more than a month in each case. That's not
because of the developers doing anything wrong; it's just the nature
of what testing is. If your reason for wanting to upgrade to testing
is because you see it as splitting the difference between stable
and unstable, you need to read about what testing is and how it's
(automatically) put together.
OTOH, if you already knew all the above, cool, sorry for being
long-winded, maybe it'll be helpful to someone else.