...that since lenny it's safe to mix the two because a. apt-get now handles (or can handle)orphans similarly thanks to autoremove and also uses the same database or better yet, has the same markings for automatically installed vs manually installed?
I've been browsing for this for a while now and the official documentation is never really clear on that and in forum there's nothing but arguing about the fact (yeah and i open yet another thread). anyway - maybe someone in here knows for sure and has some sort of convincing "proof"
If it's not I'll have to stick with apt-get but if it is that would be awesome for apt-pinning because aptitude is better at satisfying dependencies when installing packages from a lowly pinned source. apt-get only uses the highest priority version of a package unless you specify a target release with something like let's say -t experimental but then it will get ALL dependencies form experimental even if some of them could've been satisfied from unstable or testing.
so let's say "foo" from experimental needs "bar" from expreminental but for "bla" the testing version would be enough- and let's also say bar and bla are both in different versions on testing unstable and experimental. Testing is pinned very high, experimental very low.
apt-get install foo/experimental will complain about needing bar in version whatever but that's not to be installed
apt-get -t experimental install foo will install both bar and bla from experimental
now if I make aptitude install foo, it's supposed to get bar from experimental and bla from testing.
That would surely be a reason for me to switch. aptitude's snail-pace would be a small price to pay.
so - any expeerts help a pinning newbie out?