I do not follow your logic on testing being the most bug-prone branch.
In order for a package to enter testing branch, it should build for all architectures it is intended to, and it should have been in unstable for a predefined length of time (10 days?). So, while you're technically correct in saying that a bug can stay for a while in the testing branch, it can only happen when the bug existed in the unstable branch for the same amount of time, and no developer fixed it during that time.
In other words, those bugs (that get fixed within hours in unstable) simply never enter testing.
Not to mention that your statement is inconsistent with my own experience. I have been running testing for some time, and have tried to "upgrade" to unstable a few times. While the upgrades were not disasters, there were small things that stopped working. I am too lazy to figure them out all by myself, and reverted to testing which is stable enough for me.
When I want to be on the cutting edge (for example KDE and mozilla), I stay in testing and only install those packages from unstable.
I find that Woody is quite stable - not perfect but only slight glitches from time to time.
It is possible to stay with Woody and only upgrade some packages from Sid. apt-get -t unstable works if you want to upgrade just a few packages. For more substantial upgrade (such as KDE), I find that aptitude works nicely.
Slashdot Mirror
I do not follow your logic on testing being the most bug-prone branch.
In order for a package to enter testing branch, it should build for all architectures it is intended to, and it should have been in unstable for a predefined length of time (10 days?). So, while you're technically correct in saying that a bug can stay for a while in the testing branch, it can only happen when the bug existed in the unstable branch for the same amount of time, and no developer fixed it during that time.
In other words, those bugs (that get fixed within hours in unstable) simply never enter testing.
Not to mention that your statement is inconsistent with my own experience. I have been running testing for some time, and have tried to "upgrade" to unstable a few times. While the upgrades were not disasters, there were small things that stopped working. I am too lazy to figure them out all by myself, and reverted to testing which is stable enough for me.
When I want to be on the cutting edge (for example KDE and mozilla), I stay in testing and only install those packages from unstable.
I find that Woody is quite stable - not perfect but only slight glitches from time to time.
It is possible to stay with Woody and only upgrade some packages from Sid. apt-get -t unstable works if you want to upgrade just a few packages. For more substantial upgrade (such as KDE), I find that aptitude works nicely.