But then there is "testing", and "unstable" too which gets newer packages quite often. I personally run "unstable" as my main development machine and have for many years. Even "unstable" is very stable, but might require some manual work once-a-year or so, with a solution often easily found by some google-fu.
Then there is also "experimental", which might very well have broken dependencies etc.
The cool thing is you can run stable, but then pick some packages from unstable using what is called apt-pinning. This way you can get a rock solid base, but use newer packages of some software. Best of both worlds!
But obviously you try to avoid this in production environments if possible.
Debian packages aren't necessarily much (any?) more stale than they are in Arch. It depends on whether you're using stable, unstable, or testing. (If I remember correctly, "unstable" is actually fairly stable, whereas "stable" is more like an LTS. "testing" is closer to Arch, though less "bleeding edge" than Arch is.)
I run testing on my non-prod (experimental) server instances and would probably go with stable for anything production quality, although I've never experienced anything going wrong with Testing.
I have my server tied to "Testing" rather than the current name, which turns it into a rolling distribution (like Arch).
Edit: Not sure whether there was a real problem before (vs PEBKAC), but the new images _do_ reboot successfully (takes about 2 minutes). Now to get systemd running...!
I think they were just looking for a scapegoat in that particular case.
[1]: https://lists.debian.org/debian-cloud/2014/02/threads.html
On the wiki https://wiki.debian.org/Aptitude they say dist-upgrade is no longer recommended, I am a bit confused (I don't keep up with linux world very much, I am just a linux hobbyist).
Thanks.
