It's my experience as an Emacs user that Emacs (version 29.1) is one of the buggiest I ever used. (Perhaps, some time around 2015 or so was another low.) Not only is Emacs itself very buggy, but many of the packages on MELPA, so it's becoming something of a risk to update packages. I like to keep things up to date generally, but it can be quite disruptive when there are serious bugs.
Recently, I had some very puzzling problems with AUCTeX that turned out to be caused by the installation of version 14 (which I don't think is the recommended current version, and which in any case is seriously broken). I don't know how that happened, but I certainly didn't manually update AUCTeX, so I assume it was auto-package-update.
I've had a number of other similar problems over the past year, and my perception is that the problem is becoming critical, with Emacs really just holding on by a thread to keep from being essentially unusable. I, for one, am spending a lot more time than I'd like just trying to hold things together so I can keep working.
In this environment, I think it's important to have a mechanism for flagging projects that are known to have severe problems, so they are excluded from auto-package-update, and/or possibly automatically reverting them to a reliable version.
It's my experience as an Emacs user that Emacs (version 29.1) is one of the buggiest I ever used. (Perhaps, some time around 2015 or so was another low.) Not only is Emacs itself very buggy, but many of the packages on MELPA, so it's becoming something of a risk to update packages. I like to keep things up to date generally, but it can be quite disruptive when there are serious bugs.
Recently, I had some very puzzling problems with AUCTeX that turned out to be caused by the installation of version 14 (which I don't think is the recommended current version, and which in any case is seriously broken). I don't know how that happened, but I certainly didn't manually update AUCTeX, so I assume it was auto-package-update.
I've had a number of other similar problems over the past year, and my perception is that the problem is becoming critical, with Emacs really just holding on by a thread to keep from being essentially unusable. I, for one, am spending a lot more time than I'd like just trying to hold things together so I can keep working.
In this environment, I think it's important to have a mechanism for flagging projects that are known to have severe problems, so they are excluded from auto-package-update, and/or possibly automatically reverting them to a reliable version.