3 ms·

Chrome, no. But I am running FRR on Ubuntu server, and it's also distributed as a Snap - in fact, that has spurred the most known discussion about Snap autoupda

by kzemek 6y ago

Chrome, no. But I am running FRR on Ubuntu server, and it's also distributed as a Snap - in fact, that has spurred the most known discussion about Snap autoupdates[1]. Of course I can - and do - use the deb version, but it's just one of the critical-always-on things that can creep onto a system as a Snap. For example, LXD is moving with Snap as the default way to distribute on Ubuntu[2]. [1]: https://forum.snapcraft.io/t/disabling-automatic-refresh-for-snap-from-store/707 https://forum.snapcraft.io/t/disabling-automatic-refresh-for... [2]: https://discuss.linuxcontainers.org/t/is-the-snap-the-recommended-way-to-install-lxd-in-ubuntu-18-04/2248 https://discuss.linuxcontainers.org/t/is-the-snap-the-recomm...

ordinaryperson 6y ago
OK, but why is the problem FRR not addressing bugs it may be auto releasing and instead the entire principle of auto-updates in general? If we agree that auto-updates on the whole improve security for the platform, why is that not a goal worth pursuing? Why are application devs totally blameless in releasing buggy software?
- Shared404 6y ago
  All software has bugs. Also, how are the application developers supposed to test for every single environment? In an ideal world, you could get away with pushing all updates automatically, but I for one would rather not get my production server get totalled because of a botched update. *edit to fix question mark
  - GordonS 6y ago
    I think Windows updates illustrates your point nicely. Microsoft are a vast organisation and do a huge amount of testing before pushing updates, yet time after time, there are reports of serious issues with updates. As a software developer myself, I completely understand the desire to have consumers running the latest version - but I also recognise that real world users have different workloads, different levels of acceptable risk, and different consequences when things go wrong. I think automatic updates probably are the best thing for most desktop users, and for some server users but certainly not for everyone all of the time - I don't even mind if auto update is the default, but make it clear and give people a config option where they can control updates themselves!

by kzemek 6y ago

ordinaryperson 6y ago

OK, but why is the problem FRR not addressing bugs it may be auto releasing and instead the entire principle of auto-updates in general? If we agree that auto-updates on the whole improve security for the platform, why is that not a goal worth pursuing? Why are application devs totally blameless in releasing buggy software?

Shared404 6y ago
All software has bugs. Also, how are the application developers supposed to test for every single environment? In an ideal world, you could get away with pushing all updates automatically, but I for one would rather not get my production server get totalled because of a botched update. *edit to fix question mark
- GordonS 6y ago
  I think Windows updates illustrates your point nicely. Microsoft are a vast organisation and do a huge amount of testing before pushing updates, yet time after time, there are reports of serious issues with updates. As a software developer myself, I completely understand the desire to have consumers running the latest version - but I also recognise that real world users have different workloads, different levels of acceptable risk, and different consequences when things go wrong. I think automatic updates probably are the best thing for most desktop users, and for some server users but certainly not for everyone all of the time - I don't even mind if auto update is the default, but make it clear and give people a config option where they can control updates themselves!