I’d like to kick off a new effort aimed at making small but meaningful quality-of-life improvements to snapd, and I’m calling it the snapd papercuts initiative.
Over time, we’ve all encountered little frustrations. Maybe the snap command didn’t do quite what you expected, required a workaround script, or didn’t provide helpful output when things went wrong. Maybe snapd did something unexpected or confusing, snap installation or refresh failed and it was hard to find out why. These kinds of issues can make using snaps or managing ones that are installed in your system more difficult than it needs to be, and I want to help change that.
The goal of this initiative is quite simple:
Collect feedback from the community, track these usability issues, and work together to fix them.
Whether it’s a confusing message, a missing flag, or something that just feels harder than it should be, I’m eager to hear about it. These issues will be logged either in the Launchpad snapd project or in our internal tracker, and I’ll be looking for people, both from the snapd team and the wider community, who are interested in helping resolve them. I’m happy to mentor and support anyone who wants to get involved!
I’ve also been working on packaging snapd for other distributions, so if you’ve hit papercuts there, feel free to include those as well.
How you can help:
Reply here with issues you’ve encountered
Link to any existing Launchpad bugs you’ve filed
Share any ideas for small but impactful improvements
When installing say chromium, it pulls the cups snap which I don’t want. Purging the cups snap is only working until chromium is next refreshed which causes cups to be installed again.
I’d like to not have to constantly state that I don’t need/want an optional dependency snap to be installed
I think it’d be neat if the snap install command was improved so that it would clearly state e.g., downloading $content_snap (XX%). A lot of the time the install command completes quickly and the automatically connecting... feedback only shows on screen for a few seconds. However on a slow internet connection with a new content snap the user might not already have available, this process that is often fast can instead look broken because it’s not clear what’s actually going on is a separate download that could realistically take several minutes on some connections.
Well, it would be nice if we could have a way to tell snapd that there is already a deb based cups that fulfills the need on classic systems … (in fact firefox seems to be fine with the snapd provided cups-control interface provided by the OS while chromium pulls in the cups snap for this it seems)
It’s worth bearing in mind Ubuntu wants the CUPS snap to be the default provider of CUPS; and that the CUPS snap is considered preferable since it reduces the scope of access vs cups-control. It’s more likely Firefox would end up using the CUPS snap than it is for Chromium to transition to cups-control.
The request and intention still makes sense, but there won’t be a .deb based CUPS for 90% of snaps users in the long term.
Not sure whether this counts as a snapd issue but I have experienced this on 24.04, and the provided solution solved it. I had a friend try installing VLC and he didn’t come across the issue.
https://code.videolan.org/videolan/vlc/-/issues/28047
Thanks, this has indeed been a problem in the past. However, at this stage it would be best for vlc snap to replace its current base core18 with one that is more recent.
Hi everyone! I’ve been following this usability feedback initiative and love the direction it’s going.
I come from a product/project coordination background and would love to support by helping track feedback trends and sharing simple, periodic summaries in the forum highlighting top issues, progress updates, and any calls to action.
Let me know if this is helpful, or if there’s already something in motion I can support.
Thanks again to everyone contributing it’s inspiring to see how this community comes together to solve real usability challenges.
It would be nice if we could remove a snap and its base at the same time, assuming that the base is no longer needed on the system after the snap is removed. As an example:
ubuntu@core24:~$ sudo snap remove jq core
error: cannot remove "jq", "core": snap "core" is not removable: snap is being used by snap jq.
The core snap is only being used by jq, which is also being removed. It’d be nice if this didn’t take multiple steps from the user. Right now, you have to snap remove jq and then snap remove core.
One thing which bothers me: for the freeorion snap we did not find a good solution for save games yet. We do not have a cloud save option or similar, so we rely on the local filesystem.
If somebody removes the snap, the save games are gone without a warning. This will probably break expectations in a really bad way. The best case if someone did this is that they find this out and are able to figure out that they able to restore it, before it gets permanently deleted after the default 30~ day timeout. I am not sure what to do about it, it may be mostly an UX issue/managing the exjpectations of the users. One simple fix would be to give some info on removal, maybe project defined.
E.g.: “Your save games get deleted when you remove this snap. You have 30 days to restore the snap before the data is deleted permanently. Consider to copy the save games from … to another location.”.
And some info when getting info about a removed snap.
E.g. “You removed this snap 27days ago, you have 3 more days to restore your related local user data”.
