The KeePassXC snap listing page is simply promoting other method of distributions. I highly object this. Such listings should be banned. If a distributor is unhappy with their snaps, just remove it. Snap Store should not allow such promotions at any cost. I request immediate address to this as this is a big concern for users who might get tempted to such things. Such listings can also be used to scam people. @alexmurray @cav @sahnaseredini
2 Likes
On the other hand, the KeepassXC snap has had some serious problems with Snap historically, and the outreach they’ve been given wasn’t necessarily the best at the time. When you’re presented with what appears as a format with nothing but issues, I can understand them wanting to go to pastures new. We’re talking bugs bad enough that it damages their brand because their software looks unreliable, I’ve seen the Reddit and Twitter threads. Of course, it damages our brand too, it doesn’t look good for anyone.
Ultimately, whilst it’s clear they don’t like us as much as we’d like, KeepassXC does actually still take bug fixes and updates the snap for security and maintenance. Personally having recently forced myself to actually move back to Linux for daily use (Did people know I’ve been publishing snaps for 5 years despite not using Linux on real hardware in 10?) , I want that snap to work as good as it should be able to because I actually want to use it myself, and it’d be nice if we had more dogfooding like that going on in general. The recent proposal to get the browser extension working automatically was done in a few hours despite being a problem for years. How many problems are trivial but being perceived as insurmountable because we’re simply not bothering to lend a hand?
I don’t personally think the Snap is in a bad state anymore. Historically there were major bugs like not even being able to open files, or periods of time were the app didn’t even install properly because the content snaps it relies on were removed from under it. People don’t seem to stop using it whether that message appears or not.
Whilst it’s their project and ultimately their call, I’d like to think a little more community outreach would do better for both sides without much effort. A good example would be if we could be proactive on Core24 with the new KDE extensions, rather than leaving several months of an uninstallable snap as with the Core22 upgrade & etc.
So I’m not expecting that message to be removed anytime soon because historically the KeepassXC snap has had a deservedly mixed reputation. But I think most people here agree we’d want a KeepassXC snap of high quality, and if that takes having our ego bruised a little (fairly) and some actual graft to get there then so be it.
2 Likes
There is my objection. Don’t serve it until it’s ready. I got into some fights in the past, where my manifest was and published in a half baked manner to the store. The upstream had to intervene into the matter. But, to be honest. If things were reversed, would it have happened? There are many apps in the other platforms that doesn’t work as expected or work with false hope of security “Cough Cough flatpak spawn”, still no one came up to raise their voices.
It also seems like using snaps popularity to serve other purposes (i.e may be to grow other formats). Is that a good practice? You want to use the popularity and user base of a format but you don’t want to invest in that?
1 Like
In principal I agree, but I do think the KeepassXC snap was served ready. It was just served years and years ago, and that’s the issue, there’s been years for problems to be fixed that haven’t been fixed because we can’t expect every maintainer to be an expert in the art of making snaps. That snaps problems actually got worse over time, not better (through a mixture of ecosystem changes and to some extent bad luck). There’s a difference between releasing something that never worked to begin with and a genuine attempt at trying that turned sour.
I’d imagine many of us in these forums have a password manager of some form, if not potentially a lot of us KeepassXC itself. I don’t think it’s unreasonable that for something as foundational in our ecosystems as KeepassXC that Canonical should have an expectation to outreach better than it currently is, and if Canonical can’t provide the resources, well, as I say, I want that snap to work and I’d rather it stayed keepassxc than keepassxc-unofficial. I’ll dogfood it myself.
I understand the sentiment, but is releasing the snap not investing in our format? People do actively use that snap whether upstream wants them to or not. Rather than keeping them here by guilt over their existing userbase, I’d love to see that snap actually live up to its potential. It’s really not far off IMO.
But wiping off years of stigma will take time. I remember 3 other one liners I’d encouraged their snap to adopt that were seen as massive fundamental issues that really weren’t. For example, the autostart field, the password-manager-service interface, and a super specific one liner that was designed to improve the apps resistant to attack, was causing it to be unable to open web links or files for months. We’re talking problems that can be fixed trivially going unfixed for years because people aren’t lending a hand.
Ultimately, there used to be a Snapcraft Developer Advocate position. I believe Alan Pope used to hold that title for a while. I’m assuming there isn’t one currently, and it seems like there really should be, because there’s specific situations like this where we could do good for our own communities and theirs but simply aren’t.
There’s plenty of technical problems, but there’s social ones we need to fix too.
1 Like
Can’t agree more with this.
Regarding password-manager-service and other super privileged plugs, well, I think portals don’t work everywhere. And I think there should be catagories, like there are for classic snaps.
This thread doesn’t really fit the store-requests category, nor is it a productive discussion - the keepassxc snap does not appear to violate the terms of service of the store so I am closing this thread.
1 Like