Memory usage in 1Password for Safari

Options
This discussion was created from comments split from: Electron.

Comments

  • gussic
    gussic
    Community Member
    Options
  • @gussic oh yeah, I was going to address that, thank you for the reminder. I am not sure if the post belongs in this thread because there is no Electron in the browser extension.

    Some of the increase in the memory usage is related to the fact that the extension is now standalone and does not require the app to function. There are certainly some pros and cons to that. At the same time, the memory usage seems too large for me and I asked the browser extension team to dig into it.

  • gussic
    gussic
    Community Member
    Options

    @roustem yes I get its probably not strictly on topic for this thread - but it was replying to your post about how embarrassed you were that 1P7 used so much memory. The new 1P8 + the required browser extensions use significantly more memory than 1P7.

    Happy to be split off into a seperate thread, but can I ask for a genuine/honest answer - why did you proceed with a standalone extension at the expense of a dependent one, if the memory sacrifices are so large? Surely a >10x increase in memory usage is not satisfactory in anyone's book?

  • @gussic We had two different extensions for a long time now. One of them was getting all the love and attention and another one was stagnating.

    I agree that it would be nice to improve memory usage and we are looking into that. At same time, on the grand scheme of things it is not as outrageous as you might think. A single YouTube page open in Safari uses more memory:

  • gussic
    gussic
    Community Member
    Options

    @roustem the difference is YouTube videos don't stay open all day, an extension is always running (as long as the browser is) - why can't we go back to the old system, two extensions to suit both use cases? If that means you have to hire more developers then why is that such an issue? It works for your users much better, and shouldn't that be the primary goal?

  • @gussic That's an interesting idea but I have never seen it work well in practice. It is like if we asked Apple to keep supporting iTunes in addition to the Music app (which by the way also uses more memory than 1Password 😅)

  • gussic
    gussic
    Community Member
    Options

    @roustem whaat do you mean you haven't seen it work well in practice? It worked fabulously with 1P7 - it was fantastically lightweight and fast AF. The new extension is so slow, it borders on unusable.

    Music uses less memory on my machine than the combination of 1P + Safari Web Extension (new) so I'm not quite sure you're right on that? Music is using ~200 mb of memory right now on my machine.

  • Ben
    Options

    Apologies for copy and pasting from another post, but it has been a very long day. I wrote this earlier:

    One of the motivators for going in the direction that we are is to get away from maintaining a bunch of different codebases, freeing up development resources to all collaborate on moving one codebase forward that much quicker and more efficiently. As such I don't see us offering two entirely separate Mac apps.

    We used to have five different apps with five different teams developing them in five different repos. Rolling out even minor changes across the entire 1Password platform was a nightmare. Now we have one app, and one repo, and everyone is working together. The idea is it'll enable us to move faster.

    This was in response to a request that we offer and maintain two separate Mac apps, but the same concept applies to two separate browser extensions, and I think hits the point Roustem was trying to get at when he said it doesn't work well.

    Ben

  • gussic
    gussic
    Community Member
    Options

    @Ben

    Thanks for the post - I understand the reasons for trying to simplify, but when it hurts the user experience like this does, then perhaps you need to reconsider? Like the only advantage I can see for doing this is simplified development for your team, and being able to roll out new features faster. From a customer perspective its almost entirely bad, more system usage, more energy usage - I just truly can't understand why you'd let down your loyal users like this.

    Whatever way you cut it, the new browser extension is objectively worse than the previous dependent version, for those with a 1P desktop App. It only improves things for those without 1P Desktop App. For those with the Desktop App it is a step backwards.

    Given performance is going backwards, and features are being stripped out (local vaults etc) will you adjust the subscription prices down to compensate? From a customer experience perspective perhaps it would make this regression more palatable.

  • gussic
    gussic
    Community Member
    Options

    @roustem @Ben

    Just downloaded Beta 3 of Safari 15 for Big Sur - can confirm the issue still present itself here as well, it's not as bad as Safari 14 for sure, but there is still a noticeable pause/hang between clicking on the extension and the window/pop up opening. Tested on both Intel and AS Macs.

  • @gussic - thanks for bringing this delayed opening of the browser extension to our attention, and for the specifics about your current setup. I'll follow up with our development team about this.

    ref: dev/core/core#7342

  • gussic
    gussic
    Community Member
    Options

    @PeterG_1P Thank you for the reply.

    Would it assist if I sent diagnostics, or is that not a thing for the browser extension?

This discussion has been closed.