Firefox connector for Zotero

Unless you have a computer dedicated to research, you are forced by the current Zotero to do your day-to-day browsing, googling, facebooking, twitting on a Firefox loaded with your hundreds of academic papers. Any crash or exploit that kills the browser could ruin your library. Switching libraries means restarting the browser for no particular reason, and manually setting Zotfile and Better Bibtex to other folders.
The absence of a Firefox connector makes Firefox both slow and risky.
A few alternatives exist. Write a post-it note to start Firefox after Zotero, and not the other way around. If you want to switch libraries, you refer to the post-it. Or use the bookmarklet that uploads my research to GMU servers. Or the Firefox addons that open the page in a different browser - one that uploads all my browsing to this or that web behemoth.
I hacked into the Firefox extension source to make always assume that it should work in connector mode. But in the last months, Firefox began rejecting hacked extensions.
Please make a real Firefox connector.
  • I'm not really sure what you're talking about here.

    1) Zotero for Firefox functions as a connector whenever Zotero Standalone is open — the order you open the two doesn't matter.

    2) Your data in Zotero for Firefox is no less safe than your data in Zotero Standalone. It's just a database on your hard drive. You can store it in your Firefox profile or somewhere else — by default it will just use the location corresponding to the version you used first. And you should have an automated backup of your entire hard drive, including your Zotero data, wherever it is.

    3) You can still run modified extensions in Firefox — you just need to toggle a pref. When Firefox 46 comes out, that will no longer work, but you'll be able to run an unbranded build of Firefox that accepts unsigned extensions or just sign the modified version yourself. (Still not sure why this would be necessary, though.)

    4) The full version of Zotero for Firefox is going away, because Mozilla is discontinuing the technical infrastructure it requires. So Zotero for Firefox will indeed just be a connector in a future version. But there's very little difference between the full version of Zotero for Firefox and Zotero Standalone as they exist now, and the future connector-only version will function the same as Zotero for Firefox in connector mode now.
  • 0. You don't address the issue of slowness. If you have a decent library with a few hundred records, you add hundreds of MB of RAM to Firefox. Perhaps some users would prefer to browse the net more efficiently.

    1. If you want to switch libraries while using Standalone and the "connector", the order matters. Unless you shut down Firefox, restart Standalone, and then start Firefox, you get "an error has occurred, please restart (doh!) Firefox".
    With Chrome, I can switch libraries by merely restarting Standalone. I wish that wasn't needed either, but it's still simpler than with Firefox.

    2. I know a few Firefox extensions that lose their settings upon Firefox crashes (for example, Mind the Time). I wouldn't be surprised if Zotero does the same one day, so I do backup. This doesn't mean it has to be this way.

    You have a dichotomous view of exploits. In your example, exploits and bugs either eat all your hard drive, or they are completely ineffective. You exclude the likelihood of a crash affecting solely Firefox settings. My example shows that isn't true. If middle-of-the-way exploits do not exist, no one would be upset to see Windows running drivers code in ring zero, because, you know, either the viruses are impotent jokes, or they will eat hard drives regardless of the CPU ring one uses. No one would separate administrators from regular system users, because, you know, malicious users could always escalate their rights to hard-drive-destroying powers. Better to backup and ask admin-users to be nice, right?

    Sorry for the irony, but it seems you are making an active effort to ignore my point.

    3. In the long run, I could even get around to writing my own browser and my own reference manager. That was a marginal point, and I hope I won't have to do it.

    4. People are happily installing hundreds of "apps", although those are, in fact, glorified web browser windows locked onto a single website. I am sure users will survive separating browser from bibliography management, IRC, WebRTC, egg boiling, and anything else one could have built using XUL. If one writes a book or a dissertation using Zotero, one should be happier to see Zotero a true standalone program.
  • (this is a completely pointless discussion given the future trajectory of Zotero. I'd just close it).
  • (this is a completely pointless discussion given the future trajectory of Zotero. I'd just close it).
    Yup.

    I'd just add that switching data directories regularly is well outside normal Zotero usage patterns, so it's reasonable to expect that you might need to adjust your workflow slightly in that case or use a custom build. No one has ever even reported that problem with connector mode and switching data directories before.

    Anyway, as adamsmith says, no point in spending more time on this. Closing.

This is an old discussion that has not been active in a long time. Before commenting here, you should strongly consider starting a new discussion instead. If you think the content of this discussion is still relevant, you can link to it from your new discussion.

This discussion has been closed.