Upcoming Changes in Compatibility Features

Firefox 57 is now on the Nightly channel (along with a shiny new logo!). And while it isn’t disabling legacy add-ons just yet, it will soon. There should be no expectation of legacy add-on support on this or later versions. In preparation for Firefox 57, a number of compatibility changes are being implemented on addons.mozilla.org (AMO) to support this transition.

Upcoming Compatibility Changes

  • All legacy add-ons will have strict compatibility set, with a maximum version of 56.*. This is the end of the line for legacy add-on compatibility. They can still be installed on Nightly with some preference changes, but may break due to other changes happening in Firefox.
  • Related to this, you won’t be able to upload legacy add-ons that have a maximum version set higher than 56.*.
  • It will be easier to find older versions of add-ons when the latest one isn’t compatible. Some developers will be submitting ports to the WebExtensions API that depend on very recent API developments, so they may need to set a minimum version of 56.0 or 57.0. That can make it difficult for users of older versions of Firefox to find a compatible version. To address this, compatibility filters on search will be off by default. Also, we will give more prominence to the All Versions page, where older versions of the add-on are available.
  • Add-ons built with WebExtensions APIs will eventually show up higher on search rankings. This is meant to reduce instances of users installing add-ons that will break within a few weeks.

We will be rolling out these changes in the coming weeks.

Add-on compatibility is one of the most complex AMO features, so it’s possible that some things won’t work exactly right at first. If you run into any compatibility issues, please file them here.

49 responses

  1. Anders wrote on :

    I am currently on version 56. If i look in the Add-ons Manager there is 1 out of 12 enabled add-ons that is *not* marked Legacy (1 out out 21 if I include disabled add-ons).
    – Will all Legacy add-ons be disabled in 57?
    – Is it possible that some of them have a version ready for 57 and above, and if so is there an easy way to see which? Are you going to add this information to the Add-ons Manager?
    – Are you going to add a “find alternative” feature to the add-ons manager, such that it will be easy to get suggestions for alternatives for add-ons that is not going to get updated?

    1. Jorge Villalobos wrote on :

      > – Will all Legacy add-ons be disabled in 57?

      Yes. There will be a preference to turn them back on, but it will only work on Nightly.

      > – Is it possible that some of them have a version ready for 57 and above, and if so is there an easy way to see which?

      If they have a listing page on AMO, you could look now, though it’s likely that versions that only work on 57 and above won’t be submitted at least for a few weeks. There’s a green “Compatible with Firefox 57+” tag for all add-ons that use the new APIs.

      > Are you going to add this information to the Add-ons Manager?

      No, the Add-ons Manager won’t show this.

      > – Are you going to add a “find alternative” feature to the add-ons manager, such that it will be easy to get suggestions for alternatives for add-ons that is not going to get updated?

      Yes, it should be activated soon on Nightly. However, since the alternatives are hand-picked, they will be limited to very popular add-ons and a limited number of other cases.

      1. Anders wrote on :

        > There’s a green “Compatible with Firefox 57+” tag for all add-ons that use the new APIs.

        Getting to amo from the add-ons manager doesn’t seem to be easy (I think you have to double click the add-on, click the ratings, and click the add-ons name in the right column). I tried to spot check some of what I think is high-profile add-ons (greasemonkey, stylish, noscript), but didn’t see any with such a label.

        > No, the Add-ons Manager won’t show this.
        To bad.

        > Yes, it should be activated soon on Nightly
        Nice.

        > they will be limited to very popular add-ons and a limited number of other cases.
        To bad.

        Thanks for your response.

        1. Jorge Villalobos wrote on :

          We expect NoScript and GreaseMonkey to be ported, but I don’t know their timeline. Not sure about Stylish (though there’s a Chrome version that could be ported).

          1. Aidan wrote on :

            Stylus https://addons.mozilla.org/firefox/addon/styl-us/ is a possible alternative (and is apparently based on a fork of Chrome Stylish)

          2. John Duncan wrote on :

            The most useful feature of Stylish was styling internal browser chrome:// (specifically things like scrollbars which cannot be styled in the userChrome.css or userContent.css files), which is not possible in any of the Chrome/web-extensions add-ons.

            Will there be a replacement for this functionality? If not, there isn’t really a reason to even have a stylish extension replacement.

          3. Jorge Villalobos wrote on :

            Stylish did many things and it’s useful to different people for different reasons. Stylus is a good replacement for making changes to content, but not chrome. There’s a themeing API coming up that will enable add-ons to change chrome styles. Like the rest of the WebExtensions API, it will start with a small feature set and will grow gradually.

    2. Well… guess what. wrote on :

      That temporary userbase increase will be short lived with the current state of API’s, most won’t even be implemented & most big addon dev’s have already decided to permanently pull the plug on Firefox. A simple thing like making the font size for the sidebar a little bigger ain’t even possible.

      1. Ulf3000 wrote on :

        lets just all stay on nightly until someone forks firefox … looking at teh cource code there aren´t even big changes anyways , at this point its just politics

        1. Jorge Villalobos wrote on :

          Nightly is the active development version, so that’s where the changes will happen in the next few weeks.

      2. Andreas wrote on :

        > most big addon dev’s have already decided to permanently pull the plug on Firefox.

        I’d like to see the data you based this statement on.

        In fact https://docs.google.com/spreadsheets/d/1gA_Nj8SJ6ykhrqJNJqIoge2H-_CGHQc8uK5K3LYr5As/ shows that most developers of popular add-ons have already ported, are currently porting or are at least likely to port based on feedback we got from the developer.

        1. Sören Hentzschel wrote on :

          @Andreas: you can update the “WebExtension Status” of New Tab Override in this spreadsheed to “Done”. Waiting for review, but it’s done. 🙂

        2. mgol wrote on :

          1Password is already a WebExtension, that version is shipped at least to Firefox 55+. You can update the spreadsheet.

        3. qlum wrote on :

          That list seems somewhat incomplete, just going by the top 40 most popular addons I Already see two big ones that are know to not get ported missing, DownThemAll and Classic theme restorer, looking a bit further I found the following missing ones, here is what I could quickly find out about them:

          imacros – considering its also for chrome it may very well get a webextension
          DownThemAll – Not gonna happen
          Random Agent Spoofer – possibly a seperate webextension version sometime in the future
          Tab Mix Plus – in development
          Yandex.Bar – don’t know, I doubt a lot of people will care considering it has a 2 star rating.
          YouTube High Definition – don’t know, if not there are enough addons that basically do the same thing
          Element Hiding Helper for Adblock Plus – don’t know, depends on adblock plus anyway
          Flagfox – should be possible, developer said in february that he would look into it.
          DuckDuckGo Plus – don’t know
          Classic theme restorer
          Self-Destructing Cookies – already incompatible
          fireftp – don’t know
          Ant Video Downloader – don’t know, got updated recently yet changed the compatibility to 36 as a max version

          Either way the list does not tell the whole story as its quite a few of the top 40 user wise that is not in the spreadsheet at all. That being said I see quite a few addons do get updates just not all of them.

    3. Graham Perrin wrote on :

      > … add a “find alternative” feature to the add-ons manager …

      Please join the discussion at https://discourse.mozilla.org/t/-/17087/15?u=grahamperrin

      1. Anders wrote on :

        You have got to be joking. If I read the thread correctly, there is a link to a spreadsheet, that lists 180 webextensions and points out that some of them can replace 16 add-ons, and a suggestion that people post recommendations in the thread so someone can presumably add it to the sheet (double the work) “because programming for the recommendations will be done manually”. Currently there is about 9904 add-ons on amo. It is not a huge number, but it does take a serious and coordinated effort if you are going to make it.

        1. Jorge Villalobos wrote on :

          We won’t offer recommendations for all legacy add-ons, of course.

          1. Anders wrote on :

            > … of course
            Well, it seems like the very least you can do,
            Better would be if you made a serious (as in “launch of FF3”-serious) push to get legacy add-ons converted by authors, staff and, through hackathons and prices, get volunteers to help out on abandoned and would-be-abandoned add-ons. A nice way to start would perhaps be to host all add-ons on amo in its git-repo and encourage contributions (also directly from in-product devtools), not just reviews and comments.
            But it seems like you (as a company) just don’t give a fuck, of course.

          2. Jorge Villalobos wrote on :

            You don’t seem to understand how software works, how open source works, or how much work we’ve put into this for the past 2 years.

            Sorry.

  2. Lukasz Polowczyk wrote on :

    The only rescue is in WebExtensions Experiments. https://webextensions-experiments.readthedocs.io/en/latest/index.html

    This gives you access to the full XPCOM, I hope this access will stay ALWAYS.
    I just do not know how it will be with distributing Experiments? Will they be hosted on AMO?

    1. Jorge Villalobos wrote on :

      Experiments aren’t meant to be distributed to a wide audience. They will not be hosted on AMO.

      1. Lukasz Polowczyk wrote on :

        I will keep them “in the underground”.
        Sad times have come. ;(

  3. Idiot wrote on :

    Will it be possible to enable legacy addons in Nightly 58+?
    I guess all legacy addons will be broken by the time Nightly 58 is released?

    1. Jorge Villalobos wrote on :

      Yes, and quite possibly yes.

  4. rh2unx wrote on :

    Thank you, Mozilla. The recent Firefox with the broken and disabled addons, the browser I can’t customize, the one with unpolished and caustic UI, is the browser I always dreamed about. I want to thank you for that I can’t use my favorite Ubiquity anymore, for that some other addons are broken permanently and no one will fix them because this is impossible, so, I’ll stay on FF54 forever. Thank you, Mozilla for your incessant care about us, users.

  5. Aris wrote on :

    @Jorge

    It is possible to achieve many visual changes to main ui by adding custom css to userChrome.css. WEs do not allow any of them.

    http://i.imgur.com/ejPeWU3.png

    Will userChrome.css survive Firefox 57+ (beta/release)? Or will its support be dropped too?

    1. Jorge Villalobos wrote on :

      There are no plans to drop userChrome.css that I’m aware of.

    2. custom.firefox.lady wrote on :

      Apparently there are people who would like to see it gone:
      https://bugzilla.mozilla.org/show_bug.cgi?id=332529#c14

  6. Tomo wrote on :

    Mozilla, welcome to the Kodak Club.

    People don’t choose a browser, they’re pushed away from their old one.
    Once people switch browsers, they are very unlikely to try anything else at all until their new browser does something they don’t like – and can’t easily restore.

    This is that something.

    Today I found out that all of my extensions will stop working in three months time, and it seems that none of them can ever have viable replacements. Mozilla has even explicitly rejected any hope of future compatibility.

    Mozilla, you are on the verge of corporate suicide.
    I am neither a shareholder, employee or contributor – but I urge you to reconsider this course.

    I for one will be sad to see you go. It’s been a good decade.

    1. Weboh wrote on :

      Pale Moon, a Firefox fork, still works with most “legacy” Firefox extensions. It has a better UI and just about everything else too.

      1. John Duncan wrote on :

        Pale Moon has no future. It has already fallen hopelessly behind in terms of HTML5 features, security patches, etc.

        SeaMonkey is what PaleMoon should have been, but unfortunately, SeaMonkey will also die if Mozilla continues to erode the features that users care about most (complete control of browser internals, complete themes, XUL/XPCOM, low lag/latency UI, etc).

    2. SKDJ wrote on :

      I totally agree. I am very lost with these corporate, yet still so called ‘open source’ decisions. The Firefox success is finally killing itself. The intent of Firefox was to serve US, the people. And so we (the ones who knew how to use this magnificent browser & it’s addons to upgrade our productivity) made it a corporate success.

      All that ends here. In the Netherlands we call this: Eenheidsworst. Wurst! Yes, you got that right.

      I must say I do not care at all how much work you’ve put into this for the past two years and I know way much more about software development then most developers do. And I am not even a developer. I am the guy that needs to explain to users why you guys do what you do. And very often that is explaining how you help yourself and the support without thinking of them at all!
      That is why most ‘solutions’ come from psychology. Not developers or system engineers.

      I have understood there still is a version 52.3ESR that will extend the misery of needing to find a new browser. It will probably take years of stress within the Firefox organization to realize all the good things that made it great working for this browser (it’s users, red) actually died in 2017. Not long from now you guys will be zombies with their brains being eaten by ‘the industry’.

      The biggest problem is, that those who use the most and the most ‘niche’ ‘legacy’ (not in my experience but ok), are the ones that made our special browser a success. All the popular addons can easily be created for other browsers too. There is nothing that will distinguish YOUR product from anyone else’s.

      So the last time I will close my usually 125+ tabs in seven groups since around 2010 (tabsgroups is my memory but firefox could care less about it’s developer), my heart dies a little. For taking your browser into every company I’ve worked since 2004, using FEBE to be set in minutes with all other addons that I’ve gathered from true open source developers that never left me standing in the dark.

      I hope though at my age it may come too soon, I will enjoy as much productivity with the Pale Moon. But I will leave this coporate product and will never tell anyone again that Firefox beats Chrome. Or Edge. Because it doesn’t.

  7. elena wrote on :

    I returned from holidays and linux suggested a gazillion software updates. Among them, FF. I authorized them and BAM! now most of my addons don’t work anymore. The address bar simply disappeared, for instance…
    The extensions’ developers say they won’t rewrite them. Some even say that going back to a older FF won’t fix it unless one has a backup of the profile, because the new FF permanently screws the previous profile, addon-wise.
    Congratulations for the genial idea you had. Really, congratulations. You erased the one advantage you had over all other browser, the possibility to do everything you wanted with the browser, and forever. You threw away years of programming and user experience from thousands of people just like that. Don’t expect it to be an harmless move. I, for one, will start testing other browsers immediately. Netscape really didn’t teach you anything, eh?

    1. Weboh wrote on :

      Pale Moon is a good one to start on. It’s a Firefox fork that will work with most of your extensions (and doesn’t have Australis!).

      1. SH wrote on :

        > Pale Moon is a good one to start on. It’s a Firefox fork that will work with most of your extensions

        Nonsense. Pale Moon does neither support WebExtensions or SDK based add-ons. This means thousands of add-on was never supported by Pale Moon. Pale Moon is no alternative and Pale Moon is a browser with a totally outdated code base…

        1. JustOff wrote on :

          Only WebExtensions-related part of this message is true.

  8. Sean wrote on :

    Anyone know a good web browser comparison site? I’m leaving firefox for good. It’s a relic.

  9. Why This? wrote on :

    I looked down to see that my menubar is missing. Firefox has broken it. I’m not sure why, I’m just a user. Mozilla and Firefox seemed to have nefariously schemed to break the 24 addons that I’ve been using to make a better browsing experience, to control the web I see so I’m not looking at white text on black background, not being javascripted into advertisers’ joy, able to put the menu buttons I want where I want them. There are a lot more, but they are all marked with big yellow LEGACY buttons.

    I understand that it’s not really a security issue, but instead is widely known to be a result of the politics of open source software, where developers are like the pizza guy on the Simpsons, gnarly, supercilious basement denizens who act out of some unknowable reasoning to wreck convenience and continuity for plain users.

    How does that make users feel? Who at Mozilla/FF cares? There’s alway some other mythical generation of idiots who trust that the apparently good systems they grow to use will continue to be usable, instead of a false message about how it’s now better than before and in a few months will be even better, even though the good parts have already been broken. Really, why break the extension now instead of sticking to your words and breaking them in November? F***

    1. Why This? wrote on :

      I did a search on 57+ in addons.mozilla.org — I have yet to find an addon that claims to be 57+. What a bad joke this is.

      1. Jorge Villalobos wrote on :

        Here’s a list of all of them so far: https://addons.mozilla.org/firefox/tag/firefox57

  10. Why This? wrote on :

    Wait, here is a useful addon (really?)
    https://addons.mozilla.org/en-US/firefox/addon/tab-counter-webext/?src=search

  11. Assaf wrote on :

    Will it be possible to enable legacy addons in Firefox ESR 59 ?

    1. Jorge Villalobos wrote on :

      There might be a pref that lets you install them, but they probably won’t work at all.

      1. Assaf wrote on :

        Currently, I can make it work on Firefox 57 nightly channel, using the following prefs:
        extensions.legacy.enabled
        extensions.allow-non-mpc-extensions
        Will these prefs will still be avialable in the next ESR?

        1. Jorge Villalobos wrote on :

          I think so, but like I said, legacy add-ons will probably be significantly broken by then. They’re already breaking in 57 in very big ways.

  12. qlum wrote on :

    I for one completely understand you guys moving away from the old addon model and most of the addons will be ported. On my current firefox install the only important addon that will not be ported is Classic Theme Restorer. I only really care about 2 of its features namely tab on bottom and hiding the Hamburger menu. the first I will in all likelyhood be able to get via a different addon and the second may be lost to me.

    I just hope that firefox does not break my prefered layout in the future. I think firefox should generally be a lot more careful now with limiting build in ui customization like it did with australis. I would love to see a increase in build in customization options be it locked behind a about:config setting.

    I for one use kind of a odd layout namely:
    top: menu bar + bookmark icons + addons + window controls
    bellow that the nav bar without search on full with
    bellow that the tabs.

    Just one main thing I disagree with is the speed at which the change happens. For some api’s you are really not giving the addon devs a lot of time to port and firefox 57 also happens to be the lts version waiting for one more version would make things a lot easier especially as it would allow people to stay on 57 and move to lts so they can use their addons for the full lts period of 57. Instead now if people really want to stick to their old addons they have to revert to an old lts version which will break things or stay on an unsecure unsupported version.

    In my opinion firefox 58 would have been a much better version to introduce the change it being the version right after a lts release.

    1. qlum wrote on :

      edit: for some strange reason I was thinking 57 was an esr release, it is not. 59 is so my whole argument about timing does not make a lot of sense. There is however the problem of downgrades breaking with 55 so people relying on non webextension addons will be forced to scrap their profile directories which is a shame. But I guess the resources needed to support one extra esr release would complicate things as well.

  13. Robert wrote on :

    What a joke, to disable extensions that some really like and think its ok to just say try a different extension. I have been with Fox since the beginning I think its time to bail

  14. mark wrote on :

    Mozilla is working very hard to help Google. How else to explain this CONSTANT NON-STOP CHANGE because the mozilla devs think they are smarter than the rest of the world?

    Come on guys … all these vast plunders in the last year, you are doing this on purpose.