The Road to Firefox 57 – Compatibility Milestones

1. Vincent Bentley wrote on February 23, 2017 at 12:05 pm:

   I have a suggestion. When Firefox 56 updates to 57 test to see if any extensions are installed. If extensions are present, update to Firefox 56.17.11 . If none present update to Firefox 57.

   Maintain the 56 line for 24-months on a security hot-fix basis only, versioning something like 56.17.11 to 56.19.11 so users know how out of date they are at a glance of the About box. Rename the browser ‘Firefox56’, ‘FirefoxXUL’, ‘Firefox Legacy’, ‘Emberfox’ or whatever seems most appropriate. Leave the Firefox name to release 57 and onwards.

   Provide a method in Firefox 56 to update to 57 at any time but put warnings in the process to alert users about which XUL extensions they are using that do not have WE replacements before updating.

   If users/developers need more support beyond 24-months, seek crowd funding in May 2019 for an additional six months for 24 to 30. If you don’t get the funding, Emberfox becomes Ashfox on month 25.

   This is a fair method of giving Mozilla up to a year to get the new API finished! And after the new API is finished, extension developers get the remaining year to port to WE.

   The key thing is that Mozilla is unlikely to alienate users or developers by freezing further development of 56 while putting its resources into 57. Yes it sucks having old code to support but does Mozilla really want to destroy a large chunk of its fan base?

   1. Jorge Villalobos wrote on February 23, 2017 at 12:40 pm:

      The ESR branch, which works sort of how you described, will support legacy add-ons for a few more months. The timeline you describe, however, is not feasible.

2. Shin wrote on February 27, 2017 at 2:44 am:

   Let me give you the word of self-responsibility.
   Do not deprive us of freedom with an overprotective idea.
   Please truncate past assets early and do not involve us in a self-suicide suicide.

3. Alan Robertson wrote on February 27, 2017 at 8:03 am:

   I’m quite concerned with the role of Firefox, it’s relevancy, and the direction that Mozilla is taking – particularly with Addons.

   There are some people that argue that the majority of users do not use Addons, and this is only true until they are shown an alternative way of using the internet – a free internet, an internet not subverted by large advertising companies, an internet that is not a surveillance state, an internet that allows people to innovate, inspire and communicate freely without fear of repercussion. That is why I use Firefox, and in particular, the many privacy and freedom enhancing Addons.

   When you show other people the difference that Firefox can make when combined with additional privacy and security Addons , then they will always choose to use them. The only reason they are not using them is because they are not aware of them or do not know how to implement them. It is not because they do not care about Addons nor is because they choose not to implement them.

   I personally feel it is my right to be able to decide what I should and should not allow websites to gather information on my personal details. Today’s internet has lost it’s way with more and more people content to have their personal freedoms and liberties sacrificed all in the name of security, government surveillance, spying and hacking…. The Mozilla I knew would never allow this to happen.

   Don’t get me wrong – I understand the reasons why: It’s easier to adopt the Extension based way of implementing additional features. It’s also easier to base your whole underlying browser on Chromium – like Opera did. It’s also easier to just stop believing in your core values and watch your market share slide too, so let me ask you this:

   “If Opera is now based on Chromium and Firefox is now using Chromium Extensions, then why don’t I just use Chromium?”

   With this in mind I would like to remind you of one of your core Mission Objectives:

   Mozilla Mission Objective: January 23rd, 1998:

   “And we will, above all, be flexible and responsive. We realize that if we are not perceived as providing a useful service, we will become irrelevant, and someone else will take our place.”

   A lot has certainly changed in the last nineteen years but your core mission objective hasn’t changed at all. With this in mind Mozilla, today, has never been in more demand than ever before, and yet you are losing market share. You are the only internet browser today that respects personal freedoms and privacy, however, a lot of this is only possible with Addons, and yet you appear to be setting down the path of removing these.

   If you remove the ability to have those Addons then you are in effect not offering anything better than what Chromium is already offering. Considering that Chromium / Chrome is monopolising the market then the majority of users will switch and Mozilla will become irrelevant and cease to be.

   Now, while there are a lot of Addons which are also present as an Extension equivalent in Chromium, there are a lot of Addons which are not available as an equivalent Extension: No Script is a perfect example. Yes, I’m aware there is a No Script Lite, but it’s not the same and doesn’t have the same functionality. I want No Script – it works and it prevents spying, malware and it blocks a plethora of obnoxious advertising.

   As for asking the developers to update their Addons to make them into Extensions – I doubt you will find many of them willing to do so. In many cases the additional features that the Addons offer are just not able to be implemented as an Extension or (in most cases) will have to be re-written from the ground up. I don’t think you will find many developers willing to do this with a browser that is losing market share.

   So, you’re between a rock and a hard place – users like myself who believe in personal freedoms and privacy wish to use the Addons feature. Developers don’t want to spend time re-writing their Addons as Extensions. The net result? Those Addons die, some users switch to Chromium, Firefox loses more market share, then more users switch to Chromium with forum postings along the lines of “What Happened to Firefox?” / “Why I Switched From Firefox to Chrome” / “Firefox Market Share Tumbles” etc.

   I don’t want that and you certainly don’t want that. Even people who are using Chrome don’t want that – let’s face it, without Mozilla offering any real alternative browsing experience to Google’s Chrome, could you imagine how much worse Google will become with invasive privacy practices? The thought sends shudders down my spine….

   So why don’t you implement the Addons which won’t get updated when you switch to Extensions youselves? Bake “No Script” into your browser. Take a page out of Opera’s book and offer Adblocking (or malware distribution prevention as I like to call it) within the browser. Give users the choice when they install Firefox to have HTTPS Everywhere, Disconnect and Canvas Finger Print blocking. Get rid of all the weak ciphers like RC4 and any that are vulnerable to Logjam, Freak and Poodle. Implement TLS 1.3. Ask whether users want to reveal their GEO location. Allow DOM storage to be turned off – why do websites assume the right to use my computer for storage without asking? Block third party cookies. Why are all my searches going to Amazon or Twitter? Do I really need to send websites referring agents, or the status of my battery, or use my computer in a peer to peer network using my bandwidth, or allow tracking, or cross site scripting, or uniquely identifying me – isn’t my IP address enough? Since when did everyone expect email to have no privacy, but a written letter sent through the post is expected to be private.

   This is all wrong and it has to stop!

   There’s no reason why you cannot implement these features into Firefox and eliminate the need for Addons in the first place. If Firefox did this straight from the install then I wouldn’t need all these extra Addons. In turn, it wouldn’t bother me that the Addons will stop working and then I’m actually looking forward to the future with Mozilla safely steering the way.

   The solution can only come from Mozilla: Your biggest beneficial selling point is “Mozilla isn’t Google”.

   “Mozilla: Free, Secure, Private”.

   So will Mozilla implement these features directly into the browser? Or will it be a half hearted attempt which ultimately will leave the browser still vulnerable to privacy invasive techniques?

   It’s your call at the end of the day but right now all I see is irrelevancy.

4. Rob wrote on February 28, 2017 at 9:32 am:

   Hello Jorge, please give us a few simple answers:

   Are there any WebExtensions-based addons available right now? Where do we find them and how do we identify them from the legacy addons?

   Are any of those new addons ports of the most used 5-10% legacy addons on amo? Do they cover most of the functionality or at least is someone at mozilla working on the needed apis for them to be feature complete when 57 gets released?

   What is your target number/percentage for those ports before switching to WebExtensions? Will you enforce that by delaying 57 as necessary?

   What is your intended level of completeness and usage for the WebExtensions apis when 57 gets released? Will you enforce that by delaying 57 as necessary?

   1. Jorge Villalobos wrote on February 28, 2017 at 11:20 am:

      > Are there any WebExtensions-based addons available right now? Where do we find them and how do we identify them from the legacy addons?

      Yes, there are close to 2,000 WebExtensions listed on AMO, with many more that aren’t listed. We don’t expose if a version is a WebExtension or not, though many developers note it in the Version Notes. The hard way to figure it out is to download the file, unpack it, and look for a `manifest.json` file. Legacy add-ons use `install.rdf` instead.

      > Are any of those new addons ports of the most used 5-10% legacy addons on amo?

      Some big extensions have already been ported to WebExtensions, notably: Ghostery, Awesome Screenshot Plus, AdBlocker for YouTube, Evernote Web Clipper, and ColorZilla.

      > Do they cover most of the functionality or at least is someone at mozilla working on the needed apis for them to be feature complete when 57 gets released?

      API prioritization has been largely based on APIs that are popular on Firefox and Chrome add-ons, as well as developer requests. This blog post explains how to contribute to the process.

      > What is your target number/percentage for those ports before switching to WebExtensions? Will you enforce that by delaying 57 as necessary?

      The Firefox 57 deadline is not dependent on how many add-ons will be ported. We are tracking add-on migration closely, but we have no reason to believe it will be a blocker.

      > What is your intended level of completeness and usage for the WebExtensions apis when 57 gets released? Will you enforce that by delaying 57 as necessary?

      The WebExtensions API will never be “complete” and there isn’t a completeness target. See my earlier reply on prioritization regarding what will or won’t make it by 57. Similarly to migration, this appears to be on track and not a blocker.

5. Easton wrote on March 1, 2017 at 3:37 am:

   Since we can continue update existing SDK add-ons until Firefox 57.
   Would you please share if it is valid to release both WebExtension and SDK with same extension ID?
   Or we must use a new ID for WebExtension?

   1. Jorge Villalobos wrote on March 1, 2017 at 6:50 am:

      For listed versions, once you switch to WebExtensions, you can’t go back to legacy. You should definitely use the same ID when moving to WebExtensions, to ensure users are updated automatically and don’t end up stranded using a legacy version that will eventually break.

      1. Saga wrote on March 6, 2017 at 10:46 pm:

         My extension belongs XUL overlay extensions and use update.rdf to check new version add-on. If my extension switch to WebExtensions, how can i make old users to update new version automatically like before?

         1. Jorge Villalobos wrote on March 7, 2017 at 12:23 pm:

            You can continue updating users with update.rdf. The only difference is that in WebExtensions you use a different manifest key. Or you can also switch to the new update format, explained here: https://developer.mozilla.org/en-US/Add-ons/Updates

6. Nikaas wrote on March 1, 2017 at 4:31 pm:

   Sadly the extensive modding capabilities were the only thing for me that made firefox still worth using. Probably now is the time to finally get used to chrome’s ui.

7. Tony Li wrote on March 1, 2017 at 10:31 pm:

   – Please show me in FF (preferably without additional addons) which addons currently installed may become troublesome so I can start looking for alternatives/rewrite/etc.
   – Please present to the user warnings about extensions that will be disabled by updating BEFORE choosing to apply the update.
   – Please display in AMO whether or not an addon is WE or not, because as a user I do not want to invest time learning to use an addon when it will be disabled/prevent me from updating FF in a few months.

8. Dmitriy Kropivnitskiy wrote on March 2, 2017 at 3:02 pm:

   I want to add my voice to many other voices all expressing the same idea. XUL and SDK add-ons are currently one of the very few remaining reasons me and many other users stay with Firefox.

   I switched to Mozilla when it first forked off of the Netscape Navigator source code in 1998. I switched to Firefox 0.9 when it became available and was a Firefox use ever since. In the past few years, Chrome has been getting more and more popular and most people I work and socialize with switched to it. I have stayed with Firefox. The reason I stayed are the add-ons. I am writing this in my Firefox window where I have a multi-row tab bar, last used tab switching and multiple session management via Tab Mix Plus add-on and several separate tab groups via Tab Groups add-on. Neither of these functions is available in Chrome and not for the lack of trying. They are impossible to implement via the Chrome Extensions. There are open tickets with Google for all of these features. The multi-row tab bar one has been open since 2008 with no reaction whatsoever (https://bugs.chromium.org/p/chromium/issues/detail?id=3334). Same for the tab switching order (https://bugs.chromium.org/p/chromium/issues/detail?id=5569). Once Firefox 57 disables all these features, I will not gave any other reason to stay. Chrome rendering is faster, their development tools are better, a lot of sites tend to optimize for Chrome first. Goodbye Mozilla Firefox. We have been together for two decades and now you are abandoning me. It makes me sad, but I guess we need to do what we have to.

   1. Stefan wrote on March 7, 2017 at 9:22 am:

      You’re implying that these things will not be possible anymore, which need not be true. Push the developers of your addons to request the extension APIs that they need.

      1. Andrew Gillett wrote on March 8, 2017 at 9:57 am:

         The author of Tab Groups spent a year and a half rewriting the extension to be e10s compatible, only to find that this work was largely wasted. He is done writing extensions for Firefox. http://fasezero.com/lastnotice.html

         1. Jigar wrote on March 9, 2017 at 1:21 am:

            Yeh..This is sad. I loved that one. But New Test Pilot Containers look promising.

            Mozilla should really host set of hackathon. For at least top 100 addons. Addons were life for Firefox…Some of them will never make it to WebExtension due to heavy rewrite required.

            Mozilla should make some alternativeTo kind of project to help user transition to newer addons or alternative addons

9. josh wrote on March 10, 2017 at 5:40 pm:

   Is noscript able to be a web extension and still do everything it currently does?

   1. Grzesiek wrote on March 13, 2017 at 2:06 am:

      There is a work in progress to make it a WebExtension. Both on the addon side (porting) and on the Firefox side (creating necessary APIs).

      1. That Random Guy wrote on March 17, 2017 at 8:16 pm:

         I swear, if they don’t bring that guy over along with certain ad-blockers, I’m cutting off all ties and switching to PM.

10. Dark_Ronius wrote on March 12, 2017 at 11:02 am:

    What boggles my mind is that Mozilla’s response isn’t “We won’t enforce WebExtensions until they are feature equivalent to XUL/SDK Extensions”. But then I suppose that would beg the question, what is the point in changing?

    This whole so-called “consultation process” also boggles my mind. Accepting that there is usually more negative than positive reviews on anything in the world (people feel less inclined to moan about how good something is), there still seems to be an astonishing amount of denial about the amount of negative feedback these changes are creating. The fact add-on developers I almost literally grew up with are either leaving or considering to leave the development scene saddens me very much.

    Even Firefox evangelists seem to be in denial, to the point that when I suggested I’d actually rather use Edge as my main browser if old-style add-ons are removed, they replied Edge wouldn’t be as secure… in that self-righteous tone. This isn’t the 00s, when this could justifiably be the default reason for switching to Firefox. There is very little, from a security standpoint, to separate any of the browsers nowadays (with the possible exception of IE, but even then that is in a far better state than it ever was). Mozilla can certainly congratulate themselves for contributing to that, however for some people to actually fall back on the stereotypical security argument by being a Microsoft product is laughable and almost childish. Arguments ended up bogged down in technicality, almost being summed up as “What features do you even really need that won’t be available anymore?” and “There won’t be a better product than Firefox/like the other companies would be any better.”

    ARE THESE REALLY ARGUMENTS FIREFOX-EVANGELISTS SHOULD BE PROUD OF?!

    Without old style add-ons being able to reshape my experience into almost anything I want, there is little to differentiate anymore between any of the browsers. As has been mentioned time and time again, and as no one defensive of Mozilla seems to acknowledge. When WebExtensions are enforced, I will definitely be switching to Edge purely because there is no point to deviate from the default browser experience anymore (whilst potentially dabbling with Pale Moon and seeing where that goes).

    I’ve been a Mozilla obsessive since 2003. Every new feature or technology I’ve leaped on.

    When XULRunner became a thing, I switched to apps like Songbird, Instantbird, which genuinely improved my PC experience at the time. The prospect of being able to migrate from developing an add-on to migrating that to an application in itself suddenly became feasible to me. As an intermediate PC user, I never even thought I’d actually be able to write my own application. I know it’s not as simple as javascript, but all the tools were there. I was hoping for a kind of inversion of the Mozilla/Seamonkey experience, where I could have separate apps all united by the same underlying “engine” in the background. Or rather, a far more modular experience to Seamonkey, but where their resources were pooled despite being in separate interfaces. However, the enthusiasm seemed to disappear. Eventually all these other applications crumbled, and XULRunner itself became deprecated.

    When the Firefox Marketplace became a thing across platforms, I was again impressed by the potential. By modifying it to use my original Firefox profile, it meant I could use apps like uBlock to act as a privacy layer; I could have a nearly equivalent experience on my phone to the native Facebook app, but with more control over my experience. I was hoping for a Firefox home screen to tie it all in… Instead we got a developer preview of Firefox OS for Android. Which I still saw as a good stepping stone to this privacy layer idea, if a bit slow. Eventually webapps were also pulled.

    I’ve had a similar experience with Prism, Tab Groups, anything created under Mozilla Labs… There are so many things that I have felt could change the way we use technology if they were marketed right and supported/maintained for longer. Tools promoted enthusiastically by Mozilla, tools I quickly snapped up and used even where I didn’t feel I needed them… then subsequently dropped as I grew to love them. I understand Mozilla is getting harder to fund; since Chrome started becoming the top browser, Mozilla has had a limited time to use the money they had before Google renewed their search contract, inevitably smaller with Firefox’s shrinking market share. I feel like lots of plates started spinning whilst they looked for a niche.

    I’ve reluctantly accepted the ending of all of these large projects/features, and still come back to Firefox because of what it does best: at its core, Firefox is the browser you can make into everything you want. I know one of the goals of Mozilla is to be able to reshape the web, but I think the point being missed is Firefox was the browser you could completely reshape in itself, to easily turn it into whatever application you wanted beyond the confines of what can be done on a web page. This inevitably means it is also a browser you can break, through add-ons or your own coding. But this is what I loved about it. Firefox versions 3-4 were the glory years, where we finally didn’t have to accept Firefox as a slow browser anymore, where it could still be stable despite the million add ons we had installed. That was all that needed to change. Removing this concept of a reshapable browser, more restricted to the confines of this new API, is the last straw for me. You can sweet-talk through it all you want. Claim you’re listening to developers to file bug requests to add features to the API. Claim the onus is on me for not understanding the reason for changes, that as a user I don’t really need those features. Claim you’re implementing the features “most users want”.

    But, when XUL and SDK add ons are permenantly disabled, I will stop using any official version of Firefox. Because I will no longer be able to use the features and add-ons that brought me back to Firefox every time. Maybe, ironically, Seamonkey will see a resurgence by continuing to implement them?

    I expect this will go unanswered. Which again concerns me.

    1. Jorge Villalobos wrote on March 13, 2017 at 8:39 am:

       I’m not sure what kind of answer you expect besides what’s already in blog posts and other comment replies here (which are numerous and I’m trying not to repeat myself too much). You also said yourself you’re not looking to be convinced to stick with Firefox…

       I’m not sure which evangelists you’re referring to – maybe our community people -, but calling Edge unsafe seems unfair to me. At least from my perspective its main disadvantage is that it’s still behind on web standards support, which may or may not be relevant to you.

       1. Dark_Ronius wrote on March 13, 2017 at 11:31 am:

          Firstly, I appreciate the reply. None of what I say is meant as a generalisation of all Mozilla developers and volunteers, and equally I don’t mean anything personally- I know the vast majority of the team work very hard. I was paraphrasing a conversation I had on Reddit, but to be fair 1) it wasn’t a fair way to make an argument about the state of users/volunteers 2) it was a conversation on Reddit, the “nicest place on the web” ha.

          I suppose this is me trying to put my vote on how I wish things to turn out, against the current direction… I have no idea where else I could place these opinions that might make a difference or add to the final decision making process. I don’t feel I have the right to just turn up on one of the mailing lists and post that comment. I know it most likely will have the impact of peeing in the wind, but the whole situation involving organisation I loved frustrated me… indeed, an organisation I felt (indirectly) part of, that I had donated to in the past. Mozilla is an organisation I have an emotional investment in, it’s one I want to succeed in the face of Google or Microsoft. If Google pulled anything like this, I’d just leave a short angry e-mail and move on.

          It concerns me that the conversation on these decisions doesn’t *feel* more open… I know technically everything is viewable on the Mozilla forums and/or the mailing list, technically it’s a conversation I could have taken part in at any time. That half the reason is there are not enough people taking part as less and less people can be paid to work on it. But it all feels opaque, or cliquey, when you join and read through these mailing lists, attempt to take part. I understand this is how the whole open source software movement works. It just *feels* like decisions I, as a user and hobbyist developer, can’t have any say on. That they have already been made, with sanitised and whitewashed press releases like any other company would make, as if these issues are being ignored (again, I don’t blame you for that, you’re not about to write on the official blog and say it’s all a load of balls yourself!). I resent that I can only resort to voting with my feet, so to speak, which I say out of love for the organisation as I wouldn’t resent a for-profit in the same way. This is why, again, my comment is becoming so long!

          And I slate “Firefox Evangelists” knowing full well I was one myself… After my inevitable honeymoon (or is that one-night stand?) with Chrome when it launched, and generally enjoyed it, I’d always come back to Firefox. Even when I knew some features were missing, it was the potential there, for add ons, improvements, how versatile it felt that a number of applications were “powered by Mozilla”! I can no longer see that potential.

          Even with driver signing, at least there was the renamed Developer Edition where it could be disabled, which honestly seemed the perfect way to filter the computer hobbyists who may be more willing/able to take risks to security from the everyday users. I would be happy with a situation more like that, where users can be advised “complete themes” and the like are only compatible with this edition. Maybe with appropriate security warnings. I wouldn’t care then if old-style add ons were disabled for the regular Firefox. It might increase the workload for a while having two builds (but no different from at the moment?), but also might even aid that bridge building the equivalent features into the main edition, until that equillibrium is reached of the “right” amount of WebExtension APIs implemented? It would keep those add-on developers on board considering leaving, and ironically reduce some of the workload from their input…?

          1. Dark_Ronius wrote on March 13, 2017 at 11:34 am:

             ^I meant Add-on signing!! No idea why I said Driver Signing!!

          2. Jorge Villalobos wrote on March 13, 2017 at 1:43 pm:

             After 57 it will still be possible to run low-level code in Developer Edition and Nightly, via WebExtensions Experiments. It’s not quite an off switch like we have with signing, but it’s one way to run something like legacy add-ons. However, you should be aware that most legacy add-ons will break after 57, regardless. There are big changes coming to Firefox (some of which we aren’t talking that much about yet) that will break legacy add-ons anyway. 57 is not an arbitrary goal, but a fairly hard deadline.

             It’s true that communicating our plans and goals with the community and getting everyone involved has been a problem. Like you said, it’s part of the nature of open source. However, it’s also because the high-level plans for Firefox can’t be democratized effectively, and it comes down to the leadership to decide on a course and make sure it’s followed through. Another side of this is how news spread. We’ve been talking about this plan for quite a while. But, naturally, most people will find about it as deadlines approach and things get real. Also naturally, people were skeptical we would stick to the original plan. This blog post is meant to confirm this plan and reinforce the fact that we’re following it through.

             We know it will be a painful transition and we will lose valuable members of the community. But we believe it’s the best way forward for add-ons.

11. That Random Guy wrote on March 17, 2017 at 8:11 pm:

    Will the FF devs at the very least make amends or attempt at an approach for particular features that a good deal of us (100k+) see in particular current add-ons?

    I just don’t understand how something as simple as UI can’t be tweaked with natively. What possible security vulnerability is there in not allowing certain buttons or disallowing tabs? I guarantee you there’s none.

    Guy’s, 100k+ people don’t use CTR because they LIKE Australis. I’m not even willing to accept that the poll done for that was unbiased. I bet you more than half of those idiots also used Chrome.

    At this point, I don’t care if FF comes shipped with Australis or not, I just want the devs to consider and provide flexibility for those of us who obviously don’t like the “default” design.

    This isn’t about safety and this sure as hell isn’t about conformity. It’s about choice.

    If I wanted to use Chrome’s ugly ass browser, I would. It’s obviously superior to some extent.

    Firefox’s willingness to allow users to tweak around with the UI is it’s god-send–that and it’s array of well-off extensions.

    That being said, if the devs are really not even going to consider adding in extra functionality for the things they won’t let others create for, then that really shows some lack of judgement.

    In any design process, the people responsible are SUPPOSED to look out for all intended user’s wants and needs. You can’t just stick to one profile and give into all of that one group’s needs.

    Do you know what they call that?

    A BIAS.

    Pick yourselves up and cater to everyone’s needs, not just the majority.

    I will repeat myself: 100k+ people don’t use an extension for will-nilly reasons. They use it because it does something USEFUL. That means it has a PURPOSE, and it does something people WANT.

    IF you’re hell-bent on not porting some native functionality to allow for TINY rudimentary customization, then you’ve really lost out my friends. I can’t justify your future if you decide to sellout for something like speed and extension homogenization.

    It’s pathetic that I even have to post something like this. At the very least, the devs should try to include native functionality for the features they will no longer allow extensions themselves to act on.

    BE CONSIDERATE. HAVE SOME SENSE.

More comments:« Previous 1 2