A grim outlook on the future of browser add-ons
palant.info
palant.info
Relax, everybody. Fenix is a major rewrite. It takes time to reimplement all the WebExtension APIs. Nevertheless, the intent is to eventually add support for all the APIs that make sense on mobile. (See here, for example: https://bugzilla.mozilla.org/show_bug.cgi?id=1632626#c1, or just scroll through the list of bugs here: https://bugzilla.mozilla.org/buglist.cgi?product=GeckoView&c...)
If you go back and read old HN threads from the first public release of Fenix, people were afraid that the lack of extension support was a sign that Mozilla had given up on adblockers. I hope it's clear by now that's not true. (From what I recall of lunchtime conversations, back when those were a thing, the plan for extensions was "well, we're definitely not shipping without uBlock Origin, and we'll see what else we can get working from there".) Similarly, the fact that parts of the WebExtension API are not currently available doesn't mean Mozilla has given up on it. It just means that things take time. (If it were easy, other mobile browsers might support add-ons!)
It takes a lot of resources to support two parallel codebases. At some point you have to rip off the band-aid. The sooner Fennec is cut loose, the more effort we can dedicate to Fenix improvements (like improving WebExtension support). The transition was always going to be a bit bumpy, but things should hopefully get better from here.
It's not that I question the intentions. It's rather that, especially given the execution so far, I have my doubts about what these intentions will lead to in practice.
They are demonstrating they have a knack for it.
I‘m a Firefox user on desktop but I see their poor choices around extensions haven’t ended. They’ve borked the extensions on desktop before as well.
The charter should be: we support the browser for dorks first and always, and work on technical problems for dorks second.
Turns out that I was right saying that any corp supporting diversity in tech is a corp that has no business being in tech.
Stable just feels unfinished in its current state, I'm not sure about Nightly. I prefer my browser to not surprise me with much of anything, so I prefer not to be on the bleeding edge.
'a bit bumpy' doesn't begin to describe it, until this day there are many ways in which the current iteration of FireFox is behind where 56 was.
Anyway, even if I did, it probably wouldn't make any difference.
> Will add-ons not part of the Recommended Extensions program ever be supported on the new Firefox for Android?
> We would like to expand our support to other add-ons on the release channel. We’re currently evaluating how we can enable general add-on support in a secure way without creating compatibility issues.
https://blog.mozilla.org/addons/2020/02/11/faq-for-extension...
So, why relax exactly?! Even the extension that would work now are disabled.
[0] - Basically anything to do with fixing^W modifying the browser UI itself.
It's tough to read this in any way other than cynically. If Mozilla actually cared about extension support, they'd at least provide an advanced option to enable it outside of the curated list in Fenix. That they didn't speaks a lot to how much they really care.
At the risk of going too far off topic, Fenix really is a disaster. The walled garden for extensions is just one of the user hostile anti-features. I immediately uninstalled it and switched back to Fennec from f-droid. To be frank, you do not have a large enough user base that you can afford to alienate power users in this way. Pretty much the only reasons to use FF on mobile are its power user features, which have been mostly ripped out in this update.
That makes me almost incoherently furious. I understand if you rebuild the foundation of a browser for better performance and that breaks all the extensions until the developers can rewrite them for the new model. I get it if one of my favorite extensions is no longer under active development and will never have a version released for the new model. But how fucking dare you gatekeep what I install on my own browser? Freedom from this walled-garden bullshit is the only reason I run Firefox instead of Chrome. Google's twice as good at UI as you. They're faster, look better, and miraculously I never have to swipe five times trying to dismiss a Chrome tab.
I get that Mozilla is drowning, flailing around trying to create a future where they aren't irrelevant. This is just ugly, though.
This is already the case with the sole exception of Firefox, the only truly independent browser not backed by an OS/platform vendor, and it's struggling badly precisely because of that. But hopefully, it will still manage to nominally keep pace with the "living standards" and be ready to step into renewed prominence when the bigger players mess up or become too top-heavy for their own good.
Mind blowing all of them do not resign out of shame and humiliation.
- not allowing for earmarking of donations for specific projects
- breaking user trust repeatedly
- firing a large fraction of employees the evening prior to picking up funding (or because of, hard to see cause and effect separate here), almost ensuring that everybody else who is capable will jump ship or at a minimum work on polishing up their resume.
- moving income derived from one project to support others that can't stand alone
- allowing a vast gap to come into existence between how the outside world sees the company vs how the company sees itself
- chasing lofty, highbrow future goals and vague missions at the expense of squandering market share and user goodwill for existing products
- prioritizing new and shiny (but unproven) stuff over core and mission critical stuff
Do you want me to go on?
First of all, Mozilla donations never get used to fund browser development. Instead the browser is in part funded by "frivolous" projects that actually generate an income. The other major source of income is from Google web search. Firefox's dependence on this isn't ideal and has been critcised by many, hence the search for other revenue streams.
Is there anything controversial in what I wrote? If so I am not aware of it, this has all been laid out in some detail by now - repeatedly - so I don't see why the burden of providing evidence should be on me, unless you think one or more of these items are controversial or contradicting evidence already out there.
> Mozilla donations never get used to fund browser development.
Exactly. But they should be.
> Instead the browser is in part funded by "frivolous" projects that actually generate an income.
I think the burden of proof is on you now.
> The other major source of income is from Google web search.
That is income that should 100% go to FireFox development, and never to anything else.
> Firefox's dependence on this isn't ideal and has been critcised by many, hence the search for other revenue streams.
Donations to the Mozilla foundation are one such stream.
Yes this pales in comparison to the hundreds of millions from web search partnership but then so does donations. However, significantly growing revenue from paid products and services is very possible. Whereas increasing donations by two orders of magnitude is very very unlikely.
> this has all been laid out in some detail by now
Providing a link to that would be useful.
[0]: https://assets.mozilla.net/annualreport/2018/mozilla-fdn-201...
Of course it is possible, but that's called 'Accenture', not 'Mozilla'. Really, this all goes directly back to what I wrote above: Mozilla has an entirely different view of what company it wants to be from what the general public thinks it should be based on previous statements and there is no amount of creative redefinition that will solve that.
If Mozilla wants to be an ordinary software house then that's fine by me but then stop using funds pulled in through FireFox to support the rest of the org and allow FireFox to be completely independent of the foundation.
It is beyond ridiculous that an important resource like FireFox should be milked in order to start a software house that does not seem to be all that dedicated to keeping FF around for the long run.
If FF dies then Mozilla has failed, and the income from the search box alone should be enough to ensure FireFox's continued existence for a long time into the future.
That won't happen if Mozilla keeps raiding the piggy bank to cover for their mistakes, those 'millions of dollars from their profitable projects' are a drop in the proverbial bucket compared to the costs of the whole org. I'll bet you not a single sou of that effectively makes its way to FireFox due to the top heavy org structure.
Quite clearly yes, given it's all either factually incorrect or a subjective opinion.
is it?
"In March 2020 Mozilla and KaiOS Technologies announced a partnership to update KaiOS with a modern version of the Gecko browser engine and more closely aligned testing infrastructure."
Probably a little too much customization to be totally honest.
Edit: I guess this wasn't really related, but you may enjoy nonetheless. I personally found myself enjoying tech again by playing around with Linux/BSD and alternative software. YMMV
All of these things seem to be under direct attack, mostly by businesses, wantrepreneurs, and security considerations brought about by them:
Extensibility.
Manipulating files.
Open standards.
Options dialogs and deep customization.
Information density.
Consensual updates.
Privacy.
Multiple code paths.
Low-level/technical details.
Backwards compatibility.
Secrets and easter eggs.
Platform-native applications.
Ownership of the bits in your possession.
And I've been a linux geek for a while. It is still a bastion of "the old way" of doing things but there are so many people trying to rewrite the rules, pulling the platform in their own direction and writing new code when they should be contributing to existing code. Mastering Linux seems to entail mastering several sub-platforms all at once. If only they'd agree on conventions newer than 30 years old.
BSD looks interesting though.... if only it ran software I actually want to use
As an increasingly old computer nerd, I am yelling at clouds and holding on to what I have for dear life. Too much of my identity is wrapped around this idea that computers and tech are personal brain-amplification devices, and not appliances like everyone else seems to want them to be.
It’s kind of possible already, but the massive profits in per user per month billing alongside feature tiers to maximize price discrimination means we’re stuck with the status quo for a while.
> As an increasingly old computer nerd, I am holding on to what I have for dear life. Too much of my identity is wrapped around this idea that computers and tech are brain-amplification devices, and not appliances like everyone else seems to want them to be.
I feel this. I hate seeing everyone become more and more sucked into technology, but with less and less understanding.
I want to know my tech inside and out as much as possible, and be able to control it.
---
As a preface to everything below this I'm not good at programming.
I've got a theory. I think that there are two main philosophies when it comes to creating $THING. One is to focus on pumping out features, and one is to focus on correctness.
Eventually, the project focusing on features will become unmaintainable, and a new project will form from the correctness project -- focused on features. This new fork/direction will continue, until it becomes to large to deal with.
And thus the cycle repeats.
I believe Linux is currently straddling the Features/Correctness line, and that BSD will be what follows up. This is obviously years/decades in the future though. I could also see something like RedoxOS catching on instead.
Edit: Spelling. Probably still missed some. Also, image -> image/video.
The fashionable way to break this cycle is to remove old things or unpopular things. But this creates all sorts of negative externalities with your users, and in my mind is the most selfish option for that choice.
There is another option: saying 'no'. Calling something "done". Doing what you set out to do and then moving on. How to do this in a way that is sustainably profitable and prevents someone else from coming along and eating your lunch for as long as possible, is something I would very much like to learn how to do.
Is there something wrong with just doing one thing well, making some money from that, and then moving on to do something else well and making some money from that too? I know it's an old-fashioned idea that something you do in tech might not continue to bring in rental income indefinitely...
Step 2. Don’t educate users. Blame them and call them dumb.
Step 3. Remove features. Make UI worse.
Step 4. GOTO 1.
I also blame engagement metrics. “Hey look! We removed this useful widget and now users spend an extra 2 minutes in the app. They love it!”
Expecting them to spend time to learn and configure your UI is not cost free. Therefore it needs some convincing value to justify the investment.
If that can't be done for most users simplification (by way of removal) is a valid way of caring for your user ecosystem.
It's valid, but extremely selfish. I think one should strive to minimize the amount of negative externalities one creates for their users, especially by removing something that people use. I'm not against removing things -- it's just that the current development zeitgeist seems to believe in killing things as soon as they hit a growth peak, and not when that something is nearly dead.
To this end I really admire(d) Microsoft for just how much they cho(o)se to keep in their big ol enterprise platforms like Windows or Office. Hell the Mac OS versions of Office have the same menu layouts that I learned when I was in elementary school. There are so many things they could/should kill -- please God let Wordart be on that list -- but don't, because they know people have muscle memory, and because they know a platform with lots of diverse capabilities is part of how their customer base thrives.
The Office ribbon may be a fantastic UI design, but people hate it because it broke muscle memory and shattered expectations. Some of the groupings make no sense, and people constantly have trouble finding things, is at least partially attributable to just how long is takes for intuition and expectations to die.
I think it’s ok to have some expectation for the user to learn the UI. As a user though, it often feels like the designers want to impede my ability to learn the UI. I used to be able to click around and discover everything. Now I need to sort through 50 articles of blog spam to find out what gestures or hidden menus to use. Even the docs suck (or don’t exist).
It's true that the average computing device is turning into an appliance (is it a bad thing?), but it also appears the options for DIY are ever-expanding.
Yes, for those of us who mostly want the appliance, but sometimes want to tweak it. I don't want to have to build a system from the ground up to get my desired experience if it's 98% the same as the appliance would have been out of the box; I'd much rather start from the appliance and just change the 2%.
I know of one company that has a few "desktops" that are Pi4 and I've seen loads of them powering "active terminals" - like in-store menu and office status screens. Gotta start somewhere. Special case and move towards general case (some call it crossing the chasm)
As Strong Bad might say, we have too much appliance in our computers, and too much computer in our appliances.
Contrast that with Visual Studio Code, which has a restricted, well-defined API extensions may use.
Oh, and the other thing about restricted, well-defined APIs? You're not locked into your present architecture. You can rewrite the entire code base, and all extensions will Just Work provided you reimplement the extension API. Extensions in the Emacs paradigm of having all system internals exposed to them will break the moment those internals change.
Restricted, well-defined APIs are system design 101.
Granted, the latter two are for niche use cases. But, the point is they exist, and I want to use them. Losing that ability is going to be frustrating, which is one of several reasons why I do not intend to "upgrade" to the new FF Android version until they've actually fixed that and the other major issues from this release.
I beg your pardon? Literally hundreds if not thousands of add-ons worked flawlessly on Firefox Android until a few weeks ago.
> All the major addons that people use are available
You'll need to seriously find some source for that extraordinary claim. What does "major" even mean? Who cares about "major ones" if 50% of my add-ons are missing, crippling the functionality of my browser on mobile? This claims makes no sense at all.
This is actually a small portion of addons available in the desktop Firefox. Many of them somehow worked, but were never optimized for mobile. "Worked flawlessly" is in my opinion a great overestimation.
Presumably, "major" means "used by most people", and even without checking I'm quite sure that if you looks at the install numbers of old Firefox for Android Add-ons, you'll find that a very large part of those installs were of extensions that are still available.
Sure, I'd like my small Hacker News extension to be available as well, but I do care far more about uBlock and Privacy Badger being available - and judging by installs, so do loads of other people.
(Again, which is not to say that I'd like the smaller extensions to be available as well. But "the major addons that people use are available" looks like a mostly accurate claim.)
this is not true. I used to have many addons working perfectly on Android.
> All the major addons that people use are available.
Wow, all 9 of them? /s
(from Mozilla's announcement: "We’re kicking it off with the top 9 add-ons")
That doesn't contradict the statement you're replying to at all.
> Wow, all 9 of them?
Yes, the major ones that people use. More are coming but I'd guess those 9 cover the vast majority of add-on installs
The number of add-ons that used to work before, was exponentially bigger than 9. Can we agree on that?
One example of add-ons that will stop working, is the Google container. Mozilla keeps making choices on behalf of its users by releasing an official Facebook container but not an official Google container (or an Amazon or a Twitter one) even though it would be trivial to fork the Fb container to achieve that [1].
This is problematic because we know that the largest share of Mozilla's revenue is from Google.
That's like a smoking cessation products company that is funded by Philip Morris.
So, you will excuse me if I think that some or recent Mozilla's choices are not in the interest of its users.
[1] and no, the multi-account container is not the answer because it's non-trivial to set up for less technical users
Where are you getting these numbers from? Unfortunately it seems that addons.mozilla.org statistics have become private for all add-ons (it used to be the case that add-on authors could decide on making them public or keeping them private), so I can only look at my own current numbers, but something's definitively off there:
Whereas general usage data goes back slightly more than one year, for me the Android data currently only starts in June (!) and with ridiculously low numbers at that (just a handful of users). Since August they have started to pick up a little, but they're still a far cry from what I remember – https://addons.mozilla.org/android/addon/video-background-pl... shows almost 1k Android users at the moment, which is less than the amount of Windows users being reported, even though installing the addon on desktop doesn't even make all that much sense due to desktop Youtube not blocking background playback anyway. Unfortunately I hadn't looked at that statistics for quite a while so my memory is a bit hazy, but according to the Wayback machine, a year ago that add-on had around 35,000 users, with a very clear majority (that I still remember for sure) of them having been on Android.
Likewise, my memories of uBlock origin's statistics are somewhat fuzzy, but I think it had something on the order of a few 100,000 (maybe 300,000 to 400,000 users?) on Android. Adblock Plus had a similar amount, which I remember within the context of them having considerably more users on Desktop than uBlock Origin, but a relatively similar amount on Android.
I wonder how the author did that. If it was a custom build of FF, we could rely on a fork as long as FF is open source and it supports extensions. If it is a configuration, I'll be happy to apply it to my FF.
Meanwhile I used ApkExtractor to save the APK of FF 68 on my phone (no root required) just in case I inadvertently upgrade FF on it. I already upgraded FF on my tablet and my old phone to test the UI. My take is that the new collections are good but the navigation UI is really bad. Opening new tabs, tapping the URL bar to go to a new site, forward button, they are all worse than in FF 68 and before. Hopefully they'll make them work as before.
I understand that staying on version 68 is also bad because of no security fixes. Furthermore it's possible that extension authors kill the Android version of their software. The plan is to wait for a while and see what happens.
I actually used the debugging functionality to install a temporary add-on. This is a supported scenario, but it isn't exactly easy to set up.
It doesn't – it's still at the stage of only being thought about:
There is a technical reason. They need to thoroughly test the new implementation of the APIs before they enable broader support. Just because the author did some (brief?) testing of a few more addons, it doesn't mean that the APIs themselves are production ready.
The main issue is that the new Firefox was released too early so compromises had to be made. Thoroughly testing a handful of addons is easier than thoroughly testing the entire web extension surface area.
This isn't a great situation to be in but there's no need to imagine nefarious plans to explain it.
Most people would probably see this as a Very Bad Thing, for a few reasons. First of all, if we were to take the claim that it's "to ensure stability of APIs" at face value, it would indicate a staggering lack of maturity in their decision making process around releases. Next, we shouldn't take the "API" argument as being anything representative of reality: applications people develop (or browser extensions) have always been the responsibility of the developers developing them. If there were bugs in a browser, they'd discover and work around them, or they'd have a broken extension/application. There isn't a way around that. Perfect APIs don't fix it: people write bugs into their applications/extensions/platforms all the time. Things are broken all the time. Testing has never been Mozilla's responsibility. It's always been on extension authors. There are already ways to "declare" in an extension that support of a specific version exists.
There is no competent technical reason to end up in this situation. We're only left with incompetence and malice. And it's _very_ difficult to think that the people running a popular mobile browser could be so incompetent. Though I'm sure we've all said that before about other Mozilla decisions.
Secondly, Mozilla absolutely do have a responsibility to ship APIs that work to spec. It's not up to Addon authors to fix Mozilla's bugs (and that may not even be possible depending on the nature of the bug). And of course you're assuming Addon authors would be interested in going so far out of their way to support what is a relatively small user base.
It seems very funny to say "we had to give up and release a broken browser" and in the same breath take on extra responsibility for API correctness prior to letting any one who can't convince Mozilla to whitelist them to use the APIs. And take on responsibility of maintenance of a whitelist
We've been here before with desktop Firefox when the browser got a lot of flak for issues that were mainly caused by addons.
Their solution to that problem hasn't exactly led to a positive perception of the browser either, and it hasn't even left them the out of blaming it on anyone but themselves.
Besides, it would not be a good look to be blaming extension authors for Mozilla's problems. So I don't think that's a viable "out".
They did it to desktop, now they've done it to mobile.
I like using well designed webapps and I generally prefer to leave Javascript enabled. Unfortunately most content sites have become so incredibly user-hostile that the only way to read them is without Javascript. I do not care about enhanced privacy yadda yadda whatever... I just need to be able to turn off the crazy that much of the modern web has become.
I guess I'm too late? Sounds like FF is not my answer.
I recently switched to Kiwi+uBlock Origin because it has a proper night mode, but if you don't need that Brave is a good option (you can ignore/disable the crypto currency/advertisment nonsense).
The full list of supported addons are: uBlock, HTTPS Everywhere, Privacy Badger, Decentraleyes, Search by Image, Dark Reader, NoScript, YouTube High Definition, and Privacy Possum.
You can put all the scary security warnings you want to scare off the unknowledgeable masses, but at least you'll provide some vector to allow niche users with niche requirements to add and trust a niche repository of their niche add-ons.
Are you friggin' kidding me? uBlock Origin is the only reason I've switched to Firefox on my mobile. I pledged to do the same on desktop the moment Chrome disables ad blocking extensions, but cheeky bastards are only threating that for years yet.
Only grim for those who chose not to (for whatever reason short of missing hands or bashed in cranium), or cannot be bothered…