Pocket migration to Firefox accounts | Hacker News Reader

Pocket migration to Firefox accounts

support.mozilla.org

44 pointsoleskiewicz66 comments

This is such a nothing story it's surprising it's so high on hn. There's nothing new with Pocket and nothing new with Firefox, so why are we all complaining about things unrelated to this story? Almost any company will move to provide SSO for all their connected products. Having a separate ff and pocket login is silly at this point.

Because a simple, integrated part of the pocket product just turned into a dependency on a large, complex, fragile ecosystem.

Instead of improving and investing in the pocket product, firefox instead chooses to leverage it to promote their other stuff, at the expense of making pocket's user experience worse.

Because HN loves complaining about how bad Mozilla is, and that they therefore instead use Chrome...

Nevertheless, HN crowd would gladly lambast Google, too.

Mozilla is the best of the rest, and it's fine to use it instead of Chrome and still be critical of Mozilla.

Yeah, of course. I just find it tiresome hearing "Mozilla did X, therefore I use Chrome", when Google on the same metrics is orders of magnitudes worse.

I use Mozilla for the same reason I vote for a shitty political party.

Nothing better exists than the heaping turd that we have.

I noticed when my firefox upgraded itself I got yet another giant fullscreen pocket ad...

Damnit. I use a fork of FF that has all the (worthless) Pocket stuff removed. So, now I'll be stuck refactoring my patches.

I feel like every FF related post reminds me that Mozilla has lost their way. What can be done to bring it back?

I just disable pocket in about:config. It’s annoying that I have to do that at all, but it does the trick.

> I feel like every FF related post reminds me that Mozilla has lost their way. What can be done to bring it back?

cutting off the money supply or a complete cleanout of the executive team

I'm wondering if this will affect the Pocket integration on Kobo eReaders.

Absolutely the first thing I wondered and came here to comment on. Not answered in the FAQ section of the page.

Hi all! I'm the head of product at Pocket. We're in active conversations with Kobo right now to find ways to preserve that connection. It's a really important one for us.

Thank you for being proactive on this! It's one of my favorite feature on the Kobo and I'd be disappointed to see it break.

Somewhat tangentially related. Pocket seems like a dead product. I have been using it since it can out but it feels old, slow and quite behind.

I use it extensively to capture a reading list and then trim it occasionally. Any recommendations for alternatives?

Offline web-rendered article views are a must - something pocket can’t do. And their text mode misses quite a few important details when there are equations or illustrations in the content.

I am also looking for an alternative. Pocket just isn't good at what it's supposed to do: saving articles for later reading, often offline.

Same. I miss pocket's chrome app. Worked great when offline too.

I don't really care about web service. Any desktop tool that can capture websites like pocket? As in download the stripped-down article. They call it "reader view" sometimes.

There's some extensions/scripts like Markdownload that uses Mozilla's Readibility.js to download stripped-down articles. But it's a bit manual for my taste.

Instapaper is you like cross-platform, including Kindle.

GoodLinks if you only use Apple and value privacy.

Instapaper after multiple acquisitions feels even more abandoned than Pocket.

I heard Wallabag is a nice Pocket alternative. And you can host it yourself. I'll try putting it on my odroid one of this day.

I'm now using the nextcloud bookmarks extension, but it's rather slow, seems unmaintained, most android apps & browser extensions don't work anymore (way worse than pocket, but at least there is no adds)

I'd considered using Wallabag years ago. How has it evolved in recent years?

I recently tried wallabag on my home server. The web version is fine. Functional even.

But the state of the iOS app is, well, not great. Clunky design, with odd syncing. Which is unfortunate as 99% of my use case is reading it on my phone whenever I’m waiting for something.

Using a selfhosted version behind a subdomain of one of my TLDs. Something to be aware of - it can't be hosted behind a specific URI if you plan to have other stuff hosted on the same web root path (like example.com/wallabag), it always needs it's own (sub)domain (wallabag.example.com).

Other than that, it works pretty well with most text-based content. I'm using it through the Android app occasionally, and while the app won't win any design awards, it serves it's purpose.

I've noticed similar issues lately. In fact, I have resorted to keeping my my Pocket browser extension in Brave disabled because for whatever reason, Pocket seems to cause the browser to fully utilize ~1.5 cores. I only discovered this after disabling all extensions and using them one at a time. Pocket was the only one that seemed to have a consistent signal. I haven't had weird Brave CPU spikes since I disabled Pocket.

I've wondered about how to exfiltrate all of my accumulated links I've put into Pocket over the past ~7 years or so.

> I've wondered about how to exfiltrate all of my accumulated links I've put into Pocket over the past ~7 years or so.

Yeah, this is what has prevented me from moving away. I was a premium user and used to use their search. I’m not anymore but I still want to keep the list around. I still go back and find references that Google doesn’t show anymore.

There are some apps that move data from pocket to kindle or such, but having an import from pocket in a new product would be really nice.

Could you share more about what Google misses today? Does the content exist and is now effectively unlisted?

I am hopeful that the Readwise Reader product will be good. However, it is still closed beta.

As a beta user, I can say that while it's super promising and I use it daily, it does still have some rough edges to it hence the closed aspect of it. The onboardings are starting to scale up a bit I believe though. Being able to highlight and take notes on RSS items or emails is very nice.

I bought feedly just to be able to put rss and mailing lists into Readwise but it is really painful UX personally.

And Pocket keeps having trouble with doing reader mode with sites in 2022 and thus can’t do highlights.

I recently heard about Alfread on HN. [1] If you're just looking to replace the front end, you can use it to pull from Instapaper/Pocket feeds.

Not sure if it currently does all you'd want, but the founders are responsive on email so perhaps worth asking for any features that you'd like them to add!

1: https://news.ycombinator.com/item?id=29703671

I've used instapaper for a long time, and it's a better product than Pocket, but it, too, has the problem of missing images sometimes.

If you want to save the entire snapshot of the web page for archival and references purpose, I think zotero is a good choice. But it's not a good read-it-later product.

Another vote for Instapaper vs Pocket. Basic Instapaper allows you to make folders and select to which folder to save, while basic Pocket only gives you a blank, suggestionless space to write tags.

I've been pretty happy with https://pinboard.in/

It doesn't have fancy graphical previews and stuff like that, but it works and archives page contents for an extra fee.

Worldbrain Memex is incredible

I used Raindrop.io for a while after moving off Pocket, it’s much better. Now I’m trying out mymind but it’s costly so I may revert back if I don’t find it overly better.

Save as ebook extension (https://addons.mozilla.org/en-US/firefox/addon/saveasebook/), everything goes in a folder managed by syncthing so I have it on my smartphone as well.

On the scale of years, products fail, get bought, are unmaintained, become too expensive, etc... An export of your database is always a hassle to manage, and you must hope there will be an importer for the next product that will also fail a few years down the road.

On the other hand, files have always been there and will always be there. Epub is a standard-enough format that will live. The highest piece of complexity is syncthing, which can be easily replaced the day it doesn't work anymore. Bonus: I have all my data offline so I don't need connectivity, and I lower my environmental impact.

Too often products feel like a reinvention of the whole stack because it's easier to sell, rather than reusing proven tools and using them the way they were supposed to be used.

Has Pocket improved in any way since this acquisition? Or is their team tied up with integrations like this and unable to add any new features or improvements?

So, how to fully disable it via `about:config` now?

> Why is Pocket migrating to Firefox accounts?

... nonsense omitted

What's the point of acquiring something if you don't leverage it to coerce their users into using your other products?

https://en.wikipedia.org/wiki/Pocket_(service): > On February 27, 2017, Pocket announced that it had been acquired by Mozilla Corporation, the commercial arm of Firefox's non-profit development group. Mozilla staff stated that Pocket would continue to operate as an independent subsidiary but that it would be leveraged as part of an ongoing "Context Graph" project.

When a company with a huge user base (Mozilla) acquires a company with a much smaller user base (Pocket), they're probably not doing it to cross-sell the smaller user base.

I thought Mozilla was going to be investing in Pocket and giving them the resources to grow, as well as sending them Mozilla's users — a fraction of whom would become premium Pocket subscribers.

But as far as I can tell, there haven't been any notable new features in the last 5 years. In fact, some stuff like the web app have gotten noticeably worse. I used to use Pocket almost daily, but now I use it maybe once every couple months.

So what was the point?

Money laundering? Fundamental incompetence? Pick any reason honestly, they all make the same amount of sense. I've never understood why Mozilla bought Pocket.

I don’t know what the economics were, but it may have been a soft landing / acquihire. I’m sure some people believed in the business case, but the driving force for the acquisition may well have been a VC that wanted an exit, a well-connected founder, etc.

I’m not really sure. I’ve been using Pocket since it was Read It Later and it hasn’t become very much more useful to me since then. It probably could have advanced in many more ways given the right people and resources. They have access to these, but the great innovations have been in bundling the product and administering its accounts.

I’m glad they haven’t changed the Pocket experience too much. I still use and pay for Pocket but I’ll probably jump ship to pinboard or build a personal solution if they do something that diminishes the value I find in it today. The most likely ways this could happen is by reducing functionality (e.g., discontinuing clients/APIs), adding unwanted things (more advertising and social media), and doing things that seem unexplainably arbitrary (see: colorways). Pocket works everywhere I need it to, it has acceptable design and performance, and it has demonstrated some maturity and resilience as a product that’s around 15 years old.

Other solutions I’ve tried include clips in Evernote, OneNote, Zim; the app Keep Everything; the services RainDrop and Larder; and e-mailing myself. All of these except the last turned out to be more of a hassle to put up with than using Pocket.

If Pocket gets worse, I’ll probably make my own solution with e-mail. I practically never look at or for anything I save in Pocket, so it would seem my main attraction is the ease of sending things into a vault from anywhere I’m likely to use the internet and the confidence that I can find these another day. I’ll probably be able to access all my old e-mails in 10 years, there’s pretty good search, and it’s ubiquitous (although it is easier to use HTTP than SMTP in some scenarios).

I've never understood - what does Pocket do over 'save to reading list' in your browser? That's how I save things to read later - and it's free.

I don’t know. I guess I stated using it before that was a common feature, and it was useful to have a cross-platform way to save bookmarks across devices. It’s the easiest cross browser option IMO.

Fair question. For me, it was the cross-device sync and offline functionality. Also, save to reading list didn’t exist when I started using Pocket.

Some people also use Pocket as a repository.

Has anyone around here tried to run their own pocket/ff account replacement server for session management? I did some reading a couple of years ago but got caught up in other things and never even tried doing it.

I recently set up a Firefox Accounts server - basically the minimum parts necessary for a self-contained FFSync, without either calling home to Mozilla or their integrated third-parties (mostly payments and analytics; AWS, GCP, Paypal, Stripe, and a few others). A handful of microservices and a few MySQL databases and I think a Redis or two? I only recommend doing it if you enjoy it for the challenge, the stack it's clearly not engineered with an independent self-hoster in mind (though they do support it and people are helpful). It took several days to get up and running. I will share this soon, I hope...

In the meantime, if this all sounds like too much but you'd still like self-hosted FFsync, you can run the syncserver docker image standalone and piggyback on the rest of Mozilla's hosted FXA stack.

I'm not certain just how big the gap is for Pocket ATM (I know they've continuously been opening up since acquisition).

EDIT: Just to not be that guy, microservices as this isn't clearly documented anywhere:

  * pushbox (mysql)
  * fxa-auth-server (redis, 2xmysql, smtp)
  * browseridverifier
  * syncserver (actually tokenserver+syncstorage, if you're not running the docker image. db of choice)
  * fxa-contentserver
  * fxa-graphql-api
  * fxa-profile-server (mysql)

...and "just glue them together" with JWT tokens (look for the scripts in the repo instead of trying to find enlightenment through the source code, docs and issues... heh) and your reverse proxy or load balancer of choice.

Basically impossible to do without docker and even then, it feels like it would be a full time job to maintain.

I played around with the idea of writing a minimal version that cuts out all the fat but even that seems like a monstrous undertaking.

See my sibling. It’s a mess to do the initial setup but after that the maintenance is trivial and quite hands-off.

Firefox accounts are terrible. I recently thought I had a second account because I got some error about not using the primary account when trying to reset the password using that email. Turns out that Mozilla still had an old account connected even though it was removed in their interface. If you look through their bug report history they've been plagued with similar issues that they've never been able to figure out. The Firefox Sync protocol is a mess.

Works much better for me than Chrome had.

You mean regular Google accounts?

Pocket is basically unusable for me without the extension In My Pocket, which is basically a companion extension, that makes it a lot more like it used to be in the past.

I use zotero for this. I have a "interesting_reads" collection that I save stuff I see on the internet (like from HN) and want to read later. It takes 'snapshot' of the website so you can read it offline, or in case it disappears. [1] https://www.zotero.org/

Is pocket open source yet?

Hmmm, the bookmarklet recently stopped working. I wonder if this is related in any way.

We still got no news about self hosting of pocket. Does authentication work with a selfhosted Firefox Accounts instance?

I don't think I've seen anything official after this: https://discourse.mozilla.org/t/pocket-source-code/43686/12

Quite a lot has happened on their GH, though. I don't think the web front-end is there, but it seriously looks like they're getting there?

https://github.com/Pocket

It's pretty radical just how transparent Mozilla actually is, I think. Here's the start of the mentioned migration https://github.com/mozilla/fxa/issues/10635

Will it become necessary to log in to use Firefox?

You had to login to use pocket before, what is different now?

It isn't. The firefox account features things like sharing your bookmarks/history/etc between devices easily. You don't need to use it to use firefox.

Given that we've seen things like Windows take the "you need to create an online account" route, and only a decade ago that would've seemed absolutely absurd, I think your question raises a good point and the downvotes are unjustified.

The only true answer is "not currently". A "not yet" would be too pessimistic, and "never" is too optimistic. If there's one thing that stands out during the past ~decade of software, it's "things are optional... until they aren't."

That's what I was getting at.

Who would have thought you'd need an account to read a newspaper or watch broadcast TV over the internet?

Reply on news.ycombinator.com