Hacker News Reader

Comment by krn | Hacker News Reader

> Pocket is also experimenting with personalizing recommendations for certain users based partly on each person’s Firefox browsing history.

Why does an open-source web browser have a proprietary cloud synchronization service installed by default?

EDIT: I know that Pocket is owned by Mozilla. It doesn't change anything.

View on HN

Mozilla actually owns pocket now; they seem to believe that giving people recommendations from around the web is the best way to combat walled gardens who want all your content going through them. This article is pretty revealing: https://blog.mozilla.org/blog/2017/02/27/mozilla-acquires-po...

I still don't think it should be proprietary, and I don't use it, so

  extensions.pocket.enabled = false

is my friend.

I wish that where off by default, it actually makes me feel uncomfortable installing firefox on my laptop at work even though it really is better than chrome when you want to read docs and don't have a ton of free memory.

So, what do you use? Chrome is more comfortable?

Yes, chrome doesn't dump news articles on my screen every time I open a new tab.

I use firefox anyway, it just makes me uncomfortable.

Turn off site recommendations: https://help.getpocket.com/article/1147-hiding-pocket-in-you...

Disabling Pocket: https://help.getpocket.com/article/1025-disabling-save-to-po...

Chrome on Android does. Although you can now hide it since a few weeks ago.

This is weird to me; I never open a new tab in a manner that exposes me to that... it's always typing into the URL bar and then hitting alt+enter, or middle-clicking a link. (Plus session restore)

you can easily customize that

All browsing history is kept on your device and recommendations are calculated on your device. The only cloud synchronization is the recommendation engine - no different to the existing synchronizations like the update checker or the phishing sites list.

They phone home with the recommendations which leaks info about your browsing history.

Source? That's contrary to everything I've seen.

>When you use Firefox, Mozilla by default collects technical and interaction data. You can read about the data we collect and how you can choose what you share in the Firefox Privacy Notice.

>If your Firefox is configured to share this technical and interaction data with Mozilla, when Recommendations from Pocket are displayed on your new tab, we collect information about how many times they appear and whether or not you interact with them.

>This reports all the Pocket recommended articles (a list of ids) when the user opens a newtab.

https://help.getpocket.com/article/1142-firefox-new-tab-reco...

https://github.com/mozilla/activity-stream/blob/master/docs/...

The sentence following your second quote is:

>However, this information is not associated with any of the other technical and interaction information about you or your copy of Firefox.

Are you concerned about collection of anonymized metrics, or do you have some reason to believe that the data collected is not anonymized?

I'm using an open source browser because I don't trust third parties; what if they were served a NSL or similar?

Either way the common claim that it's purely local is certainly not true.

Pocket is owned by Mozilla.

Pocket is proprietary and owned by Mozilla.

The client side is fully open sourced and to my knowledge there is work ongoing to opensource the server too.

Mozilla owns pocket.

Yeah, it's Mozilla's proprietary cloud synchronization service.

Cloud Sync is not proprietary, you can run your own with all functionality that Mozilla's provides, the code is on Github.

This is about Pocket, which is what krn referred to as "a proprietary cloud synchronization service". Mozilla has been promising to make Pocket open source since it acquired it over 1.5 years ago, but haven't done it yet.

There is plenty of client source code available on github, the remaining parts are the server which understandably might pose some issues if it contains either patents or third party proprietary code.

Yeah, the clients are open source, and they're interfacing with the proprietary cloud service. Just like krn said.

Atleast part of the cloud service is open source (mozilla's side) and they are still planning to open source them. Until evidence of the contrary surfaces I do believe Mozilla when they say that.

A trustworthy open source company doesn't bundle its product with a service its users have no control over.

Reply on news.ycombinator.com