1406647 - Please do not ship the Cliqz addon (well-known adware in Germany) as experiment

Status: RESOLVED INACTIVE

(Firefox :: General, defect)

Description

[Darkspirit]

(I am just a Nightly user/community member.)

Cliqz is a well-known Adware from Germany[1] and it was a plague because it was bundled with Freeware installers (for example from chip.de).

There is massive negative feedback[2][3].

Burda (shareholder) has a bad image as german press publisher because of things like https://en.wikipedia.org/wiki/Ancillary_copyright_for_press_publishers - just because it's a german publisher.

Cliqz has a very interesting technology, but those arguments don't reach users and they wouldn't be strong enough to overcome negative concerns.
Buy them (change the name, make an own (legally public benefit) search engine), or cancel every partnership. Their commercial background plus Burda (even Cliqz itself seems to have good intentions) is source of mistrust.

It has to be Opt-In. Ask in the onboarding tour or show some bubble on about:newtab.

You don't want to ask to enable basic tracking protection while onboarding (germans would love it), instead we can read sentences like 'which evaluates the surfing activities of the users.' In the last year we had enough bad press of dangerous Firefox/Chrome Addons that sell surf history.

At the moment, some users find it useful, but most other users get a negative impression of Mozilla.

[1] https://www.google.de/search?q=cliqz+adware
[2] https://www.heise.de/newsticker/meldung/Firefox-Testlauf-fuer-neue-Empfehlungsfunktion-3852129.html
[3] http://stadt-bremerhaven.de/mozilla-firefox-mit-cliqz-experiment-welches-von-nutzern-automatisch-daten-sammelt/

Comment 1

[Darkspirit]

> Buy them (change the name, make an own (legally public benefit) search engine), or cancel every partnership. 
And it would have to be open source then. Think about crowdfunding.

Another thread:
https://www.reddit.com/r/firefox/comments/74n0b2/mozilla_ships_cliqz_experiment_in_germany_for_1/

Comment 2

[klk745]

I second the opinion of the reporter.
This would be (or already is) a complete publicity disaster for Mozilla.

Comment 3

[Darkspirit]

https://blog.fefe.de/?ts=a726167b (in Top 10 of german blogs according to Wikipedia)

Comment 4

[fff]

I definitively agree with the reporter. Very bad publicity and negative feedback. Also hit German tech newsmag: https://www.heise.de/newsticker/meldung/Firefox-Testlauf-fuer-neue-Empfehlungsfunktion-3852129.html
Firefox will lose a major market (Germany, Switzerland, Austria and other countries) with this!

Comment 5

[Joe]

Given this
https://wiki.mozilla.org/Firefox/Principles
this certainly is a bug that happened accidentally, right?!!

Comment 6

[Christian Stadelmann]

Please use the voting feature instead of spaming the comments with +1s.

(In reply to fff from comment #4)
> I definitively agree with the reporter. Very bad publicity and negative
> feedback. Also hit German tech newsmag:
> https://www.heise.de/newsticker/meldung/Firefox-Testlauf-fuer-neue-
> Empfehlungsfunktion-3852129.html
> […]

Please see comment #0 above.

Comment 7

[Darkspirit]

I think a "Sorry, we were wrong and heard your feedback" press release which refers to https://www.mozilla.org/en-US/about/manifesto/ may be helpful. A bit of humility would certainly be a good thing, even you had good intentions. You could combine this with an announcement of doing bug 1387681 (privacy.trackingprotection.ui.enabled;true) and that you want to test if you could enable it by default (bug 1377563) - at least for German users. This would show in what direction Firefox wants to go.
If you really want to promote Cliqz, you could show an "Install Cliqz" button on about:preferences#search or when asking if search suggestions should get enabled.

Comment 8

[Darkspirit]

> If you really want to promote Cliqz, you could show an "Install Cliqz" button on about:preferences#search or when asking if search suggestions should get enabled.

But this certainly would be also perceived negatively. Just showing it in the list of search engines would be the least offending.

Comment 9

[Niels]

(In reply to Christian Stadelmann from comment #6)
> Please use the voting feature instead of spaming the comments with +1s.

Just for the Users, you has to expand "Details" on Top to Vote.

Comment 10

[Alex]

(In reply to Jan Andre Ikenmeyer [:darkspirit] from comment #7)
> I think a "Sorry, we were wrong and heard your feedback" press release which
> refers to https://www.mozilla.org/en-US/about/manifesto/ may be helpful. A
> bit of humility would certainly be a good thing, even you had good
> intentions. You could combine this with an announcement of doing bug 1387681
> (privacy.trackingprotection.ui.enabled;true) and that you want to test if
> you could enable it by default (bug 1377563) - at least for German users.
> This would show in what direction Firefox wants to go.
> If you really want to promote Cliqz, you could show an "Install Cliqz"
> button on about:preferences#search or when asking if search suggestions
> should get enabled.

With a market share of 6% (world) and 20% (Germany) who would read it?
Memory leaks persist, webext addons are shipped with incomplete features, push notifications are annoying, closing button for inactive tabs is already gone for many month now, google analytics on addon page, pocket.
No, enough is enough!

Comment 11

[stmz]

mozilla has (had?) a good repution for respecting user privacy. This decision is absolutely catastrophic and I am losing the confidence that mozilla will not harm my privacy.

I am a firefox user from the very first day on. If you stick with your decision than I will switch to a different browser.

It is absolutely impossible to assure 100% anonymization. How do you know what parts of an url might contain sensitive informations (e.g. session keys, private informations)? There is no way to remove all private or sensitive informations!

Comment 12

[Daniel Rieken]

Firefox lost the browser war (https://www.heise.de/imgs/18/2/2/4/5/6/8/1/infografik_427_globale_marktanteile_webbrowser_n-031b7e88c81bd222.jpeg) and Mozilla is wondering why...
It is because of things like this!

Comment 13

[Thomas Peter]

Couldn't agree more. Why is this even a thing? It goes pretty much against everything that Mozilla is supposed to stand for. 

What about "Internet for people, not profit"?

Comment 14

[o-zilla]

I recommend rolling back this change and to apologize. This has already gone viral and it'is becoming PR desaster. And if any of the major IT-news websites/magazines picks this up… let's just say in a normal company people could get fired. You might loose your most valuable user base, the power users and advocates, and that in a country with a very high Firefox user base. 

If you want this test data, advertise on the download page “download firefox“ „download firefox with better search results by Cliqz“ with an explanation of the privacy issues (Cliqz' privacy policy doesn't look to bad and it's open source, only the owning company is questionable).

Ask the user, trust is at utmost importance!

And another thing: make your announcement quick - rumors are always bad, fight them early.

Comment 15

[andriy.komm]

I hope you did not sign any contract with them already.

Comment 16

[celtar]

I don't think Firefox can be used in the company environment anymore. Data protection is no longer guaranteed here and there are doubts as to whether the data will be used for other purposes. (Patriot Act). We will therefore have to switch to a different browser in the company environment. (The ESR version is not really updated).

Comment 17

[gabe]

The worst part of this is that no developers in that bug even brought up privacy concerns. This is the primary issue here. You're supposed to be the browser that stands up for privacy, yet you're failing that mission here, and instead of publicly addressing it, you actively attempt to hide it.

That's not acceptable. This needs to be rolled back. This needs a public apology. And most importantly, this needs to not be like Canonical's "Don't trust us? We have root!" moment.

Comment 18

[DerEnttaeuschte]

After 50++ installations of FF over years for friends, for family, for neighborhood against IE (and later Chrome) and for more privacy i will withdraw my recommendation for FF, if that "bug" happens. Pale Moon is fresh installed, passwords and bookmarks are migrated... your choice!

Sorry, i know my English is bad, but i hope you understand.

And hopefully the reporter won`t get hurt too hard, if my comment is misplaced!

Comment 19

[Darkspirit]

Please do not post further "me too"s. There are some suggestions how to solve this situation. I think it's clear enough that there is a problem (Opt-Opt, Burda, Adware). Let's wait for Mozilla's response. And please try to find some words that try to explain why Mozilla and Firefox are important to you (some motivational words!), what Mozilla should do better and maybe what you could imagine to do for the non-profit Mozilla: https://input.mozilla.org/feedback

I (just a Nightly tester) would be thankful if you could help testing the amazing Firefox Quantum (https://beta.mozilla.org) and make clear to Mozilla that it is more important than this experiment.
Let's donate (without Paypal :-) to underline this and give a small comment as purpose:
https://donate.mozilla.org/de/ways-to-give#wire
Maybe Mozilla finally thinks about memberships and making surveys among their members.

Comment 20

[mtx]

I'd like you folks to know that I've already received pretty nervous mails from several of my clients (over the weekend!), asking for advice how to proceed with this situation. Seems that this aleady went viral on IT managers in larger organisations.

Quote #1: "worse than microsoft and google together !"
Quote #2: "can we make our own browser ?"

You should know that this is a major violation of data protection laws (at least in Europe), as well as company rules.
IT manager allowing that put their head into the Guillotine (yes, talking about criminal charges).

Comment 21

[mtx]

By the way: the Burda empire was built in the national socialist Third Reich, and played an important role in its propaganda.

Comment 22

[andriy.komm]

(In reply to mtx from comment #21)
> By the way: the Burda empire was built in the national socialist Third
> Reich, and played an important role in its propaganda.

To add a source: https://en.wikipedia.org/wiki/Hubert_Burda_Media#Business_during_the_Nazi_regime

While all companies were affected in germany during this time, most just did what they had to do, instead of clapping the hands and pushing statements like burda did.

Comment 24

[Clemens Eisserer]

What a managemenet fail. How can Mozilla set such an action in times like this? Currently one one of the only major selling points of Firefox was its privacy awareness.  Please stop this desaster.

Comment 25

[tt_1]

I don't like the idea of a toolbar recommending me stuff, it costs me performance and I prefer to think for myself. However, I can find it acceptable to ship it with the browser under these circumstances: 

OPT-IN, no discussion about this

if you make it a so called "system addon", read: you include it into your source tree 

- make sure to allow to disable via compile switches such as --disable-cliqz 
- if being too lazy, blah, whatever reasons, mark it as disabled at runtime and allow to activate it with the users active consent, see opt-in before. 

This is a massive pr desaster, I spent the weekend (read: I had to spent the weekend) to elaborate and test different other browsers to switch over, in case you upstrem guys are stubborn and ignorant. In the meantime you closed #1392855 for comments because they were not of your liking, same goes for comments in this bug here. Is it too much of a burden for you to read through what people, which are users of your product, write here, free of charge for you, in their free time?

Comment 26

[Darkspirit]

Please note: I am hiding your me-toos, not Mozilla. Bugzilla is a place to report problems, not to attack people making ones. You get an assessment, mostly a fix, then you verify the fix and express some appreciation to give motivation. You tell them your perception (no me-too) and may bring up possible solutions (no me-too). If you want to express your disgust you can use the feedback formular, but don't forget to mention why Firefox is important to you. They are friendly people, they see the feedback. Firefox Quantum 57 (https://beta.mozilla.org) and regular updates (please don't disable them) would not be affected by this experiment.

Comment 27

[mtx]

(In reply to Jan Andre Ikenmeyer [:darkspirit] from comment #26)
> Please note: I am hiding your me-toos, not Mozilla. Bugzilla is a place to
> report problems, not to attack people making ones.

Some of these comments highlight the social/political background, which 
*is* an important aspect of the situation. As engineers/scientists we
need to consider that our technology can be abused (we see it every day).

Therefore we have an obligation to be very careful about how we design
our systems/machines, and whom we give power over them. In today's
society it became very hard for the common user to protect himself,
many users aren't even aware of the dangers (I even know managers, who
still belive their company's product won't ever be attacked and therefore
believe a 4-digit pin would be enough).

This puts us, the IT professionals in charge of doing the best we can
to prevent abuse of our technology (technical measures, as well as 
education) - because we know better.

I'd suggest reading Goethe's "Faust", which copes w/ that topic.

The underlying problem - the abuse - indeed isn't primarily of technical,
but political/social nature.

> You tell them your perception (no me-too) and may bring up possible
> solutions (no me-too).

For the time being, the only real solution I see, is completely dropping
the whole experiment, and not attempting any similar experiments with
uninformed users - those things should always be entirely optional,
by external (not bundled) extensions or separate builds.

Certainly, I'm very interested in an distributed - user owned - search
engine system. But we need to be very careful not to leak any personal
information, not allowing any tracking/sourveillance. This would make
up an entirely own project, far out of scope of browsers.

> They are friendly people, they see the feedback. 

I'd like to belive that, but, unfortunately, my personal experience
in the mailing lists draws a different picture.

Comment 28

[mtx]

(In reply to tt_1 from comment #25)

> In the meantime you closed #1392855 for comments because they were not of your liking

Meanwhile this bug is even completely secret.
This matches the common practise of censorship in the mailing lists.

Comment 29

[Marco Bonardo [:mak]]

(In reply to tt_1 from comment #25)
> In the meantime you closed #1392855 for comments because they were not of your liking

Hi, I closed comments in the other bugs, because differently from this one they are technical bugs. I didn't touch this bug that has far more many comments, as you can see, even if it's probably still not the best place to discuss these topics.
Please remember that for us Bugzilla is a work tool we use every day, and spread noise in multiple technical tickets just makes our lives harder for no benefit of anyone.

Comment 30

[mtx]

(In reply to Marco Bonardo [::mak] from comment #29)

> even if it's probably still not the best place to discuss these topics.

What would be a better place ? firefox-dev is censored for many years now.
I've seen similar topics, eg. despotic restriction management, that have
been silenced by censorship.

You should be aware, that discussion here is just the tip of the iceberg:
Mozilla greatly lost trust, it's reputation is in free fall, by many actions
over the recent years. The situation looks pretty similar to OOffice - and
we all know what happened then.

> Please remember that for us Bugzilla is a work tool we use every day, and
> spread noise in multiple technical tickets just makes our lives harder for
> no benefit of anyone.

Please keep in mind that most people don't see much other options to raise
their voice. They feel ingored, overrun, scared, angry. If those things
happen to countries, we would call it an pre-revolutionary atmosphere.

IMHO, Mozilla and its community is huge tension between commercial vs.
public interests, and the problem is increasing. Aspects like Cliqz,
DRM, marketing, b2g, etc. clearly point into the commercial direction.

At some point, Mozilla has to choose sides and decide whether it wants
to be a commercial project (like Chrome) or a public community - and
act that way.

Comment 31

[H.-Dirk Schmitt]

184 Votes and still "Unconfirmed" ?

Comment 32

[Paul Menzel]

Why is Firefox 57 unaffected. Could you please provide an URL or a commit, that it the add-on isn’t enabled on 10 % of the users? I only found news reporting that the integration is planned.

[1] https://www.reddit.com/r/firefox/comments/7df7n9/cliqz_in_firefox_57/

Comment 33

[H.-Dirk Schmitt]

After the next attack – „Mr. Robot“  it is time to strike back.
See e.g. the German Heise News →  https://www.heise.de/newsticker/meldung/Werbe-Add-On-fuer-Mr-Robot-Kritik-an-Mozilla-3919951.html

The following bash script fragment is an example to get rid of unwanted features and trigger the „monitoring“ if an unknown feature appears.

---- 

declare -r -g -a  _BannedFeatures=(
    [e-mail]
    [e-mail]
    [e-mail]
    [e-mail]
    [e-mail]
    [e-mail]
    [e-mail]
)

declare -r -g -A _AllowedFeatures=(
    [[e-mail]]="? newtab window"
    [[e-mail]]="✓ Application Update Service" # → https://www.ghacks.net/2016/11/04/application-update-service-helper/"
    [[e-mail]]="✓ Form Autofill"
    [[e-mail]]="✓ Screenshots"
)

declare -r -g -a  _FeatureDirs=(
    /usr/lib/firefox/browser/features/
    /usr/lib/firefox-esr/browser/features/
)
function checkFeatures() {
    (( $# == 1 )) || c42Cfg_abortCalLError --sig='<feature dir>' "$@"
    declare dir=${1}
    declare f
    declare -A features=()

    c42Cfg_info "check features in ${dir}"
    # remove banend features
    for feature in ${_BannedFeatures[*]}
    do
        if [[ -e ${dir}/${feature}.xpi ]]
        then
            c42Cfg_warn --syslog "found banned feature ${feature} in ${dir} → remove it"
            rm --force ${dir}/${feature}.xpi || c42Cfg_error "failed to remove ${dir}/${feature}"
        fi
    done

    for f in  ${dir}/*.xpi
    do
        f=${f%.xpi}; f=${f##*/}
        if [[ -z ${_AllowedFeatures[${f}]:-} ]]
        then
            c42Cfg_error "found not categorised feature ${f%.xpi}"
        else
            c42Cfg_info "$(printf %-30s ${f}) →  ${_AllowedFeatures[${f}]:-}"
        fi
    done
}

Comment 34

[Christian Holler (:decoder)]

(In reply to H.-Dirk Schmitt from comment #33)
>
> declare -r -g -a  _BannedFeatures=(
>     [e-mail]

Please note that aushelper is sometimes used to temporarily resolve topcrasher situations (e.g. by disabling a certain feature) until a fix is available or in case of a security vulnerability if immediate response is required.

Comment 35

[Justin Clift]

(In reply to H.-Dirk Schmitt from comment #33)
<snip>
> The following bash script fragment is an example to get rid of unwanted
> features and trigger the „monitoring“ if an unknown feature appears.
<snip>

Looks really useful.  Want to put that up on GitHub, so people can expand it out via Pull Requests? :)

Comment 36

[Henrik Winnemöller]

(In reply to Paul Menzel from comment #32)
> Why is Firefox 57 unaffected. Could you please provide an URL or a commit,
> that it the add-on isn’t enabled on 10 % of the users? I only found news
> reporting that the integration is planned.

A while ago I filed https://bugzilla.mozilla.org/show_bug.cgi?id=1432048 regarding the privacy statement on the about:rights page which links to the default statement even on Cliqz editions. In addition to stating the the about:rights page is a static page and does not support changing the link to the privacy statement, Mike Kaply from Mozilla stated:

(In reply to Mike Kaply [:mkaply] from comment #7)
> Those funnelcakes were disabled in early November and are no longer being
> served to users. This line:
> 
> https://github.com/mozmeao/www-config/blob/2e0339abbd2696350cb5035b11ee83db69884dce/configs/bedrock-prod.env#L91
> 
> turned them off.
> 
>  I have verified this with the mozilla.org team. I will be having those
> removed from that file.
(see https://bugzilla.mozilla.org/show_bug.cgi?id=1432048#c7)

That change Mike refers to happened on 2017-11-01 (UTC) in commit https://github.com/mozmeao/www-config/commit/32e90e844a138ad1feb3b186ef9470eda4f0e48c, deactivating the distribution of the Cliqz funnelcake experiment. 

And only some hours after Mike Kaply's cited comment the Firefox distribution configurations were updated in commit https://github.com/mozmeao/www-config/commit/03c0fca308e0321eaf120ee87f08fad9621aab65 that got merged into master later on in commit https://github.com/mozmeao/www-config/commit/eb98f086e224ef5923f1c90e878c091319a4f477 on 2018-01-29 (UTC). Visiting the links that still served the v56.0 funnelcake experiment editions before that commit, e.g. https://www.mozilla.org/de/firefox/new/?scene=2&f=120, then just served the standard installer and still do today.

Note: This does not stop already existing Cliqz installations from functioning. I have a working Cliqz edition that still updates like a regular Firefox and also receives updates for the Cliqz extension (Firefox 58.0.2, Cliqz 8.23.3 and Search Volume Modeling 0.4). This only stops new users that download Firefox in Germany from receiving the Cliqz edition.

TLDR:
- The Cliqz experiment distribution was turned off in early November 2017, stopping the distribution when trying to download Firefox using the standard links like https://www.mozilla.org/de/firefox/new/.
- On 2018-01-29 (UTC), the distribution using explicit links to the funnelcake experiment like https://www.mozilla.org/de/firefox/new/?scene=2&f=120 was also stopped.
- Existing Cliqz editions will continue to work and update like a normal Firefox does. The Cliqz extension stays active until the user uses the opt-out methods. Updating Firefox will not deactivate or remove the Cliqz extension.

Comment 37

[Darkspirit]

The chapter of Cliqz seems to have ended, so I should close my report.
Mozilla has massively improved Tracking Protection and other privacy features in the mean time. The efforts with Rust and Servo improve software security and performance to a grade we had not before. They are fighting for the good: https://www.mozilla.org/en-US/internet-health/ A big thank you for that!


But now there's a new idea: Advance by Firefox, Powered by Laserlike.
It's 100% Opt-In (advertised by only one press release and nowhere else) and they ask for feedback.

German:
https://blog.mozilla.org/press-de/2018/08/07/advance-neues-firefox-test-pilot-experiment-gibt-relevante-content-empfehlungen/
https://www.heise.de/newsticker/meldung/Firefox-Erweiterung-Advance-empfiehlt-Webseiten-4131351.html
https://www.golem.de/news/mozilla-firefox-testet-funktion-fuer-webseiten-empfehlungen-1808-135901.html
(Offtopic: Some readers might be interested in this: https://addons.mozilla.org/firefox/addon/livemarks/)

English:
https://blog.mozilla.org/blog/2018/08/07/firefox-offers-recommendations-with-latest-test-pilot-experiment-advance/
https://www.reddit.com/r/firefox/comments/95brn8/advancing_the_web_firefox_test_pilot_medium/
https://www.reddit.com/r/firefox/comments/95j92s/firefox_experiment_recommends_articles_based_on/

Some open questions:
- If I am not paying, who pays for processing my browsing history and with what intention?
- What are your thoughts about their privacy policy?
    https://laserlike.com/privacy
    "Corporate Restructuring. — We may share some or all of your Personal Data in connection with or during negotiation
    of any merger, financing, acquisition or dissolution transaction or proceeding involving sale, transfer, divestiture,
    or disclosure of all or a portion of our business or assets. In the event of an insolvency, bankruptcy, or receivership,
    Personal Data may also be transferred as a business asset."
  Especially the last five words are worse than, for example, Google - which is well-funded, regulated and were you just
  need to turn off personalization. Do you think it's realistic that Google wouldn't want to pay anymore
  sometime in the future (unless you have destroyed market share or blind trust before)?
- Are you financially desperate or do you just offering irresponsible users to be analyzed a bit more?
  You have half a billion dollar income, more than ever before. Please don't mess with your brand and keep your identity.
  (Two suggestions for saving a bit money: Why can't Focus just be the Private Mode of Firefox? Why do you have
  extremely wasteful banking conditions in Germany? https://donate.mozilla.org/de/ways-to-give#wire (5 EUR/SEPA transfer)
  Please Move to https://bit.ly/2IpqkTe (0.12 EUR/transfer), Netzpolitik.org is customer there, Moz Berlin will know them.)
- Mozilla Foundation: https://blog.mozilla.org/berlin/wie-privatsphaere-immer-exklusiver-wird-und-zwar-ueberall-auf-der-welt/
  Is privacy becoming a luxury? If the Mozilla Corporation advertises a start-up to us where we can upload all our
  history to develop a new revenue stream, then the answer might be yes. I am missing Mozilla's advocacy for data reduction.

(Please do not post feedback here, visit Reddit or https://discourse.mozilla.org/.)