Identifying the Opera browser unique identification strng

Mad Roger

unread,

Jul 20, 2017, 9:45:18 PM7/20/17

to

The Opera browser "contains a unique ID that is linked to your device"
http://www.opera.com/privacy/computers

On Ubuntu 16.04, is there a way for a non programmer to "trap" this unique
browser ID to see what it looks like?

The ultimate goal will be to spoof the Opera browser unique ID but the
first goal is just to see what it looks like and whether just reinstalling
the app daily will generate a new unique id each time it's installed.

William Unruh

unread,

Jul 20, 2017, 10:07:47 PM7/20/17

to

On 2017-07-21, Mad Roger <roge...@yahoo.com> wrote:
> The Opera browser "contains a unique ID that is linked to your device"
> http://www.opera.com/privacy/computers

Why not just stop using Opera?

Mad Roger

unread,

Jul 20, 2017, 10:50:22 PM7/20/17

to

On Fri, 21 Jul 2017 02:03:45 -0000 (UTC),
William Unruh wrote:

> Why not just stop using Opera?

This is a technical question, so the answer you seek is below, but the
answer doesn't change the need for the technical answer to the question.

That point being made that the question remains a technical question no
matter what the answer to your psychological question is, to answer your
rationale question directly, unfortunately, Opera has value as a rare free
vpn service that isn't blocked by almost all major forums and major web
sites.

By way of complete contract, the just-as-free FF-based tbb exit nodes are
almost always blocked (and tbb is far slower than is Opera).

You might suggest a free vpn combined with ff but most of them seem to be
blocked also by most web sites (security kiss, hidemyass, cyberghost,
etc.).

The Opera privacy policy isn't all that scary
http://www.opera.com/privacy/computers

But this is a technical question so it really doesn't matter why.

It's a Linux question of how to "trap" the unique id transmission or
creation or how to find it inside the Opera binary.

Mad Roger

unread,

Jul 21, 2017, 12:25:51 AM7/21/17

to

On Fri, 21 Jul 2017 01:45:14 -0000 (UTC),
Mad Roger wrote:

> The ultimate goal will be to spoof the Opera browser unique ID but the
> first goal is just to see what it looks like and whether just reinstalling
> the app daily will generate a new unique id each time it's installed.

If it helps, here is how I installed Opera via PPA on Ubuntu:
1. Set up the key:
wget -O - http://deb.opera.com/archive.key | sudo apt-key add -
2. Set up the repository:
sudo sh -c 'echo "deb http://deb.opera.com/opera-stable/ stable non-free" >> /etc/apt/sources.list.d/opera.list'
3. Set up the package:
sudo apt-get update
sudo apt-get install opera

William Unruh

unread,

Jul 21, 2017, 3:11:56 AM7/21/17

to

On 2017-07-21, Mad Roger <roge...@yahoo.com> wrote:

> On Fri, 21 Jul 2017 02:03:45 -0000 (UTC),
> William Unruh wrote:
>
>> Why not just stop using Opera?
>
> This is a technical question, so the answer you seek is below, but the
> answer doesn't change the need for the technical answer to the question.

No it was not "below", unless you meant that the answer to my question
was below, not the OP's question.
...

> The Opera privacy policy isn't all that scary
> http://www.opera.com/privacy/computers

But the fact of the "Unique ID" IS scary. The assurance that this is not
linkedto the uaser is of very little worth, because the browser history,
and their mainainig a database of everything you do with the browser
means that the link between the number and you would not be hard o make.
It would be far more useful for oprivacy if instead they told you how to
opt out of thi unique ID. "just believe us-- we are doing this for your
own good" is not very comforting from anyone. Their privacy policy
basically says "We do not give a damn about your privacy".

>
> But this is a technical question so it really doesn't matter why.
>
> It's a Linux question of how to "trap" the unique id transmission or
> creation or how to find it inside the Opera binary.

I do not believe it is in the binary. To make a unique id for each
person means that they either generate on the fly or do it once and save
it somewhere, perhaps in encrypted form so you cannot find it.

Melzzzzz

unread,

Jul 21, 2017, 3:14:38 AM7/21/17

to

On 2017-07-21, William Unruh <un...@invalid.ca> wrote:
> On 2017-07-21, Mad Roger <roge...@yahoo.com> wrote:
>> On Fri, 21 Jul 2017 02:03:45 -0000 (UTC),
>> William Unruh wrote:
>>
>>> Why not just stop using Opera?
>>
>> This is a technical question, so the answer you seek is below, but the
>> answer doesn't change the need for the technical answer to the question.
>
> No it was not "below", unless you meant that the answer to my question
> was below, not the OP's question.
> ...
>> The Opera privacy policy isn't all that scary
>> http://www.opera.com/privacy/computers
>
> But the fact of the "Unique ID" IS scary. The assurance that this is not
> linkedto the uaser is of very little worth, because the browser history,
> and their mainainig a database of everything you do with the browser
> means that the link between the number and you would not be hard o make.
> It would be far more useful for oprivacy if instead they told you how to
> opt out of thi unique ID. "just believe us-- we are doing this for your
> own good" is not very comforting from anyone. Their privacy policy
> basically says "We do not give a damn about your privacy".

Using closed source blob for web surfing is not wise in any way...

>
>
>
>
>>
>> But this is a technical question so it really doesn't matter why.
>>
>> It's a Linux question of how to "trap" the unique id transmission or
>> creation or how to find it inside the Opera binary.
>
> I do not believe it is in the binary. To make a unique id for each
> person means that they either generate on the fly or do it once and save
> it somewhere, perhaps in encrypted form so you cannot find it.

--
press any key to continue or any other to quit...

Mad Roger

unread,

Jul 21, 2017, 5:06:57 AM7/21/17

to

On Fri, 21 Jul 2017 07:07:53 -0000 (UTC),
William Unruh wrote:

> But the fact of the "Unique ID" IS scary. The assurance that this is not
> linkedto the uaser is of very little worth, because the browser history,
> and their mainainig a database of everything you do with the browser
> means that the link between the number and you would not be hard o make.
> It would be far more useful for oprivacy if instead they told you how to
> opt out of thi unique ID. "just believe us-- we are doing this for your
> own good" is not very comforting from anyone. Their privacy policy
> basically says "We do not give a damn about your privacy".

I'm sorry for saying it's not scary.

My way of answering you wasn't meant to minimize the privacy issue as much
as it was to try to focus you on the first step of figuring out what this
unique id looks like.

We can talk about the issue and wave our arms around until the cows come
home but I'm hoping to somehow trap the unique id to take a look at it -
which is why I'm asking for your help.

I need to know if you know of a tool that we can use on Ubuntu to figure
out what the heck Opera is sending home about each of its users.

Something like wireshark spits out so much that I need help in narrowing it
down to just the communication with the opera servers.

Mad Roger

unread,

Jul 21, 2017, 5:08:07 AM7/21/17

to

On Fri, 21 Jul 2017 07:14:36 +0000 (UTC),
Melzzzzz wrote:

> Using closed source blob for web surfing is not wise in any way...

We can wax prolifically but I don't think anyone knows exactly what Opera
is sending home to the mother ship.

So we'd break new ground if we knew of a tool that we could use to identify
what is being sent home.

Caver1

unread,

Jul 21, 2017, 7:47:23 AM7/21/17

to

On 7/20/17 10:50 PM, Mad Roger wrote:
> On Fri, 21 Jul 2017 02:03:45 -0000 (UTC),
> William Unruh wrote:
>
>> Why not just stop using Opera?
>
> This is a technical question, so the answer you seek is below, but the
> answer doesn't change the need for the technical answer to the question.
>
> That point being made that the question remains a technical question no
> matter what the answer to your psychological question is, to answer your
> rationale question directly, unfortunately, Opera has value as a rare free
> vpn service that isn't blocked by almost all major forums and major web
> sites.
>
> By way of complete contract, the just-as-free FF-based tbb exit nodes are
> almost always blocked (and tbb is far slower than is Opera).
>
> You might suggest a free vpn combined with ff but most of them seem to be
> blocked also by most web sites (security kiss, hidemyass, cyberghost,
> etc.).
>

If you did some research Opera's free VPN isn't a VPN at all, it's more
of a proxy.

--
Caver1

Mad Roger

unread,

Jul 21, 2017, 11:25:52 AM7/21/17

to

On Fri, 21 Jul 2017 14:52:17 +0200,
Yrrah wrote:

>> On Ubuntu 16.04, is there a way for a non programmer to "trap" this unique
>> browser ID to see what it looks like?
>

> FYI, there is a Ubuntu newsgroup: alt.os.linux.ubuntu

Thanks for the admonition that there is a ubuntu-specific ng.

It might have been best to originally crosspost to a.o.l & a.o.l.u; but
there's little utility in starting a new thread because there is no
indication that this is an operating-system specific quest.

Given that Opera, on all operating systems, generates and transmits a
unique id to someone somehow, where would you hazard a guess as to
specifically how the unique id is generated (so that we may spoof its
parameters perhaps) and where the ID is stored (so that we may look at it
perhaps) & in what form do you think it's transmitted & to whom (so that we
may watch and intercept it perhaps)?

Mad Roger

unread,

Jul 21, 2017, 11:25:53 AM7/21/17

to

On Fri, 21 Jul 2017 07:47:16 -0400,
Caver1 wrote:

> If you did some research Opera's free VPN isn't a VPN at all, it's more
> of a proxy.

Where do you think the ID is stored & in what form do you think it's
transmitted & to whom?

I did plenty of research (but I am not an expert!), where four things are
of note, two of which everyone knows (so it has no value in this thread),
and two of which are not posted anywhere in the web so you could sort of
say only we know.

Everyone knows this:
1. Opera is a fast free unlimited (pseudo)proxy service for web browsing
2. It's based on Chromium (which may mean it uses Chrome-like unique iDs)

Only we seem to know this:
3. Opera is one of the rare (psuedo)vpn's that work with most web forums
4. There must be a way to trap & identify the unique install ID assigned

If this thread was about the first two items, the thread wouldn't exist.
The thread is about the latter two items, specifically that last item.

If that last item were obvious, then it would have been resolved already.

The question at hand is how to trap that unique id with the end goal in
mind being to spoof it eventually; but you can't spoof that which you can't
see.

There's either a unique static binary ID, or a static config file ID, or
the unique ID is generated on the fly.

Where do you think the ID is stored & in what form do you think it's
transmitted and to whom?

Michael Black

unread,

Jul 21, 2017, 1:41:25 PM7/21/17

to

On Fri, 21 Jul 2017, Mad Roger wrote:

> On Fri, 21 Jul 2017 14:52:17 +0200,
> Yrrah wrote:
>
>>> On Ubuntu 16.04, is there a way for a non programmer to "trap" this unique
>>> browser ID to see what it looks like?
>>
>> FYI, there is a Ubuntu newsgroup: alt.os.linux.ubuntu
>
> Thanks for the admonition that there is a ubuntu-specific ng.
>
> It might have been best to originally crosspost to a.o.l & a.o.l.u; but
> there's little utility in starting a new thread because there is no
> indication that this is an operating-system specific quest.
>

But this has nothing to do with the repair of electronic devices, which is
what sci.electronics.repair is about.

I'm tired of fools who somehow think talking about computers fits in here.

Unless your problem can be fixed by replacing a component, then it doesn't
belong here.

Michael

William Unruh

unread,

Jul 21, 2017, 3:10:07 PM7/21/17

to

On 2017-07-21, Mad Roger <roge...@yahoo.com> wrote:

> On Fri, 21 Jul 2017 14:52:17 +0200,
> Yrrah wrote:
>
>>> On Ubuntu 16.04, is there a way for a non programmer to "trap" this unique
>>> browser ID to see what it looks like?
>>
>> FYI, there is a Ubuntu newsgroup: alt.os.linux.ubuntu
>
> Thanks for the admonition that there is a ubuntu-specific ng.

I am virtually certain that there is nothing special about ubuntu in any
of this. tools available on Ubuntu are also there on other versions of
Linux. This is about opera (which operates on all versions of linux).

>
> It might have been best to originally crosspost to a.o.l & a.o.l.u; but
> there's little utility in starting a new thread because there is no
> indication that this is an operating-system specific quest.
>
> Given that Opera, on all operating systems, generates and transmits a
> unique id to someone somehow, where would you hazard a guess as to
> specifically how the unique id is generated (so that we may spoof its
> parameters perhaps) and where the ID is stored (so that we may look at it
> perhaps) & in what form do you think it's transmitted & to whom (so that we
> may watch and intercept it perhaps)?

This is not a question in which wild guesses are going to be of much
use.

IF you know where the uniqID is sent to you could use tcpdump with that
IP as the host. But since we have no idea where it is sent to, that is
probably pointless.

You have the source for Opera, and you could look through that to find
if there are any indicators to "unique ID"

Mad Roger

unread,

Jul 25, 2017, 11:01:51 AM7/25/17

to

Just starting Opera to a blank page makes connections, depending on the
setup, where the current set of connections includes a few things that we
can discount, such as an autoupdate of some sort, and speedials and some
search engine updates.

Slowly I'm turning off each of these below, using the Opera settings,
and/or the /etc/hosts file so that I can narrow down what it connects to
for sending the Subscriber-ID (from Opera Inc) and Device-ID (from
SurfEasy).

# Result Protocol Host URL Body Caching Content-Type Process Comments Custom
37 200 HTTP Tunnel to autoupdate.geo.opera.com:443 0 opera_autoupdate:10512

# Result Protocol Host URL Body Caching Content-Type Process Comments Custom
38 200 HTTP www.ecb.europa.eu /stats/eurofxref/eurofxref-daily.xml 1,624 text/xml opera:7576

# Result Protocol Host URL Body Caching Content-Type Process Comments Custom
24 200 HTTPS speeddials.opera.com /api/v1/suggestions?country=US&language=en-US&uuid=ae57db72-45af-4547-8714-4d0686940320&type=desktop-suggestions 1,496 max-age=3600 application/json opera:9080

# Result Protocol Host URL Body Caching Content-Type Process Comments Custom
5 304 HTTP crl.globalsign.net /root-r2.crl 0 Expires: Tue, 25 Jul 2017 13:23:06 GMT svchost:1668

# Result Protocol Host URL Body Caching Content-Type Process Comments Custom
6 304 HTTP pki.google.com /GIAG2.crl 0 Expires: Tue, 25 Jul 2017 14:45:30 GMT svchost:1668

Still working on it though so take this with a huge grain of salt because
I'm only looking at the http protocol right now (so that I can narrow down
the data to search through).

Mad Roger

unread,

Jul 26, 2017, 8:05:40 PM7/26/17

to

On Tue, 25 Jul 2017 15:01:45 -0000 (UTC),
Mad Roger wrote:

> Just starting Opera to a blank page makes connections, depending on the
> setup, where the current set of connections includes a few things that we
> can discount, such as an autoupdate of some sort, and speedials and some
> search engine updates.

Just for the record, here are the startup calls Opera makes to the search
engines, even with everything possible in the settings turned off.

These are in order, with, I think, all but the last two not being
important, I don't think, for our purpose, and hence can be filtered out.