Browser Extension - Swash - Streamr Dev Community Forum

Browser Extension - Swash

peter

Project goal: integrate user data generated via the Firefox browser with Streamr pub/sub and Marketplace.
Team composition: three team members working on it as a side project, with experience in software security and full stack development.

My teammates(Mehrdad & Vahid) and I are planning to build, initially, a Firefox extension to allow users an easy way to monetize their internet browsing data anonymously. The overarching goals for the project are:

Create a simple to use browser extension/plugin that most internet users would be able to install without too much effort Empower users to have a stake in their data, by giving them a way to monetize what it is righteously theirs Build it with user privacy as a core focus and allowing users full control of the level of privacy applied to their shared data Integrate with Streamr tech stack and Marketplace to offer a monetization platform and allow an automated, decentralized and fair distribution of revenue.

This extension has been designed in a way that users have full control of what would be sent to Streamr and they can easily choose which data is permitted to send by leveraging consent, filter, and, privacy capability of this extension. Also, we considered a modular engine for this product to make it easy to extend and update. Currently, it supports six modules:

Amazon Module: This module look through all the user activities on Amazon and capture those activities that users have permitted
Facebook Module: This module look through all the user activities on Facebook and capture those activities that users have permitted
Search Module: This module captures all user searching queries
Surfing Module: This module captures visited pages and bookmarks created by user
Twitter Module: This module captures users activities on twitter
Youtube Module: This module captures users youtube activities

We belive that transparency is a key point for success of our product, so we are going to keep the project open-source when it would be ready for release.

Edits by Wei: added images below for reference.

peter

We developed a backend application for the extension to let all website/product owners create their own module for the extension. Newly created modules can be added to the extension from the marketplace menu. By this new feature, all end-users can search in the extension marketplace to find newly created modules and add them to the extension to earn Data Coin by sending website/product owners requested data to Streamr marketplace.
In this new business model, the owner of a module is the buyer of data. We tried to identify interesting data for website/product owners and at this step, we can let website/product owners collect these type of data:
1- web site errors, warnings, and info logs
2- web site network traffic
3- Task Results(Task is a combination of actions on a web application that is defined by a website owner and asks end-user for completing the task)
4- website/product survey
5- user context attributes
All these data types are helpful for troubleshooting, creating a better UX, identifying unknown errors and bugs and zero-days vulnerabilities

tomhamilton

peter Can't wait to see it in action!

peter

Thank you tomhamilton

Remy

That seems awesome! looking forward to try it out 🙂

peter

New Update for our Community Product:
After getting some interesting feedbacks from @wei we improved the extension user experience by developing new features. Our extension comprises a number of modules. Each module is a combination of some relative data collectors that collects user browsing activities. In this new update, we added these features:

Tooltips for collectors: on collectors mouseover, a short description will appear.
Privacy enforcement guide modal: to show the user who privacy enforcement affects data sent to the marketplace we created a sample privacy enforcement tools
Surfing module whitelist: Surfing module collects the user visited pages and pages visiting graph. In the previous version, it had been collecting data for all URLs. In this version, the user can create a white list for this module to let it collects data just for some specific URLs.
Time picker for delay message sending: there is a feature in the extension that let the user make delay on sending data to the marketplace. We added a time picker to let the user choose how long it would make delay.
Progress bar for delayed messages: the user can see the percentage of time is remaining before the messages be sent to the marketplace
User identity privacy model: by installing the extension, a unique identifier will be generated for each module(mId), and a unique identifier will be generated for the extension(id). In each message is sending to the marketplace, there is a parameter for identifying the user: this identifier will be generated based on the privacy level of the module that generated the message:
Lowest privacy level: userid = hash(id)
Low privacy level: userid = hash(mId)
Medium privacy level: userid = hash(mId || date)
High privacy level: userid = hash(mId || date || hours)
Highest privacy level userid = hash(mId || timestamp)
Ads search result: We added a collector to the Searching module to collect ads search results
Ads click links: We added a collector to the Searching module to collect ads clicked by the user
Rank for search result and clicked links: We added a rank property to each item of search result and each clicked links

peter

New Updates:
I welcome Masoud as our new teammate. He has a strong background in DevOps and Security Standards auditing. At this time he helps us to develop backend application for surf-streamr(UI & Services).
I'm delighted to tell you that we could successfully port the extension to the chrome. We built a wrapper around firefox webextension APIs that automatically converts webextension APIs to chromium APIs. We have also extended the Search module to support DuckDuckGo search engine. @wei

Rdodge3

peter how soon can we get access to this extension? Are you thinking of creating an open source or white label version of it as well?

Remy

peter That sounds great Peter. Good to hear that!

Remy

Hi peter, this post is more to give some transparency to all readers on the amount of tokens you and your team receive for the awesome extensions for Firefox and later on Chromium browsers!

All the community fund admins talked about the funding part which was not clear yet before. The current funding was decided upon 350K DATAcoin for the Firefox plugin and the Chromium plugin.

As i understood from @tomhamilton and @wei, is that the first half (175K DATAcoin) will then be transferred to you and your team very soon. The other half would then be sent upon completion of the Chromium Plugin.

If you like, you can message me the wallet address you like to receive the DATAcoin on. Or send it to Tom or Wei ofc 🙂

we can’t wait to try it out 🙂
good luck!

peter

Rdodge3 I'm preparing a beta version. After finishing Streamr Core App Beta test, we release the first beta version. We are planning to make it open-source after publishing it in firefox marketplace.

Rdodge3

peter That would be great!! Looking forward to it!

Ana

Will the tokens be earned without having to work through an intermediary? This is my chief criticism of Brave, which requires the user to sign-up with and verify their identity via an uphold account, making the project far more centralised and rendering the user subject to platform risk (such as having their Uphold account cancelled without reason or redress). By not following this route you would have a significant competitive advantage over Brave, which has received a fair bit of criticism over this issue.

Remy

Ana My understanding is that that is not necessary. Users will be able to enable the data that they want to share. This will get published in a Data Union on the marketplace where it can be bought. The user is the only one that makes the decision if they want to participate or not. If the Data Union is then bought - you as a participant will earn from it.

But i could be mistaken ofc. What is your view on this @wei?

Shiv

Heya @Ana !

So the user journey for a data producer (an ordinary person in most cases) should look something like this.

So let's break this down:

They go to the app which they either have a hardware of software interaction with. Something like say Fitbit or Spotify. Or it is a piece of software which say is run and operated locally like the Firefox plugin.
The user then opts to push the real-time data which they are creating/generating through those interactions, to somewhere else (the Streamr Marketplace). In order to do this in most apps, they will need some kind of permission from that app. Either that will be offered through a third party integration (like say the 3rd party Spotify demo we have here or through a native integration on the app. If it is native then it might look something like this: Spotify has a "send your data to Streamr" section into its existing app. With the Firefox plugin, since Firefox doesn't control your browser, nor what plugins you can install, then there is no external permission needed.

(There's another section here I could write about why we believe GDPR will force all companies into permissioning this data i.e. they won't be able to say no).

Now once you get past that stage, you will be sending that data to a type of digital Data Union, we're calling a Community Product. For v1.0 of Community Products, these will be created and run by "admins".

Who are admins? Admins will have significant control over "their" products. Their job will be to; ensure that the data is useful and that malicious users aren't trying to game the system, market that data to buyers, create larger data sets by appealing to more users. For this, admins will be able to take a cut of the sale of a Community Product. So from the perspective of an ordinary user, the ordinary user could be kicked off by a product admin because that admin no longer finds the data which the user is generating useful/genuine etc.

Here's an exclusive peak btw at the product flow creation for admins.

However there is no reason why other Community Products can not be established and set. And, even more excitingly, in something like v2.0 of Community Products we are hoping to integrate governance tooling from the Colony team so that ordinary users and admins can have a far less hierarchical relationship. This is really exciting because you end up with data cooperatives where data producers, data managers, and data buyers start working together to create value for everyone in digital systems that really take account of those different interests and can scale.
Of course Streamr's Marketplace will be decentralized - open and permissionless - so admins, buyers and users can use the system without fear that a single central actor will be able to shut them down.

Finally when it comes to earning tokens - phew I finally got there! - the data producers who are part of a Community Product will get the tokens directly from the buyers of those products the instant a payment is made. It will not via Community Product admins, Streamr, or anyone else. The payment system is bespoke and is called Monoplasma. You can read more here and you can watch Henri Pihkala talk about it here.

I hope that answers the question! @Ana

Ana

@Shiv Thanks for the comprehensive breakdown.

peter

Hi Remy, I really thank you and other community fund admins. My teammates and I appreciate it a lot. Personally, it's my honor to be a small part of a movement(or maybe a revolution, @Shiv should help me to choose the right word 😀 ) to place power and value back into the hands of data owners. 🙂
Btw we've tested the extension on chrome and brave, and it works well. 🙂

Remy

Awesome! Good to hear that about the chromium extension too! Could you perhaps also test it on the just newly released Microsoft Edge browser (chromium version - https://www.microsoftedgeinsider.com/en-us/). Not sure how many people will use that version, especially at the beginning but I think it would be very powerful if eventually we could say at release that 4 browsers are supported 🙂

peter

Remy