[Feature Request] Add option to toggle off news fetching
See original GitHub issueFeature Request
Is this feature request related to a problem?
DarkReader silently sends a POST request to darkreader.github.io
which is redirected to darkreader.org
a proprietry website as far as I know.
This poses a privacy issue, For the sake of the argument lets suppose that the owner of darkreader.org
decides to take nefarious actions and track users without consent,
In that case, I could not know about that unless im told to.
So as far as things currently are, DarkReader could be classified as a spyware, unfortunately. It collects user’s data (including IP which is not anonymous).
Screenshots
Location in code: ./src/background/newsmaker.ts:56
https://github.com/darkreader/darkreader/blob/0097a6d4c32fae10f13bbe8236d89e53a25191a8/src/background/newsmaker.ts#L56
Additional Context
I could suggest few options:
- Transparency over
darkreader.org
backend - Toggle switch option for connecting to
darkreader.org
- Bonus: disable news fetching by default and issue a request only upon clicking “Get news!” or “What’s new?”
A similar discussion took place on Etcher’s github page: https://github.com/balena-io/etcher/issues/2977
Please enlighten me if there was any mistake in my statements. I’d like to amphasize that im a big fan of the work done on this project, it is really outstanding and i wish for it to become much more popular, please give users the privacy that they deserve.
EDIT: Typo and link to Etcher issue
Issue Analytics
- State:
- Created 2 years ago
- Comments:7 (3 by maintainers)
Top GitHub Comments
@codeswhite I’m pretty confident that commit which makes news fetching optional (#4472) will be included in the upcoming Dark Reader release.
@codeswhite wrote:
Browser extensions are in a sense the worst XSS threat that that web has ever faced. If you installed an extenison with
<all_urls>
andtabs
permissions, than you have a much bigger problem on your hands than a benign server ping for recent blog post titles.I glanced over that PR and it looks reasonable. It just needs to be updated to use
chrome.alarms
instead ofclearInterval
andclearInterval
. If you have time to resolve merge conflicts and migrate tochrome.alarms
, I can do a proper review, test this manually and add unit tests.Also, we should make sure Dark Reader does not store extra uniquely identifiers like user id/session cookies, uniquie HTTP ETags, etc. The fetched URL contains
(new Date()).toISOString()
: https://github.com/darkreader/darkreader/blob/0097a6d4c32fae10f13bbe8236d89e53a25191a8/src/background/newsmaker.ts#L56 Is this really necessary? Off the top of my head, I do not see any legitimate reason for it. If user’s and server’s time is accurate, this value should match. If user’s time is not accurate, it might be somewhat uniquelly identifying. If this is used to force cache off, it is probably redundant.