anewguy9900

to be less disingenuous, the review process seems to be limited to looking for specific known attack vectors, rather than a full review or evaluation of the sourcecode (which would be impractical). that said, yes, someone at mozilla does at least eyeball it

nightlight is "dynamic" by using a simple heuristic after pageload to deinvert if necessary. but the crowdsourced filter is a great idea! might borrow it :)

does your extension run on android? there are two motivations for nightlight aside from its obvious purpose:

1) be fast on android 2) keep source code simple as possible so everything the extension does is obvious. when i wrote nightlight it was difficult to evaluate the safety of the other extensions at the time, something important to me for a plugin which has full access to all browsing

i made a similar extension some years ago -- it tries to be fast and injects very little code. i'm curious to know how the performance compares:

https://github.com/conceptualspace/nightlight