3 ms·
> If chromium is open-source, why not strip it from telemetry and other “badness” and use it as a base to build a browser That's exactly how qt-webengine is pa
by coffeecat 5y ago
> If chromium is open-source, why not strip it from telemetry and other “badness” and use it as a base to build a browser
That's exactly how qt-webengine is packaged. And qutebrowser (based on qt-webengine) is fantastic for a volunteer effort, but memory leaks and other issues, too much stuff for one volunteer to realistically address, make it difficult to use as a daily driver. And that's just building a UI around an existing rendering engine!
The root of the issue is: W3C specifications are >100 million words, and growing at a rate of hundreds of new standards per year. So even if the Chromium code base were forked from its current state, a team of full-time developers would be required just to keep up with new W3C standards.
What I find tragic about the Mozilla mismanagement is, if one guy is willing to develop and maintain qutebrowser as a volunteer, then surely Mozilla can find developers who would be willing to work on Firefox at sub-FAANG-but-reasonable salaries without the waterfront San Francisco office.
- wruza 5y agoI see. But did anyone try to fix that >100M words issue? Is there even a point to have a standard that only a couple of direct-profit bigcorps can implement in reality? Even MS could not and gave up, as I understand it.
- arp242 5y agoThe standards you need to implement a functioning web browser are far less than ">100M words". W3 has published a lot of stuff in the last 20+ years, but you don't need much of it.