3 ms·
As a counterpoint, this does make it so that one group has disproportionate power over what features make it into that engine, or how they are implemented. What
by lucasban 1y ago
As a counterpoint, this does make it so that one group has disproportionate power over what features make it into that engine, or how they are implemented. What if their incentives change over time and are no longer aligned with what we might consider the success of the web?
- charcircuit 1y agoThen it can be forked because it's open source.
- cosmic_cheese 1y agoThe problem with forking Blink/Chromium is that in order to be able to counter Google, the organization maintaining the fork is going to need dev manpower on the order of Google’s to be able to keep up with upstream patches, which is prohibitively expensive for all but a handful of orgs (not to mention, skilled talent capable of working on web engines doesn’t grow on trees). Without that any fork that differs substantially from mainline is eventually doomed as the divergence grows and overwhelms the team.
- charcircuit 1y agoIt does not take the same order of engineers for only integration. That is false as you can see by the existing integration teams for forks that exist. And I'm sure Mozilla is able to find talent capable of working on web engines.
- cosmic_cheese 1y agoThere are no Blink forks with appreciably large differences yet, though. In the aftermath of Google turning “evil” and working against the better interests of the web in multiple ways, you’re looking at a fork with divergences as large or larger than those that prompted Google to fork Blink from WebKit, making integration of patches from mainline Blink a full time non-trivial job. Personally I’d rather see Mozilla working on Gecko or maybe consider switching to Servo or something instead.
- charcircuit 1y agoThe software is very modular. I don't see what could make it so hard to remove "evil" modules.