On Android, although a built-in isolatedProcess API [1] is available for them to use, there is no sandboxing. No sandboxing on the web in 2025 (!!!). This has been an issue for so many years, yet Mozilla refuses to address it [2]. Chromium does do proper sandboxing on Android, and additionally restricts what syscalls a process can access. Other alternatives, such as Vanadium have even stronger sandbox implementations [3]
On desktop, it's a similar story. Site isolation has had numerous bad issues that haven't been fixed for many years [4][5][6], and especially the Linux builds have had bad sandbox escape vulnerabilities that Chromium is not susceptible to. This is mostly due to architectural differences, like [7] and [8].
The idea of someone being able to take over your computer by just visiting a site is scary. It's beyond me why Mozilla does not prioritise security over yet another sidequest that will slowly bankrupt them.
[1] https://developer.android.com/guide/topics/manifest/service-...
[2] https://bugzilla.mozilla.org/show_bug.cgi?id=1565196
[3] https://grapheneos.org/usage#web-browsing
[4] https://bugzilla.mozilla.org/show_bug.cgi?id=1505832
[5] https://bugzilla.mozilla.org/show_bug.cgi?id=1484019
[6] https://bugzilla.mozilla.org/show_bug.cgi?id=1707955
Your complaints about Android are valid (I should know, I used to work on trying to get Android sandboxed), but site isolation on desktop has been out for a long time.
Respectfully, posting a bunch of bug numbers whose context you aren't familiar with is not a valid representation of the state of things.
"Young people are just smarter" and so on...
2. Canadian Liberals aren't US MAGA, when they win an election they don't spend six months in caucus to figure out how they can do their best to punish the provinces and people that didn't vote for them.
There's a lot of far-right propaganda in Alberta that implies #2 is happening, but it's not actually factual. Its oil & gas sector has reached record output under the Trudeau government, and Carney is not exactly looking to kill it, either.
Transfer payments are really the only legitimate grievance Alberta should have with the federal government. All of its other problems are either imagined, self-inflicted, are caused by other provinces, or are caused by the US.
I'm going to gently push back on that one a bit. Partially, yes, but also in part due to the federal government deferring to provinces in cases where it actually has the constitutional authority to override them.
— Firefox OS would've worked
— Mozilla leadership has always been interested in chasing fads
— Letting smart, motivated people work on cool shit is necessary to create the future.
cries in Bell Labs
I think the fundamental problem then (and still a problem today) is that Mozilla only ever had the budget to do Desktop XOR Mobile. Not both.
They basically mortgaged Desktop Firefox to build Firefox OS, but when things didn't pan out was quickly as they liked, they had to revert back to Desktop to save the cash cow.
After that experience, upper management became averse to mobile, to the extent that it was (and probably still is) very difficult to get resources allocated to mobile implementations of Gecko features that are supposed to be cross-platform. In practice many of those features are never implemented with mobile in mind.
That said I think Mozilla were right to try. The phone ecosystem we live in today is locked down and tightly controlled by tech giants. Leveraging web tech to bridge that gap made sense, even if it had drawbacks. But it was a competitive environment where they needed to nail absolutely everything first time and just couldn't pull it off.
I was at Mozilla during this time, and I remember a huge blocker was WhatsApp. They were not interested in porting, but at the time it was essentially the killer app in the same emerging markets that Firefox OS was targeting.
It blows me away an issue like this could take weeks to track down. If I were in any leadership position at this company I'd be rolling heads with the lack of telemetry or domain knowledge for these systems.
Concurrent programming has been mainstream for some time now, but I don't think the level of expertise of most engineers has kept up. That becomes most apparent when software starts hitting concurrency pitfalls: performance problems, deadlocks, UAFs, and so on...
At this point I plan to donate to Ladybird instead. Excited for that project.