More than any other change in the car industry, autonomous taxi services would make things massively better. Once you make it cheaper and more convenient for most people to use the taxi service than to own your own car, then:
You can customize your car to the journey. One-person 3km trip? Hail a short-range one-person electric car. No need to drive a big car just because you need it once in a while. This leads to a massive reduction in average vehicle size on the roads, and big reductions in energy consumption.
Most people won't need driveways and garages anymore. Extend your house or put in a garden.
Large reduction in the need for on-street and off-street parking. Replace on-street parking with pickup/dropoff bays and/or bike lanes and/or more general lanes.
On one hand, this would reduce public transport usage by providing a more convenient alternative for many people. On the other hand, this would increase public transport usage by reducing instances where people drive from A to B only because they'll need their car to get from B to C. Both of those cases are wins.
Mozilla doesn't distribute any closed-source binary blobs. Firefox downloads the CDM separately. You can configure it to not do that. There's a "don't play DRMed content" checkbox in preferences; I'm not sure if that stops the download, but it will stop the blob being loaded.
What evidence do you have that Mozilla wants to spy on people?
Mozilla bent over backwards to make it possible for you to run your own Firefox Accounts and Sync servers: https://docs.services.mozilla.... Is that the action of an organization that is determined to spy on you?
Four, actually. (Five if you count Opera.) Three of which are open-source and have reasonably open development models. Plus the APIs are governed by an open-standards process that anyone can participate in.
I'm not sure what the point of your message is actually. Are you arguing that NPAPI was a valuable extension point to let people democratically extend the Web platform? Because in practice it was not; the only NPAPI plugin developers could use in practice was Flash. And Flash was 100% Adobe, so did not lend itself to "freedom and control of your own systems".
I'm not sure what you're implying here. This is about removing a large chunk of very old cruft and replacing it with something much smaller and much more tightly constrained.
Browser vendors have been working very hard on that goal for a long time and are almost as close as they're ever going to get.
DRM is "special" in that Hollywood won't let you play their movies in any full-open-source no-binary-blobs browser. You can opt out of DRM in the Firefox preferences and that gets you as close as you're going to get until you persuade Hollywood to alter their licensing terms.
Mozilla's implementation choices were very good. The CDM sandbox is extremely restrictive --- far more restrictive than the Flash sandbox could ever be, since the CDM needs access to far fewer OS APIs. That's why the move from NPAPI to CDM is good.
Use a Web site to set up a WebRTC peer-to-peer session. I like talky.io, which uses peer-to-peer for one-to-one chats. There are many others, and if you don't like them or don't trust them, you could pretty easily build your own.
The security properties of peer-to-peer WebRTC are pretty good: -- end-to-end DTLS with perfect forward secrecy -- all protocols involved are IETF standards and have had a decent amount of public security review -- Firefox/Chromium implementations are fully open source that you can build yourself and run on Windows/Mac/Linux/Android -- the Web site that sets up the connection could MITM you, but there are many WebRTC sites to choose from and it's pretty easy for anyone to set up more.
I kinda wonder why governments aren't complaining about WebRTC. It's probably just not popular enough yet.
Indeed, Rust needs to make it easier to discover "trusted" libraries for your tasks at hand.
However, cargo makes it very easy to import libraries and manage their dependencies. Rust does not need the usual massive monolithic centrally-managed "standard library" that is versioned with the language.
Chrome and Firefox open PDFs that you browse to in those browsers with their PDF readers, not any PDF you get via email or whatever. So that's less attack surface.
Furthermore, Firefox uses pdf.js which is basically a Web app, so there's almost no additional attack surface over just visiting a Web page... which you were already using Firefox to do.
That doesn't follow. Sometimes you lose market share for reasons entirely outside your control.
Even when you lose market share due to factors under your control, it doesn't necessarily follow that you made bad decisions. You have to make decisions under uncertainty and sometimes the best decisions given the data available don't turn out to be the best in hindsight.
And even when you lose market share because of decisions you should have known are bad, it doesn't necessarily follow that those are the same decisions that some subset of people ranted about on the Internet.
There are many possible causes of marketshare decline, some of which are not under Mozilla's control... Google properties saying "works best with Chrome!", for example. So it's difficult to know how much of the decline is due to, say, browser UI changes.
It's also difficult to know how what would have happened had no changes been made. It's pretty clear that *some* users would only be satisfied if Mozilla had made no UI changes since 2001, which would clearly not be winning.
Slashdot Mirror
More than any other change in the car industry, autonomous taxi services would make things massively better. Once you make it cheaper and more convenient for most people to use the taxi service than to own your own car, then:
You can customize your car to the journey. One-person 3km trip? Hail a short-range one-person electric car. No need to drive a big car just because you need it once in a while. This leads to a massive reduction in average vehicle size on the roads, and big reductions in energy consumption.
Most people won't need driveways and garages anymore. Extend your house or put in a garden.
Large reduction in the need for on-street and off-street parking. Replace on-street parking with pickup/dropoff bays and/or bike lanes and/or more general lanes.
On one hand, this would reduce public transport usage by providing a more convenient alternative for many people. On the other hand, this would increase public transport usage by reducing instances where people drive from A to B only because they'll need their car to get from B to C. Both of those cases are wins.
Because the code that loads the CDM blob is open source. If you don't trust the binaries distributed by Mozilla you can build it yourself.
Mozilla doesn't distribute any closed-source binary blobs. Firefox downloads the CDM separately. You can configure it to not do that. There's a "don't play DRMed content" checkbox in preferences; I'm not sure if that stops the download, but it will stop the blob being loaded.
What are you talking about? Firefox already implemented the relevant WebRTC standards; they're what Hello was based on.
You can disable DRM by clicking one checkbox in preferences.
What evidence do you have that Mozilla wants to spy on people?
Mozilla bent over backwards to make it possible for you to run your own Firefox Accounts and Sync servers:
https://docs.services.mozilla....
Is that the action of an organization that is determined to spy on you?
In other words, "haha I lied"
> enjoy your free, mandatory DRM in the browser.
Not mandatory. Disable it with one checkbox in preferences.
> get rid of pocket
Being moved to a disable-able extension.
> please work on fixing the bugs youre assigned
Of course they're always doing that. To insinuate they don't is just gratuitously insulting.
> hustle up and get that godforsaken voice chat program out of the browser
Already done.
> quit mandating signed plugins to curtail adblock users
No connection between signed addons and adblock. The suggestion there is is something you just made up.
> ditch the targeted advertising tiles
Already done a long time ago.
Firefox does do "on the fly decompression of images so off screen images are not stored in memory uncompressed".
When complaining about Firefox memory usage, make sure your understanding of the issues is up-to-date.
Removing NPAPI in favour of the much more limited CDM interface is a major net reduction in code and attack surface.
Your hate reflex is definitely misplaced this time.
Four, actually. (Five if you count Opera.) Three of which are open-source and have reasonably open development models. Plus the APIs are governed by an open-standards process that anyone can participate in.
I'm not sure what the point of your message is actually. Are you arguing that NPAPI was a valuable extension point to let people democratically extend the Web platform? Because in practice it was not; the only NPAPI plugin developers could use in practice was Flash. And Flash was 100% Adobe, so did not lend itself to "freedom and control of your own systems".
I'm not sure what you're implying here. This is about removing a large chunk of very old cruft and replacing it with something much smaller and much more tightly constrained.
Browser vendors have been working very hard on that goal for a long time and are almost as close as they're ever going to get.
DRM is "special" in that Hollywood won't let you play their movies in any full-open-source no-binary-blobs browser. You can opt out of DRM in the Firefox preferences and that gets you as close as you're going to get until you persuade Hollywood to alter their licensing terms.
Mozilla's implementation choices were very good. The CDM sandbox is extremely restrictive --- far more restrictive than the Flash sandbox could ever be, since the CDM needs access to far fewer OS APIs. That's why the move from NPAPI to CDM is good.
"The browser should be..." Sure! But it turns out that Hollywood won't let you play their movies in such a browser :-(.
If you want Firefox to be that browser, just uncheck "play DRM content" in the preferences and you're all good.
Use a Web site to set up a WebRTC peer-to-peer session. I like talky.io, which uses peer-to-peer for one-to-one chats. There are many others, and if you don't like them or don't trust them, you could pretty easily build your own.
The security properties of peer-to-peer WebRTC are pretty good:
-- end-to-end DTLS with perfect forward secrecy
-- all protocols involved are IETF standards and have had a decent amount of public security review
-- Firefox/Chromium implementations are fully open source that you can build yourself and run on Windows/Mac/Linux/Android
-- the Web site that sets up the connection could MITM you, but there are many WebRTC sites to choose from and it's pretty easy for anyone to set up more.
I kinda wonder why governments aren't complaining about WebRTC. It's probably just not popular enough yet.
Indeed, Rust needs to make it easier to discover "trusted" libraries for your tasks at hand.
However, cargo makes it very easy to import libraries and manage their dependencies. Rust does not need the usual massive monolithic centrally-managed "standard library" that is versioned with the language.
The reason is pretty obvious, and it's not convenience. Microsoft needs to increase Edge usage however they can, and this is one way.
Chrome and Firefox open PDFs that you browse to in those browsers with their PDF readers, not any PDF you get via email or whatever. So that's less attack surface.
Furthermore, Firefox uses pdf.js which is basically a Web app, so there's almost no additional attack surface over just visiting a Web page ... which you were already using Firefox to do.
Other than pdf.js.
If you spent the same resources to fool a human driver, how hard would that be?
"if you aren't experiencing it, it's not a real problem" is a fallacy. So is "if you are experiencing it, it must be a huge problem".
That doesn't follow. Sometimes you lose market share for reasons entirely outside your control.
Even when you lose market share due to factors under your control, it doesn't necessarily follow that you made bad decisions. You have to make decisions under uncertainty and sometimes the best decisions given the data available don't turn out to be the best in hindsight.
And even when you lose market share because of decisions you should have known are bad, it doesn't necessarily follow that those are the same decisions that some subset of people ranted about on the Internet.
And how many of those complaints are about bugs that have been fixed?
There are many possible causes of marketshare decline, some of which are not under Mozilla's control ... Google properties saying "works best with Chrome!", for example. So it's difficult to know how much of the decline is due to, say, browser UI changes.
It's also difficult to know how what would have happened had no changes been made. It's pretty clear that *some* users would only be satisfied if Mozilla had made no UI changes since 2001, which would clearly not be winning.