Browsers are pretty complicated, yes. Things like low-latency high-performance VMs, hardware-accelerated video pipelines, plus the details, like actual HTML parsing, CSS layout, a network stack, and so forth. Also, what matters is not just the complication but how fast you're trying to change things, and people are adding new things (flexbox, more complicated CSS layout modes, mode DOM APIs, etc) faster than ever before.

But also, in addition to a browser Mozilla is working on FirefoxOS, which involves a whole separate bunch of developers, since it's not like the browser developers are writing things like the dialer app for FirefoxOS. Also, you need QA, not just developers.

And yes, Mozilla has 1000-ish employees, for what it's worth.

It's not just Mozilla. If I look at https://www.openhub.net/p/chro... I see on the order of 600 committers with commits in the last month. And that's not even counting whoever is working on the non-open-source parts of Chrome. And not counting, again, QA and so forth.

And the worst part is, this is not a new development. Microsoft had over 1000 people working on IE6 in 1999, according to http://ericsink.com/Browser_Wa...

So yes, browsers, complicated.

The "let" keyword is not the same thing as "let blocks" and "let expressions".

The keyword looks like this:

    let x = 5;

and is in ES6. A let block or let expression (neither of which is in ES6) looks like this:

    let (x = 5) alert(x);

so that "x" is only in scope for the duration of the let block. It's syntactic sugar for:

{
    let x = 5;
    alert(x);
}

> So you still have to buy an iPhone, an iPad, an
> Android phone, and an Android tablet to test on them,

Sure. The point here is to allow you to use the devtools of your choice, not to create a test environment.

> So, they're running Android and iOS on your
> computer to run the same binaries as those
> platforms?

No. "They" are allowing you to connect your Android or iOS device to your computer (likely via USB), then debugging the on-device browser using the Firefox debugger running on your computer. That way you're debugging the thing you actually want to debug, but using the same developer tools you're using for your other debugging, and which therefore you're already familiar with.

He also (correctly) warned about the entanglement of government and scientific research, in the same speach as the military-industrial complex warning. It's funny that people remember one warning, but not the other.

Ukrainian tanks don't have reactive armor, as the article points out.

And sure, no one is suggesting launching nukes at Russia based on the evidence we have right now.

That's pretty normal for press coverage, for what it's worth: they have to fill up space, so will throw in unrelated pictures all the time...

Getting clear close-ups of stuff in a war zone is hard, of course, especially if the stuff is being hidden.

http://www.bbc.com/news/world-... has some photographs if you care.

> [citation please]

http://www.charlesmann.org/art... has a good summary.

CrimsonAvenger's point was that we've had evidence since the early 1800s that humans (and probably other hominids, in fact) ate mammoths. Nowhere did he say that humans were eating mammoths in the 1800s.

It's really not that complicated. Firefox releases work like this: 6 weeks of development, 12 weeks of testing and stabilization (split up into two 6-week phases called "aurora" and "beta"; the former corresponds more or less to feature freeze and the latter more or less to "code freeze unless we discover a stop-ship issue"), then release.

So right now 31 is released, 32 is beta, 33 is aurora, and development is happening on 34.

I suggest you take that browser from the old days, run it on today's web sites, and see how many hundreds of MB it takes. Assuming it loads them at all.

It's a matter of funding.

Looking at the chart at http://en.wikipedia.org/wiki/F... and in particular the inflation-adjusted line there tells you pretty much what the story was: at the peak of the Apollo program NASA's budget was about $40 billion/year in today's dollars (the red line in that graph is in 1996 dollars). NASA's budget today is less than $18 billion/year.

Or to put it in relative-to-the-economy terms, in 1966 NASA was 4% of Federal budget expenditures. 4% of the 2013 US expenditures (actual, not requested) would be $138 billion, according to http://en.wikipedia.org/wiki/2...

I bet if you funded NASA at that level (even just the inflation-adjusted one; I understand that the overall budget structure is quite different now from what it was in 1966, so the $138 billion number is pretty much meaningless), I bet it could produce results a lot quicker than it can at current funding levels...

That would take a lot more development effort, since plug-ins depend on a lot of functionality being present in-process with them that's based on libraries that make up a good bit of that 54MB.

On Mac, the plugin process is the same binary as the 32-bit Firefox process...

You mean have it download both versions, on 64-bit, right? It's not a matter of choosing: you need a 32-bit process to run the plug-ins in, and a 64-bit one for the actual browsing.

This is doable, and being worked on; it's just not been a top priority for various reasons.