Hacker News Reader

I thought that they introduced.

Sandboxing for individual tabs

Ability to drag tabs between windows

Isn't each tab a process or something so when you close it memory is freed?

I am sure there are some other things. This is all I can think of off the top my head.

One process per tab is not a feature. It is (or should be) invisible for me, the user, and from my experience it creates more problems than it's worth (first time I tried Chrome it simply stopped my computer dead by running 10 heavy processes on a single-core CPU - so much for performance). It's a bad design decision that will only pay off when I have a 20-core CPU. Sandboxing for individual tabs? Just about the same.

My point is that saying "Chrome came and shook things up introducing features no existing browsers had" is a bit of a stretch. When it came out Chrome was basically not very different from Internet Explorer 6 (feature-wise), and it seems to have stayed that way (I like full page translation but that's it).

We should be fair and say that Google Chrome has a large market share because it has Google behind it. It doesn't have much else going for it.

Back when I switched from Firefox to Chrome, FF was a very heavy browser that took up a lot of memory. Chrome was very light-weight and nimble. Clearly, that has changed since then but it was a big competitive advantage.

One process per tab is a feature. If a tab crashed in FF, typically the whole browser would need to restart whereas in Chrome the tab could be closed and re-opened without affecting the other tabs. I think that's a pretty significant feature.

Auto-updating every 6 weeks is another feature. As a developer, I don't have to worry when an update comes out since I know 90%+ will be on the new version.

While I think part of the reason Chrome has a large market share is clearly because of Google, I don't think it's the main reason. There are many other factors.

> I think that's a pretty significant feature.

The problem here is that you trade a lot of performance for stability that you should have to begin with, so in the end you have worse overall performance just so that a fringe case is covered. What's the worse that can happen if Chrome crashed? You could lose all your tabs because the default setting is "open homepage" instead of "reopen the pages that were open last", but that's your fault.

But let's investigate this more deeply - browser crashes have two major causes - internal bugs and plugins. What used to happen was that Flash (or Acrobat Reader, or whatever) would crash and it would bring down the whole browser. The solution is (of course) one separate process for the plugins. Problem solved. Going beyond that and fitting every tab with its own process to account for the fact that your business isn't really building browsers is of little comfort to me; you shouldn't be dereferencing that NULL pointer to begin with.

Opera has auto-update as well. I guess we'll have to disagree on the reason Chrome has its market share.

Opera 7 supported dragging tabs between windows back in 2003.

omnibox

I'll use this page as reference for omnibox - http://support.google.com/chrome/bin/answer.py?hl=en&ans... .

Opera has custom inline searches. I type in the address bar "eng cacophony" and I am being redirected here www.thefreedictionary.com/cacophony . "w Opera browser features" takes me to the wikipedia page. I can define how many custom keywords as I please. Again, long before Chrome existed.

I could do this in the year 2000.

Firefox did that too. It's very nice but not quite the same as Omnibox. Also, it's a bit of a superuser feature IMO, with Omnibox being more accessible.

Fast JS (it took Firefox forever to catch up) and SPDY.

Making Javascript a first-class citizen can't be understated. Before V8 came along, Javascript was good for doing form validations and not a lot more. Chrome sparked a Javascript engine arms race that has tremendously benefited web developers and users.

That's nice and everything except now they want to kill it and replace it with Dart. Basically Google hated Javascript (and rightfully so) but used it to gain market share, and now they can't really get rid of it. To add insult to injury, now you also have server-side Javascript.

They stopped WebKit from being owned by Apple. Webkit supports lots of new CSS3 & HTML5 stuff, and for a while was the only way to use lots of the CSS3 stuff. This means you can get advantages of the modern web without relying on an Apple browser.