Slashdot Mirror
Google Chrome, the Google Browser
Philipp Lenssen writes "Google announced their very own browser project called Google Chrome — an announcement in the form of a comic book drawn by Scott McCloud, no less. Google says Google Chrome will be open source, include a new JavaScript virtual machine, include the Google Gears add-on by default, and put the tabs above the address bar (not below), among other things. I've also uploaded Google's comic book with all the details (details given from Google's perspective, anyway... let's see how this holds up). While Google provided the URL www.google.com/chrome there's nothing up there yet."
Found the comic link: http://blogoscoped.com/google-chrome/
I believe you're confused as to what "404 Not Found" means. It means the page you're looking for isn't there, not that the server is overloaded or can't handle the request. It's not slashdotted.
However, this is not Google's normal 404 page. They've definitely configured www.google.com/chrome differently than the rest of the site, so they're obviously planning to put something there.
$x='S24;r)>63/* h@<5+oZ)32"5cz';$me='phroggy'x$];
$x=~y+ -xz+\0-Tx+;print$_^chop$me for split'',$x;
Apple chose KHTML as the foundation of WebKit for the size and quality of the codebase compared with Gecko, despite having Gecko experts working on the project. It makes sense that others would choose WebKit for the same reasons.
Bogtha Bogtha Bogtha
They're not building the whole thing, but it's a bit more than just a rebranding. They're using Webkit (Safari, Konqueror) rather than Gecko (Firefox), but adding a new Javascript engine and UI, and building in Google Gears.
Chernobyl 'not a wildlife haven' - BBC News
You've never used Opera have you?
Default look is tabs (well, more like mini windows unlike binder tabs) over the adress bar. =/
Google doesn't pay Mozilla because they like firefox. They pay because Mozilla drives millions of hits to Google's search engine. As long as firefox is doing that, Google will pay (although, I'm sure they will only freely advertise their own browser now).
"Knowledge is the only instrument of production that is not subject to diminishing returns" -Journal of Political Econom
It won't in the medium-term, because Google just extended its investment in Mozilla through 2011.
The Rise and Fall of Online Community
ECMAScript 4.0 is a syntactical change. It does not offer new features like multi-threading. Multi-threading is more the domain of the WHATWG APIs. There has recently been quite a bit of talk over a "Worker" API that would allow threads to be spun off into the background.
FWIW, Gears is effectively a Google-specific implementation of many of the WHATWG APIs. It appears to be the intent for Google to get the technology out now, then Gears can be the optional basis for implementing storage and multi-threading features when other browser makers are ready.
Javascript + Nintendo DSi = DSiCade
Read the cartoon. You'll find a lot of interesting ideas there. It doesn't sound at all like Firefox with a few default extensions and a custom theme.
All of mozilla's code is tri-licensed -- MPL, GPL, and LGPL. Unless Google releases it under all 3 licenses (or a BSD/MIT license), they won't use it.
Do you even lift?
These aren't the 'roids you're looking for.
Have you read the comic? Google has some really cool ideas for Chrome, making it quite different from another browsers, so you won't get 'another firefox' but an entire new browser. I hope it won't end up like another google products which never go past beta.
If you think Google's www.google.com address just goes to one server that picks out different content by file name, you're in for a surprise. Try the http://www.google.com/chrome address and the http://www.google.com/chrome1 address with a tool that lets you look at the HTTP headers. Look at the "Server" header. Different server code. Google runs a high performance, massively load balanced, widely geographically distributed, HTTP front end that figures out what server to pass things to based on the URI part of the URL. They don't need to do separate hostnames (although they can still do that, too, such as http://maps.google.com/ and http://mail.google.com/).
now we need to go OSS in diesel cars
I find Firefox a pig, especially on G4 Macs. So I'm glad for the competition. Anyway, the new browser war is about speed and compliance so oddly enough it will lead to fewer incompatibles, not more. Finally, read the comic - there are some useful technical motivations to the browser which I think are compelling: Sandboxing, process pre-page, plugins in separate process, javascript compiled to native machine-code, fast garbage collection, etc...
Some privacy policy Slashdot.
Scott McCloud is a comic book artist who authored "Zot!" He also wrote several comic books about comic books as a medium, the most famous of which being "Understanding Comics."
Curiously, he was born Scott McLeod, and the Highlander's last name was MacLeod, and he was Scottish. However, Scott McCloud was born in 1960, and the Highlander film was released in 1986, so if it's anything more than coincidence, the Highlander was named after the cartoonist.
"I'm only 9 pages into the comic but the fact that every tab and plugin will run as a separate process seems significant to me and something more than just a rebranding."
Which is good because current plugins in firefox (they add up) will freeze/slowdown/crash the browser, and I hate that, it's Firefox 3 too.
modded troll? well, that's life. What I mean is that, even if it opens in another process or thread, you cannot be sure (without the code) that the processes are truly sandboxed and won't interfere with one another.
There are several answers for that question in the comic. Most of them can't be implemented in Firefox (in fact, any of the other browsers afaik) without a major rewrite, and you could want some of the features they are introducing there.
This isnt about "just another open source browser", it goes to the core of several problems that browsers have with today's web requirements.
Try Tree Style Tabs as a Firefox addon -https://addons.mozilla.org/en-US/firefox/addon/4287?addons-author-addons-select=5890
I've been using it for a few months and it works pretty well on a 22" monitor, as someone else has also mentioned.
Does that mean that their relationship with Mozilla will be ending?
Google recently renewed their monetary agreement with Mozilla for 3 more years.
So, no, it seems their relationship remains strong. Google Chrome sounds like a very cool project, but I'm thinking that it'll be more of an experiment than an actual product, just like most things Google make.
Also, I doubt Mozilla would have a hard time finding funding even if Google pulls the plug on them.
This has been in the works since 2006.
Firefox isn't the only browser funded primarily by Google; Opera is as well.
GLaDOS for President 2016! "Well here we are again. It's always such a pleasure." -- GLaDOS, 2011
To most people this would mean reporting every URL visited, possibly even what you're typing into forms. To you it means making google.com the default homepage, and OMG if you type something into that box, it's gonna report what you type to Google!! Duh, it's gonna report it to Google because google.com is in the address bar.
You really think this is the only information sent to Google? Really?
Put on your tinfoil hat kiddies, you need to actually research this...
That's because Flash 9 received a brand new Virtual Machine. FireFox was given the code for it (it's called Tamarin), but it has not yet made it into a release. Once it does, FireFox and Flash 9 should show similar performance profiles.
Previous versions of Flash were absolutely terrible from a performance perspective. So the entire JS-language community is slowly moving forward. :-)
Javascript + Nintendo DSi = DSiCade
The point is that when you click on a tab, stuff _below_ it changes. A tab is not an element "click me and all around the screen stuff changes". A tab is an element that tells you "if you switch to a different tab, stuff below me changes".
Now look what happens when you currently click on a different tab in Firefox, Safari or IE: stuff changes below it (the page) and above it (the URL in the address bar). This is illogical! It dillutes the meaning of a tab. And it makes it difficult for normal computer users to understand the concept of a tab. Placing it the way Google does does now fix this.
This is first and foremost an issue of correctness and preserving the concept of a UI element, not a question of taste.
(ugh!) reference counting
Why all the hate for reference counting? Automatic reference counting with cycle detection is a nice way of doing accurate GC. It has more deterministic performance when done with a generational cycle detector than a pure tracing collector, and works better in a number of distributed computing settings. Presumably you've read the Unified Theory of Garbage Collection paper from TJW if you're commenting on this topic, so I'm interested in what you see as the problem with reference counting, since tracing and reference counting plus cycle detection are both special cases of the same general algorithm.
I am TheRaven on Soylent News
You mean the canvas element? Firefox and Opera are already working on 3D drawing contexts.
Opera build, Opera Code example
Firefox Addon
Another advantage to giving web apps this power- it makes learning programming (especially the flashy bits) easier. Elementry-schoolers needn't worry about configuring compilers, managing imports, window handles, etc; the browser does it all. HTML and parts of Javascript are simple enough to explain with a good teacher; gloss over the trickier bits at first with a voodoo var artist = getElementById('canvas').getContext('2d'); line, and drawing becomes much more accessable.
I don't think it's ideal either. What google are really doing is every tab is a new instance of a browser
Firefox managed to produce tabs within a single browser. Google is going the other way. Shades of old IE where every page was another browser.
Why? Google's sandboxing technique calls for it. But is it that important to isolate each page from other pages like that? The whole thing will be a monsterous load on resources. Can you imagine what 10 'pages' running 10 instances of V8 java be like?
Now I know why I'd need a 3GHZ quad core and 4GB of memory. And it's early days yet!
Don't be apathetic. Procrastinate!
IE8 is a multi-process browser.
Not only does it do it automatically, the per-tab process is just the rendering engine, so there's some overhead in repeating it over and over, but not nearly as much as an entire firefox process. If they use webkit as a dynamic library, the OS will almost certainly have the code loaded only once.
Google has no problem recruiting people - who hasn't heard of Google? In fact as far as I hear they've cut down on recruiting lately (as on other things) because they have too many engineers and still only one money-making product.
Sounds like a classic Unix architecture to me! Before threads, there were processes and signals, and it still works pretty well... I wonder how it will affect memory usage though.
The comic addresses this.
There is an upfront memory overhead in starting a new process. However, there are long term memory usage gains. The comic suggests that fragmentation and memory leakage occurs in the browser's memory space.
Sure, you could work to improve the browser's memory management - but by moving to process-per-tab, you make it a non-problem. When you close a tab, the OS reclaims that process's memory. You'd hope that the OS has rock solid memory management code.
would be an *online browser*. Like Google docs.
Oh, that. Already been done.
There's an extension for Firefox which does tree style tabs (called Tree Style Tab). It's not perfect (sometimes it loses track of which tabs are inside others when you restart Firefox) but it works pretty well, I've found!
It has lead to me having a lot more tabs open than I used to though - when you can find them so easily, there's less of an incentive to close them as you go. Whether this is a good thing or not is open to debate!
When you fork a process, the new process gets a new virtual memory space of its own (including mapped out kernel space and everything else...like any other process). Initially everything will just be pointers to the parent process's memory locations. As things change (load a new page), the processes will start to branch off as the child starts to behave differently and has to begin using its own memory instead of pointing to the parent (the pointers to the parent's page are moved to empty memory blocks which are then filled with the new page). Because of this, there doesnt have to be a big overhead for running tabs as processes--libraries, functions, etc do not have to be loaded again.
Bottles.