Sep 022008

Google announced Chrome, last night, and promised to deliver tonight IST. In case you still are unaware read the official google blog. In case you have any questions about the motive for this game, they have tried to cool the hot heads and foxes through a comic book strip. Its a a couple of pages of scriptures, and will take some time to go through. I liked the idea of marketing their initiative though a comic strip. It answers and lot of questions you might have, and is like FAQ 2.0. It focuses around Why, How and What. A new browser on the block was bound to get some curved eyebrows. After all, google has widely endorsed firefox (and continues to do so).

Following is the summary :-

1. It is grandchild of webkit.

2. It has tabs.

3. Each tab runs in separate process, which means, firefox’s memory leak problems might go away. In case of a crash, only one tab dies.

4. Javascript also gets a virtual machine model, no more asynchronous stuff.

5. Homepage has been improved a lot.

6. Gears supported.

7. Multithreaded.

8. Detachable tabs.

9.  Mozilla prism like functionality.

10. Tab bound sandboxes.

11. Plugin support (pligins not sandboxed)

12. Open source.

Looks like an impressive list to me. Combines the experience of desktop apps and power of web apps. They have used the word process so many times, that i am a bit curious about what happens when i open standard 35 tabs like firefox. Our friendly fox is single threaded, which means that all tabs share same memory space. I suspect that initial memory footprint is a bit high in case of chrome, which can lead to a bit of problems in case of memory starved PCs. I see a lot of context switiching between processes.

The more active apps you use, the more is the active work done by each process. For example, i use youtube to listen to music instead of watching the videos. i build a playlist for this. Thus, one tab is reserved for youtube. In case you use many bandwidth consuming, rendering heavy apps, this might slow down chrome a bit, but it would scale well when majority of your tabs are mere text based static web pages.

The good side is that if one rouge javascript is not sticking to its business, the whole browser might not come down (and you will not loose your tabs and your research on fixing the latest bug).

The sandboxing is still not very clear, thus, it will be interesting to see how cookies and all work (I see people using cookies for interprocess communication more and more).

Gears, though has been around since some time, but is still to take off. So i do not see that as a killer feature.

More juice after some chroming!!

ashish

  • Anon
    It's not a browser, IMO, it's something else. It goes way beyond the ideas that lead to the development of browsers, because it will create an instance of each accessed website as a separate process, each having its interface, which in some cases is specific, like that of Gmail.

    It resembles a desktop application in its design. It might be a springboard for the online OS that Google is trying to build, an integrated environment which allows data manipulation without buying licensed applications. The target is bigger: it's Windows and its suites, not IE or Firefox.
blog comments powered by Disqus

Switch to our mobile site