EntrepreneurFix Is a Free Entrepreneur, Business and Start Up Forum

Connect With Other Entrepreneurs Online Today!
Dismiss Notice
Welcome to Our Community
Wanting to join the rest of our members? Feel free to sign up today.

Google's new algorithm will make Chrome run much faster

Discussion in 'General Discussions and Lounge' started by EF-Roger, Jan 20, 2016.

  1. EF-Roger

    EF-Roger Entrepreneur

    Chrome is about to load web pages a lot faster than you've experienced up until now. Thanks to a new compression algorithm called Brotli, which Google introduced last September, Chrome will be able to compress data up to 26 percent more than its existing compression engine, Zopfli, which is an impressive jump.

    Read more Google's new algorithm will make Chrome run much faster
    setupdisc likes this.

  2. setupdisc

    setupdisc Moderator Entrepreneur

    Data compression is my speciality. Brotli is good for compression of html and some php streams, but it does only about average when compared to any other LZ variant that already exists. It is fast, which is its highlight, but it compresses less than some other alternatives when combined with arithmetic encoding and statistical algorithms, preprocessing, move to front algorithms, and other techniques to get the most out of it. It isn't going to do even half as well as 7zip on the same types of data, and will only do about the same or slightly better than gzip or zlib. It's important to remember that a lot of data being sent over the web is usually compressed already as closely as it can be to the limits of Kolmogrov Complexity.

    So for html pages, bitmaps, and wave or raw audio streams, it will do well with that. As for media that is already compressed like mp3s, mp4 files, jpeg images, and precompressed archives...it isn't going to give you extra there. In fact, if it doesn't identify it and tries to compress it anyway, it will expand it if it encounters it in the stream. That doesn't seem to be an issue, since the major focus of Brotli is for html text and prebuffering common tags.

    Most of the advantage it has comes from using predefined dictionary (a preloader of common html tags) that takes extra space and ram. You can replicate and get the same or better results if you design or use an LZMA plug-in for Firefox, Chrome, or any other browser if you chose to approach it. Most of the attention it is given is because "Google" is doing it, whereas people who have done it privately and for smaller companies 11 to 12 years ago were either blackballed or got no mention to where it could be used commercially or otherwise.

    It is going to help the speed of Chrome with page loads significantly whether you're using mobile bandwidth or broadband connections, but the approach and algorithm they are using is not new...just a variant that they're going to make popular, and make the public be aware of to solidify it as "theirs" when they do.

    If you're interested to see it, here's the algorithm for it and technical details from Google (it's a little long and technical for some, but it will help to give the big picture on what's the same, and what things are different): http://www.ietf.org/id/draft-alakuijala-brotli-08.txt
    Last edited: Jan 20, 2016
  3. Federico.Gimeno

    Federico.Gimeno Entrepreneur

    Will it have a boost of speed when executing JavaScript code (I have several applications written purely in JavaScript, which some are performance critical, like some old game consoles emulators).
  4. ProxyRadar

    ProxyRadar Entrepreneur

    Chrome is my favorite browser for page loading speed, usability and web-developing standards support

Featured Businesses (View All)

Share This Page