Google.com source code (not simple anymore) | Hacker News Reader

Google.com source code (not simple anymore)

15 pointsrk198710 comments

view-source:https://www.google.com/

Maybe it's just me, but realized google.com source code (HTML) used to be very simple but it looks very different now.

Try navigating HTML elements using dev tools.

And here I thought this was a post about the google.com source code being leaked. :-)

I'd imagine that the rendered HTML is probably produced by a legacy system that stitches together a lot of different templates, some maybe even unintentionally.

No surprise, they have to compress like mad and avoid superfluous http requests. That sort of optimisation will never look fancy

I wonder why there are multiple style elements, in div#footcnt there are 2 right next to each other, they could have saved a whole "</style><style>" there, think of the bandwidth!

Quantified: https://news.ycombinator.com/item?id=4647834

I wonder if the giant script was an external file, and now it's being splat on the page?

I'd believe it. As someone else mentioned above, they've also started inlining image data, so I wouldn't be surprised if these changes all fell under the goal of minimizing HTTP requests.

The other day I noticed that they're inlining image data in the src attribute instead of pointing at an external URL. Really taking things to the next level all-around - I wonder when that will become the norm instead of css image sprites.

Hopefully never, and certainly not with http 2.0 around the corner (this adoption wont be so slow, things are already starting to fall neatly into place). It's a micro-optimization Google can afford to do, others will just rely on optimize images size and a CDN instead (possibly an extra hostname) and vast majority wont/don't even bother..

Click-bait ;)

I don't get it, what's the point of this post? How is it adding any new knowledge to the readers?

Reply on news.ycombinator.com