Chrome shows tricks
The other day I discovered a rather interesting fact related to the passage by browsers of various tests, and in particular - the popular Peacekeeper . Checked three browsers - Opera, Firefox and Chrome. Everything is in Linux. Everything went as usual, but when I got to Chrome (version 5.0.307.7 beta), I noticed that something was going wrong. I understood the reason only after a few seconds of observing the balls (there is such a test in Peacekeeper): they behave differently from all other browsers! See for yourself. Here is Opera 10.50:
And here is Chrome:
')
If you have not noticed the difference - give a hint. In all browsers (and in Opera, of course) the balls repel each other in a collision (well, or simply interact), but in Chrome they don’t notice each other at close range and just fly through!
And here comes the question. Even a few.
1. Why does the test work out differently in Chrome?
And this is not idle curiosity. The fact is that even with my meager knowledge of programming it is clear that adding one more physical property to an object affects speed - it's one thing just to move the balls according to the algorithm, and another thing is to additionally ensure their “tangibility” in relation to each other, so that achieve a repulsive effect.
2. If this is the result of some kind of "optimization" by the Chrome developers of the code of their browser, then what other similar "improvements" were made to improve the results of the Peacekeeper test?
In general, if someone can explain this phenomenon - welcome to the comments. ;)
PS Sorry for the video brakes - the laptop is no longer top-notch, and at the same time as the video is captured, the tests start to slow down in a godless way
And here is Chrome:
')
If you have not noticed the difference - give a hint. In all browsers (and in Opera, of course) the balls repel each other in a collision (well, or simply interact), but in Chrome they don’t notice each other at close range and just fly through!
And here comes the question. Even a few.
1. Why does the test work out differently in Chrome?
And this is not idle curiosity. The fact is that even with my meager knowledge of programming it is clear that adding one more physical property to an object affects speed - it's one thing just to move the balls according to the algorithm, and another thing is to additionally ensure their “tangibility” in relation to each other, so that achieve a repulsive effect.
2. If this is the result of some kind of "optimization" by the Chrome developers of the code of their browser, then what other similar "improvements" were made to improve the results of the Peacekeeper test?
In general, if someone can explain this phenomenon - welcome to the comments. ;)
PS Sorry for the video brakes - the laptop is no longer top-notch, and at the same time as the video is captured, the tests start to slow down in a godless way
Source: https://habr.com/ru/post/84135/
All Articles