In April 2014, Chrome gave the best performance of JSmol: smoothest rotation and few if any long (~ one minute) pauses during loading of some modest PDB files (pauses seen in Firefox). I configured FirstGlance in Jmol, when using JSmol, to recommend Chrome over Firefox. Safari was good. Internet Explorer was unusably slow with JSmol.

Recently I began to notice jumpier rotation in Chrome, while Firefox and Safari continue to perform well.

Today I found Chrome to be unable to rotate any molecule (even caffeine) in JSmol. Although there are brief moments when rotation can be seen, these are broken by long freezes (> 5 sec) when no rotation can be accomplished.

The problem is worst when the mouse is moved quickly. The molecule freezes until the mouse is stopped. If you keep moving the mouse, the molecule may freeze for more than 10 sec. The molecule does rotate with very slow movements of the mouse.

The problem occurs in OS X, where Chrome has updated to version 47.

Rotation is still quite good in Chrome version 46, which happens to remain in my Windows 7 and 10 machines, despite version 47 becoming the stable release on December 1 (http://googlechromereleases.blogspot.com/search?updated-max=2015-12-01T12:43:00-08:00&max-results=10). Presumably my Windows Chrome will shortly auto-update to 47 and then I will be very curious about performance.

I tested the OS X beta version of Chrome, version 48, in hopes the problem would be fixed -- but it remains in Chrome 48 in OS X.

Since Chrome stopped supporting Java on September 1, 2015, Chrome is now useless for JSmol websites.

We can hope this is a mouse-related bug in Chrome that will be fixed, but as I mentioned, a fix is not present in the current beta version 48.

This is a sad turn of events! Comments? Suggestions? Insights?

-Eric


------------------------------------------------------------------------------
_______________________________________________
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users

Reply via email to