Hi,
Did you try ClosureCompiler?
http://closure-compiler.appspot.com/home
It will remove whitespaces and rename local variables.
In case of GLmol, it reduced the size more than 40%.
Best regards,
Takanori Nakane
--
Bob,
Chrome Windows and Mac are about 7 seconds. Firefox Windows and Mac results in
the following pop-up and no load. Am I possibly missing a required FF setting?
Otis
3554.666748046875 function (text, font3d, g3d) {
var pixels = g3d.apiPlatform.getTextPixels(text, font3d,
Otis, almost certainly that's just a caching problem.
On Fri, Nov 2, 2012 at 7:46 AM, Otis Rothenberger osrot...@chemagic.comwrote:
Bob,
Chrome Windows and Mac are about 7 seconds. Firefox Windows and Mac
results in the following pop-up and no load. Am I possibly missing a
required FF
Seven seconds on iPad Safari. Fourteen seconds on iPad Chrome. Interestingly,
iPad Chrome is also more sluggish on model manipulation.
I still don't understand the load problem on FF, but it means I can't give a
Java/HTML5 load comparison. While the HTML5 worked on Chrome, I can no longer
use
By the way, this is interesting.
3554.666748046875
is width x height
Apparently there is some sort of bug in Firefox that results in the width
and height becoming non-integer values after repeated use of CTRL-[+] and
CTRL-[-]. Use View...reset and then it should be OK, and I believe I fixed
Bob,
On both Mac and Windows, I cleared cache and rebooted. The FF problem persists.
Here's the Windows error log:
Error in parsing vale for value for width. Declaration dropped.
Error in parsing vale for value for height. Declaration dropped.
RangeError: invalid array length.
Otis
--
Otis
Bob,
You probably know this, but this is the code block that FF does not like - Otis
Clazz.newArray32 = function(f, args) {
var dim = args[0];
if (typeof dim == string) {
dim = dim.charCodeAt (0); // char
}
var len = args.length - 1;
var
you're getting non-integer values for those widths and heights. This is my
page or yours?
Can you check:
Jmol._getElement(jmolApplet0,canvas2d).width
Jmol._getElement(jmolApplet0,canvas2d).height
? and possibly
Jmol._getElement(jmolApplet0,appletdiv).clientWidth
?
On Fri, Nov 2, 2012 at
it's being sent a non-integer value, that's all.
On Fri, Nov 2, 2012 at 1:17 PM, Otis Rothenberger osrot...@chemagic.comwrote:
Bob,
You probably know this, but this is the code block that FF does not like -
Otis
Clazz.newArray32 = function(f, args) {
var dim = args[0];
if
OK, JSmol is loading from one core file now, with additional file loading
as necessary. See
http://chemapps.stolaf.edu/jmol/jsmol/test2.htm?USE=HTML5
I'm interested in which is faster, this JavaScript loading or the standard
Java applet loading:
I have to give a slight edge to Java on my system! Mac OS X 10.8, Safari (Java
7). Of course, once the files have been cached, both are lightning…but my
perception is that Java is slightly faster. Something like a spinning loader
image would probably help minimize the difference.
Cheers, Mike
How long does it take for the load, roughly, in seconds?
On Fri, Nov 2, 2012 at 12:25 AM, Evans, Michael James
evan...@illinois.eduwrote:
I have to give a slight edge to Java on my system! Mac OS X 10.8, Safari
(Java 7). Of course, once the files have been cached, both are
lightning…but my
Very non-scientific study done with a stopwatch, but it confirms:
Javascript: 14.1 seconds
Java: 11.8 seconds
That's the total load time for the page, from hitting Enter until caffeine
appeared.
Cheers, Mike
On Nov 2, 2012, at 12:34 AM, Robert Hanson
13 matches
Mail list logo
