Changes between Version 3 and Version 4 of Ticket #15478, comment 5


Ignore:
Timestamp:
Jun 10, 2012, 4:59:31 AM (8 years ago)
Author:
Kenneth G. Franqueiro
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #15478, comment 5

    v3 v4  
    33HOWEVER, while this does seem to considerably improve performance in browsers which don't need the extra logic, it actually noticeably '''worsens''' performance in browsers which do need it (I tested IE6-7; not sure if 8 does as well, 9 doesn't).  I'm talking a significant 33% increase in execution time.
    44
    5 IMO this consequence outweighs the benefit, since the non-problem browsers are significantly faster to begin with.
     5IMO this consequence outweighs the benefit, since the non-problem browsers are significantly faster to begin with.  Perhaps this would be a better idea for 2.0?
    66
    77I'm attaching a standalone test page for reference (not using jslitmus to avoid IE's "script taking too long" warnings which will totally throw off the results).