jQuery Grid Plugin - jqGrid - Topic: Performance improvements for remapColumns http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns Simple:Press Version 5.7.5.3 2late2die on Performance improvements for remapColumns http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns#p24310 Discussion http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns#p24310 Oh I already saw those and have them in my code, and they do improve the performance but they're different from my modifications above.

]]> Tue, 16 Aug 2011 20:31:24 +0300 OlegK on Performance improvements for remapColumns http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns#p24307 Discussion http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns#p24307 If you made some reset with ColumnChooser try to repeat the test after the simple fix described here. The fix is already in the github code.

]]> Tue, 16 Aug 2011 18:36:43 +0300 2late2die on Performance improvements for remapColumns http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns#p24304 Discussion http://www.trirand.com/blog/?page_id=393/discussion/performance-improvements-for-remapcolumns#p24304 I've implemented some performance improvements to remapColumns function, but I couldn't do properly exhaustive testing, and I'm also not very deeply familiar with jqGrid. So what I'm saying is, these changes might negatively affect something I didn't think of, so if anyone has the inclination to do some further testing I'd appreciate it. Or even just take a look at the code and double check I'm not doing something stupid.

In any case, I've forked jqGrid on github and here are the changes to grid.base.js. They are fairly straitforward but here's a quick outline.

  • Put a check to make sure the remap is actually needed. remapColumns is called by columnChooser (and possibly other functions) with original order, so I run a quick loop and make sure to only proceed if there's been a change in the order of columns.
  • Replaced $.each loops in both resortArray and resortRows with for loops.
  • Changed rows selectors to use parent.find instead having parent as context (tiny improvement but also costs nothing).

The actual improvements in speed ended up being only about 10% (my data has 345 rows and 13 columns). However, the number of calls made was reduced almost five times. (I used firebug's console.profile to get these numbers)

]]> Tue, 16 Aug 2011 17:24:58 +0300