本人前年写的东西

连续插入节点性能



起始数 插入节点数 IE耗时(秒) FF耗时(秒)

11,111 1000 2.469 0.344

12,111 1000 2.891 0.406

13,111 1000 3.484 0.469

14,111 1000 3.984 0.921

15,111 1000 4.468 0.562



当前总结点数16,111

将从第二层开始的全部收起,性能大有提升



起始数 插入节点数 IE耗时(秒) FF耗时(秒)

16,111 1000 2.375 0.219

17,111 1000 2.5 0.25





插入节点时,性能实际上是很好的，只是因为插入节点后要重新计算所有展开的节点，

这是一个相当耖时的工作,如果改进算法，将展开的节点数控制在1000以内，那么不管是插入或是删除，

性能都会大有提升，这样即使是在IE下生成10W节点都是可能的，当然在 FF下100万节点也许都可能

在这个测试中，FF表现很好，性能差不多达到了IE的10倍-Insert a row node performance



Starting IE several time-consuming to insert nodes (sec) FF time (seconds)

11,111 1,000 2.469 0.344

12,111 1,000 2.891 0.406

13,111 1,000 3.484 0.469

14,111 1,000 3.984 0.921

15,111 1,000 4.468 0.562



Summing up the current 16,111 points

From the beginning of the second level all away, go a long way to enhance performance



Starting IE several time-consuming to insert nodes (sec) FF time (seconds)

16,111 1,000 2.375 0.219

17,111 1,000 2.5 0.25





　Insert node, the performance is actually very good, but because of the node to insert after the start to a recalculation of all the nodes,

This is a working center, if the improved algorithm, the number of nodes to be launched in 1000 within the control, then either insert or delete,

Will greatly enhance performance, so even if IE is the next generation of 10W in the node are possible, of course, in the FF may be the next one million nodes can be

In this test, FF good performance, the performa