Performing the test with the list in the middle muddies the waters.
Removing the list from the middle of the #2 and then reloading each time, I
would get times like:
Wiki 218,188,192
Set 296,228,285
Linux,FF 64
There's a lot of variation in both numbers. The Set numbers do seem 10-30 %
@Mark:
The log also contains other measurements, but I do not know exactly what
they measure. Anyway, even after refresh, Set seems slower on my machine
(in Vivaldi, a Chromium based browser).
Opening wikify example:
performance: styleRefresh: 6.60ms
performance: +filter: 1.20ms
Hi Bimlas
I think the tests should be structured so that the set widget vs. wikify widget
are within the list widget, so that we'd be testing a few hundred iterations.
Here we're only testing a single instance of each construction.
The main refresh time is the important timing; the “+filter”
On mine, the Set did slightly better than Wikify. Consider refreshing the
browser between tests. I think the browser caches stuff in its memory which
may impact performance.
Why are there list widgets inside of the Wikify and Set widgets? It seems
to me that they may impact the performance
4 matches
Mail list logo
