Contributors/Projects/Editing performance 2

Report publication date: ???

Contact: Dan Garry

In order to better understand the performance of the 2017 wikitext editor relative to the 2010 wikitext editor, the Editing team has collected performance metrics for these editing environments. Definition of the metrics, explanations of how the data was collected, and an analysis of the data to recommend a course of action, is documented in Contributors/Projects/Editing performance. This page documents improvements that were made to the 2017 wikitext editor to reduce the amount of time it takes for it to become interactive.

Definitions

edit

"Time to interactive" and "time to fully loaded", the two metrics used here, are defined in Contributors/Projects/Editing performance#Definitions.

Methods

edit

The data collection methods are defined in Contributors/Projects/Editing performance#Methods. The same data collection method is used here.

The time to interactive was mostly improved by implementing T184614, which shows a temporary plain wikitext editor to the user while the rest of the toolbar code is still loading.

Results

edit

No two real-world edits are exactly identical, and a single average provides very little information. To compare performance, the data set has been sorted by speed. This allows us to compare the faster (or slower) edits in each editing environment against the faster (or slower) edits in the other editing environments. Results are given below for particular percentiles. The number given at the 50th percentile (median) means that 50% of the edits using that editing environment were faster than the number given, and 50% were slower. Similarly, the number given at the 90th percentile means that 90% of edits in that editing environment were faster than this, and 10% were slower.

Time to interactive for 2017 editor before and after performance improvements

edit


Time to interactive for all editors after performance improvements

edit

Data table

edit
Performance data by percentile (after performance improvements)
Time (seconds) to interactive 1st 5th 10th 20th 25th 50th

(median)

75th 80th 90th 95th 99th
Visual editor 0.613 0.817 0.972 1.225 1.356 2.156 3.747 4.403 7.014 10.853 39.080
2017 wikitext 0.202 0.264 0.310 0.393 0.431 0.642 1.026 1.232 1.868 2.748 8.570
2010 wikitext 0.616 0.798 0.915 1.102 1.188 1.694 2.721 3.159 5.215 11.295 117.292
Time (seconds) to fully loaded 1st 5th 10th 20th 25th 50th

(median)

75th 80th 90th 95th 99th
Visual editor 0.614 0.818 0.973 1.226 1.358 2.159 3.750 4.405 7.013 10.831 39.048
2017 wikitext 0.430 0.604 0.726 0.917 1.012 1.603 3.003 3.522 5.486 8.879 39.167
2010 wikitext 0.717 0.924 1.054 1.264 1.362 1.933 3.102 3.597 5.875 12.087 118.806

Conclusions

edit

Caveats

edit

Assignment between conditions in this observational experiment is not completely random. Most users of the 2017 wikitext editor have made a deliberate choice to use that editing environment. We expect that users who experienced significant performance problems, within their personal computing environment, would normally stop using the 2017 wikitext editor. Similarly, we expect people who perceived the 2017 wikitext editor to be the faster option, within their personal computing environment, would normally be less likely to return to the 2010 wikitext editor.

Next steps

edit