Troubleshooting RIDE 2.0 big projects performance

I plan to use this thread to report and discuss the problems with RIDE 2.0, regarding slowness to load big projects. It is even worse, that after loading them, RIDE is unusable because it gets blocked/unresponsive.

We will use a big project that is available here.

The following graphic shows the time to:

  1. Start RIDE, which will load the last project used (benchmark/)
  2. When available on Tree (Test Suites), expand first test suite folder and select to run first test case.
  3. Select to Editor the first test case
  4. On the Editor Control-Click the single keyword to open its definition
  5. Close RIDE

This was done with Python 2.7.18 and RIDE 1.7.4.2, and with Python 3.10.9, RIDE development version from today, 2.0b3.dev4. With this one, was only possible to perform up to step 3, and RIDE was forcibly closed.

The chart:

Conclusions:
We can see that the test was completed in 40 seconds with RIDE 1.7.4.2, and around 200 seconds with RIDE 2.0.
RIDE 2.0 uses less memory than RIDE 1.7.4.2.
RIDE 2.0 took around five more times to reach readiness, than RIDE 1.7.4.2. Considering last memory plateau, RIDE 2.0 at 100 seconds, and RIDE 1.7.4.2 at 20 seconds.

Next experiments will be done with a known RIDE 2.0 code commit, to see if it can be restored.