# Recommendations for bindingtester usage in CI

**URL:** <https://forums.foundationdb.org/t/recommendations-for-bindingtester-usage-in-ci/2989>\
**Category:** Development\
**Tags:** bindings\
**Created:** [November 3, 2021, 9:27pm UTC](https://forums.foundationdb.org/t/recommendations-for-bindingtester-usage-in-ci/2989 "2021-11-03T21:27:20Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![kocolosk](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/kocolosk/32/412_2.png) [@kocolosk](https://forums.foundationdb.org/u/kocolosk)\
**Post date:** [November 3, 2021, 9:27pm UTC](https://forums.foundationdb.org/t/recommendations-for-bindingtester-usage-in-ci/2989/1 "2021-11-03T21:27:20Z")

</div>

Hi, I was doing some work to improve the CI system for our Erlang bindings and wondered what the best practice might be regarding the use of the `bindingtester.py` code in an automated CI run. I see that there’s a deterministic `scripted` test that’s the default, and other tests like `api` that pick a random api\_version and seed on startup. I also seemed to find that the `scripted` test only supports the latest API version (at least, it still seemed to be testing newer functionality regardless of the value of `--api-version` that I passed to the script).

Are there any pointers to how the in-tree bindings go about their testing that might be useful to replicate?