# Inconsistent sprite behaviour with on wall settings
**URL:** https://forum.makecode.com/t/inconsistent-sprite-behaviour-with-on-wall-settings/832
**Category:** Arcade
**Created:** [December 1, 2019, 11:29pm UTC](https://forum.makecode.com/t/inconsistent-sprite-behaviour-with-on-wall-settings/832 "2019-12-01T23:29:36Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@kjw](https://forum.makecode.com/u/kjw)
#### Post date: [December 1, 2019, 11:29pm UTC](https://forum.makecode.com/t/inconsistent-sprite-behaviour-with-on-wall-settings/832/1 "2019-12-01T23:29:36Z")
The sprite settings `bounce on wall` and `destroy on wall` have _inconsistent_ behaviour when there’s no tilemap and the screen edge becomes an “implicit wall” in some cases:
- sprites bounce off the edge of the screen with `bounce on wall` ,
- sprites travel off screen and continue to exist with `destroy on wall`.
This can be seen in this small demonstration program where the cyan duck should die but makes a fleeting comeback proving there’s life after going off screen: [https://makecode.com/\_8tED843tp1ih](https://makecode.com/_8tED843tp1ih)
There’s also a change in behaviour from `production beta` to `beta beta` with a sprite with an acceleration value with is then later set to follow another sprite. Watch the _green_ duck. Which one is correct?
---
### Author:  [@kjw](https://forum.makecode.com/u/kjw)
#### Post date: [December 2, 2019, 4:46pm UTC](https://forum.makecode.com/t/inconsistent-sprite-behaviour-with-on-wall-settings/832/2 "2019-12-02T16:46:41Z")
On reflection the following on 0.14.45 must be broken. I put in [GH: microsoft/pxt-arcade: follow sprite behaviour looks broken in 0.14.45 and absence of unfollow block #1505](https://github.com/microsoft/pxt-arcade/issues/1505) for this.