CT: Assertion failed: scale should be finite #321
Assignees
Comments
|
Not immediately reproducible after 5 minutes running unbuilt with Manually mapping the FramedImageNode.js assertion to FramedImageNode.ts, it's failing at line 88: const scaleX = ( viewBounds.width / initialWidth ) || GOConstants.MIN_SCALE; // prevent zero scale
const scaleY = ( viewBounds.height / initialHeight ) || GOConstants.MIN_SCALE; // prevent zero scale
88 parentNode.scale( scaleX, scaleY );Node.js is failing at line 2120: // eslint-disable-line
scale(x, y, prependInstead) {
// eslint-disable-line
if (typeof x === 'number') {
2120 assert && assert(isFinite(x), 'scales should be finite');... which maps to Node.ts line 2285: scale( s: number, prependInstead?: boolean ): void;
scale( s: Vector2, prependInstead?: boolean ): void; // eslint-disable-line
scale( x: number, y: number, prependInstead?: boolean ): void; // eslint-disable-line
scale( x: number | Vector2, y?: number | boolean, prependInstead?: boolean ) { // eslint-disable-line
if ( typeof x === 'number' ) {
2285 assert && assert( isFinite( x ), 'scales should be finite' ); |
|
This has not occurred in the 4 most-recent CT cycles. It has not occurred since 2/20/2022 @ 1:30:34 AM. And CT seems to have restarted, so I can't tell how many times it occurred since the first occurrence on 2/19/2022 @ 3:31:37. So before spending more time on this, I'm going to wait to see if it reoccurs, or was perhaps some transient common-code issue. |
|
Still nothing in CT, so closing. |
|
Reopening. This occurred once: Manually mapping from .js to .ts, this is occurring on line 87 of FramedImageNode.ts: const initialWidth = parentNode.width;
const initialHeight = parentNode.height;
const scaleX = ( viewBounds.width / initialWidth ) || GOConstants.MIN_SCALE; // prevent zero scale
const scaleY = ( viewBounds.height / initialHeight ) || GOConstants.MIN_SCALE; // prevent zero scale
87 parentNode.scale( scaleX, scaleY ); |
pixelzoom
added a commit
that referenced
this issue
Mar 2, 2022
|
I suspect that parentNode has zero bounds because the Image node that it wraps does not have valid bounds -- for some reason. In the above commit, I've added assertions with messages that will help me (dis)prove this theory. EDIT: I've tried to reproduce this locally with |
pixelzoom
added a commit
that referenced
this issue
Mar 2, 2022
|
I haven't seen this failure since 3/1/2022. |
|
I fuzzed tested the unbuilt sim for 30 minutes and did not encounter this problem. |
|
This is a non-blocking error. It won't cause the sim to crash, and is unlikely to result in any noticeable problems in practice. 1.1 was published with this as a known issue. |
|
Can you confirm if this issue is still needed? I don't see it in CT right now, but it is noted as rare. For phetsims/qa#832 |
|
Yes, please keep this open. |
|
It is not necessary to address this for #278, the 1.3 PhET-iO release. |
|
Still occurring occassionally, last seen 5/4/23 @ 10:56AM. |
|
Still occurring occassionally, last seen 9/25/2023, 11:16:26 PM. I still cannot reproduce locally. |
pixelzoom
added a commit
that referenced
this issue
Sep 26, 2023
pixelzoom
changed the title
|
After addressing #487, this is occurring again in CT, see below. The problem is that |
pixelzoom
added a commit
that referenced
this issue
Dec 27, 2023
|
Addressed in the above commit by testing for both zero and non-finite scale. I'll leave this open for a few cycles to verify that CT is happy. |
|
CT is happy. Closing. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The first time that this occurred in CT was Saturday 2/19/2022 @ 3:31:37 AM. I hadn't touched the code since Thu 2/18/2022 @ 8:57PM. So I'm wondering if this is a common-code regression.
The text was updated successfully, but these errors were encountered: