[ID-4275394][Bug] Share for review updating issue (There Was an Error Updating the Artifact) with hint to solution
In addition to https://community.adobe.com/questions-671/persistent-error-in-adobe-indesign-there-was-an-error-updating-the-artifact-887573 i'd like to mention the bug here too, as it is present in the current beta as well.
I've investigated the "there was an error updating the artifact"-error arising since 2026 versions of InDesign.
It's a reproducible issue (tested on different machines) depending on the complexity of linked files and a certain quantity of elements/pixels/rendered areas.
I investigated the issue and identified the lowest common denominator — it was reproducible across multiple systems, tested in both InDesign 21.2 x64 and InDesign 21.3 x64 Beta. To set up the test, I only needed 1 PDF and 1 JPG file on, for example, a two-page spread in (e.g.) DIN A5.
The error must lie in a certain quantity of elements/pixels/rendered areas — I’ve created a short 3:45-minute video that clearly demonstrates and analyzes the symptoms and variations. (may be more understandable than my write-up below ;))
https://www.youtube.com/watch?v=n-IwO87yLLI
Here is a brief summary of the findings:
- Once the content reaches a certain level of complexity/"volume", the error "There Was an Error Updating the Artifact" occurs when the review-link is updated.
- If you reduce the number of (or more precise: area of rendered) objects in the file, the error disappears.
- In my example, I’m using a somewhat complex PDF with some QGIS vector elements—just to trigger the error more quickly. With mixed content and simpler embedded PDFs/images, the error occurs later/with more objects in the project.
- The PDF is placed full-page on a spread and then duplicated to a total of 8 instances, with the copies simply stacked on top of each other. (The number of duplicate layers required may vary depending on your machine.)
- A review link is then generated -> which works.
- Immediately after that, click "Update Content" -> Error: "There Was an Error Updating the Artifact"
- If you then hide one of the layers, updating the content works again without errors.
- If you show the layer again, the error reliably reappears.
Now it gets weird:
- With all 8 layers visible: If I reduce the width of all frames from (in my example) 296 mm to 281.163 mm (determined by trial and error), the content can be updated without errors.
- At 281.164 mm, the "Updating the Artifact" error is generated. (+0.001 mm)!
- At 281.163 mm, any image can be added in a separate frame on the same spread — provided the image isn’t too large (this refers not to the file size but to the area it occupies on the page in the layout), the content will update without errors.
- If you then enlarge the frame and the JPG, the error is triggered again.
- If you reduce the size again, the update goes through.
- You can also push the limits to the point where individual text fields containing placeholder text trigger the error mentioned above. If you then reduce the visible area of the placed image (or the PDFs behind it), the update will complete successfully again.
The update also works with any number of placed images, provided you hide one or more of the PDFs below.
So there must be some kind of buffer overflow (or similar) that causes the updates to fail.
If you are in a state where an error occurs (e.g., 15 duplicate PDFs + 10 images, etc.) and delete the review link, you can regenerate it once (without errors); however, if you click "Update" immediately afterwards, the known error occurs again. If you then hide 8–9 of the PDFs again, for example, the update will work without any issues.
Comments or pasted graphics were not present during the tests.
The test files shown in the video can be downloaded here [https://download.age-pi.de/fragment.issue.zip] (just a zip) if anyone would like to experiment.
Finally, it’s worth mentioning again: The error isn’t caused by the PDF file used — it merely serves as a diagnostic tool (besides, it wouldn’t work with 7 instances of it and would stop working with 8). With enough other "normal content", the error occurs anyway; it’s just harder to reproduce.