Hi everyone,
I’ve been using DEVONthink To Go (DTTG) on my iPad Pro for my engineering studies and professional workflow. While the document management is excellent, I find the current implementation of Apple Pencil annotations to be a significant bottleneck due to its rudimentary nature.
The Core Issue: Destructive Pixel-based Erasing
The eraser tool in DTTG currently operates on a pixel/bitmap level. This leads to a highly frustrating and inefficient workflow, especially when combining handwriting with highlighting:
-
Destruction of handwritten notes: The biggest issue occurs when I highlight my own handwritten words or headings. If I make a mistake with the highlighter and try to erase it, the pixel-based eraser destroys my handwritten letters as well. Because both are treated as simple pixels on the same layer, it is impossible to remove the highlight without wiping out the underlying text. I then have to rewrite everything from scratch.
-
Inefficiency: Even for simple notes, I have to manually “scrub” over every single part of a line to remove it. There is no way to just “tap” a stroke to delete it.
-
Comparison: Most professional iPadOS apps (Apple Notes, PDF Expert, GoodNotes, etc.) treat Pencil input as vector-based paths. This allows for a “Stroke Eraser” that removes an entire line or highlight with a single gesture without affecting other nearby or underlying strokes.
Technical Environment:
-
iPadOS: 18.7.3
-
Hardware: iPad Pro (12.9-inch, 6th Generation)
-
DTTG Version: 3.8.16
Research and Settings
I have checked the app settings and the iOS system settings, but as noted in previous community discussions, there are no configuration options for the Apple Pencil or the eraser behavior (e.g., toggling between “Pixel Eraser” and “Object/Stroke Eraser”).
My Question / Suggestion to the Community and the Team:
-
Is there any hidden setting or a specific workflow that allows for path-based erasing or at least protects handwritten strokes from the highlighter’s eraser?
-
Are there plans to modernize the annotation engine to support vector-based strokes and a non-destructive “Stroke Eraser”?
In my opinion, the current “pixel-scrubbing” approach is a major regression compared to dedicated annotation tools. Implementing a proper Stroke Eraser would be a massive quality-of-life improvement for power users who use DTTG as their primary workspace for active reading and note-taking.
I’m curious to hear if others share this frustration or if there are any plans for improvement.
Best regards,
Daniel