w3c
Textual chunk which is not `safe to copy`
Raised by: Adrian Custer Original date: 12 Jun 2012
A second question regards the need for a 'chunk' for textual metadata which is not 'safe-to-copy'.
If I understand the standard correctly, I conceive of the need for an 'inFO' chunk that would act like the newest chunk for text (iTXt) but would explicitly not be safe-to-copy. My lack of certainty about the need for this element comes from being unsure as to whether the modifications allowed for 'safe-to-copy' chunks include cropping the image which would invalidate the information I hope to store in the 'inFO' chunk.
The issue makes sense to me.
If I were to open an image in an editor and change parts of it, safe-to-copy chunks would be preserved when I save the edited image.
Suppose I opened a satellite image of a city and zoomed in & translated to a particular building. Any geospacial information about the original building would no longer be accurate.
However, I'm not sure a new inFO chunk is required for this.
The original post suggests the inFO chunk would work much like the iTXt chunk. If all that is needed is not-safe-to-copy, it could simply be a iTXT chunk.
This opens up other questions though. Should the spec be flexible about known chunks matching case? Should new keywords be added to iTXt for geospacial data? Should a new chunk be introduced just to avoid these complications?
