Post body with full uppercase and images discussion revolves around a loophole in the uppercase check. marianord discovered that posts with all caps text and images don’t trigger any issue with this check. sam explains that the image in the post causes a bypass and suggests that a partial fix is not ideal.
To resolve this, sam proposes a test on Raw -> Cooked -> converted back to text without images, but acknowledges it would add cost to the pipeline. codinghorror considers this case too rare to warrant significant engineering work, pointing out that other bypass methods exist.
tophee suggests an alternative solution, triggering the setting at 90% uppercase instead. marianord likes this idea and proposes using a regex to exclude common use cases like emojis, links, and images from the parsing.
The problem is that we need to do this particular test on Raw -> Cooked -> converted back to text without images. This adds definite cost into the pipeline.
I don’t feel this rare case is worth doing a lot of extra engineering work for. There are lots of ways to bypass. Putting in your post would work too, for example.
I was thinking maybe in a regex, to avoid taking into account [ * ]( * ) something like that. So the most common use cases, without trying to really bypass the limitation like emojis, links, images, are excluded from the parsing.