Editable vs Visual

editable and visual should be treated as architecture terms, not marketing labels.

They describe a real system tradeoff: presentation content often cannot be optimized for maximum editability and maximum visual fidelity at the same time.

The Constraint

Some content maps well to native presentation objects:

• text blocks
• shapes with native editable text bodies
• simple shapes
• standard images
• basic grouped structures

Other content is much harder to preserve as fully editable native objects without losing appearance:

• complex SVG effects
• browser-like layout behavior
• transforms that do not map cleanly to DrawingML
• advanced styling or unsupported visual effects

Strategy Modes

PPTKit should use editable and visual as strategy terminology that future import, render, and export layers may rely on.

editable Strategy

Favor native presentation objects wherever possible.

This strategy should prefer:

• editable text
• one editable shape object when geometry and text belong together
• editable shapes
• native grouping where feasible
• structure that survives post-export authoring

visual Strategy

Favor closer visual fidelity when native object mapping would degrade the result too much.

This strategy may rely more heavily on:

• SVG-backed output
• rasterized fallbacks
• grouped visual approximations
• explicit loss of some object-level editability

Mixed Fallback Behavior

Some workflows will need mixed behavior rather than a single pure mode.

For example:

• preserve editable text but flatten one unsupported effect region
• export most elements as native objects but keep one visual cluster as SVG-backed output
• parse a deck into normalized editable structures while preserving unknown content for later export

The architecture should allow these mixed outcomes instead of pretending one global mode always solves everything.

Documentation Rule

PPTKit should prefer explicit degradation over implicit false promises.

That means the system should be able to say:

• this part stays editable
• this part stays visually closer
• this part was preserved but not fully understood

That is more trustworthy than claiming both properties can always reach 100%.