Repository navigation
Conversation
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
There was a problem hiding this comment.
🟢 Approval recommended
The documentation clearly separates current capabilities from proposed work and unresolved decisions, with no identified blocking issues.
No open issues 🎉
What changed in this PR
Adds a Phase 0 investigation documenting the current MapWinGIS6 capabilities and the work needed for rendering and an ActiveX wrapper. This is documentation only; no code or dependencies change.
Changes:
- Records existing core capabilities, build coverage, and legacy control compatibility references.
- Outlines a CPU-first rendering path while leaving ATL versus MFC and the initial COM surface as explicit design decisions.
| File | Description |
|---|---|
| docs/ACTIVEX-INVESTIGATION.md | Documents findings, architectural gaps, and a staged rendering and ActiveX implementation path. |
Review effort: Balanced
pmeems
left a comment
There was a problem hiding this comment.
Reviewed the Phase 0 investigation against the repository guidance and the PR diff. I found no blocking defects in this documentation-only change. The architecture direction is sound: keep COM outside the core, preserve the legacy contract deliberately, and establish a CPU-rendered vertical slice before introducing OpenGL. I left two non-blocking suggestions to make the investigation easier to verify and turn into implementation work.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Records Phase 0 findings before implementation. The current core provides GDAL/OGR vector data access but has no map model, renderer, or ActiveX target; the report also notes the legacy _DMap/event contract and paint lifecycle. It recommends a CPU-rendered vector slice before OpenGL. ATL vs. MFC and the initial compatible COM surface remain explicit design decisions. No dependencies added.