Ideas for Major Changes

This is a page to list ideas for changes we want to make to Overviewer, but may be very large or may break backwards compatibility with old map caches. With the new Minecraft save file format on the horizon, now may be a good time to start writing some of these changes.

Note: Breaking backwards compatibility means that existing caches will not be compatible with future versions. For small worlds, this shouldn't be a problem, since the time to re-render and entire map will be small. However, for larger maps, the time to re-render may be prohibitively long.

(Discussion spawned from Issue #242)

render multiple map types (day/night/spawn/...) at once

I'm thinking we have each type write out different chunk names in the cache ("night.###.png" and "day.###.png"), sorta like how cavemode used to do. Then we can just change bits of chunk.py and quadtree.py to handle these new names, and move the google map generation code out of quadtree.py so it only runs once. Overviewer is already nice and object-oriented, so this shouldn't be too complicated. It might need a config file though. -- agrif

I still have yet to push my config file code. It's nearly done and works fairly well. In order to give it wider exposure, I'll push it to a 'testing' branch of Minecraft-Overviewer -- eminence

Get python setup.py install working

This would involve putting all the free-floating python files into a python package, and moving the front-end code out of gmap.py and into its own script (like overviewer.py or something). This is a minor feature, but a large code rearrangement, so this might not get done anyway. -- agrif

No intermediate cache, use tiles and re-render only needed tiles. (as pigmap does)

If chunk render speed is improved we can render tiles on the fly, that would simplify the way to keep updated a map. I think at the current speed doesn't worth the try, but if we get Minecraft-Overviewer faster it can make a difference. -- Fenixin

Bigger cache files (if we keep cache as it is now).

From profiling data a ~15% of the time is spent in encoding and saving cache files, we can generate bigger cache images (2x2 or 3x3 chunks) maybe this helps to reduce the encoding and saving time. (idea seen in Issue #52 )