Rebased Rescaling Experiment IV

[mcarans]

Rebased version of PR: #287

Relates to: #584

[mcarans]

@Redspear-of-Oolite @oocube @LW-archlinux How do you find this rescaling?
(Packages here: https://github.com/mcarans/oolite/releases/tag/1.93.1-PullRequest287.7994)

[oocube]

I ran ./Oolite-1.93.1-PullRequest738.3157-dev-x86_64.AppImage. When exiting the station the planet is indeed much bigger and the sun is further away. In this respect I like the change. When flying towards the sun, eventually I looked back to see the station and the plant. And I got to see all three moons - and they are absolutely unrealistic. Or is that just my impression?

[phkb]

That would be an issue with the OXP, rather than the rescaling code itself. Several OXPs will need to be tweaked to match the size changes, particularly ships.

[phkb]

But one of the primary things we would need to test in this PR is the number of encounters ie. how many ships do you meet between the witchpoint and the station, with the goal of having a similar gameplay experience to 1.92.1. That would mean testing with no OXPs installed, or in strict mode.

[mcarans]

That would be an issue with the OXP, rather than the rescaling code itself. Several OXPs will need to be tweaked to match the size changes, particularly ships.

Hmm, that sounds like a major piece of work involving coordinating changes with OXP authors. On the other hand, it's an opportunity to get OXP authors interested in this feature and get some Commanders to try it - which hasn't happened so far even though I've asked on Discord, Reddit and the BB.

I looked more closely at how this has been implemented. It could be beneficial to have many of the hard coded values exposed and changeable by OXP instead. I see some multipliers like sun_distance_multiplier and planet_distance_multiplier already exist in the Obj-C as values in a dictionary but it looks like they aren't changed anywhere and always use their default values. I'm unfamiliar with the mechanics of how they might be changed by OXP and how then other OXPs could be coded so that they respect and adjust to the values that have been changed to avoid the issue in oocube's screenshot.

How would you suggest taking this forward so it doesn't block other work like the long overdue reformatting of the code to modern standards?