Draw hints behaviour for Comets and Asteroids

Bah, stupid sign reversal... I did mean to increase visibility for faint objects.
Fixed in trunk r7217.

Georg, your latest commit make visibility of the hints and names worse. Seems we should use +4.f (maybe +5.f) for comets/asteroids and +2.f for planets.

HuH?

I first tried with plus5, this was even worse than the plus2, though. minus2 made comets visible for a longer time (still visible beyond the smallest stars).

Ah, hints and names? No, I fixed draw(). The hints/name balancing is elsewhere, but I am blind currently?

drawHints() called within draw() ;) Current implementation will be not reaches this call.

if ((getVMagnitude(core)-2.0f) > core->getSkyDrawer()->getLimitMagnitude()) return.

assume sky limit magnitude=8, comet magnitude=9.5, and I still want it displayed. 7.5>8=false, so we do not return but draw it.
Previously, I had 9.5+2 > 8 = true, therefore comet was not visible, and this was indeed a silly late-night error.

The sign must be negative, and the comets are visible for 4 more magnitude steps now than before, or for 2 mag steps beyond general sky brightness. We can increase this further of course, if required, but I think it is now well-balanced.
I cannot reproduce how hints visibility can be now worse than before?

getLabelsAmount() broken now as fact, because drawHints() call after checks for comet visibility. From your example all hints and labels will not be displayed for objects dimmer than magnitude 8. If we set amount for labels near magnitude 13 then we don't get any changes. For example in attachment screenshot you can see visible of hints and labels for planets in version 0.12.5 when slider for planets labels and markers set in extreme right position (all comets from MPC's list of observable comets was added). In the current trunk you get visibility of hint and label only for Jupiter.

Actually for my example, mag9.9 comets will be shown in a sky showing mag 8 stars. Indeed, hints for mag11 comets are not drawn here. Do you want to have hints for mag15 objects in a 60°FOV mag6 sky? Then we can just increase the limit, but then we may again have a problem with too many dim comets slowing down fps. We can have hundreds of comets computed in the background when we only display what's visible "and slightly more".

I just tested loading all MPC comets and displaying without this early-out test, and with a mag difference of 8. The latter is acceptable, without the test frame rate dropped too much IMHO. Or maybe introduce two more config variables, astro/comets_hints_magdiff(default 5?), and astro/planets_hints_magdiff (default 3?), that can be adjusted to the computer's CPU power?

Georg, showing comets and hint/label of those comets is different features :)

But they were always linked like this, or not? If you want to show all possible comets as hints, we can move the early-out test (Comet.cpp line 455) to after drawing the hints (l. 499), but then we will likely have again conplaints about slow fps because all the tests must be made also for mag20 comets. What do we do?

Alexander, can you test if you like this behaviour of comets now, and if it causes too much loss in frame rate.
A similar change can be made in Planet.cpp, but if somebody again has 11000 NEAs, fps loss is massive, i.e. the same issue we had before 0.13.1.

Current behaviour is very good for me. I'll prepare package 0.13.1.91 and ask the original reporters for testing.

Wait, I did not change behaviour for the planet(oid)s yet. Do you want to do that yourself? Just remove the test in Planet.cpp line 970. But then we have the "11000 NEAs slowdown" problem again if people install all those elements. OTOH, then occultation by dim objects works again. Not easy to fulfill all desires...

Or maybe leave the line in, but increase mag difference test to 4-5? But test that with all NEAS! Or add a config option, because this test may prevent occultation simulation.

I'm building 0.13.1.91 now - IMHO version with increase mag difference test to 4-5 should be better than without any versions - it's a good compromise between visibility and performance.

Georg, can you increase mag difference test to 4-5 for comets and asteroids (and disable current behaviour)? Maybe we should introduce limit magnitudes for SSO also, like for DSO and stars?

I have now magnitude difference of 5 for both planets and comets. Dimmer objects will not get a hint circle unless you zoom in further to increase sky limit magnitude. Still, as I pointed out for 0.13.1, having the test in Planet.cpp will exclude dim asteroids from stellar occultation, not sure if this can be accepted. I hope it's a good compromise now.

(This was commited as r7230).