Talk:StarChart

From OLPC
Jump to navigation Jump to search

This program can be much more than just a map of the night sky. Offhand, I can think of a whole bunch of things it can demonstrate. A couple of less-obvious ones are:

How the human mind sees patterns in what is actually random arrangements of points. (The constellations.)

That bright stars are less common than dim ones. (Because to be bright, a star has to be either unusually luminous or unusually close.)

Which pairs of stars rise or set at the same time depends on your latitude. (This is how the Polynesians navigated, I'm told.)

Some more obvious ones:

The earth rotates once in 24 hours. Which makes the sky appear to move at a rate of 15 degrees per hour from east to west.

The earth goes around the sun once in 365.242 days. Which makes the sky appear to move at a rate of about one degree per day from east to west.

The sun, the moon and the planets follow roughly the same path against the stars (the ecliptic), but they move at different rates. And because the earth goes around the sun, too, some planets appear to reverse direction at times.

Changing to a different longitude is like changing your clock.

Unless you're on the equator, some stars never set.

Davewa 16:07, 22 May 2008 (EDT)

Constellations and Cultural History

Each culture has its own set of constellations and stories about how the sky and the universe was formed. StarChart can be used to introduce other cultures via their constellations and stories. And can be an adjunct to local cultural history. --User:AuntiMame 22 may 08 16:14

Developer's reply: Beta 4 has broken out the constellation table from the code. So in theory a local version of "constellations.py" could be created to show the local culture's figures today. Of course, today nothing is labeled -- as was noted below, that needs to be changed, too. The constellation table organization would need a tweak to support adding a name and an abbreviation. The code would have to be modified to support this change to the structure of the data (presently simply an array of line segments; after the change an array of structures consisting of two strings and an array of line segments). This is all perfectly feasible to do.

But to create a local constellations.py will require a local expert who both knows how to draw his culture's constellations and who knows enough Python to be able to successfully modify the data for the stick figures (each line of which requires specifying two equatorial coordinate values) and for the labels. Davewa 11:08, 23 May 2008 (EDT)

Labeling

How about labeling planets, bright stars, constellations, etc. as you hover over them? The Astrological symbols for the planets are not that commonly known, hovering and seeing a label 'Mars' pop up could be a great help. Same with bright star names Rmyers 17:27, 22 May 2008 (EDT)

Developer's reply: Other than quibbling about how the label happens and whether it pops up, gets drawn over the chart or gets displayed in one corner of the margin area, I completely agree. My plan had been to detect right-click and figure out the brightest object at or nearest to where the user clicked. Then I'd show the object's name, the name of the constellation it is in (for stars and deep-sky objects), the object's brightness (magnitude) and catalog designation (Bayer number for stars; Messier (or occasionally NGC number) for DSOs) and for DSOs additional data such as the object's type (e.g.: "galaxy") and angular size.

I hadn't planned to separately identify constellations -- getting the constellation name would be a side effect of getting the star name

But "terse label on hover-over" in addition might be useful. -- Davewa 11:18, 23 May 2008 (EDT)

Milky Way

What about a Milky way graphic kicking in at an appropriate brightness -- say about mag 4? With a dark sky this is a major orientation feature. Rmyers 17:27, 22 May 2008 (EDT)

Developer's reply: Living under severely light-polluted skies, I never even thought about that! But you're right. It's worth adding. -- Davewa 11:21, 23 May 2008 (EDT)

Low magnification points of interest

What about adding a layer with points of interest not visible with the unaided eye, but which can be seen with a small telescope or binoculars? M catalog objects, prominent double stars and the like. Rmyers 17:27, 22 May 2008 (EDT)

Developer's reply: That one's already in my project plan for version 2. I expect to add a feature whereby a left-click will bring up a six-degree-square view centered on the nearest, brightest object at the clicked coordinate. (Left-clicking again restores the full-sky view.) Six degrees is the field of view of good binoculars (magnitude of about 7X).

As for a smaller field of view, whereas that would support a telescope better my assumption is that any school which could provide a small telescope for an astronomy class could probably also have a better planetarium program on a more elaborate computer to use with it. (I use Cartes du Ciel, myself, for my astronomy -- it's free and you can download the source, but I'm not sure if it's FOSS, technically.) So I'd rather target naked-eye and binocular observers with this program and let the telescope-equipped find a better tool if they need one.

I also plan to expand the star catalog to a depth of magnitude 8.0 and plot these dimmer stars in the magnified view only. I don't want to go dimmer than magnitude eight because as you probably know, the catalog size grows exponentially as you add dimmer-magnitude objects. The smaller, dimmer DSOs will probably not get added: my rule was not dimmer than magnitude 8 nor smaller than Mars in angular diameter -- this is consistent with binoculars or a small telescope. -- Davewa 11:35, 23 May 2008 (EDT)

'Sky View'

I don't know if this is computationally feasible but...

What about a view that more closely resembles looking at the sky? Show a view from horizon to zenith looking in a specific direction with a view 90° (?) left and right. That is if someone is facing west he'd see from south to north, horizon to zenith -- more closely resembling his eye view. Panning left and right would change the direction of the center point.

In keeping with the education mission, this would assume less 'mapping' skills on the part of a young user. Mapping a flat sky chart to a hemispherical bowl over your head is not a simple task, particularly with objects near the horizon. This could give an alternative view with a less distorted view for objects lower in the sky. Rmyers 11:18, 23 May 2008 (EDT)

Developer's reply: I've seen this view in Cartes du Ciel and find the one that CdC provides to be harder (for me) to use than the whole-sky chart. The mapping of a quadrant of sky results in an onion-shaped figure that just looks weird to me. I don't know enough about mapping theory to produce a view of the sky which would look enough like what you actually see to be intuitive. My current mapping is pretty simple: azimuth maps to location around the circumference and altitude maps inversely to radius.

I find if you simply hold the XO so that the display is perpendicular to the ground and the direction you're facing is down then the map is pretty close to what you see from zero to 45 degrees altitude. Above 45 degrees, the best thing to do is lie down with your head facing north and hold the XO over your head with south at the bottom. -- Davewa 11:45, 23 May 2008 (EDT)