johanley/planisphere

Generate a precise, customizable planisphere as two PDF files (a star chart and a transparency).

planisphere

Generate a precise, customizable planisphere as two PDF files (a star chart and a transparency).

Main documentation.

planisphere's Issues

Error in calculating proper motion for the trivial case when year is 2000.

The Comparator for the Stats object blows up. There is no proper motion when year = 2000.

Delete GreenwichSiderealTime

It's been obsolete for a while.

Fix error near pole in PrecessionOLD

Code for basic star charts, for demos not attached to the planisphere.

Transparency: remove the affordance for centering.

The outermost circles act as a better affordance because they are much bigger. In addition, if the transparency and paper are the same size, then they can also be centered just by placing the transparency on top of the paper.

Remove the DoubleTrouble class

The idea of fusing two close stars into one object, in the context of rendering a chart for the human eye, is a valid one.

But in the current context, it makes less sense. The planisphere makes no fine distinction between magnitudes. Moreover, when the code is used to generate charts over long time scales, the fusion makes less and less sense. That's because the fusion depends on separation, and separation depends on time. For optical doubles, with independent proper motions, pairs can gradually separate over long time scales. The DoubleTrouble class is hard-coded to the present day.

Removing the DoubleTrouble class will make the task of migrating to a new data source easier.

In short, the DoubleTrouble class, and the idea of amalgamating close doubles for this project, no longer seems to be worthwhile to me.

SiderealTime - move 20h to the constants file.

Stray constellation line crossing the full chart, for year -2796

Latitude 46.0, longitude -63.0.
Visible in this range of years: -2755..-2835 (approx).
Straight line moves across the map from 4 o'clock to 10 o'clock.

Use an improved source of data for star positions.

I will be using the output of another github project of mine:
https://github.com/johanley/star-catalog

This won't typically change much in the final output.
But the calculations of proper motion over millennia will certainly be improved.

I've been frustrated with the Yale Bright Star catalog for a while, so I'll be happy to replace it with something I consider to be more robust.

Precession demo chart: separate the code cleanly from the core.

The lunar transit table has the rivet in its centre.

It might be possible to split the table of lunar transits in two, leaving a small space for the rivet.

Remove items in the Star class that aren't used.

There are copy-paste artifacts from a previous project that aren't used in the current one.

Moon table: include the fraction illuminated.

This immediately tells the user an important fact about the Moon - how bright it is.

Revisit all docs and outputs, using the new star catalog data.

Don't filter the list of stars by geometry.

Because of precession and proper motion, it's not logical to subset the stars on any specific date (J2000) using, for example, a range of declination.

Add a central dot to the transparency, for centering

Drilling holes separately in the two pieces is a useful option. This central dot will allow that.

Projection is poor for places in the tropics

Add affordances in the corners that correspond to the seasons/months.

The planisphere needs to rotated to be used. You hold it by a corner of the card stock. You hold one of 4 corners. Put icons for the 4 seasons in each corner, such that you hold it in your left hand, and your left thumb covers the current season, and the chart is roughly oriented for typical use for an evening. That is, with the current month near/at the bottom, ready to go.

Code repetition for drawing tasks

There is some copy-pasta repetition for low-level drawing operations on the PDF's drawing context.

By default, suppress Polaris, to make it easier to center and align the chart with the transparency.

On the chart, add horizontal lines as an aid to cutting the card stock.

For 8.5x11 card stock, add horizontal lines above and below the chart.

Or, apply any time the width is less than the height.

Deploy in a servlet, so that people can generate the PDFs using a simple web form.

Not necessary to allow config of all configurable items. Just the basics should be fine: lat/long, time zone, year, declination gap, location name.

Put the year in the PDF metadata.

Reinstate affordances on the transparency for centering

There are two separate issues: the centering of the transparency on the chart, and the placement of the rivet. Those are two different tasks.

The centering of the transparency can be done using BOTH the outermost circles (for fine tuning) and the central affordance.

The placement of the rivet can ONLY be done with the central affordance.

The central affordance needs to allow for rivets of different sizes, large and small.

Add config for the grey-scale value of the alt-az circles.

Using the same grey as the constellation lines is just barely visible. Using full black overwhelms the underlying star chart.

Need to experiment with a spectrum of values.

This setting might be very sensitive to the printer/transparency being used.

Transparency: the alt-az circles are probably too faint.

Printing on a transparency seems to render quite a bit differently than printing on paper. It seems to take more emphasis (darker color) in order to make the same visual impression.