<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 27/02/2014 7:36 a.m., bnorman wrote:<br>
</div>
<blockquote cite="mid:1393439766688-5106188.post@n6.nabble.com"
type="cite">
<p>I would like to use openlayers to display a map of the sky
using custom imagery. I plan to use geoserver on the back end.</p>
<p>Ideally, I would like:</p>
<ul>
<li> Display units of right ascension (in hours 0 to 24 hours),
declination (-90 to 90 degrees) </li>
<li> Seamless scrolling across the right ascension boundary
(0/24 hours) </li>
<li> Seamless scrolling across the poles </li>
<li> Use of image tiles (ultimately converted from FITS files)
in quadrilateralized spherical cube (QSC) or HEALPIX
projection. Both of these are available in PROJ4. I'm not sure
how to store these tiles, because those projections don't
necessarily work with GEOTIFF. </li>
<li> Vector overlays (star maps, etc). </li>
</ul>
</blockquote>
<br>
If you settled for scrolling across the RA boundary but not the
scrolling across the poles, it would make life simpler. I am not
sure how you would do both without modifying OL code. The
projections also look a little challenging. The forward projection
from RA,D -> x,y is straightforward enough but as users pan the
map, you are have to do the reverse projection all the time and with
an interrupted projection, that sounds "interesting". The absolute
key to this is going to the processing that inverse projection. <br>
<br>
I'm interested. I have battled away at OL on the dateline issues and
at Geoserver on the same thing. As far as I can see, they both have
fundamental concept of "Extent" rectangle with maps the mapframe
rectangle in screen coordinates to projection coordinates. There is
assumption that<br>
RH>LH, and inequality equation for top/bottom is constant.
Various checks in the code cope with the 180 line. Inhouse software
used for geophysical mapping deals with various cases of the Extent
rectangle defined by the inequality relations between left/right
top/bottom. These different cases generated different "working
coordinates" for the Extent used to the drawing maths, but panning
etc carefully adjusts the Extent rectangle to working on sphere
rather than flat earth assumption. It would interesting to see what
would be involved in working that into OL. This is something that I
think you need.<br>
<br>
I would rather like to be able to pan seamlessly down from NZ to the
Ross dependency and across the pole. This would be same issue that
you are dealing with.<br>
<div>
<div style="border-top: solid 1px black; border-bottom: solid 1px black;
padding: 10px 0; margin: 20px 0; font-size: 8pt;
font-family: Verdana, Arial, Helvetica, sans-serif;">Notice: This email and any attachments are confidential. If received in error please destroy and immediately notify us. Do not copy or disclose the contents.<br></div>
</div>
</body>
</html>