MSUM Archive of the MAPS Catalog News
I fixed the communication issues with the SIMBAD name resolver (in fact, it might be a tad bit faster with the streamlined code). I have re-enabled object name based searches on the Sky Searches and Finder Charts.
Due to data privacy rules, I had to move my entire web server to a separate server from my desktop machine. We took advantage of the move to improve security of the server. One advantage of this is I was able to get the Finder Charts service repaired. However, I discovered the new server is having issues with communicating with the SIMBAD name resolver server. As such, I have disabled object name resolving on the Sky Searches and Finder Charts until I can figure out the problem.
I upgraded the server this website resides on to MacOS 10.6.1. I am using the built-in Apache2 and PHP (5.3.0). I have disabled the Finder Charts feature on this site because it is not used much and I don't have time anymore to custom compile PHP to support JpGraph's requirement for of FreeType support to be compiled in.
As I previously noted, when I updated to using Leopard, I lost the ability to use JpGraph because of a problem with Apple's built-in PHP installation. After waiting to see if Apple could resolve this, I decided to roll my own PHP install. This proved troublesome, at least if you wanted to use the built-in Apache2 server. As such, I switched to using MacPorts version of Apache2 (2.2.9) and PHP (5.2.6), hacking the version of FreeType (2.3.7) included with MacPorts, and integrating it with my current MySQL server (5.0.67) as well as updating the JpGraph and NuSOAP libraries got everything working again.
I have updated the operating system on my server to MacOS
X 10.5 (aka Leopard). I have also upgraded MySQL to
version 5.0.51a and am using this OS's built-in PHP 5.2.4
hacked to support GD. This is working, however, any routines that call JpGraph crash out in an error of the form:
Break on __THE_PROCESS_HAS_FORKED_AND_YOU_CANNOT_USE_THIS_COREFOUNDATION_FUNCTIONALITY___YOU_MUST_EXEC__() to debug. The process has forked and you cannot use this CoreFoundation functionality safely. You MUST exec().The specific call in JpGraph is
$graph->Strokeand I suspect that this command forks some processes to speed up drawing and it triggers this (new) MacOS X error. This is new to Leopard as noted in this Apple TechNote.
I updated the menus to a pure CSS driven menu system and added a sitemap for ADA compliance. Also switched some layout tasks to Dreamweaver for easier code maintence.
I had to tweak the code for the Finder Charts to work with PHP 5.2.0.
I have changed jobs to a tenure-track position at Minnesota State Unviversity Moorhead, so I am taking the MAPS Catalog with me.
I am working on upgrading the MAPS database code to run on MySQL 5.0.18 and PHP 5.1.2. The beta server is now running the binary distribution of MySQL 5.0.18 for MacOS 10.4 (MySQL.com now produces this with each build) and I compiled PHP 5.1.2 from the source code. The operating system on the beta server has been updated to the latest version of MacOS X (version 10.4.4) with all the security patches.
I have updated the MySQL database that manages the MAPS data to version 4.1.3.
I have updated the hosting computer for this service to the latest version of MacOS 10.4.1. This upgrade allowed me to update MySQL to version 4.1.12 and PHP to version 4.3.11. It has also sped up the computer a bit.
Queries with circular bounds have now been implemented in the Sky Query (renamed from "Skybox Query") form. This allows you to limit a query to all objects within a given distance of a target position (useful for cases were positions are uncertain). Circular queries also output the angular separation between the searched position and the object position as an additional output column.
Position angles in the database are now all stored as degrees East of North ranging from 0 to 180 degrees, regardless of coordinate system. This should make comparisons to other catalogs much simpler.
There was a problem resolving positions near the equator
(This is a negative zero degrees declination begin
interpreted as positive zero degrees issue). This has been
I am also dealing with the discovery that the MAPS position angles are measured W to N (from the "x" axis) instead of the standard N to E. For now I have made a change to the Finder Chart code to reflect this and noted the change in the documentation.
Since this project and the MAPS
Project were basically replicating efforts, we agreed to settle
on the same code base. Today the project switched to using the same
source code as the MAPS project. The most obvious side effect of this is some minor changes in
the layout of this site.
I am still working on implementation of "circular" querying about a given location and am about 90% there.
I have cleaned up some of the code. This is in preparation for the eventually sharing of the PHP and MySQL codebase between the SCSU Archive of the MAPS Catalog and the original MAPS catalog website at UMN. I am working on adding positional searches that return everything within a given angular separation instead of just a square skybox, a process which required some major behind the scenes changes to the MySQL database.
I have also upgraded to PHP 4.3.9 and MySQL 4.0.21. I moved away from using Fink's implementation of MySQL instead choosing the Complete MySQL implementation from Server Logistics, which is not only easier to install, but also offers better compatibility with applications expecting MySQL to be in /usr/local/bin/mysql instead /sw/bin/mysql. This wasn't without pitfalls, as their code for autostarting MySQL at boot up didn't install with the proper permissions, so I had to edit the permissions of the scripts in /Library/StartupItems/MySQL/MySQL to be executable.
I had never used Fink's PHP, prefering to upgrade the version of PHP Apple installs under MacOS X using the PHP source code for maximum flexibility, so that upgrade was relatively painless. I do use Fink's versions of variouos graphics and XML libraries, it's just that I needed flexibility that a package manager doesn't offer in my PHP and MySQL installs.
The MAPS Project went live with their implementation of my MAPS site code!
I have finished the implementation of the new Finder Chart code. The interface is the same, but the code now uses gnomonic projection to plot up the objects, so it can accurately represent any small field on the sky! The code seems to run about as fast as the original, not so sophisticated Finder Chart code. I have also added object labels as requested by one user.
Apparently when I updated to PHP 4.3.8 there was a change to how formatting of printf() statements behave in PHP. This broke the output for this site. This is a big change and I am surprised the PHP team would have snuck it in an incremental change (instead of just moving this to PHP 5). Apparently a few other people are upset about this change. I am working on revisions to fix this and have updated to PHP 4.3.9-RC2-Dev in the meanwhile to deal with this change.
I have also worked out how to resolve the issue of the Finder Charts becoming distorted at high latitudes, but the fix will take a bit of time to implement.
With today's updates to the documentation and viewer for the ancillary data, I have now gotten 'feature complete' for this site. I consider the SCSU Archive of the MAPS Catalog to have entered 'beta' status where all the features I intended to have in the catalog have been added, but debugging may not be complete.
All the ancillary data from the MAPS Catalog into the database for interactive access, including:
- Astrometric fit parameters for each POSS I field
- Lists of which regions of each plate were cut due to flaring around bright sources, nebulousity, etc.
- Details on the stellar magnitude diameter to magnitude relation including the list of standards used to accomplish the diameter-to-magnitude fit on each plate.
- Details on the integrated magnitude density to intensity relation for each plate.
Thanks to Laura McDonald of SkyView for adding an inverse black and white color map to their color map choices. It means I can now present SkyView images of the POSS I in a form that looks like a photographic negative (much easier to pick out faint objects on than a "positive" image).
Also, as I work toward placing all the ancillary data online, I have caught and fixed two duplicate entries in the photometry standards information.
- P211.DM.standards.txt: Duplicate entry for cap-23
- P324.DM.standards.txt: Duplicate entry for cap-778
- P777.cuts.txt: Was missing estimated RA and Decs for two circular cuts.
- I also fixed many duplicate entries in cuts files. These were due to multiple runs being recorded in the cuts log when MAPS projects had to be manually re-processed.
All of these changes have been reflected in the online MAPS catalog database and changed files (relative to the DVD distribution) are available in the downloads area.
I have cleaned up the Finder Chart code to provide a more accurate representation of the field, notably using position angle and ellipticity information to better draw the galaxies in the field. However, the code will still fail at the poles. I will eventually have to use some sort of spherical projection algorithm, but that will require a significant re-write of the code, which I do not have time for at the moment.
I have placed all the files originally distributed on the MAPS Catalog DVD set online and available for download here. I don't expect many people will need access to the raw datafiles now that many tools for accessing the MAPS Catalog are online here, but just in case you need them, they are online, including any of the updates.
It looks like the original (diameter-based) magnitude limits listed in the MAPS Catalog datafiles were corrupted in some cases (e.g. - P058, P243, P402, P586, P610, P885, and P879 have a limiting E magntiude of 0.00). As such, I decided to redetermine these limits using the existing MAPS database, and while I was at it, I computed separate limits for diameter-based (stellar) and integrated magnitudes on each plate (luckily this took only about 30 minutes with the new database).
As a result, I have dropped the maglo and magle parameters from the plates table and added magdomax, magdemax, magiomax, and magiemax to track all these parameters.
In this process, I also discovered three objects (out my 89 million) had bad magde values, which
I traced to bad colord values in the original files on the DVD. Darn. Oh well, I used the archived
photometry on the DVDs and the stored E plate diameter and ellipticity information to reconstruct the magdE and colord
values for these objects:
In the good news front, I have implemented a basic finder chart service. It's a bit touchy if you reload the page (I think there is caching issue here), but otherwise it seems to work quite well.
SkyBox Queries of the MAPS Catalog also return links to other online catalogs and imaging (if requested, its on by default). This is meant to allow quick querying of multiple services by researchers.
After much pain and anguish learning the basics of SOAP, I have implemented SIMBAD's name resolution service in PHP. This allowed queries to be done on objects by name instead of position.
All of the query modes that the original MAPS Project has have now been implemented here. In addition to their query modes, I have added an advanced query interface that allows relatively complete control of the output version of MAPS Catalog searches.
On the agenda for the short term is some cleanup of the output to be better formatted and then possibly a finder chart interface. After that, I have to get back to my other research for the moment.
An interesting tidbit is that the MAPS Project folks have taken a strong interest in my reimplementation of the MAPS Catalog and will likely be using my source code shortly for their own website.
I have fleshed out this website a lot more, instead of it being just a copy of the MAPS Project. It now conforms with St. Cloud State University's web design preferences and has its own cheesy logo so no one confuses it with the MAPS Catalog stored at the U of Minnesota.
I am now working on getting the actual querying code for online queries working.
I have completed the creation of a SCSU archive of the Minnesota Automated Plate Scanner (MAPS) Catalog of the POSS I. This local version has several "tweaks" compared to the version that has been online at the APS Lab at University of Minnesota, mostly designed to help me get my research done. However, you may find these "tweaks" useful as well. The differences between this database and the version at the U of Minnesota include:
- The SCSU archive adds J2000 and Galactic coordinates to all objects directly in the database computed using the same algorithms as the NASA Extragalactic Database (NED). This makes searches in J2000.0 and Galactic coordinates much, much quicker.
- The SCSU archive adds estimates of Galactic extinction E(B-V) based on bilinear interpolation of extinction estimates from the dust maps from Schlegel, Finkbeiner, & Davis (1998).
- The SCSU archive is indexed on all the coordinate positions (B1950, J2000, Galactic) as well as all magnitudes, diameters, and E(B-V). This provides much faster querying where these variables are concerned.
- The SCSU archive is implemented in MySQL 4.0.15 running on a Power Macintosh G5 (2x2GHz) with 1GB of RAM located in my office on the campus of Saint Cloud State University.
This archive of the MAPS Catalog is completely independent of the U of Minnesota effort, although I am grateful to Matt Gray of the U of MN for allowing me to retrieve some code they used to implement their MySQL version of the MAPS database online in Spring of 2004. Any questions about this database should be addressed to me (Juan Cabanela) and NOT the kind folks at the U of Minnesota's APS Project.