Oracle to Buy Sun, a second look

For some time now, I've been learning parts of the Java Enterprise Edition platform, and what I've seen is a great platform that can be the best-in-class if used properly and if it was easier to learn.

But what I find so much interesting is the multitude of frameworks in the market, two different presentation layer standards (JSP and its successor JSF), many different JSF extension frameworks (Apache MyFaces Tomahawk, Sun Woodstock, ICEfaces, JBoss RichFaces, Apache MyFaces Trinidad, Oracle ADF Faces), other not-in-the-specification frameworks (Tapestry, Struts, Apache Wicket, GWT, Spring MVC and others), many implementations of the JPA standard (the de-facto standard Hibernate, the JPA2 reference implementation EclipseLink, the Oracle TopLink, OpenJPA, and others), many Servlets containers (Tomcat, Jetty and commercial offerings) and various EJB containers, the complete enterprise applications building frameworks as Spring, Seam, and many others.

And that's just for the parts I explored so far.

Also the desktop front suffers from less overlap/duplication with 2 GUI widgets framework (the Java standard Swing, and the IBM/Eclipse influenced SWT framework) and their respective applications framework (the Swing Application Framework and the Eclipse Rich Client Platform).

The same is true in the IDEs front, we just don't have 2 major IDEs (Eclipse and NetBeans), but we have many commercial IDEs built over the Eclipse foundation (many of the IBM Rational offerings, RedHat/JBoss Developer Studio, Embarcadero/Borland JBuilder), some commercial offerings (most notably the IDEA IntelliJ) and the Oracle free (but less fair to play with the market standards) JDeveloper.

At first I just thought that is a waste of brain cycles and time of a talented and very smart community, and not just how many things has been done and redone and then-again redone, it is just plain confusing to try to learn or even design an application while taking in consideration the pros and cons of each framework/standard/application/library and presents a real challenge for a topic that could be easier to penetrate and learn, also such duplication creates many different-and-separated camps of experienced developers in the market.

But then I found that the real reason all these people are still in the Java camp is because of the ability to do things just as you want it (just think of any barely-reasonable way to do enterprise applications and chances are that a framework is out there waiting for you to be used) and the many different ideologies behind each framework (ranging from heavily-closed source and commercial frameworks to BSD-style open source frameworks) and that the ability of the Java community to re-invent the wheel when desired easily a key feature for the backing of virtually all the big software companies in the world (except Microsoft of course, not because it was uninterested, but because it wanted to play it too-hard with the market).

How such community will be affected by the Oracle acquisition is yet to be seen, but I'm not so optimistic as before, the Java EE may grow better under the Oracle umbrella, but it may be less free and less open to the community, which is exactly what the Java EE community doesn't need, the backing of the various software companies and open source communities what makes Java a market leader and the biggest player in the enterprise applications market.

I hope that Oracle understands the chemistry between Java and freedom (or it keeps enough Sun executives to keep such freedom possible) and that only a multiplayer game is a game worth playing.

New Global DEM released from METI and NASA

The Ministry of Economy, Trade, and Industry (METI) of Japan and the United States National Aeronautics and Space Administration (NASA) at 29th of June jointly released Version 1 of the ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) Global Digital Elevation Model (GDEM).

Some information about the ASTER instrument:

• Building Agency: METI of Japan
• Carrier Spacecraft: USA NASA’s Terra
• Launch date: December 1999
• Stereoscopic capability: Along-track (no need to pass multiple times over the same area to collect stereo image data)
• Bands available for DEM generation: VNIR (near infrared spectral band)
• DEM base-to-height ratio: 0.6
• Spatial resolution: 15m in the horizontal plane
• Single ASTER VNIR scene size: About 60 km-by-60 km ground area (4,100 samples by 4,200 lines)

Some information about the ASTER GDEM:

• Land surface coverage: between 83°N and 83°S
• Grid cell size: 1 arc-second (30 m)
• Tiles Format: GeoTIFF
• Number of tiles: 22,600
• Single tile size: 1°-by-1°
• Coordinate System Type: Geographic (Lat/Long coordinates)
• Earth Model: WGS84/EGM96 geoid
• Collection Date: 2000-2009
• Vertical accuracy estimate: 20 meters (at 95 % confidence)
• Horizontal accuracy estimate: 30 meters (at 95 % confidence)
• Area of missing data: Areas with constant cloud coverage

This mission is remarkable because it involved the automated processing of the entire 1.5-million-scene ASTER archive, it took approximately one year to complete production of the beta version of the ASTER GDEM using a fully automated approach.

The process included:

1. Stereo-correlation to produce 1,264,118 individual scene-based ASTER DEMs
2. Cloud masking to remove cloudy pixels
3. Stacking all cloud-screened DEMs
4. Removing residual bad values and outliers
5. Averaging selected data to create final pixel values
6. Correcting residual anomalies
7. Partitioning the data into 1°-by-1° tiles.

Previously, the most complete Global DEM publicly available was the SRTM (Shuttle Radar Topography Mission), it has an 80% land coverage (between 60°N and 57°S). The ASTER GDEM expands coverage to 99%; also the SRTM data has a grid cell size of 3 arc-seconds (90 m) outside of the USA borders, despite being captured at 1 arc-second, it was later averaged before being publicly released.

I hope that NASA would release the full resolution 1 arc-second data of the SRTM DEM globally to augment the ASTER GDEM data and allow for a unified DEM which benefits from the ability of the Radar to collect data in the places constantly covered by clouds.

I currently don't know the shoreline data used to outline the DEM data, but I hope they release it as well or use a publicly available one, such as the NGA Prototype Global Shoreline (PGS) data.

A good technical page about the new ASTER GDEM can be found here.

METI and NASA marks version 1 of the ASTER GDEM as “experimental” or “research grade”, but the size and accuracy of the data makes such an initial release a very good step toward a highly enhanced and accurate Global DEM very near.

Download sources:

Earth Remote Sensing Data Analysis Center (ERSDAC) of Japan

NASA’s Land Processes Distributed Active Archive Center (LP DAAC)

I will post shortly about my experience with the download of the new GDEM and its estimated size and usage characteristics.

Sources and News Coverage:

NASA’s Land Processes Distributed Active Archive Center

The Map Room, Astronomy Magazine, Universe Today, BBC News