VisAD library update
Tom Whittaker
tomw at ssec.wisc.edu
Mon Jan 29 12:48:41 MST 2007
The VisAD library has been updated. The changes since the last update
in Feb, 2006, are listed below. I apologize for the length of time
between updates -- I had computer hardware and software issues that
kept getting in the way...
Note that there is now a "build.xml" file in the main 'visad'
directory that can be used with 'ant' for doing builds. Thanks,
Bruce!
You may pick up the latest by clicking the links off the VisAD homepage
<http://www.ssec.wisc.edu/~billh/visad.html>
or via FTP from:
<ftp://ftp.ssec.wisc.edu/pub/visad-2.0/visad.jar>
<ftp://ftp.ssec.wisc.edu/pub/visad-2.0/visad_doc-2.0.jar>
<ftp://ftp.ssec.wisc.edu/pub/visad-2.0/visad_src-2.0.jar>
If you have any issues with the new JAR files, please let me know.
These were all built
using 'ant' this time.
Now that my system is pretty much working again, the updates will be
more frequent....I promise!
tom
------------------------------------------------ChangeLog-------------------------------------------------
2007-01-25 17:35 jeffmc
* visad/: util/ThreadPool.java, ActionImpl.java: Add a getTaskCount
method
2007-01-25 09:57 brucef
* visad/build.xml: Added jar.src target to jar up the distrubution
source files. It jars the source files corresponding to the files
that are compiled in the jar and dist targets.
2007-01-24 11:05 brucef
* visad/build.xml: You should always test before commiting, right?
Fixed itty-bitty syntax error.
2007-01-24 10:59 brucef
* visad/build.xml: Packages/directories to compile are now
explicitly listed at the top of the build file as patternsets so
the compiled classes now matches what was compiled by the
Makefile. Unfortunately I was not able to find a way to use the
same listings for creating the javadocs, so there is still a
seperate doc.packages patternset.
The dist target used to be the same as the jar target. This has
been modified so the dist target now cleans first.
2007-01-23 17:28 donm
* visad/data/vis5d/V5DStruct.java: don't check for max row and
column size. keep checks for mins though
2007-01-23 16:24 rink
* visad/: java3d/DisplayRendererJ3D.java,
bom/ScreenLockedRendererJ3D.java: Add a couple of new signatures
for setting the locked trans proj to the current one - this
proved useful in locking text in HYDRA displays - no existing
logic changed.
2007-01-23 12:01 rink
* visad/BaseColorControl.java: add a convenience signature for
initTableGreyWedge - inverted grey scale
2007-01-17 16:57 curtis
* visad/: all: 2007 copyright.
2007-01-17 16:49 curtis
* edu/wisc/ssec/mcidas/: all:2007 copyright.
2007-01-08 12:14 donm
* visad/data/units/: UnitParser.java, UnitParser.jj: handle
positive exponents for units that have "+" in them.
2007-01-03 14:45 brucef
* visad/build.xml: Tom W. discovered that the jar target for making
the distribution jar file did not work under windows, so I had to
change the incredibly cumbersome manner in which the dist jar
file was created.
There is now a patternset at the top where the file patterns for
all the files to be included in the jar are defined. This may
need to be fine tuned some because now it's pretty general,
including just about all class files.
Also, I should note that the changes were tested on my MacBook
and a Parallels VM running Windows XP both running Ant 1.6.2 on
top of Java 1.5 and everything worked fine.
2006-12-28 13:44 rink
* visad/Contour2D.java: Fix last fix for an
ArrayOutOfBoundsException; make sure contours are labeled if only
one contour level.
2006-12-27 10:54 donm
* visad/Contour2D.java: fix an ArrayIndexOutOfBoundsException when
there is only one contour level
2006-12-18 16:42 donm
* visad/Gridded3DSet.java: allow alpha for color filled contours
2006-12-15 11:09 rink
* visad/java3d/: DefaultRendererJ3D.java,
ShadowFunctionOrSetTypeJ3D.java: More work on Animation
rendering: if any ScalarMap to Animation, make the Renderer's top
branch live, ie setBranchEarly, otherwise do things as before.
IDV and visad/examples look OK.
2006-12-14 08:54 rink
* visad/Contour2D.java: remove some debug stuff.
2006-12-14 00:09 donm
* visad/: FieldImpl.java, FlatField.java, GriddedDoubleSet.java,
Linear1DSet.java: add a resampleDouble method to FieldImpl and
FlatField (with appropriate comments). Make GriddedDoubleSet
extend GriddedSetIface and have Linear1DSet implement
GriddedDoubleSet. If the sampling domain of the Field is a
GriddedDoubleSet, we use resampleDouble to maintain precision.
2006-12-13 09:57 rink
* visad/Contour2D.java: Improve detection of EASY/HARD case,
rearrange some internal code for this. Fixed a performance
problem, and removed some 'static' declared test code and
comments.
2006-12-04 14:49 donm
* visad/Gridded1DDoubleSet.java: fix a typo in an exception
2006-11-17 16:24 rink
* visad/Contour2D.java: implement different strategy for labelling
contour lines on very fine terrain-like grids.
2006-11-16 12:05 donm
* visad/AxisScale.java: add get/setTicksVisible methods
2006-11-15 11:45 donm
* visad/Set.java: fix a problem with __getitem__
2006-11-03 12:28 rink
* visad/java3d/ShadowFunctionOrSetTypeJ3D.java: animation logic:
check MathType of recursed data, not top-level.
2006-10-31 16:03 donm
* visad/georef/EarthLocationLite.java: return a LatLonTuple for
getLatLonPoint() instead of this
2006-10-29 12:00 donm
* visad/data/mcidas/PointDataAdapter.java: only set the realtype
for Z to Altitude if we are using aliases. Tom W, can you check
some other apps that use PointDataAdapter to make sure this
doesn't break them?
2006-10-29 10:00 donm
* visad/ShadowType.java: make sure we pick up the right contour
control
2006-10-26 15:27 brucef
* visad/java3d/ShadowFunctionOrSetTypeJ3D.java: Added logic to
doTransform(...) to determine whether to do the animation code or
default code, thereby bringing the AnimationRenererJ3D
funtionality to the default.
Also, using the
RendererJ3D.setBranchEarly(...) logic, the main group is now made
live before rendering has completed.
2006-10-26 15:23 brucef
* visad/java3d/DefaultRendererJ3D.java: added extend capabiility to
branch in doTransform(...)
added main for testing.
2006-10-20 10:16 rink
* edu/wisc/ssec/mcidas/adde/AddeServerInfo.java: add capability for
application to specify whether the ADDE server is an 'archive'
server along with an archive date to satisty 'DAY='.
Note: the application needs to know when this is necessary. ADDE
provides no query mechanism, other than an Error, to determine
this.
2006-10-18 13:23 brucef
* visad/ShadowType.java: in makeContour(...) removed buffered
previous values for polygon offsets in leu of cloning the
GMC.
If filling clone the GMC and set the polygon offsets and pass
that
GMC to the shadow_api.add*ToGroup(...) methods
2006-10-17 22:15 rink
* visad/Stream2D.java: initial commit for problem with missing
spatial samples.
2006-10-17 22:02 rink
* visad/: Gridded3DSet.java, ShadowFunctionOrSetType.java,
ShadowType.java, java2d/ShadowTypeJ2D.java,
java3d/ShadowTypeJ3D.java: Re-order Data transform path for
isosurface rendering *only* when spatial_set cannot be formed as
a result of a display coord_sys transform, eg. a global model
mapped to a GVAR navigated display.
This involves adding a new signature,
makeIsoSurfaceMissingSpatial, which is invoked by makeContour for
the case above. The new method is essentially the same as
'makeIsoSurface' except that the isosurface is computed in data
coordinates and then map scaling, coordinate transforms and
spatial offsets are applied to the geometry rather than the
sampling domain.
This new method is necessary for two reasons: (1) testing
revealed that the marching-cube algorithm will not work
consistently if there are missing samples. I wish it would
have.
(2) The scaling and transforms of the resulting geometry must
be applied before the normals are computed for consistency.
I've tested this pretty thoroughly in the IDV and the resulting
isosurfaces look consistent with those generated via the normal
path.
2006-10-13 13:05 brucef
* visad/: ContourControl.java, ShadowType.java: === ISSUES
============== On Oct 13, 2006, at 8:48 AM, Don Murray wrote: -
if I run Test05 in the visad/examples, I always have labels
whether the checkbox is checked or not. - if I plot a
color-filled contour of sea level pressure (2D field) in the IDV,
I automatically get labels as well. I might have to tweak
something on my end, but if possible, I think the default should
be no labels if using color filled for backward compatibility. -
also, with the previous plot, only some of the labels show up
unless I tilt the box slightly. I saw in one of your checkins
that you were offsetting the labels slightly, but perhaps, it
needs even more of an offset as in your comment in ShadowType:
// BMF 2006-10-05 set offset to make labels more clear
// FIXME: There may be a better value to use here
Also, you might want to change this only temporarily and set it
back afterwards in case someone has explicitly set it through a
constant map.
=== FIXES ==== I have made the following modifications :
* fixed the fact that the labels were always drawn It was
just a piece of code I forgot to remove
* the label color will be white for filled contours if it has
not been previously set via a call to one of the
ContourControl.setLabelColor(...) signatures. If the color has
been set the filled contours and contour line will have that
color.
The polygon offset is now buffered when contours are filled and
reverted when contours are turned off.
2006-10-12 15:29 brucef
* visad/java3d/ShadowTypeJ3D.java: ProjectionControlListener
Changed static array iteration to a loop
addLabelsToGroup(...)
Sperated stetchy line segment transforms from label
transform code
to enable by-passing segment code for filled
contours
2006-10-12 15:23 brucef
* visad/ShadowType.java: in makeContour(...):
Set the polygon
offset factor to an appripriate value if filling and 0 otherwise.
Add code to get the iso contour map for
Set.makeIsoLines(...)
Ensure labels are drawn whether filling
or not
2006-10-12 14:23 donm
* visad/data/vis5d/Vis5DVerticalSystem.java: add in float
to/fromReference signatures for the pressure to height CS
2006-10-12 13:57 brucef
* visad/Gridded3DSet.java: Added code to static contour(...) to get
the color from the control.
Moved some code arround in
makeIsoLines(...) to ensure lables are still generated when doing
a filled contour
2006-10-12 13:53 brucef
* visad/ContourControl.java: Added getter/setters for the label
color.
In the case of a call to setContourFilled(...) the label
color is set to white.
2006-10-12 13:50 brucef
* visad/Contour2D.java: Modified necessary method signatures for
propagating the label color.
2006-10-10 08:34 donm
* visad/CartesianProductCoordinateSystem.java: every now and then,
you should go back and look at your old code and see how bad a
programmer you once were. Remove lots of unnecessary memory
allocation and don't clone values. Let the underlying CS's
handle the copying or not copying just as they would if you
called them directly.
2006-10-04 15:07 brucef
* visad/Contour2D.java: ContourStrip line arrays with point
quantities below a set threshhold are interpolated to increase
the number of points. This allows labels to show on short strips.
Moved algm. for displaying dashes from Contour2D.contour(...) to
ContourStrip.getDashedLineArray(...). Contour2D.contour(...) now
creates and passes a boolean flag array, indicating which levels
are to be dashed, to ContourStripSet.getLineColorArrays(...)
which in-turn sets if the ContourStrip will be dashed or not.
2006-10-04 15:03 donm
* visad/georef/EarthLocationLite.java: compare apples and oranges
2006-10-04 14:49 donm
* visad/georef/EarthLocationLite.java: lightweight Earthlocation
class
2006-10-03 11:53 donm
* visad/FlatField.java: fix a problem with clone when the RangeSets
are doubles
2006-09-20 16:26 donm
* visad/FlatField.java: make cloneFloat and cloneDouble methods
protected (only the ones that are used). Pass copy == false for
CS.transformCoordinates in resample
2006-09-14 00:02 brucef
* visad/build.xml: I wrote the build script with the assumption
that all the packages would be checked out. It appears that's
not a valid assumption, so here's what I did:
1. removed all the dependencies from the visad target 2. added a
deps target that will compile all the dependencies 3. added a
library directory to the classpath
If you wish to compile visad against your libraries rather than
the cvs source you can do the following:
ant -Dlib.dir=$LIB_DIR visad
Where $LIB_DIR is the directory, either full-qualified or
relative to the visad's parent directory, where your libraries
are kept. All jar files in that dir will be added to the
classpath. If you want to use the cvs source for the
dependencies you will have to do an "ant deps" to compile the
dependencies first.
It seemed appropriate, from my humble little corner anyways,
to still require the dependencies cvs source to compile the
distribution since I figured that is what was 'official'.
It's still not clear to me what happens when you specify an
lib.dir and also have the deps. from cvs. I would think that it
would take what is first on the classpath, which is the contents
of the lib.dir, but I have not tested to be sure.
Sorry about the failonerror, I missed that one. It's in
there now.
2006-09-13 13:04 brucef
* visad/build.xml: build script for VisAD core packages
init -- print ant configuration
visad -- compile the VisAD package
visad.examples -- compile the visad examples
javadoc -- generate javadoc for all packages
jar -- jar the entire VisAD distribution
jar.docs -- jar javadoc
jar.examples -- jar examples (binary and source)
clean -- clean all files generated by this
build script
visad.clean -- clean only visad package binaries
dods -- compile the dods package
loci -- compile the loci package
HTTPClient -- compile the HTTPClient package
Jama -- compile the Jama package
gnu -- compile the gnu package
ncsa -- compile the ncsa package
nom -- compile the nom package
ucar -- compile the ucar package
edu -- compile the edu package
2006-09-11 12:00 donm
* visad/FieldImpl.java: some memory savings in combine: - check to
see if RangeSets are all FloatSets. If so, call getFloats
instead of get values. - add a new signature with a copy flag
to retain original arrays from combined fields instead of
making new arrays. Default is true (as before)
2006-09-06 10:04 donm
* edu/wisc/ssec/mcidas/: AREAnav.java, GMSXnav.java, GOESnav.java,
GVARnav.java, LALOnav.java, LAMBnav.java, MERCnav.java,
MOLLnav.java, MSATnav.java, MSGTnav.java, MSGnav.java,
PSnav.java, RADRnav.java, RECTnav.java, TANCnav.java: add new
imgCoordToAreaCoord and areaCoordToImageCoord methods that take
an input array to reduce memory footprint. Pass in the same
array in subclasses. Add float toLinEle and toLatLon methods in
Meteosat classes. Map projection classes still need float
methods
2006-09-05 10:54 curtis
* visad/jmet/DumpType.java: Change PrintWriters to PrintStreams
(fix a problem with no output).
2006-08-31 14:14 donm
* visad/ShadowFunctionOrSetType.java: fix a problem with the recent
changes when the coordinate system is a CachingCoordinateSystem
2006-08-30 08:47 donm
* visad/: DoubleStringTuple.java, DoubleTuple.java: new lightweight
Tuple classes
2006-08-29 16:45 donm
* visad/util/Util.java: fix the arraysEqual method to check for
null element. This should probably just call
java.util.Arrays.equal().
Add some printArray methods
2006-08-29 14:19 donm
* visad/: RealTuple.java, Tuple.java: use getComponents() and
getComponent(int) instead of accessing tupleComponents directly
so it's easier to subclass
2006-08-29 14:06 donm
* visad/bom/ImageRendererJ3D.java: check for a TupleType of
RealTupleTypes as the range and throw an exception in
isRendererUseable.
2006-08-29 11:50 jeffmc
* edu/wisc/ssec/mcidas/McIDASUtil.java: Refactor mcDayTimeToSecs
method to allow client code to pass in the calendar
2006-08-28 16:14 jeffmc
* visad/DateTime.java: Add a dateFormatClass static member that, if
set, is used to instatiate an application specific date
formatter.
Make the creation of the utcCalender be lazy.
2006-08-24 15:03 tomw
* visad/bom/radar.dat: File needed to test visad.bom.RadarDisplay
2006-08-23 17:28 donm
* visad/: ShadowFunctionOrSetType.java, ShadowType.java: clean up
Trace calls
2006-08-23 14:22 donm
* visad/: CoordinateSystem.java, ShadowFunctionOrSetType.java,
bom/Radar2DCoordinateSystem.java,
bom/Radar3DCoordinateSystem.java: check in latest performance
improvements
2006-08-22 17:19 donm
* visad/ShadowType.java: if mapValues copy == false and a Real is
mapped to more than one ScalarMap, we need to set copy == true or
we munge the data
2006-08-21 17:05 donm
* visad/: DerivedUnit.java, OffsetUnit.java, ScaledUnit.java,
Unit.java: check for NaN's when doing conversion. Also fix some
issues where the copy parameter was not being correctly handled
when units were null
2006-08-21 16:52 donm
* visad/ScalarMap.java: new signatures for scaleValues and
inverseScaleValues which take a copy argument
2006-08-21 16:51 donm
* visad/: ShadowType.java, java2d/ShadowTypeJ2D.java,
java3d/ShadowTypeJ3D.java: add a new method signature for
mapValues that takes a copy argument
2006-08-21 16:34 curtis
* visad/Makefile: Include ucar.netcdf in visad-lite
(ucar.netcdf.RandomAccessFile is needed by
visad.data.visad.BinaryReader).
2006-08-21 15:52 donm
* visad/Gridded2DSet.java: fix a problem with the 1D manifold ctor
which was not passing along the copy flag
2006-08-18 19:18 curtis
* visad/Makefile: Add "lite" target for building visad-lite.jar
with a subset of packages.
2006-08-18 16:29 curtis
* visad/AxisScale.java: Eliminate setSide2 in favor of TERTIARY and
QUATERNARY parameters as Don suggested.
2006-08-18 15:23 curtis
* visad/AxisScale.java: Add setSide2 method to allow scales on the
other two edges in 3-D.
2006-08-16 15:55 jeffmc
* visad/util/Trace.java: Wrapper around the ucar.unidata.util.Trace
facility. Uses reflection so no dependency on the ucar package
2006-08-15 08:55 donm
* visad/ShadowType.java: fix a problem in colorSum method
2006-08-14 16:56 jeffmc
* visad/meteorology/SingleBandedImageImpl.java: Tune up the
setMaxMinValues loop. Call it lazily
2006-08-11 11:08 donm
* visad/FlowControl.java: add autoScaling method for zooming, add
all the streamline params to the saveString, javadoc
2006-08-10 12:13 curtis
* visad/java3d/ShadowFunctionOrSetTypeJ3D.java: Add commented out
code and notes on construction of ImageComponent2D by reference,
for use at some point in the future with Java3D texturing by
reference (but currently screws up Java2D -- need to make
construction of the ImageComponent2D by reference contingent upon
such factors).
2006-08-10 12:07 curtis
* visad/benjamin/.cvsignore: Ignore
visad_benjamin_Galaxy_ContourCell.h as well.
2006-08-10 12:05 curtis
* visad/ImageFlatField.java: "Slow" color lookup in getBytes
returns color values for only as many range components as there
are (rather than always 4).
2006-08-10 12:03 curtis
* visad/bom/ShadowImageFunctionTypeJ3D.java: Use unsigned byte
values instead of signed for color table lookup with
ImageFlatFields.
2006-08-09 15:40 brucef
* visad/examples/Test05.java: fixed fill button on contour widget
and eliminated exception: visad.VisADError: Action.run:
visad.DisplayException: IsoContour color-fill is enabled, so
vis_radiance must also be mapped to Display.RGB
2006-08-09 10:44 tomw
* edu/wisc/ssec/mcidas/EnhancementTable.java: Add signature to look
up unique RGB value and return an index...if possible.
2006-08-07 11:21 brucef
* visad/.cvsignore: [no log message]
2006-08-01 15:10 donm
* visad/data/mcidas/PointDataAdapter.java: fix the memory
"enhancement" put in last year which used numObs*numParams more
memory than the previous version.
2006-07-31 12:21 donm
* edu/wisc/ssec/mcidas/adde/AddeURLConnection.java: allow for EU
loads in aget. It was hard wired for EC for some reason.
2006-07-30 16:06 donm
* visad/data/mcidas/PointDataAdapter.java: use the new
setSample(index, data, copy, checkType) to cut clock time in half
2006-07-30 16:06 donm
* visad/FieldImpl.java: add a setSample(index, data, copy,
checkType) method to go along with Jeff's recent change to
setSamples. Use the new setSamples method on resample and
domainFactor. Need to go through FieldImpl and see if other
places could benefit from this savings also
2006-07-26 13:43 curtis
* visad/data/DefaultFamily.java: Use GIFForm before LociForm. The
rationale is as follows:
Many GIFs, such as those in visad/ss, use 8-bit indexed color.
LociForm uses ImageIO to read the pixel values correctly,
returning type ((ImageElement, ImageLine) -> value0). But the
associated indexed color table is unfortunately ignored in the
conversion from BufferedImage to VisAD object.
GIFForm, on the other hand, uses an old-school Toolkit.getImage
call, which retrieves a non-buffered java.awt.Image object, which
is then translated into a VisAD object with DataUtility.makeImage
(which assumes the image is RGB color).
So, GIFForm results in a larger data object, but which displays
correctly in the VisAD SpreadSheet with default mappings to Red,
Green and Blue. For compatibility and consistency, we will use
GIFForm.
2006-07-26 13:24 curtis
* visad/data/gif/TestGIF.java: Update URL to SSEC logo.
2006-07-19 14:17 jeffmc
* visad/FieldImpl.java: Add a new parameter to setSamples:
checkAllRangeTypes. If false then we only check the type of the
first range element. If true (the default) then operate as
usual, checking the type of every element
2006-07-17 10:45 jeffmc
* visad/data/gis/ArcAsciiGridAdapter.java: Make a number of
performance and memory fixes. e.g., old way for a big file was 60
seconds. new way is 20 seconds
2006-07-14 13:56 rink
* visad/: Gridded3DSet.java, ShadowType.java: final, hopefully, fix
for a cntr label initialization problem
2006-07-05 12:39 curtis
* visad/util/Util.java: Add getVersionDate method for determining
date and time of VisAD's build.
2006-06-29 12:34 billh
* visad/paoloa/GoesCollaboration.java: change ftp server from
www.ssec to ftp.ssec, in print statement
2006-06-28 10:51 donm
* visad/georef/: EarthLocationTuple.java, LatLonPoint.java,
LatLonTuple.java: add ctors for passing in units for memory
efficiency (mimic signatures in RealTuple)
2006-06-27 16:27 donm
* visad/data/gis/UsgsDemAdapter.java: fix some more issues
2006-06-21 15:51 rink
* visad/ContourControl.java: separate init_scale and current scale
2006-06-21 10:34 donm
* visad/data/gis/UsgsDemAdapter.java: fix some bugs and ommissions
2006-06-15 09:01 donm
* visad/java3d/ShadowAnimationFunctionTypeJ3D.java: implement the
same memory savings in here as in ImageRendererJ3D
2006-06-13 17:38 curtis
* visad/jmet/DumpType.java: Fix bug Ken Evans found.
2006-06-13 14:32 donm
* edu/wisc/ssec/mcidas/GMSXnav.java: bring into line with latest
mcidas version. Fixes a problem with data not being displayed
correctly.
2006-06-06 15:48 curtis
* visad/util/BarGraph.java: Class providing a simple API for 2D and
3D bar graph plotting in VisAD.
2006-05-30 09:33 donm
* visad/examples/Test61.java: add in an option for texture3d mode.
Either run as "Test61 t" for Texture3D or "Test61 x t" NICEST
transparency and Texture3D
2006-05-30 07:47 donm
* visad/: DataDisplayLink.java, Display.java, DisplayImpl.java,
GraphicsModeControl.java, ShadowFunctionOrSetType.java,
java2d/GraphicsModeControlJ2D.java,
java3d/GraphicsModeControlJ3D.java: support a Texture3DMode
DisplayRealType for ConstantMaps and a GraphicsModeControl
set/getTexture3DMode
2006-05-24 17:56 donm
* visad/ShadowFunctionOrSetType.java: remove check for alpha when
determining curvedTexture
2006-05-24 11:45 rink
* visad/Stream2D.java: fix indexing problem JamesK found
2006-05-18 13:00 curtis
* visad/: Makefile, Makefile.WinNT: Add loci/formats/readers.txt
and loci/formats/writers.txt resource files to visad.jar build
task.
2006-05-18 09:55 donm
* visad/Contour2D.java: fix a ArrayOutOfBounds exception from
yesterday's change. Shows up in color filled contours
2006-05-17 16:21 donm
* visad/Contour2D.java: move allocation of the aux* arrays outside
the loop, fix the numc calculation logic
2006-05-15 10:56 tomw
* visad/: Makefile, Makefile.WinNT: Update the compiler flags to
use the -source option. Set the class compatibility to Java 1.4.
2006-05-15 10:25 rink
* visad/: FlowControl.java, ShadowType.java, Stream2D.java,
examples/Stream2DTest.java: improvements for streamline
rendering.
2006-05-15 09:56 donm
* visad/Set.java: check in the change to throw an exception if the
type is null.
2006-04-20 17:06 donm
* visad/Gridded3DSet.java: some more tries at memory improvement
2006-04-20 17:06 donm
* visad/Contour2D.java: some tries at memory improvement
2006-04-19 11:27 donm
* visad/DisplayImpl.java: fix a null pointer exception when using
offscreen
2006-03-30 17:59 curtis
* visad/data/: bio/LociForm.java, tiff/LegacyTiffTools.java: Latest
work on loci.formats package.
2006-03-30 13:53 donm
* visad/: DisplayEvent.java, DisplayImpl.java: add in code to
support a COMPONENT_RESIZED display event.
2006-03-30 05:49 billh
* visad/Stream2D.java: Here's a few changes to hopefully help with
James's streamline density problem. I can't test it because I
can't compile VisAD, because I haven't figured out how to check
out the 'loci' packages (I tried 'cvs update -d' but that didn't
do it).
2006-03-28 10:51 rink
* visad/Contour2D.java: fix bug DonM found.
2006-03-08 16:36 curtis
* visad/: Makefile, Makefile.WinNT: Add loci.formats package to
Makefiles.
2006-03-08 16:32 curtis
* visad/: ImageFlatField.java, VisADException.java,
data/BadFormException.java, data/DefaultFamily.java,
data/avi/AVIForm.java, data/bio/BioRadForm.java,
data/bio/BioRadNote.java, data/bio/DeltavisionForm.java,
data/bio/FluoviewTiffForm.java, data/bio/GatanForm.java,
data/bio/ICSForm.java, data/bio/IPLabForm.java,
data/bio/IPWForm.java, data/bio/ImageProSeqForm.java,
data/bio/LegacyZVIForm.java, data/bio/LeicaForm.java,
data/bio/LociForm.java, data/bio/MetamorphForm.java,
data/bio/OMETools.java, data/bio/OpenlabForm.java,
data/bio/PerkinElmerForm.java, data/bio/ZVIForm.java,
data/bio/ZeissForm.java, data/biorad/BioRadNote.java,
data/ij/ImageJForm.java, data/qt/PictForm.java,
data/qt/QTForm.java, data/tiff/BaseTiffForm.java,
data/tiff/BitBuffer.java, data/tiff/BitWriter.java,
data/tiff/ByteVector.java, data/tiff/LZWTreeNode.java,
data/tiff/LegacyBitBuffer.java, data/tiff/LegacyTiffForm.java,
data/tiff/LegacyTiffTools.java, data/tiff/RandomAccessArray.java,
data/tiff/TiffForm.java, data/tiff/TiffRational.java,
data/tiff/TiffTools.java, examples/Test73.java,
util/ComboFileFilter.java, util/DataConverter.java,
util/ExtensionFileFilter.java, util/FormFileFilter.java,
util/McIDASFileFilter.java, util/ReflectedUniverse.java,
util/TextEditor.java, util/Util.java: Major reorganization of
biological file formats. The intent is to distribute loci.formats
both as a standalone package, and as part of VisAD (similar to
the edu.wisc.ssec.mcidas and nom.tam.fits packages, so that no
reflection is necessary).
Now visad.data.bio.LociForm is the master file adapter for all
image formats supported within the loci.formats package. The old
file adapters are trivial extensions of LociForm. Except for
TiffForm, they have been deprecated in favor of LociForm or
direct use of loci.formats with visad.ImageFlatField.
Some classes in visad.data.bio and visad.data.tiff have been
removed, since they were largely helper classes for the file
adapters. If you need one of these classes, you will find it in
the loci.formats package.
The loci.formats package also provides better support for
FileFilters and JFileChoosers. The visad.util.*FileFilter classes
have been deprecated in favor of the corresponding classes in
loci.formats.
Also added (unrelated to the above) are exception constructors
for creating chained exceptions in the VisAD exception hierarchy.
2006-03-06 11:40 donm
* visad/ScalarMap.java: fix a problem caused by recent change when
max/min range are equal
2006-02-23 10:43 tomw
* edu/wisc/ssec/mcidas/AREAnav.java: Remove import for and
dependency on visad.Set. Easier just to convert the values
herein.
2006-02-22 14:12 curtis
* visad/: ImageFlatField.java, examples/Test73.java: Add
ImageFlatField(BufferedImage) constructor; use it in Test73.
2006-02-21 16:22 curtis
* visad/examples/Test73.java: Add comment illustrating other
approaches to ImageFlatField construction.
2006-02-21 13:47 curtis
* visad/data/mcidas/AreaForm.java: In isThisType(String), recognize
files with relative or absolute path prefixes (e.g.,
/home/curtis/data/AREA0007).
2006-02-21 10:22 tomw
* edu/wisc/ssec/mcidas/: AreaFile.java, GMSXnav.java: Fix problem
with GMSX navigation for MTSAT. Nav bytes should not be flipped
when reading a local, big-endian file....strange...
2006-02-20 20:00 donm
* visad/bom/ShadowImageFunctionTypeJ3D.java: null out the old_node
arrays after we're done with them
2006-02-17 16:43 donm
* visad/ScalarMap.java: a fix for problem of datarange changing
when override unit was set
2006-02-16 16:27 jeffmc
* visad/AxisScale.java: Fix a problem with multiple threads
reading/writing the labelTable Hashtable
--
Tom Whittaker
University of Wisconsin-Madison
Space Science & Engineering Center (SSEC)
Cooperative Institute for Meteorological Satellite Studies (CIMSS)
1225 W. Dayton Street
Madison, WI 53706 USA
ph: +1 608 262 2759
==============================================================================
To unsubscribe visad, visit:
http://www.unidata.ucar.edu/mailing-list-delete-form.html
==============================================================================
More information about the visad
mailing list