| Cloud-sis.apache.org |
Inbox (2) | New Cloud Notification |
Wed, 02 Dec, 03:06 |
| Martin Desruisseaux (Jira) |
[jira] [Created] (SIS-505) Upgrade OpenOffice/LibreOffice dependency |
Mon, 07 Dec, 12:23 |
| Martin Desruisseaux (Jira) |
[jira] [Created] (SIS-506) Compute isolines from a coverage |
Fri, 11 Dec, 17:33 |
| Martin Desruisseaux (Jira) |
[jira] [Created] (SIS-507) Create modules for visual tests / integration tests |
Fri, 18 Dec, 14:59 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Most accurate selection of best grid. In addition of maximal intersection area, try also to minimize area outside the Area Of Interest (AOI) to to prefer grids more centered on the AOI. |
Tue, 01 Dec, 17:40 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Take a more representative center of data. |
Wed, 02 Dec, 18:07 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Take temporal extent in account in addition of geographic extent. This is an anticipation for the support of dynamic datum in a future SIS version. It is also useful for other code using `ExtentSelector` for making a choice between different rasters (for example). |
Fri, 04 Dec, 15:15 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (6d0bce8 -> 78cb599) |
Tue, 08 Dec, 10:07 |
| desruisse...@apache.org |
[sis] 01/02: When `PropertyNotFoundException` are reported as warnings, keep the exception message instead than using the `getValueOrFallback(…)` method. The reason is that the exception message provides a deeper analysis, telling for example if the property can not be provided because it does not exist or because the given name is ambiguous. |
Tue, 08 Dec, 10:07 |
| desruisse...@apache.org |
[sis] 02/02: Documentation fix: the "Supported by" column is not relevant anymore. |
Tue, 08 Dec, 10:07 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (89d2a24 -> 4f223fe) |
Fri, 11 Dec, 17:37 |
| desruisse...@apache.org |
[sis] 01/03: Update in NetBeans project configuration introduced by upgrade to NetBeans 12.2. |
Fri, 11 Dec, 17:37 |
| desruisse...@apache.org |
[sis] 02/03: Gives more details when à `MathTransform` does not have the expected number of dimensions. |
Fri, 11 Dec, 17:37 |
| desruisse...@apache.org |
[sis] 03/03: First version of a contouring algorithm, adapted from Johann Sorel work based on "marching squares" algorithm. This implementation builds isolines for all bands and all levels in a single iteration over image pixels. Java2D shapes are constructed on-the-fly during iteration. |
Fri, 11 Dec, 17:37 |
| desruisse...@apache.org |
svn commit: r1884563 [2/4] - in /sis/ip-review: ./ rev/01091/ rev/01094/ rev/01096/ rev/01097/ rev/01188/ rev/05052/ rev/05999/ rev/06057/ rev/06067/ rev/06068/ rev/06410/ rev/10796/ rev/11086/ rev/12288/ rev/17452/ rev/20874/ rev/20883/ rev/24476/ |
Thu, 17 Dec, 17:18 |
| desruisse...@apache.org |
svn commit: r1884563 [4/4] - in /sis/ip-review: ./ rev/01091/ rev/01094/ rev/01096/ rev/01097/ rev/01188/ rev/05052/ rev/05999/ rev/06057/ rev/06067/ rev/06068/ rev/06410/ rev/10796/ rev/11086/ rev/12288/ rev/17452/ rev/20874/ rev/20883/ rev/24476/ |
Thu, 17 Dec, 17:18 |
| desruisse...@apache.org |
svn commit: r1884563 [1/4] - in /sis/ip-review: ./ rev/01091/ rev/01094/ rev/01096/ rev/01097/ rev/01188/ rev/05052/ rev/05999/ rev/06057/ rev/06067/ rev/06068/ rev/06410/ rev/10796/ rev/11086/ rev/12288/ rev/17452/ rev/20874/ rev/20883/ rev/24476/ |
Thu, 17 Dec, 17:18 |
| desruisse...@apache.org |
svn commit: r1884563 [3/4] - in /sis/ip-review: ./ rev/01091/ rev/01094/ rev/01096/ rev/01097/ rev/01188/ rev/05052/ rev/05999/ rev/06057/ rev/06067/ rev/06068/ rev/06410/ rev/10796/ rev/11086/ rev/12288/ rev/17452/ rev/20874/ rev/20883/ rev/24476/ |
Thu, 17 Dec, 17:18 |
| desruisse...@apache.org |
[sis] branch visual-test created (now 6162dbc) |
Fri, 18 Dec, 15:55 |
| desruisse...@apache.org |
[sis] 01/02: Swing components as helper tools for visual tests of some Apache SIS classes. Those widgets are not part of main Apache SIS code base. We do not intent to build any application on top of them (we use JavaFX instead). |
Fri, 18 Dec, 15:55 |
| desruisse...@apache.org |
[sis] 02/02: Add a visual test for `org.apache.sis.internal.processing.image.Isolines`. |
Fri, 18 Dec, 15:55 |
| desruisse...@apache.org |
[sis] branch visual-test updated: Bug fix in event handling when magnifier glass dragging or resizing stops. Adjust windows size and location. |
Fri, 18 Dec, 17:29 |
| desruisse...@apache.org |
[sis] branch visual-test updated: Make isolines visual test more useful: |
Sun, 20 Dec, 23:13 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Fix an error in determination of when a polygon is closed. Reduce some code duplication by taking advantage of recurrent patterns in interpolations. |
Sun, 20 Dec, 23:21 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (cf5739d -> 02f0113) |
Wed, 23 Dec, 17:23 |
| desruisse...@apache.org |
[sis] 01/02: Remove some spikes which appear when pixel values are strictly equal to isoline values. |
Wed, 23 Dec, 17:23 |
| desruisse...@apache.org |
[sis] 02/02: When a single `Polyline` instance can not be closed as a polygon, keep that instance in a temporary map. As additional instances are added in the map, some `Polyline` instances get concatenated together. It allows the creation of "real" closed polygons instead than juxtaposition of polylines that looks like polygons. |
Wed, 23 Dec, 17:23 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Implementation of a java.awt.Shape backed by an array of coordinates that we can handle more easily than with java.awt.geom.Path2D. It introduce a class hierarchy similar to the OGC geometry classes (Polyline, Polygon, MultiPolygons), but it is not the intent to provide an OGC geometries implementation now. Those classes are needed for simplifying `IsolineTracer` (in a next commit). |
Sat, 26 Dec, 00:01 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (728f70e -> 11c6b74) |
Sat, 26 Dec, 14:31 |
| desruisse...@apache.org |
[sis] 01/02: Replace the use of `java.awt.geom.Path2D` by `PathBuilder` in `IsolineTracer`. It simplifies the algorithm and allow us to resolve the remaining spikes that we saw in isolines. |
Sat, 26 Dec, 14:31 |
| desruisse...@apache.org |
[sis] 02/02: Avoid the call to `Arrays.binarySearch(…)`. It has been identified by profiling as a performance bottleneck (in the context of `Isolines`). |
Sat, 26 Dec, 14:31 |
| desruisse...@apache.org |
[sis] branch visual-test updated: Write coordinates as text (for debugging) when zoomed enough. Update for change in Isolines API. |
Sun, 27 Dec, 13:12 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (11c6b74 -> eceea7f) |
Tue, 29 Dec, 17:27 |
| desruisse...@apache.org |
[sis] 01/03: Retrofit `DefaultIterator` into `PixelIterator`. This change allows slightly better encapsulation (more fields can be made private) and reduce the need to create `WritablePixelIterator` when only read operations are desired. The previous abstraction level was apparently not needed in practice. |
Tue, 29 Dec, 17:27 |
| desruisse...@apache.org |
[sis] 02/03: Add `BandedIterator`, a specialization of `PixelIterator` for the case of images using a `BandedSampleModel`. This is an attempt to resolve a performance bootleneck observed during `Isolines` calculations. |
Tue, 29 Dec, 17:27 |
| desruisse...@apache.org |
[sis] 03/03: Make possible to use `BandedIterator` with multi-tiled images. It required to retrofit `LinearIterator` into `PixelIterator`. |
Tue, 29 Dec, 17:27 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (eceea7f -> eac2307) |
Wed, 30 Dec, 19:23 |
| desruisse...@apache.org |
[sis] 01/02: Add more tests and add comments about optimization strategies that have been discarded. Two strategies were tried and failed to provide significantly better performance: |
Wed, 30 Dec, 19:23 |
| desruisse...@apache.org |
[sis] 02/02: Replace the use of `PixelIterator.Window` by cached values of previous row. It make easier to provide a special case for one-banded images. |
Wed, 30 Dec, 19:23 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Add a test case for isolines on multi-banded image. |
Wed, 30 Dec, 21:51 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated: Add a note about another failed attempt to optimize. |
Thu, 31 Dec, 09:57 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (b9edb1f -> ce75340) |
Thu, 31 Dec, 09:58 |
| desruisse...@apache.org |
[sis] 01/01: Add a note about another failed attempt to optimize. |
Thu, 31 Dec, 09:58 |
| desruisse...@apache.org |
[sis] branch geoapi-4.0 updated (ce75340 -> d33c28c) |
Thu, 31 Dec, 16:17 |
| desruisse...@apache.org |
[sis] 01/03: Store isoline coordinates as single-precision floating point numbers, but with a translation for attenuating the precision lost. |
Thu, 31 Dec, 16:17 |
| desruisse...@apache.org |
[sis] 02/03: Delegate `contains(…)` and `intersects(…)` implementations to Path2D. |
Thu, 31 Dec, 16:17 |
| desruisse...@apache.org |
[sis] 03/03: Move part of `CategoryColorCell` into a new `ColorCell` base class. The intent is to share this code with other kinds of color tables, to be created later for isolines. |
Thu, 31 Dec, 16:17 |
| jso...@apache.org |
[sis] branch geoapi-4.0 updated: Use new feature.getValueOrFallback method in filters and expressions to avoid exception creation when exception stack is not required |
Mon, 07 Dec, 09:25 |
| jso...@apache.org |
[sis] branch geoapi-4.0 updated: Portrayal : initial work on OGC symbology encoding portrayal engine |
Tue, 08 Dec, 16:08 |