Le dimanche 25 août 2013 02:34:21, Kurt Schwehr a écrit : > Since I haven't check on any prior update, I consider this just a baseline > and don't see this as a reason to hold up the update.
Yes, that doesn't look criticial to me either, apart the failures with libpng 1.6.3 > > Notes inline. backing off on png gave a passing test. backing off on > jpeg... no change. Other issues, I'll see if I can look into later. I've tested against libpng 1.6.3 on Linux 64bit, and I don't get any failed tests on png.py. Strange. Do the png utilities of libpng 1.6.3 work on autotest/gdrivers/data/test.png ? I've also tested libjpeg-9a and could reproduce the failure on gdalwarp_33. Visual inspectiong of the result looks OK, so that's just a difference in the maths involved in JPEG decompression between libjpeg versions. I'll increase the tolerance to 40. > > -kurt > > On Aug 24, 2013, at 11:26 AM, Even Rouault <[email protected]> wrote: > > Le samedi 24 août 2013 19:57:47, Kurt Schwehr a écrit : > >> LGTM with respect to fink package on mac osx 10.8 > >> > >> - Looks like I have a todo on jasper. > >> - not sure on http_4 > >> - png.py and test_gdalwarp below. > >> - rest are flagged as know issues or pre-release things > > > > Hi Kurt, > > > > Thanks for your testing. Comments below > > > >> -kurt > >> > >> > >> ------------ Failures ------------ > >> Script: ogr/ogr_shape.py > >> > >> TEST: ogr_shape_68 ... fail > >> > >> line 3249: fail > > > > Yes, I observed that it fails on Travis too on the Mac. Not sure why > > being not in a position to debug on the Mac. Advanced stuff, not > > critical if it doesn't work > > > >> Script: gcore/basic_test.py > >> > >> TEST: basic_test_8 ... fail > > > > Same, fails on Travis for unknown reason on the Mac. The test is likely a > > bit fragile. > > > >> line 161: fail > >> > >> Script: gcore/gdal_api_proxy.py > >> > >> TEST: gdal_api_proxy_3 ... fail > > > > Same, fails on Travis for unknown reason on the Mac. Advanced stuff, not > > critical if it doesn't work > > > >> Script: gdrivers/gdalhttp.py > >> > >> TEST: http_4 ... fail > > > > Might be linked to a temporary unavailability of the remote resource. I > > somehow remember that it can fail sometimes, but I have never tested on > > the Mac, so it might not be just an interminent failure > > > > Does "gdalinfo > > /vsicurl/ftp://ftp2.cits.rncan.gc.ca/pub/cantopo/250k_tif/MCR2010_01.tif"; > > repeatedly fail ? > > > > You can try adding "--debug CPL_CURL_VERBOSE YES" for more info. > > I think you called it with just being a network issue > > ./gdalhttp.py > TEST: http_1 ... success > TEST: http_2 ... success > TEST: http_3 ... success > TEST: http_4 ... success > TEST: http_cleanup ... success > > Test Script: http > Succeeded: 5 > Failed: 0 (0 blew exceptions) > Skipped: 0 > Expected fail:0 > Duration: 11.35s > > >> Script: gdrivers/jpeg2000.py > >> > >> TEST: jpeg2000_8 ... fail > >> > >> line 189: Jasper library would need patches > > > > Yes, this is a defect in upstream jasper that most linux distro have > > patched. Applying Debian patches might help you fix that > > > >> Script: gdrivers/png.py > >> > >> TEST: png_1 ... fail > >> > >> line 429: Checksum for band 1 in "test.png" is 6846, but expected > >> > >> 57921. TEST: png_2 ... fail > >> > >> line 484: Failed to create test file using CreateCopy method. > >> > >> libpng: No IDATs written into file > > > > Hum, this is new to me. Which libpng version do you use ? > > otool -L /sw/bin/gdalinfo | grep png > /sw/lib/libpng16.16.dylib (compatibility version 20.0.0, current version > 20.0.0) > > libpng16 1.6.3-1 > > backing off to libpng15 1.5.17-1 gives success: > > otool -L /sw/bin/gdalinfo | grep png > /sw/lib/libpng15.15.dylib (compatibility version 33.0.0, current version > 33.0.0) > > ./png.py > TEST: png_1 ... success > TEST: png_2 ... success > TEST: png_3 ... success > TEST: png_4 ... success > TEST: png_5 ... success > TEST: png_6 ... success > TEST: png_7 ... success > TEST: png_8 ... success > TEST: png_9 ... success > TEST: png_10 ... success > TEST: png_11 ... success > > Test Script: png > Succeeded: 11 > Failed: 0 (0 blew exceptions) > Skipped: 0 > Expected fail:0 > Duration: 0.05s > > >> Script: utilities/test_gdalwarp.py > >> > >> TEST: test_gdalwarp_33 ... fail > > > > The test might be a bit fragile indeed. It might depend on the libjpeg > > version you are using. There's a tolerance at 37, and you have reached > > 40. Could likely be changed to 40. > > otool -L /sw/bin/gdalinfo | grep jp > /sw/lib/libjpeg.9.dylib (compatibility version 10.0.0, current version > 10.0.0) > > i libjpeg9-shlibs 9-3 Shared libraries > for libjpeg package > > Backing off still has a test that fails > > otool -L /sw/bin/gdalinfo | grep jp > /sw/lib/libjpeg.8.dylib (compatibility version 13.0.0, current version > 13.0.0) > > i libjpeg8-shlibs 8d-3 > > TEST: test_gdalwarp_33 ... Diff at pixel (160, 0) : 7.000000 > Diff at pixel (161, 0) : 1.000000 > Diff at pixel (162, 0) : 6.000000 > Diff at pixel (165, 0) : 2.000000 > Diff at pixel (167, 0) : 8.000000 > Diff at pixel (169, 0) : 8.000000 > Diff at pixel (173, 0) : 4.000000 > Diff at pixel (174, 0) : 9.000000 > Diff at pixel (176, 0) : 9.000000 > Diff at pixel (189, 0) : 10.000000 > Diff at pixel (266, 0) : 13.000000 > Diff at pixel (284, 0) : 17.000000 > Diff at pixel (172, 1) : 22.000000 > Diff at pixel (265, 1) : 31.000000 > Diff at pixel (301, 6) : 32.000000 > Diff at pixel (388, 37) : 38.000000 > Diff at pixel (118, 319) : 40.000000 > Max diff : 40 > Number of diffs : 3683 > fail > TEST: test_gdalwarp_34 ... success > TEST: test_gdalwarp_35 ... success > TEST: test_gdalwarp_36 ... success > TEST: test_gdalwarp_cleanup ... success > > Test Script: test_gdalwarp > Succeeded: 37 > Failed: 1 (0 blew exceptions) > Skipped: 0 > Expected fail:0 > Duration: 1.80s > > >> ---------------------------------- > >> > >> Succeeded: 2608 > >> Failed: 8 (0 blew exceptions) > >> Skipped: 723 > >> Expected fail:12 > >> Duration: 06m4.3s > >> As GDAL_DOWNLOAD_TEST_DATA environment variable is not defined, 107 > >> tests relying on data to downloaded from the Web have been skipped As > >> GDAL_RUN_SLOW_TESTS environment variable is not defined, 22 "slow" > >> tests have been skipped > >> > >> > >> png.py: > >> > >> python png.py > >> > >> TEST: png_1 ... ERROR 1: libpng: IDAT: invalid distance too far back > >> > >> ERROR 1: data/test.png, band 1: IReadBlock failed at X offset 0, Y > >> offset 29 ERROR 1: GetBlockRef failed at X block offset 0, Y block > >> offset 29 ERROR 3: Checksum value couldn't be computed due to I/O read > >> error. > >> > >> fail > >> > >> line 429: Checksum for band 1 in "test.png" is 6846, but expected > >> > >> 57921. TEST: png_2 ... ERROR 1: libpng: IDAT: invalid distance too far > >> back ERROR 1: data/test.png, band 1: IReadBlock failed at X offset 0, Y > >> offset 29 ERROR 1: libpng: IDAT: invalid distance too far back > >> ERROR 1: data/test.png, band 1: IReadBlock failed at X offset 0, Y > >> offset 30 ERROR 1: libpng: IDAT: invalid distance too far back > >> > >> TEST: test_gdalwarp_32 ... success > >> TEST: test_gdalwarp_33 ... Diff at pixel (160, 0) : 7.000000 > >> > >> Diff at pixel (161, 0) : 1.000000 > >> Diff at pixel (162, 0) : 6.000000 > >> Diff at pixel (165, 0) : 2.000000 > >> Diff at pixel (167, 0) : 8.000000 > >> Diff at pixel (169, 0) : 8.000000 > >> Diff at pixel (173, 0) : 4.000000 > >> Diff at pixel (174, 0) : 9.000000 > >> Diff at pixel (176, 0) : 9.000000 > >> Diff at pixel (189, 0) : 10.000000 > >> Diff at pixel (266, 0) : 13.000000 > >> Diff at pixel (284, 0) : 17.000000 > >> Diff at pixel (172, 1) : 22.000000 > >> Diff at pixel (265, 1) : 31.000000 > >> Diff at pixel (301, 6) : 32.000000 > >> Diff at pixel (388, 37) : 38.000000 > >> Diff at pixel (118, 319) : 40.000000 > >> Max diff : 40 > >> Number of diffs : 3683 > >> fail > >> > >> TEST: test_gdalwarp_34 ... success > >> > >> On Aug 23, 2013, at 11:27 AM, Even Rouault > >> <[email protected]> > > > > wrote: > >>> Hi, > >>> > >>> PSC members, please vote for the following motion after conducting your > >>> own testing. > >>> Non-PSC members please let us know if you discover problems. > >>> > >>> --- > >>> > >>> Motion: The GDAL/OGR 1.10.1RC1 package is promoted as the > >>> final GDAL/OGR 1.10.1 release. > >>> > >>> +1 Even > >> > >> _______________________________________________ > >> gdal-dev mailing list > >> [email protected] > >> http://lists.osgeo.org/mailman/listinfo/gdal-dev -- Geospatial professional services http://even.rouault.free.fr/services.html _______________________________________________ gdal-dev mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/gdal-dev
