Hi Hervé, Cameron, I've cut and pasted the failed tests from my regard run. The fact that others don't see the change is probably enough for me. It's good to have answers for these sorts of changes but it seems pretty clear that what ever small regression this _might_ possibly be is much less than the advances/
samples/tests/resources/wmf/imageWMF.wmf (WMFAccuracyTest.error.generated.svg.inaccurate) (org.apache.batik.transcoder.wmf.WMFAccuracyTest) Line number 7 Column number 35 Expected column value 4 Found column value 3 Reference line ><g fill="white" stroke-width="0.4805" stroke="white" stroke-linejoin="round" stroke-linecap="butt" New line ><g fill="white" stroke-width="0.3604" stroke="white" stroke-linejoin="round" stroke-linecap="butt" samples/tests/resources/wmf/negApmText1.wmf (WMFAccuracyTest.error.generated.svg.inaccurate) (org.apache.batik.transcoder.wmf.WMFAccuracyTest) Line number 9 Column number 17 Expected column value 9 Found column value 1 Reference line ><path d="M9 9 L473 9 L473 45 L9 45 L9 9 Z" New line ><path d="M10 9 L529 9 L529 45 L10 45 L10 9 Z" samples/tests/resources/wmf/negApmText2.wmf (WMFAccuracyTest.error.generated.svg.inaccurate) (org.apache.batik.transcoder.wmf.WMFAccuracyTest) Line number 9 Column number 17 Expected column value 9 Found column value 2 Reference line ><path d="M9 9 L516 9 L516 45 L9 45 L9 9 Z" New line ><path d="M23 9 L1347 9 L1347 45 L23 45 L23 9 Z" samples/tests/resources/wmf/testChart.wmf (WMFAccuracyTest.error.generated.svg.inaccurate) (org.apache.batik.transcoder.wmf.WMFAccuracyTest) Line number 64 Column number 70 Expected column value 5 Found column value 1 Reference line ><g stroke-linecap="butt" stroke-linejoin="round" stroke-width="1.5118" New line ><g stroke-linecap="butt" stroke-linejoin="round" stroke-width="1.1687" samples/tests/resources/wmf/textGreek.wmf (WMFAccuracyTest.error.generated.svg.inaccurate) (org.apache.batik.transcoder.wmf.WMFAccuracyTest) Line number 7 Column number 36 Expected column value 7 Found column value 5 Reference line ><g fill="white" stroke-width="0.0756" stroke="white" stroke-linejoin="round" stroke-linecap="butt" New line ><g fill="white" stroke-width="0.0584" stroke="white" stroke-linejoin="round" stroke-linecap="butt" samples/tests/resources/wmf/textGreek.wmf (WMFAccuracyTest.error.generated.svg.inaccurate) (org.apache.batik.transcoder.wmf.WMFAccuracyTest) Line number 7 Column number 35 Expected column value 0 Found column value 3 Reference line ><g fill="white" stroke-width="0.0756" stroke="white" stroke-linejoin="round" stroke-linecap="butt" New line ><g fill="white" stroke-width="0.3333" stroke="white" stroke-linejoin="round" stroke-linecap="butt" Hervé Girod <[EMAIL PROTECTED]> wrote on 01/08/2008 06:46:07 PM: > Hello, > > For me the WMF rendering with Batik 1.7 seems OK, and I did not found > any differences in the rendering but I don't think what improvements in > the WMF transcoder could have done anything for the line width ;-); > In the only case were it could have changed anything in the rendering in > a while, the WMF could not be rendered before (non-APM header files, > were the transcoder try to guess the dimensions of the WMF image). > > To be sure, I converted the WMF chart test case, and all test cases for > bugs 40594, 41205, 42961, 43195, and 43595 + various other complex test > examples (some WMF files generated with PPT). All seemed OK too. > > Herve Girod > > Cameron McCormack wrote: > > Thomas DeWeese: > > > >>> Second a bunch of the WMF tests have slightly different > >>> renderings (in particular line widths have changed). I > >>> suspect that these are due to the various WMF reader > >>> improvements but I thought it would be worth getting > >>> confirmation. (I wouldn't -1 the release for this in the future > >>> it's good to 'fix' the references for these sorts of changes > >>> at least when the release is made). > >>> > > > > Cameron McCormack: > > > >> I?ll investigate this later today, but maybe Hervé knows? > >> > > > > Actually, can you point out which WMF tests have different renderings? > > I can?t see it in my test report (apart from the usual platform path > > rendering differences that cause mostly grey difference images on my > > system). > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] >
