Ok, that's what is was wondering, sounds like it'll work fine within the limits of 1.8's color model.
Good luck! -- John > On 7 Aug 2014, at 13:38, Tilman Hausherr <thaush...@t-online.de> wrote: > > It would use the old method that is already used now. It is a mix of toRGB() > and shadingtinttransform function. I will not copy the 2.0 files 1:1, I need > to make these changes, and changes in matrix transformations (there's a flip). > > Tilman > > Am 07.08.2014 22:28, schrieb John Hewson: >> So you’ll switch out the calls to toRGB() for something else? >> >> -- John >> >>> On 7 Aug 2014, at 13:19, Tilman Hausherr <thaush...@t-online.de> wrote: >>> >>> Color spaces are the way they have always been in 1.8, i.e. "not always >>> correct". Anything drawn in a colorspace that isn't looking correct in 1.8 >>> won't be correct in a shading too. >>> >>> And no, they are no shading API changes in 1.8. >>> >>> Tilman >>> >>> Am 07.08.2014 20:22, schrieb John Hewson: >>>> Tilman, >>>> >>>> That’s very compelling, with tests like that the only concern would be not >>>> making breaking API changes, which shouldn’t be too hard. >>>> >>>> What about colour spaces though? 1.8 doesn’t have toRGB() built-in to the >>>> colour spaces unlike 2.0. >>>> >>>> -- John >>>> >>>>> On 7 Aug 2014, at 10:25, Tilman Hausherr <thaush...@t-online.de> wrote: >>>>> >>>>> The original shading packages 1-5 were also added to 1.8.5 (PDFBOX-615 >>>>> <https://issues.apache.org/jira/browse/PDFBOX-615>) and the changes are >>>>> well tested. I have now over 250 test files (about 50-100 of them deal >>>>> with shading), and I've also tested thousands of files from the >>>>> digitalcorpora site without any shading related exception that was a >>>>> regression, the only exception was a bug that has been there the whole >>>>> time. The optimizations include many bugfixes as well, that came up in >>>>> the work. >>>>> >>>>> Tilman >>>>> >>>>> Am 07.08.2014 19:00, schrieb John Hewson: >>>>>> There's still new stuff in there, and optimisations are not bug fixes. >>>>>> It just seems like a lot of new code for a stable release. >>>>>> >>>>>> -- John >>>>>> >>>>>>> On 7 Aug 2014, at 09:52, Tilman Hausherr <thaush...@t-online.de> wrote: >>>>>>> >>>>>>> I disagree. I always try to add the new features to 1.8 when possible, >>>>>>> i.e. when there's no change of structure needed. >>>>>>> >>>>>>> Besides, the shadings 1,2,3,4,5 are already part of 1.8, whats new is >>>>>>> the optimizations, support of BBox and types 6 and 7. >>>>>>> >>>>>>> Tilman >>>>>>> >>>>>>> Am 07.08.2014 18:48, schrieb John Hewson: >>>>>>>> Perhaps we should stop adding new features to 1.8, and only fix the >>>>>>>> most problematic bugs? >>>>>>>> >>>>>>>> -- John >>>>>>>> >>>>>>>>> On 7 Aug 2014, at 09:11, Tilman Hausherr <thaush...@t-online.de> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>> +1 >>>>>>>>> >>>>>>>>> but after I've ported the GSoC2014-improved shading package to 1.8 >>>>>>>>> >>>>>>>>> Tilman >>>>>>>>> >>>>>>>>> Am 07.08.2014 12:35, schrieb Andreas Lehmkühler: >>>>>>>>>> Hi, >>>>>>>>>> >>>>>>>>>> there is already a number of solved issues and I guess it's >>>>>>>>>> time for a new bugfix release. >>>>>>>>>> >>>>>>>>>> I'm working on PDFBOX-2250 and I'd like to finish that >>>>>>>>>> first but how about a new release in 2 or 3 weeks from now? >>>>>>>>>> >>>>>>>>>> WDYT? >>>>>>>>>> >>>>>>>>>> BR >>>>>>>>>> Andreas Lehmkühler >