On 11 June 2012 09:39, Jasper van de Gronde <th.v.d.gro...@hccnet.nl> wrote: > On 08-06-12 23:43, Jon Nordby wrote: >> ... >> For instance adding a layer in GIMP may result in 5, 10 or even more >> changes to the GEGL graph. Grouping these into one transaction then >> allows GIMP to treat them as one logical change. >> ... >>> Which cool things do you think of? >> >> - Sharing the undo/redo history across applications using GEGL >> - Allowing the undo/redo history to be saved together with a document. >> When you open an old file, you have all the history of it, since it >> started existing. >> - Working on the same document from two different applications at the same >> time. >> - Preserving non-linear histories >> - Browsing through the entire timeline of a document > > This would imply that the GEGL "model" would have to be translatable > (ideally one-on-one) to the document model. Or where you just referring > to the case when people are actually editing GEGL graphs, rather than > some other image format?
Yes, this implies that a mapping between the GEGL graph and the internal document model is possible. That same problem comes up when talking about structured import/export of documents (ref: "[Gegl-developer] Possible strategy for improving file format support in GEGL" thread). How feasible this is for actual applications out there remains to be seen. I think it should be doable for MyPaint at least, and more GEGL/graph centric applications like a node-based compositor or apps that just do "operation stacking" will likely be simpler. Once we get some real-life experience in this problem area we can hopefully establish some best practices. -- Jon Nordby - www.jonnor.com _______________________________________________ gegl-developer-list mailing list gegl-developer-list@gnome.org https://mail.gnome.org/mailman/listinfo/gegl-developer-list
