So this put me back into research mode and I noticed Mozilla’s FileReader documentation and it would simplify things a great deal: I no longer would have to have a separate “upload” phase to allow map editing to be possible. Instead I could just let users select a few maps, have an instant display and analysis and only after that go for any upload requirement. Also thanks to Mozilla’s pages I now know that browser support is good enough to go with these new “HTML5” technologies. The biggest issue going with this is that I need to rethink a lot of things again. Like, the component that I initially used for file upload will be of no use. Instead of sending a big chunk of data I should send multiple smaller chunks, validate them on server side and then construct the final file. Even this way some Return to the Roots 1024 x 1024 map files may cause problems and I need to implement some form of compression.
The question at this point is: is it worth the time to redesign again?