The idea from here is then to surf through the link errors and, for a while, to let them drive the development. The biggest problem I can see here is that the dead-code stripping of gcc is not so good and my previous experience of the linker is that it will fail with link errors for code that is referenced but never executed.
After this my development plan is simple.
- Once Cello links I can then work through a set of sample files testing the read.
- When the read appears to work I can look at the file generation code and work through that.
- When the file generation works I will work on the saving of the files. At this point I will pick up the pieces of the structured document and file compression that I have postponed.
No comments:
Post a Comment