This is a really big issue for us too.
Using `.symlink` files can be tricky. There is always (an whether small or not) possibility that the `.symlink` file already exists....
Why not using another archiver instead of `zip`?
For example `tar` + `gzip` is supported on most platforms.
Another safe procedure is to create the `zip` archive as follows:
- Add a subfolder and store the actual Job files here instead of in the archive root (excluding the symlinks!)
- Add a file containing all symlinks (relative to the Job root folder)
| +- dir1
| +- dir2
| +- file1
| +- file2
| +- ...
where symlinks.map contains all the symlinks (relative to <actual_archive_contents>)