Quote:
Originally Posted by Terisa de morgan
I've tested it at Windows7 64 bits: - Corrupt file according to Winrar and only two files inside (neither libadobe nor librmsdk were included).
- OK file according to 7zip (extraction in two steps) and all the files inside
- Tar file ok according to Winrar after extracting the tar from the tgz with 7zip
So, perhaps we can think that the problem is not the tar itself but the tgz (the compression).
Additional info: The initial KoboRoot.tar.gz has a size of 14.614.52 bytes, and after using gunzip and gzip (I have cygwin installed in my computer), the size is 14.311.424 bytes and Winrar doesn't complain about the file. |
Ok, that helps. I guess WinRar might just have a buggy implementation, as all tools on Linux which I know about work fine with it. This also means it should work fine on the Kobo. For peace of mind, I could set the compression level to 0 so it is still a tgz, but the only compression would be the Huffman dict. This should fix any issues WinRAR is having if it is about the compression.
Does WinRAR have any sort of verbose mode or log I can look at?
Edit: can you also try using
https://github.com/mholt/archiver to create a tar.gz out of some files and see if it also gives you an issue with WinRAR.