While capturing the C:\Windows\Inf folder for testing purposes, I saw your tool has recorded in the XML data just about 19 MiB of HARDLINKBYTES... my ImagePyX says zero, instead: who is wrong?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Microsoft does not provide any documentation as to how the HARDLINKBYTES and similar statistics in the XML data actually should be calculated. However, at some point I had done a few tests to determine the behavior of their ImageX, and I tried to make wimlib produce the same behavior. This is done in xml_update_image_info() and calculate_dentry_statistics() in src/xml.c, and there are some comments about the exact behavior I am trying to reproduce. Basically, assuming no alternate data streams, the "hard link bytes" seems to be the number of additional bytes of data you would have to extract if you extracted full file copies instead of creating hard links when appropriate.
Regardless, I suggest you capture the same data with Microsoft's imagex.exe and see how the values compare to wimlib's calculation and ImagePyX's calculation.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You're right: I tought it were "Total bytes for captured files represented by hard links", but I see that ImageX records the same number in that field... or, at least, it does in its last version.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
While capturing the C:\Windows\Inf folder for testing purposes, I saw your tool has recorded in the XML data just about 19 MiB of HARDLINKBYTES... my ImagePyX says zero, instead: who is wrong?
Hi,
Microsoft does not provide any documentation as to how the HARDLINKBYTES and similar statistics in the XML data actually should be calculated. However, at some point I had done a few tests to determine the behavior of their ImageX, and I tried to make wimlib produce the same behavior. This is done in xml_update_image_info() and calculate_dentry_statistics() in src/xml.c, and there are some comments about the exact behavior I am trying to reproduce. Basically, assuming no alternate data streams, the "hard link bytes" seems to be the number of additional bytes of data you would have to extract if you extracted full file copies instead of creating hard links when appropriate.
Regardless, I suggest you capture the same data with Microsoft's imagex.exe and see how the values compare to wimlib's calculation and ImagePyX's calculation.
You're right: I tought it were "Total bytes for captured files represented by hard links", but I see that ImageX records the same number in that field... or, at least, it does in its last version.