the changes to the VM necessary to support compressed images ended up adding a bunch of complexity for little benefit. the binary is bigger, adding other compression formats is an annoying process, etc, etc. pick all of that out. better idea would be to build a mapping of extensions to decompressors and pipe the image through them, like should have been done in the first place.
the changes to the VM necessary to support compressed images ended up adding a bunch of complexity for little benefit. the binary is bigger, adding other compression formats is an annoying process, etc, etc. pick all of that out. better idea would be to build a mapping of extensions to decompressors and pipe the image through them, like should have been done in the first place.