Sergey Poznyakoff <g...@gnu.org.ua> ha escrit: > Before attempting compression format identification, tar tests if > the first block read is a valid tar header block. However, the > test itself is stricter than necessary. The following patch should > fix this:
This, however, will mistakingly qualify some compressed archives as uncompressed (see commit 1847ec67cec36a17354115374954fea211d1f0da), so definitely the test must be more sophisticated. Regards, Sergey
