A system and method for lossless data compression of byte oriented digital data, including but not exclusively executable program code and related data. This lossless data compression system and method when used to process sets of similar but not identical input data, will generate compressed forms of that data which also have a high degree of similarity. The present invention utilizes an update package generated based on the differences between the compressed forms of data, whereby the update package can be applied to the original compressed form to create the new form. A typical change to create the new form from the original would be when making bug fixes to deployed software.