I hadn't answered your questions, so here you go.


That assumption would, of course, only be valid if you believe the article that you refer to.


Chi's code does not contain a totally original work. The basic CRC code and the table were extracted from this site's Code Library, and as such, proper attribution would go to John Wren. Since Chi's code is essentially the same as John's, it stands to reason that Chi's would behave the same way.


That particular CRC implementation (code and table) is correct for data that is transmitted and received MSbit-first. It is also the implementation that is (mis)used for the XModem protocol. I say misused because the vast majority of the time XModem is used on UART-driven serial links, which we know transmit and receive LSbit-first. That said, the initial CRC value (typically all zeroes or all ones) and whether or not the remainder is complemented before transmission is beyond the scope of the code offered and would be up to you to determine as appropriate for your application or protocol.