| PLEX86 | ||
|
Minicom with Xmodem continuedAs it turns out, my buttumption that minicom (with lrzsz) doesn't support xmodem checksum was wrong. It supports both checksum and CRC. My connection just didn't work for some reason. The connection works fine with XP and Hyperterminal with both Xmodem Checksum and CRC. It fails on both using minicom. No software or hardware control is used. So I inserted a third box so as to monitor everything that happened on the serial cable and after staring for hours at the traces and all the Xmodem specifications that I have downloaded, I can't see anything wrong with the data stream. A proper packet is received with the SOH, number and complement number, data, and crc. An ack (0x6) is sent back. Hyperterminal loves it and sends the next packet. Minicom says that a nak was received and starts over. Mozilla Thunderbird Linux Paragon NTFS for Linux 1.1 - Soft.comParagon NTFS for Linux is designed to mount NTFS parbreastions under Linux operating ... System Requirements: Linux kernel 2.4.X - Microsoft operating system: ... www.soft30.com-softdown-45-3559.htm - 31k... Then I hardcoded a program that would receive a data stream as per the exact Xmodem specifications and fake a good reply with an ack. Again, Hyperterminal accepted it but minicom said that I was nak'ing the transmission even though this program can't even send a nak. buffering on named pipes Captain Dondo When process A writes to the pipe (named or not), it goes into a buffer inside the kernel. Yes, from the... There are a billion copies of Hyperterminal around so someone would have noticed if it didn't follow standards. But by the same token the lrzsz module is used in most Linux comm programs and someone long before me would have noticed any bug in the file transfer protocols. Can anybody think of what could be triggering a false nak in minicom? CptKrf
|
||||
Linux groups from Newsgroups The #1 Usenet Provider on the Internet
|
||||