1TIMEOUT doesn't always work 2 If a shell is invoked then procmail may wait while executing a 3 command for longer than TIMEOUT specifies. 4 5regexp matching bug 6 Some regexps may return an incorrect value in the MATCH variable. 7 In particular, this can happen when 'redundant' * or + operators 8 appear on the lefthand side of the \/ token. (The current best 9 guess is that fixing this would require moving the "beginning of 10 match" pointer into the "per-task" regexp structure.) 11 12Incorrect usage of -lnsl and -lsocket 13 libsocket and libnsl should be avoided if not needed as they're 14 broken under at least one version of IRIX. If your procmail 15 binary doesn't reliably find user's home directories, or otherwise 16 appears to have problems accessing the passwd file, try removing 17 -lnsl and -lsocket from the SEARCHLIBS variable in the Makefile, 18 then recompile. 19 20No "$@" in logged abstract 21 When passing "$@" to a command, the "Folder:" logged does not 22 include any of arguments passed via the "$@" 23 24Custom delimiter lossage 25 When using a custom message delimiter (like MMDF's ^A^A^A^A\n) 26 procmail fails to escape the delimiter in incoming messages, 27 resulting in corrupted mailboxes. Best current workaround is 28 to put a recipe in the /etc/procmailrc file that reads something 29 like 30 :0 fw 31 * ^A^A^A^A$ 32 |perl -pe 's:\001\001\001\001$:\002\002\002\002:' 33 The "^A"s in the condition need to be real control-A characters. 34 35Lost value on failed chdir() 36 If the user assigns a value to MAILDIR and the chdir() fails, 37 the previous value of the variable (but not the process's cwd) 38 is lost and replaced with "." 39 40Control-M isn't whitespace 41 Every so often someone copies an rcfile from a Windows box and 42 it ends up with CRs on the end of every line. They should be 43 treated just like spaces and tabs are. As is, the results are 44 really confusing. 45 46Shell Expansion 47 Shell expansion of conditions via the '$' special treats 48 double-quotes weirdly. They should not be considered special 49 at all there. 50 51Backslash-newline inconsistencies 52 Backslash-newline removal is almost completely inconsistent and 53 should be straightened out some how, but without breaking 54 anything that's in use. 55