8434 | 09-Oct-1982 |
eric |
Handle Reverse-Path in some reasonable way. You must put the 'X' flag (M_FULLSMTP) in the local mailer's flags to get this line. |
8432 | 09-Oct-1982 |
eric |
print a status indication when a message is autoqueued, either from a -bq flag or a NoConnect |
8431 | 09-Oct-1982 |
eric |
print a status indication when a message is autoqueued, either from a -bq flag or a NoConnect |
8430 | 09-Oct-1982 |
eric |
print a status indication when a message is autoqueued, either from a -bq flag or a NoConnect |
8429 | 09-Oct-1982 |
eric |
Force -v override of deferred connect |
8428 | 09-Oct-1982 |
eric |
Force -v override of deferred connect |
8427 | 09-Oct-1982 |
eric |
Only put the "from" person on the error queue if an error actually occured -- prevents many "duplicate supressed" messages, etc. |
8426 | 09-Oct-1982 |
eric |
Only put the "from" person on the error queue if an error actually occured -- prevents many "duplicate supressed" messages, etc. |
8424 | 09-Oct-1982 |
eric |
implement an experimental forward-path algorithm by stripping off a leading "@..." component and dropping it on the end. |
8423 | 09-Oct-1982 |
eric |
implement an experimental forward-path algorithm by stripping off a leading "@..." component and dropping it on the end. |
8422 | 09-Oct-1982 |
eric |
Extend test mode to be able to handle multiple addresses, etc.; this lets me test forward-path specs more easily |
8355 | 07-Oct-1982 |
eric |
try to come up with a workaround on some of the IPC problems relating to interrupted accepts -- all this can go when everyone at Berkeley runs unflakey IPC. Also, never use ruleset three alone -- in
try to come up with a workaround on some of the IPC problems relating to interrupted accepts -- all this can go when everyone at Berkeley runs unflakey IPC. Also, never use ruleset three alone -- involves changing canonname again. Probably shouldn't do an auto-call of ruleset three now.
show more ...
|
8354 | 07-Oct-1982 |
eric |
never use ruleset 3 alone -- perhaps the standard use of ruleset three should go now, since noone uses it alone. |
8353 | 07-Oct-1982 |
eric |
never use ruleset 3 alone -- perhaps the standard use of ruleset three should go now, since noone uses it alone. |
8352 | 07-Oct-1982 |
eric |
try again to work around the kernel problems in IPC connects when interrupted. |
8351 | 07-Oct-1982 |
eric |
bother.... it takes time to close a socket |
8350 | 07-Oct-1982 |
eric |
allow passing through multiple rewriting sets in test mode; <ndir.h> => <dir.h> for maximum portability |
8349 | 07-Oct-1982 |
eric |
allow passing through multiple rewriting sets in test mode |
8348 | 07-Oct-1982 |
eric |
<ndir.h> => <dir.h> for maximum portability |
8347 | 06-Oct-1982 |
eric |
add test mode (MD_TEST, -bt) |
8337 | 06-Oct-1982 |
eric |
add test mode (MD_TEST, -bt) |
8336 | 06-Oct-1982 |
eric |
add test mode (MD_TEST, -bt) |
8335 | 06-Oct-1982 |
eric |
add test mode (MD_TEST, -bt) |
8334 | 06-Oct-1982 |
eric |
add MD_TEST mode |
8276 | 30-Sep-1982 |
eric |
workaround in daemon accept code for 4.1a kernel bug: if an accept gets an EINTR and then a valid accept before the interrupt returns, the connection will be lost forever (at least, as I understand i
workaround in daemon accept code for 4.1a kernel bug: if an accept gets an EINTR and then a valid accept before the interrupt returns, the connection will be lost forever (at least, as I understand it).
show more ...
|