Re: [fetchmail] The 5.9.14 release of fetchmail is available2002-09-06 07:14:30Quoting from Eric S. Raymond's mail on Fri, Sep 06, 2002 at 05:57:20AM -0400: The 5.9.14 release of fetchmail is now available at the usual locations, including <URL:http://www.tuxedo.org/~esr/fetchmail>. The following patches (sent before, now remade with 5.9.14) should be considered (important points are marked with a `*'): Patch 1: fetchmail-5.9.14-ifelse.patch - to prevent gcc from complaining about the `if ... if .. else' construct. Patch 2: fetchmail-5.9.14-minor.patch * do not convert file named "-" to an absolute path. "-" is interpretated as stdin/stdout in other places. - sslcertpath could be relative. Patch 3: fetchmail-5.9.14-report.patch * if an untagged NO is received with imap, wait for a tagged response before continuing. otherwise, the transaction goes out of sync. - initialise the size (important when a mail is not readable). - do not print size of oversized mails twice Patch 4: fetchmail-5.9.14-socketclose.patch * fetchmail hangs on a socket error with an imap server. this error occurs in this fashion: - the headers have been read successfully, - the smtp socket has been opened successfully - DATA command has been sent to the smtp socket - fetchmail tries to receive the body - a socket error occurs on read (no signal is received) - fetchmail sends a QUIT to the smtp server. however, as the smtp socket is waiting for DATA, the QUIT is interpreted by the smtp server as part of DATA. (if protocol is set to auto, fetchmail sends the next MAIL FROM instead) - smtp server waits for more DATA, fetchmail waits for response to QUIT (or MAIL FROM). - status quo remains until the smtp server times out. * the mailserver_socket is not closed on a timeout (except a connect timeout). - do not close stdout while releasing the sink. Sunil Shetye.
fetchmail-5.9.14-ifelse.patch
fetchmail-5.9.14-minor.patch
fetchmail-5.9.14-report.patch
fetchmail-5.9.14-socketclose.patch
|
|