dcc77d71c4
* New variables: + vm-folder-file-precious-flag * added CRAM-MD5 as an authentication method for IMAP. * vm-su-do-date: interpret 2-digit years in the RFC-822 matching case as 20XX if year starts with 0-6. * vm-rfc1153-or-rfc934-burst-message: skip spaces in addition to newlines that occur after a separator line. A digest has been observed with that kind of deformity. * treat enable-local-eval as we do enable-local-variables--- always bind it to nil. * vm: don't bind vm-auto-decode-mime-messages non-nil during initial message preview if it is nil. * vm-mime-display-internal-text/html: dropped (sleep-for 2). No one cares enough about the "Need W3 to inline HTML" message to wait 2 seconds afterward. * added menu entry to allow MIME objects to be converted to another type and displayed. The new type is determined by vm-mime-type-converter-alist. * added koi8-r to vm-mime-mule-charset-to-coding-alist (XEmacs only). * vm-pop-read-list-response: check for nil return of vm-pop-read-response before using return value. * vm-pop-read-stat-response: check for nil return of vm-pop-read-response before using return value. * vm-encode-coding-region: use unwind-protect to make sure (well more likely) that the work buffer always gets killed if it has been created. * vm-decode-coding-region: use unwind-protect to make sure (well more likely) that the work buffer always gets killed if it has been created. * vm-mime-convert-undisplayable-layout: put object buffer on garbage list sooner to make rarer the situation where the buffer never gets deleted. * Makefile: remove function definition of vm-its-such-a-cruel-world after it is run. * vm-md5-region: if vm-pop-md5-program exits non-zero, signal an error. Also if the work buffer is not at least 32 bytes long, signal an error. This prevents naive callers from assumption all is well and using a possibly empty string as an MD5 hash. * vm-md5-region: check the MD5 digest returned for non-hex-digit characters and signal an error if any are found. * vm-get-file-buffer: use find-buffer-visiting if it is fbound. * vm-build-threads: fixed loop that removed child messages from a parent when better information about a child's parent is found. Previously the loop attempted to remove the same message from the parent over and over. * vm-build-threads: gather thread data using References and In-Reply-To for all messages before using the Subject header. This helps prevent the case where References says A is the parent of B but because of clock skew B is older than A, which can lead to B being considered the parent of A if A and B have the same subject and vm-thread-using-subject is non-nil. |
||
---|---|---|
.. | ||
patch-aa | ||
patch-ab |