On Wed, 5 Nov 2003, Aidan Evans wrote: > >> for it in Cyrus 2.1 beyond 2.1.13? There does not seem to be a mention of > >> it in 2.1.15 changes.html, but maybe it's covered under "Minor other bug > >> fixes". If there was a correction, is it easy to incorporate into 2.1.11? > >> I don't want to update Cyrus completely in the middle of term, but a small > >> code change and recompile may be acceptable. > > > >No, there is no fix for this problem currently. > > Would it make sense to have the detection of a "sieve parse error" in > lmtpd set a global flag to cause the lmtpd to do a "shut_down" after > processing the message, so that the corrupt state would not be propagated > to the handling of other messages? Alternatively, and perhaps simpler, > would it be bad to have this error cause a "fatal" in the lmtpd?
As I noted, this isn't a problem with the 2.2 bytecode system (except in timsieved, where I'd argue that having it fatal() is probably the wrong solution). -Rob -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Rob Siemborski * Andrew Systems Group * Cyert Hall 207 * 412-268-7456 Research Systems Programmer * /usr/contributed Gatekeeper