>>> We're looking into it. >> Right, this is not reproducible in the test suite, but I can reproduce >> it when I replicate your setup. > > I created a special test suite that reproduces the problem reliably. > This occurs when Sieve editheader, Sieve redirect and LMTP are > combined. > > Now to fix it...Hello Stephan! Thank you for your quick reply and the informations. If a fix is available, I will test it. Regards, Jan
Op 3/3/2018 om 6:17 PM schreef Jan:>>>> We're looking into it. >>> Right, this is not reproducible in the test suite, but I can reproduce >>> it when I replicate your setup. >> >> I created a special test suite that reproduces the problem reliably. >> This occurs when Sieve editheader, Sieve redirect and LMTP are combined. >> >> Now to fix it... > > Hello Stephan! > > Thank you for your quick reply and the informations. > If a fix is available, I will test it.There is a fix. Pending internal review, I will submit it here. Regards, Stephan.
Op 3/3/2018 om 6:29 PM schreef Stephan Bosch:> Op 3/3/2018 om 6:17 PM schreef Jan: >>>>> We're looking into it. >>>> Right, this is not reproducible in the test suite, but I can reproduce >>>> it when I replicate your setup. >>> I created a special test suite that reproduces the problem reliably. >>> This occurs when Sieve editheader, Sieve redirect and LMTP are combined. >>> >>> Now to fix it... >> Hello Stephan! >> >> Thank you for your quick reply and the informations. >> If a fix is available, I will test it. > There is a fix. Pending internal review, I will submit it here.The actual fix (for v2.2/v0.4): https://github.com/dovecot/pigeonhole/commit/14ec6fe861403aa09a3b5790d113247fc1e1f94d Some related changes before and after this change (last commit is unrelated), mainly relating to creating a regression test for this particular problem. https://github.com/dovecot/pigeonhole/commits/master-0.4 Regards, Stephan.