| commit | 3dc1d4ee624204aa6e7903f31e7cd5971238d4a2 | |
| author | Steffen Nurpmeso <steffen@sdaoden.eu> | |
| Thu, 26 Oct 2017 14:20:00 +0000 (26 16:20 +0200) | ||
| committer | Steffen Nurpmeso <steffen@sdaoden.eu> | |
| Thu, 26 Oct 2017 14:40:05 +0000 (26 16:40 +0200) | ||
| tree | bc4369671be859572b9e7099ed7568fd47a267ab | treesnapshot (tar.gz zip) |
| parent | 041b2934cb95d3807f8ae8129493aefcd0ea7c8e | commitdiff |
FIX debug assertion (Thomas Dickey)..
It turns out that [bf908249] (a_crese_polite_rt_mft_move(): if
upgrading all to To: - simply redo!) was _too simple_ instead,
because it forgot to update the type of the node if we do not
redoing the list, but simply assigning to To:, triggering a debug
assertion later on caused by Cc: entries ending up in To: but
nonetheless proclaiming they were in Cc:.
One more such missing-concentration errata, but also one more
because of the terrible way we do things. These nodes only need
to know where they reside because they become joined into a single
list for easier iteration and then splitted into their sublists
over and over again. That is not the way this should be, even
though that is the historic approach.
It turns out that [bf908249] (a_crese_polite_rt_mft_move(): if
upgrading all to To: - simply redo!) was _too simple_ instead,
because it forgot to update the type of the node if we do not
redoing the list, but simply assigning to To:, triggering a debug
assertion later on caused by Cc: entries ending up in To: but
nonetheless proclaiming they were in Cc:.
One more such missing-concentration errata, but also one more
because of the terrible way we do things. These nodes only need
to know where they reside because they become joined into a single
list for easier iteration and then splitted into their sublists
over and over again. That is not the way this should be, even
though that is the historic approach.
| cc-test.sh | diffblobblamehistory | |
| cmd-resend.c | diffblobblamehistory |