[P.S. I'm subscribed to the bug; you may drop me from explicit To or CC.] * Joerg Dorchain <jo...@dorchain.net> [200302 12:18]: > Then please include > https://github.com/neomutt/neomutt/commit/f197ab93c8436a39fc511f396acde24f90389f20?diff=unified > > in a new package, maybe with a Debian-specific change of the abort_backspace > default to false.
I'm not a DD or DM and have no upload rights, and don't have the time at the moment to prepare and test an upload properly (which is more than just applying the patch and building a package). I do appreciate the people who have in the past and are currently working on the Debian packaging of neomutt. Looking at the github issue, it looks like backspace is supposed to cancel the command, and the behavior that the bug reporter wants is to remain on the empty command line and ignore the backspace. But neither is the behavior I am seeing. First, the behavior in the github issue is specific to editing email headers while composing a message. For me, that specific case seems to abort the edit, and it seems to be an intentional change. I don't have a strong opinion about the github issue, but it is definitely not the issue in this Debian bug, although the code that implemented the intentional change may be the cause of this Debian bug. The situation that I feel is a grave bug is when in the message index, using certain commands that require more information (save-message is one such command), pressing backspace when the prompt is empty (only the ":" showing) causes the _default action_ to occur (e.g. save to the folder specified by the save-hook for that message) rather than aborting the command. As long as backspace does not have a binding after the prompt is aborted (e.g. in the index), I can probably live with either the abort or ignore behavior, but activating the default action is highly broken. I cannot tell whether the patch referenced in the previous message will fix this Debian bug. If the cause of the Debian bug has to do with mutt_enter_string_full being called for both completing a command in the index and for editing message headers, but the result of the function is being handled differently in the two cases, then perhaps the patch will allow the bug to be masked by unsetting the option, but with the option set the bug will still be present. I can't tell without looking at the rest of the code. I don't have a github account, and do not wish to get one for this. Will someone (Debian maintainer for neomutt, or someone else interested in this bug) please file this with upstream as a separate bug, pointing out that the other github issue is not the same as this bug, even though some of the same code may be involved. Thanks...Marvin
