On Sat 2016-04-09 07:31:47 -0400, David Bremner <david@tethera.net> wrote: > Daniel Kahn Gillmor <dkg@fifthhorseman.net> writes: > > >> + status = _notmuch_message_delete (message); >> + if (status) /* we'll report the last failure we see; >> + * if there is more than one failure, we >> + * forget about previous ones */ >> + last_error = status; > > I was initially worried/paranoid that there might be some risk of data > loss by continuing deleting after the first bad status; that doesn't > seem to be the case, but there doesn't seem to be much advantage in > continuing either, since the only error currently returned from > _notmuch_message_delete is from _notmuch_database_ensure_writable, > which seems likely to persist. So perhaps exiting the loop on the > first error might be less confusing. At the moment, that's the only possible error, but maybe there will be more errors as notmuch grows/changes in the future? I figure we should try once to delete each message we know we want to delete, regardless of the success in deleting other message. So i'm inclined to keep it as-is, but if someone wants to prepare a patch for the other direction i wouldn't object strongly. > Other than that, and my bug in ghost-report, the series looks good to > me. cool! do you need another rev of the series from me, or is that something you're up for applying directly? --dkg