Hello again Martin. Martin Lucina wrote in <[email protected]>: |On Tuesday, 26.03.2019 at 15:24, Steffen Nurpmeso wrote: |> This should not be needed, you should be able to say |> |> $ make SHELL=/what/you/want VAL_SHELL=/what/you/always/want |> |> and get along with it. (As long as it is a somewhat compatible |> POSIX shell.) | |No luck: | |--- cut here --- |Operating system is Linux |Checking for basic utility set |Setting up configuration options ... done |Joining in ./make.rc ... ERROR: invalid syntax in: # Add support for \ |history management. ... |make: *** [makefile:22: config] Error 1 |--- cut here --- | |Tried this both with and without a symlink of /bin/sh -> /bin/bash, and |also with "env - make config" to rule out any nonsense polluting the |environment.
Hmmm. Could you please run $ make VERBOSE=y and give me everything before "Joining in ./make.rc", please? The error either points to a shell or an awk error. I have commented out the shell case to test awks, and i cannot reproduce with all of nawk, gawk, busybox awk and also mawk, which i think (have heard) is the default awk on Debian. So dead end also there. |> Now you are funny.. A.K.A.: this is mysterious. Are you saying |> that dovecot sends complete MBOX- rather than RFC 5322 messages |> over IMAP? You see me entirely perplex. I must admit i have not |> tested for some time now, as nothing has changed in this area. | |Ha. Well, it may be that this is some years-old artefact from my migrating |from mbox to Maildir on the server, I mean really, like >10 years ago. |Having said that, neither mutt, Thunderbird, nor Roundcube have any |(visible) problem with my mailboxes. Ok, well i see i have a "XXX" comment saying that the error log should possibly be restricted to occur with -t and in compose-mode only, and i will look into this (we can now pass flags down to this place which turns that log off on request, which i think would be even better) once my new box is fully setup. But it is not in v14.9.13. The change will be in the next release, but will land on stable/* and master as soon as possible. (Not right now, sorry, the keyboard controller of my used MacBook Air died, and i am writing this via wireless keyboard, logitech K400+, and you better do not buy that as it seems to use some kind of Nagle's algorithm to safe battery, it is ridiculous! And i have bought the first all-new computer in a decade, an IdeaPad, but missing network connection driver yet. Or have to build my own kernel. Switch to CRUX-Linux on bare metal. I am undecided. Immense change to hardware in a decade, that IdeaPad can btrfs scrub 70+GB in 45 seconds, it is unbelievable. Anyway, hard working here at the moment. I report on this as soon as possible, hopefully this week!) |> that is really terrible. But Gunnar Ritters IMAP cache is really |> a smart idea, is it? | |Yes, I've been using Mutt's implementation of this for ages. Hm, i do not know mutt that much, despite having used it for some years. I know about a "header cache" they have, but do not know exactly what they "cache". I personally am in favourite of Zawinskis "index" that had been implemented for Netscape Mail in the 90s, which i phantasize to be implemented for this MUA some future day. Our IMAP cache can more, it is something like a usable local mirror of all downloaded data. E.g., "disconnect*" downloads the entire mailbox, on which you can operate locally in many ways. When connected again, local modifications will be synchronized with the server. Ciao, --steffen | |Der Kragenbaer, The moon bear, |der holt sich munter he cheerfully and one by one |einen nach dem anderen runter wa.ks himself off |(By Robert Gernhardt)
