Hello Johannes! [email protected] wrote in <0fe73f17e93f889c6f6acd123e8b3bfc@schoe\ pfer.info>: |> [email protected] wrote in |> <[email protected]>: |>| I fetched 2 messages from my POP server. .. |>|which worked properly. When I tried to save the second to nytimes |>|(real name!) I got: |>| |>| ? s2 nytimes mailx: nytimes: Snorkeling on empty pipe |>| ERROR# ? errors |>| 1. nytimes: Snorkeling on empty pipe |>| |>|and it didn't save. I could save it to a different file name. I |>|could open that file with mail and save it to nytimes. I could open |>|nytimes. The error seems to require both messages. The word |>|Snorkeling does not appear in either. Snorkeling is in nytimes, but |> |> What you see is the S-nail specific errno(3) EPIPE error string; | |Maybe a missing user folder in /var/spool/mail ?
It seems nytimes is a pipe/fifo, and S-nail will block signals while in this task -- C-c interruption will not do, all data needs to become consumed. Maybe i can do something about this for the next update release, add one more special case for writing to files which happen to be pipes or fifos etc. The real healing for all this not before v15, then this is "mailbox->append_msg()" and "mailbox" used to know what it is, and how to handle itself. I hope. |I don't know how heirloom-mailx handeld this. I use s-nail for sending ..actually it also hangs, and is not interruptible. I am now a bit surprised, in general the old code did not care for memory leaks and crash protection, but assumes "the user knows". I will look into this topic as soon as i can compile the development branch again, hopefully next week! We should be interruptible here. |cron/command-line mails, We will get better, especially regarding what -t can swallow, i promise. It unfortunately takes time. |i never used it as MUA. This is .. understandable. Some things will improve here, too. |>|not that phrase. I switched from mailx 12.5 recently, when Slackware |>|upgraded to it. |>| mailx v14.9.11, 2018-08-08 (build for linux) |>| Slackware, latest; kernel 4.19.4 | |Where does the string "mailx v14.9.11, 2018-08-08 (build for linux)" |come from? This is mutilated `version' command output: $ mailx -Xversion -Xx | head -n1 |Is this a prebuilt binary? |Or from slackbuilds.org, or from slackware-current? | |> Oh really! Please let me assume it is only because of OpenSSL, |> rather than because of the software as such, right? :-) | |I submitted the slackbuild, but aren't responsible for what or why |something lands in slackware, |but i proposed to replace heirloom-mailx. |OpenSSL didn't make headache i think, until now heirloom could be |patched. |For me it seemed, that heirloom-mailx development is now at s-nail, so |why not go with it. Thank you! v14.9.11 is 4674 commits ahead of Heirloom, and i hope for the better. |> The .SlackBuild script of Johannes Schöpfer is the most |> complicated think i have ever seen! I really had to think about |> it, but luckily and successfully avoided that, then. | |Well, it's basically the original slackware heirloom-mailx buildscript, |adapted to s-nail. Oh. :) |> We are at Slackware, too! Thank you, this is really nice!! |> I did not even know that Slackware is still truly alive, but |> i think: good that it is. | |:-) Thanks, Johannes! 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)
