Re: [PATCH] maintenance: specify explicit stdin for crontab
flat view
From: Derrick Stolee <hidden>
Date: 2021-03-30 19:38:46
On 3/30/2021 1:43 PM, Todd Zullinger wrote:
Hi, Derrick Stolee wrote:quoted
On 3/30/2021 1:41 AM, Martin Ågren wrote:quoted
On Mon, 29 Mar 2021 at 23:23, Kevin Daudt [off-list ref] wrote:quoted
There are multiple crontab implementations that require stdin for editing a crontab to be explicitly specified as '-'.Amusingly, I wrote the exact same patch 2 weeks ago (including not dropping the `argc == 2` which Martin mentioned). That was in response to a report in the Fedora bugzilla: https://bugzilla.redhat.com/1939930 I thought cronie might be rather rare with it's non-POSIX handling of crontab without arguments.
Thanks for this link! I appreciate the context.
In the end, the cronie folks upstream adjusted things so that crontab behaves as defined by POSIX if stdin is not a TTY: https://github.com/cronie-crond/cronie/commit/8b0241fquoted
That allows cronie to behave more sensibly for interactiveuse without breaking tools like git maintenance. And it let me sidestep proposing a patch to git (or worse, maintaining it in the Fedora packages).
Nice!
But I didn't dig in to find out whether or how many other crontab implemntations had also eschewed the (rather poor) POSIX-confirming behavior. Knowing there are several among popular OS's makes it easy to see something like this patch being generally useful. Though, as Derrick notes below, we would break systems which implement crontab strictly per the POSIX spec. I don't know how many crontab's don't accept `-`. At the time, I checked on an older OmniOS system I had access to (based on Illumos/OpenSolaris) and it did not accept `-`. So my quick sample size of 3 (Fedora, CentOS, and OmniOS) I had a 1/3 failure rate.quoted
Thank you for reporting this, especially with a patch! However, I'm not sure about this adding of '-' being something that crontab ignores so commonly. My Ubuntu machine reports this: $ crontab -e - crontab: usage error: no arguments permitted after this option usage: crontab [-u user] file crontab [ -u user ] [ -i ] { -e | -l | -r } (default operation is replace, per 1003.2) -e (edit user's crontab) -l (list user's crontab) -r (delete user's crontab) -i (prompt before deleting user's crontab) Is there a way we could attempt writing over stdin, notice the failure, then retry with the '-' option?You'd skip the `-e` there, no? Running `crontab -` in a current ubuntu container with the cron package installed (what looks like vixie-cron-3.0pl1) works as expected.
Yes, this is my mistake. My machine supports "crontab -".
Perhaps a Makefile knob to allow systems with such a crontab to adjust the behavior would be an alternative to detecting and retrying? NEEDS_CRONTAB_STDIN_OPT or something like that, with config.mak.uname to override whichever default is chosen. Whether that's a better option really depends on how much effort it is to add and maintain the detection in the code weighed against how many systems would need to have the default changed. Mildy related, I wonder whether we'll eventually see a patch to use systemd timers instead of cron (optionally, of course). Fedora, for example, doesn't install crond by default anymore. (Though, warts and all, I still prefer crond myself.)
Perhaps the best way to approach this is to try adding '-' by default, and remove it and try again on a failure. If the usage is actually a problem, that first command should fail quickly. I'm not sure if we could rely upon a specific category of error codes or if we should just say "non-zero exit means retry". Thanks! -Stolee