control: severity -1 important On 22.03.19 12:52, Antti Salmela wrote: > On Fri, Mar 22, 2019 at 11:38:44AM +0100, Christian Kastner wrote: >> I'm wondering as to the practical significance, though. Were you >> affected by this limit (in which case I would increase the severity), >> or is this more of an observation? > > I was affected, in one instance programmatically generated crontab > file was over limit.
I'm sorry to hear that you were affected, I underestimated the impact of this change. That's a strong argument for implementing the patch you have provided. The reason I'm still hesitant to implement this is because we don't necessarily trust the data even in the system crontab directories. In #378153, someone manage to force a core dump to /etc/cron.d, for example. On the other hand, we've added numerous checks to these directories in the meantime, so perhaps this is would be too overly cautious. Another idea I'm toying with is to make this limit configurable through /etc/default/cron. I'll put some thought into this over the weekend, and I'll try to get a fix out as soon as I can. >> I picked 1.000 out of a hat, because I never thought that any >> reasonable crontab would reach this limit, with maybe the >> exception of a heavily commented one. > > I don't know if this is reasonable use of cron, but I do know > that I would be either reimplementing cron or adding complexity > and fragility to other parts of the system if doing this without > cron. Well, I'd never thought 1.000 lines would be an issue, but if you have a use case, you have a use case. > Nowadays systemd timers would probably be a better option, but systemd > was not yet available when this was developed.
