Re: Draft spec for new dpkg "triggers" feature

[Ian Jackson <ian@davenant.greenend.org.uk>]

Florian Weimer writes ("Re: Draft spec for new dpkg "triggers" feature"):
> [Ian Jackson:]
> > 2. That's irrelevant.  dpkg does not need to run on every
> > POSIX-confirming system.
> 
> I'm more concerned by people who insist on running maintainer scripts
> with castrated, yet POSIX-conforming shells.

Ah, yes, good point.

> Personally, I wouldn't object to something in Policy which mandates
> that /bin/sh must support the system's ARG_MAX value, and must not
> impose a different limit.

Do any of our shells currently impose a different limit ?  The SuSv3
limit is 4096 bytes for arguments and environment, which is already
getting into the kind of range where C programmers usually use dynamic
allocation, so if they do it's probably easy to fix.

Ian.