Re: bash exorcism experiment ('bug' 762923 & 763012)
- To: email@example.com
- Subject: Re: bash exorcism experiment ('bug' 762923 & 763012)
- From: Russ Allbery <firstname.lastname@example.org>
- Date: Tue, 30 Sep 2014 18:23:22 -0700
- Message-id: <[🔎] email@example.com>
- In-reply-to: <firstname.lastname@example.org> (Russell Stuart's message of "Wed, 01 Oct 2014 09:43:06 +1000")
- References: <20140927153218.GB1755@nl.grid.coop> <20140927163017.GA7300@grep.be> <20140927181845.GA2816@gaara.hadrons.org> <20140927184257.GA26444@x230-buxy.home.ouaza.com> <20140928021144.GE1755@nl.grid.coop> <20140928083350.GA4738@riva.ucam.org> <email@example.com> <alpine.DEB.firstname.lastname@example.org> <email@example.com>
Russell Stuart <firstname.lastname@example.org> writes:
> The only reason I ported things to dash is /bin/sh is now linked to it,
> which in view makes it the standard shell. Every script starting with
> #!/bin/sh must work with. If I can't get it working because of a
> missing feature like arrays then I have to change it to #!/bin/bash or
> something, and add an explicit dependency.
bash is essential, so from a Debian perspective, you don't need to add an
extra dependency. Of course, that's exactly what this thread is about,
but that's why we're unlikely to ever remove it from the essential set.
It's a lot of work and archive churn to add all those dependencies, and
it's not at all clear that we're better off in the end, or at least not
sufficiently better off to warrant the effort.
Targetted removal of uses of bash where they're not required is, of
course, still useful, and I've been in favor of that going all the way
back to the days of active checkbashisms development and various Lintian
Russ Allbery (email@example.com) <http://www.eyrie.org/~eagle/>