the future of ii (was: [hackers] [PATCH] [ii] get rid of EXIT_bla crap)

From: Nico Golde <>
Date: Fri, 14 Oct 2016 12:52:27 +0100

* Markus Teich <> [2016-10-14 12:45]:
> Laslo Hunhold wrote:
> > Oh, and before you inform me. I know that OpenVMS has the roles
> > inverted (exit(1) means success). However, there are two aspects we
> > should keep in mind:
> >
> > 1) OpenVMS is not the most popular OS any more.
> > 2) Using EXIT_* we cannot write reliable manuals, as we do
> > use 0, 1, 2, ... there.
> adding to that, when reading code with EXIT_bla, you don't know immediately how
> it is defined. Therefore it can easily be confused when also using C-functions
> which return 1 as "true" and 0 as "false". I get that abstracting the value away
> and using a name instead can make sense, but when using `return` or `exit()`,
> you don't have to give meaning to the `0` or `1` while in a line like `ret =
> calloc(5 * sizeof(struct bla) + 7);` it is very helpful to know what those
> numbers refer to semantically.

FWIW, I do think that EXIT_FAILURE and EXIT_SUCCESS is more readable
than 0/1 and well defined. Ofc, this is not true for EXIT_TIMEOUT
necessarily (but that hardly makes a difference for practical purposes here

Anyhow, I don't think I have the mental capacity to have a purely academic
discussion about this topic and I find it for the most part very subjective.
I also realize that I lack the time and am emotionally not attached to ii
anymore as I used to be in the past. I'm essentially just a bottleneck to its

Hiltjo Posthuma was kind enough to offer taking over the maintenance for this
project. Thanks for that, I really appreciate this! So I'm leaving this change
to his discretion.

Nico Golde - XMPP: - GPG: 0xA0A0AAAA

Received on Fri Oct 14 2016 - 13:52:27 CEST

This archive was generated by hypermail 2.3.0 : Fri Oct 14 2016 - 14:00:20 CEST