[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: parsetexi warnings
From: |
Gavin Smith |
Subject: |
Re: parsetexi warnings |
Date: |
Thu, 18 Feb 2021 22:06:11 +0000 |
User-agent: |
Mutt/1.9.4 (2018-02-28) |
On Tue, Feb 16, 2021 at 02:48:03PM -0800, Per Bothner wrote:
> On 2/16/21 2:39 PM, Gavin Smith wrote:
> > It's the C version of the Texinfo::Parser module. I will look at the
> > warnings (I think most of them were due to ignoring the return
> > value of asprintf).
>
> I see a lot of warnings - see attachment (edited down compile log).
> Most but not all from parsetexi.
>
> I'm sure most or all are harmless, but I think it's a good discipline
> to compile without warnings, at least when using "our own" compiler, gcc.
(I don't who this "we" is - saying that "we should do this" sounds a lot to
me like "you should do this".)
The previous maintainer wrote the following in the README-hacking
file and I still agree with this:
"Other -W options can be useful too, and patches are welcome to resolve
diagnostics; however, removing all possible warning messages, or
warnings with nonfree compilers, is explicitly not a goal."
Fixing warnings isn't a goal in itself, although obviously warnings
can show real problems, which was the case among the warnings you
posted and it's good to compile with -Wall occasionally. I've tried
to fix some of these problems. I still have a couple to look at.
> warning: suggest parentheses around ‘&&’ within ‘||’
That is not a real warning as people should know the operator
precedence, I feel. This makes the other warnings less conspicuous.