Re: [dev] libutil/eprintf clarification patch (to the Evil_Bob's request)

From: FRIGN <dev_AT_frign.de>
Date: Mon, 21 Dec 2015 10:10:35 +0100

On Mon, 21 Dec 2015 09:45:25 +0100
"e_AT_bestmx.net" <e_AT_bestmx.net> wrote:

> > Now, what you did here is rip out this status code handling
>
> No, i didn't

My bad, I shouldn't look only at the diffs, might be confusing.

> > In sbase, we have this more or less "scheme" how function naming
> > works.
>
> how is it relevant to the topic?

venprintf is giving a false impression, because it neither exists now
nor does it take a status code as an argument.
I reworked eprintf.c to get rid of the duplicate code:
http://git.2f30.org/sbase/tree/libutil/eprintf.c

After further thoughts, I think we should keep this "usage"-check in.
Not only are we talking about error-functions here, we also already
apply a heuristic when the format-string ends with ':' to also print
the error-string afterwards.
I have got the feeling that if we go forward and start splitting up
eprintf into its individual subfunctionalities, we might as well
just write "fprintf(stderr, ...); exit(status)" everywhere in the
code, but we don't want that.

Also, who do you expect to add eprintu everywhere? I didn't see that
in your patch.
Additionally, next time, please take care of trailing whitespace.
If you miss it, highlight it in vim by adding to .vimrc:

highlight ExtraWhitespace ctermbg=red guibg=red
match ExtraWhitespace /\s\+$/

Cheers

FRIGN

-- 
FRIGN <dev_AT_frign.de>
Received on Mon Dec 21 2015 - 10:10:35 CET

This archive was generated by hypermail 2.3.0 : Mon Dec 21 2015 - 10:12:10 CET