couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Robert Samuel Newson <>
Subject Re: [DISCUSS] Erlang whitespace standards (was: [POLL])
Date Fri, 04 Apr 2014 20:52:56 GMT
I’d ask what problem it solves to reformat all the things versus the benefit of not doing

Maybe I’ve just been worn down over time but I don’t struggle to read the code however
it’s indented. New code, and modifications to existing code, are generally tidier because
we’re all more familiar with typical erlang indentation styles.

The survey clarified a few practices that had more than one popular style, and nailed down
the group opinion on them. That’s valuable in itself. New modules should certainly follow
these guidelines.


On 4 Apr 2014, at 21:18, Joan Touzet <> wrote:

> I intentionally defer here to people actually working in each file. There's nothing worse
than someone coming in and excreting all over your favourite thing in the name of some arbitrary
standard. That said, if our refactors/rewrites lag too much, we may never get around to fixing
things, and that'd be disappointing.
> Would it be fair to put a line in the sand, say, 12 months out, and strongly suggest
that all files be reformatted by then? Or we could tie it to a 2.0 release or similar.
> -Joan
> ----- Original Message -----
> From: "Robert Samuel Newson" <>
> To:
> Sent: Friday, April 4, 2014 10:33:06 AM
> Subject: Re: [DISCUSS] Erlang whitespace standards (was: [POLL])
> I appreciate firming up a consensus on indentation styles but I want to be clearly -1
on a codebase-wide reformatting for the foreseeable future. Beyond the merges, we have active
branches for older releases, the more reformatting we do, the harder back- and forward-porting
becomes. I like the idea of being more consistent for future work and, where code is particularly
crufty, refactoring before making a change. The "worst" formatted code in couchdb is generally
the oldest, and much of that needs a refactor/rewrite as we get to it.
> B.

View raw message