Boost logo

Boost :

Subject: Re: [boost] [Intrusive] Cleaning up trailing white space
From: Olaf van der Spek (ml_at_[hidden])
Date: 2012-05-29 03:56:39


On Tue, May 29, 2012 at 9:52 AM, Lars Viklund <zao_at_[hidden]> wrote:
> On Tue, May 29, 2012 at 09:34:02AM +0200, Olaf van der Spek wrote:
>> On Tue, May 29, 2012 at 9:27 AM, Lars Viklund <zao_at_[hidden]> wrote:
>> >> Can't the script be ran once for the entire repo?
>> >
>> > The downside of making "pointless" changes like that is that you get a
>> > commit that will look like the file was changed, effectively making all
>> > diffs across that commit have lots of modifications obscuring any real
>> > changes.
>>
>> Tell diff to ignore whitespace (changes).
>> But yeah, it might be a bit problematic.
>
> I did intentionally not mention that, as there's a lot of tools out
> there, many of which doesn't have/want such a feature.
>
> It still doesn't change that there would be a spurious commit out there
> muddling the waters, particularly regarding the quite important "last
> touched" date.

True

> I'd expect such a commit to put a fair amount of strain on the
> repository infrastructure, considering that it'll have over nine
> thousand source files touched, with a change to possibly every line of
> every file.

You could do it multiple commits if that's the problem.

> Not to mention that such a script cannot deduce the intended size and
> purpose of tabs, so the process/output would have to be vetted by a
> human to ensure that it has not screwed up indentation/alignment.

It's about trailing whitespace, so that doesn't apply.

-- 
Olaf

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk