# Re: [PATCH 3/3] diff --stat: sometimes use non-linear scaling.

1 messages from 2006-10-12 to 2006-10-12. Participants: apodtele.
Thread: https://gitlist.dev/t/5903

## apodtele, 2006-10-12 15:04

Subject: Re: [PATCH 3/3] diff --stat: sometimes use non-linear scaling.
Message-ID: <d620685f0610120804l59c38e07uf7b624b2a0026aea@mail.gmail.com>
URL: https://gitlist.dev/e/d620685f0610120804l59c38e07uf7b624b2a0026aea%40mail.gmail.com

```
Sublinear solution without patch is below.

On Sept 28, 2006 Martin Waitz wrote:
> On Wed, Sep 27, 2006 at 08:12:49AM -0700, Linus Torvalds wrote:
>> No _way_ is it correct to show more than three characters if there were
>> three lines of changes.
>>
>> I think "nonlinear" is fine, but this is something that is "superlinear"
>> in small changes, and then sublinear in bigger ones (and then apparently
>> totally wrong for one-line changes).
>>
>> It should at least never be superlinear, I believe.
>
> So if we want to keep the logarithmic scale we can do some maths:
> Assume we use a formula ala
>         length = a log(change + b) + c

You are probably looking for much simpler, log-less, and pure integer:

        length = width * change / (width + change) + 1

Assuming target witdth of 40, for example, it will produce

Change    Length
1         1
2         2
3         3
4         4
10        9
20        14
30        18
50        23
100       29
1000      39
10000     40
1000000   40

--
Alexei

```