Re: git: 1181d23c283c - main - =?utf-8?Q?textproc/riffdiff: update 3.4.1 =E2=86=92 3.4.2; update pager dependency =Q?textproc/riffdiff: update 3.4.1 =E2=86=92 3.4.2; update pager dependency =

From: Adam Weinberger <adamw_at_adamw.org>
Date: Sat, 18 Oct 2025 22:30:31 UTC
Git handles unicode characters like the arrow you put in, but mlmmj
doesn't. I thought we'd decided a while back to stick with ASCII whenever
possible in commit messages? Maybe I'm remembering wrong, but if I'm
correct about the decision we should probably try not to go outside the
box. You can see that mlmmj had trouble parsing other headers here as a
result.

Personally I'm a fan of UTF-8 in commit messages, but IIRC Bapt explained a
while back why mlmmj has to assume otherwise.



On Sat, Oct 18, 2025 at 3:41 PM Yuri Victorovich <yuri@freebsd.org> wrote:

> to textproc/moor?=
> MIME-Version: 1.0
> Content-Type: text/plain; charset=utf-8
> Content-Transfer-Encoding: 8bit
> X-Git-Committer: yuri
> X-Git-Repository: ports
> X-Git-Refname: refs/heads/main
> X-Git-Reftype: branch
> X-Git-Commit: 1181d23c283cd7e7cb4a0b3e636892a7f566e1a0
> Auto-Submitted: auto-generated
>
> The branch main has been updated by yuri:
>
> URL:
> https://cgit.FreeBSD.org/ports/commit/?id=1181d23c283cd7e7cb4a0b3e636892a7f566e1a0
>
> commit 1181d23c283cd7e7cb4a0b3e636892a7f566e1a0
> Author:     mew14930xvi <mew14930xvi@inbox.lv>
> AuthorDate: 2025-10-18 18:48:47 +0000
> Commit:     Yuri Victorovich <yuri@FreeBSD.org>
> CommitDate: 2025-10-18 19:41:09 +0000
>
>     textproc/riffdiff: update 3.4.1 → 3.4.2; update pager dependency to
> textproc/moor
>
>     PR:     290183
>     Submitted by:   Hiroki Tagato <tagattie@FreeBSD.org> (initial patch
> changing dependency)
> ---
>  textproc/riffdiff/Makefile  | 5 ++---
>  textproc/riffdiff/distinfo  | 6 +++---
>  textproc/riffdiff/pkg-descr | 2 +-
>  3 files changed, 6 insertions(+), 7 deletions(-)
>
> diff --git a/textproc/riffdiff/Makefile b/textproc/riffdiff/Makefile
> index 05b8b8806630..dad5455a555e 100644
> --- a/textproc/riffdiff/Makefile
> +++ b/textproc/riffdiff/Makefile
> @@ -1,6 +1,5 @@
>  PORTNAME=      riffdiff
> -DISTVERSION=   3.4.1
> -PORTREVISION=  1
> +DISTVERSION=   3.4.2
>  CATEGORIES=    textproc
>
>  MAINTAINER=    yuri@FreeBSD.org
> @@ -10,7 +9,7 @@ WWW=           https://github.com/walles/riff
>  LICENSE=       MIT
>  LICENSE_FILE=  ${WRKSRC}/LICENSE
>
> -RUN_DEPENDS=   moar:textproc/moar
> +RUN_DEPENDS=   moor:textproc/moor
>
>  USES=          cargo
>
> diff --git a/textproc/riffdiff/distinfo b/textproc/riffdiff/distinfo
> index 0cbf546b5803..2e2f49bf6b3c 100644
> --- a/textproc/riffdiff/distinfo
> +++ b/textproc/riffdiff/distinfo
> @@ -1,4 +1,4 @@
> -TIMESTAMP = 1758123882
> +TIMESTAMP = 1760752177
>  SHA256 (rust/crates/addr2line-0.14.0.crate) =
> 7c0929d69e78dd9bf5408269919fcbcaeb2e35e5d43e5815517cdc6a8e11a423
>  SIZE (rust/crates/addr2line-0.14.0.crate) = 30065
>  SHA256 (rust/crates/adler-0.2.3.crate) =
> ee2a4ec343196209d6594e19543ae87a39f96d5534d7174822a3ad825dd6ed7e
> @@ -247,5 +247,5 @@ SHA256 (rust/crates/zerovec-0.11.2.crate) =
> 4a05eb080e015ba39cc9e23bbe5e7fb04d5f
>  SIZE (rust/crates/zerovec-0.11.2.crate) = 124500
>  SHA256 (rust/crates/zerovec-derive-0.11.1.crate) =
> 5b96237efa0c878c64bd89c436f661be4e46b2f3eff1ebb976f7ef2321d2f58f
>  SIZE (rust/crates/zerovec-derive-0.11.1.crate) = 21294
> -SHA256 (walles-riff-3.4.1_GH0.tar.gz) =
> d209e5b5a68907382cc91061d2e0570789293214b402c1b344008760fe298302
> -SIZE (walles-riff-3.4.1_GH0.tar.gz) = 2074368
> +SHA256 (walles-riff-3.4.2_GH0.tar.gz) =
> 51f41141af4d7b2b60010fc165cf5ff42a79fa50a6996ba6f8771ecc8f9e525a
> +SIZE (walles-riff-3.4.2_GH0.tar.gz) = 2075149
> diff --git a/textproc/riffdiff/pkg-descr b/textproc/riffdiff/pkg-descr
> index 57bf7b116c04..74a5e4669659 100644
> --- a/textproc/riffdiff/pkg-descr
> +++ b/textproc/riffdiff/pkg-descr
> @@ -3,5 +3,5 @@ changed.
>
>  Much like 'git', riff sends its output to a pager, trying these in order:
>  1. whatever is specified in the '$PAGER' environment variable
> -2. 'moar' because it is nice
> +2. 'moor' because it is nice
>  3. 'less' because it is ubiquitous
>


-- 
Adam Weinberger
adamw@adamw.org