Thanks @koobs.

In D16483#350723, @koobs wrote:

Use MASTER_SITES=CHEESESHOP by default for Python package distribution files when the source distribution (sdist) is published there (as this one is), unless there is a compelling temporary case to use an alternative source location (like GitHub); such as missing test files / important data files, etc.

PORTVERSION => DISTVERSION

Okay, thanks @koobs.

No problem. :-)

In D15771#343959, @koobs wrote:

I can arc create the current version of the diff if you'd like?

Thanks @koobs.

In D16208#343830, @ygy wrote:

@delphij @loader One question: Looking at the diff of built website before and after this change (https://github.com/yzgyyang/freebsd-doc-web-deploy/commit/dec5f088af404cbdd5dca78bbd21f4acedf56576#diff-effdc1d6d041099f8126e5625ae4f8f3R2), it seems that the last person that changed contents in releases/12.0R/ became me as well (A diff from releases/12.0R/: https://www.diffchecker.com/iWeGDYXR).

Is this expected?

Thanks @koobs.

In D16041#339926, @ygy wrote:

It's updated (accidently) in r51920. Sorry!

Please update the Original revision:
<!-- Original revision: r51696 -->

In D15771#338288, @koobs wrote:

Please include root cause / explanations in the commit log message as well.

Am I understanding correctly and completely that:

1. MACHINE is equal to ARCH is only true on i386/amd64, AND
2. MACHINE is not equal to ARCH for other (!i386/amd64) architectures, AND

Thanks @koobs.

In D15850#335277, @feld wrote:

Let me fire off a build against CURRENT as a sanity check (this might also fix the breakage there)

Are you interested in becoming MAINTAINER as well?

Looks good, thanks.

In D15731#333307, @martymac wrote:

Hi,

Regarding the missing -f argument, this is something to patch in m4/lftp_lib_readline.m4. I'll report that upstream, thanks!

Meanwhile, to set a specific LOCALBASE only for readline, using a post-patch target seems a bit overkill to me.

Would adding the following line to the port :

CONFIGURE_ARGS+= --with-readline=${LOCALBASE}

work for you ? It should skip the wrong test and set the correct path. Can you try that ?

If it works, I can add that temporary workaround to the port while waiting for the configure script to be fixed upstream.

Look good, thanks. :)

This review is to merge the latest r51696 from English version:
https://svnweb.freebsd.org/doc/head/en_US.ISO8859-1/htdocs/index.xsl?r1=51605&r2=51696

I assume 49485:51605 will be merged later, right?

Thanks @koobs, @mat.

Bump up PORTREVISION

spaces => tab

update for OPTIONS_DEFINE variable order and remove -lunwind overlinking

In D15425#325629, @koobs wrote:

LGTM, if this change needs to be merged (so that the quarterly version can be built without root), please add relevant MFH: YYYYQX line to commit log message

Add upstream pull request link to files/patch-src_common.c

In D15425#325266, @koobs wrote:

Bump PORTREVISION does not sufficiently describe the summary of changes. Since this is related to Issue 228023, use it's summary as the shortlog

remote the regression-test targe