[pkgsrc/trunk]: pkgsrc/lang/tcl-expect lang/tcl-expect: Add best-guess patch ...

[gdt <gdt%pkgsrc.org@localhost>]

details:   https://anonhg.NetBSD.org/pkgsrc/rev/633b9b62bf68
branches:  trunk
changeset: 380758:633b9b62bf68
user:      gdt <gdt%pkgsrc.org@localhost>
date:      Tue Jun 14 10:56:15 2022 +0000

description:
lang/tcl-expect: Add best-guess patch comments

The patches had a comment from 2012 that I cannot understand at all.
This replaces them with my best guess at the point of the patches.

Add users to Makefile.common, apparently only this package.

diffstat:

 lang/tcl-expect/Makefile.common           |   5 ++++-
 lang/tcl-expect/distinfo                  |   6 +++---
 lang/tcl-expect/patches/patch-Makefile.in |  18 ++++++++++++++++--
 lang/tcl-expect/patches/patch-configure   |  17 +++++++++++++++--
 4 files changed, 38 insertions(+), 8 deletions(-)

diffs (78 lines):

diff -r 54d4989761fb -r 633b9b62bf68 lang/tcl-expect/Makefile.common
--- a/lang/tcl-expect/Makefile.common   Tue Jun 14 09:18:51 2022 +0000
+++ b/lang/tcl-expect/Makefile.common   Tue Jun 14 10:56:15 2022 +0000
@@ -1,4 +1,7 @@
-# $NetBSD: Makefile.common,v 1.13 2022/05/03 23:38:38 gdt Exp $
+# $NetBSD: Makefile.common,v 1.14 2022/06/14 10:56:15 gdt Exp $
+
+# used by lang/tcl-expect/Makefile
+# \todo Explain why there is only one user, or fold into Makefile.
 
 DISTNAME=      expect${EXPECT_FNAME}
 MASTER_SITES=  ${MASTER_SITE_SOURCEFORGE:=expect/}
diff -r 54d4989761fb -r 633b9b62bf68 lang/tcl-expect/distinfo
--- a/lang/tcl-expect/distinfo  Tue Jun 14 09:18:51 2022 +0000
+++ b/lang/tcl-expect/distinfo  Tue Jun 14 10:56:15 2022 +0000
@@ -1,8 +1,8 @@
-$NetBSD: distinfo,v 1.23 2022/05/15 01:56:24 dholland Exp $
+$NetBSD: distinfo,v 1.24 2022/06/14 10:56:15 gdt Exp $
 
 BLAKE2s (expect5.45.tar.gz) = 38c440ed0868322b6ec7438decaa036a636e7d700191b754c1e4a7dc0fdd0c46
 SHA512 (expect5.45.tar.gz) = be991c68241e607b3a689eae7e7966056dbfb577e857331d54a4911bd178c1816425217603b43918ad1b6d2e966271a0f01e79d7028a22e223562d59d10c8c51
 Size (expect5.45.tar.gz) = 628808 bytes
-SHA1 (patch-Makefile.in) = 24efcbdb0a21985cb8a0fb7812c5f45f67de7c30
-SHA1 (patch-configure) = a16c833e61d23f63abff1413522b99cd2251bc34
+SHA1 (patch-Makefile.in) = 3b234d232a47adc64b3a4d275237b64bbbf7fdf6
+SHA1 (patch-configure) = 843b330b46c7dd17deab1b2c133e5ae52d609ae9
 SHA1 (patch-exp__chan.c) = 5af6bf31fa87a4d9951be6bc978ede5440380981
diff -r 54d4989761fb -r 633b9b62bf68 lang/tcl-expect/patches/patch-Makefile.in
--- a/lang/tcl-expect/patches/patch-Makefile.in Tue Jun 14 09:18:51 2022 +0000
+++ b/lang/tcl-expect/patches/patch-Makefile.in Tue Jun 14 10:56:15 2022 +0000
@@ -1,6 +1,20 @@
-$NetBSD: patch-Makefile.in,v 1.4 2018/08/27 07:12:08 wiz Exp $
+$NetBSD: patch-Makefile.in,v 1.5 2022/06/14 10:56:15 gdt Exp $
+
+(This patch was created in 2012 and not adequately commented.  The
+comment included "no libtool", but there is no hint of libtool
+upstream and this patch does not remove it.)
+
+Define SHLIB_VERSION, mysteriously.
 
-New Makefile.in patch, no libtool
+Add a variable for a static library version, build and install it.
+
+Use standard rather than namespaced directories.
+
+Include LDFLAGS when linking (for RELRO).
+
+Version the shlib and the expect binary.
+
+\todo Push changes upstream, or explain why that's not appropriate.
 
 --- Makefile.in.orig   2010-09-30 17:14:07.000000000 +0000
 +++ Makefile.in
diff -r 54d4989761fb -r 633b9b62bf68 lang/tcl-expect/patches/patch-configure
--- a/lang/tcl-expect/patches/patch-configure   Tue Jun 14 09:18:51 2022 +0000
+++ b/lang/tcl-expect/patches/patch-configure   Tue Jun 14 10:56:15 2022 +0000
@@ -1,6 +1,19 @@
-$NetBSD: patch-configure,v 1.1 2012/08/21 21:33:58 marino Exp $
+$NetBSD: patch-configure,v 1.2 2022/06/14 10:56:15 gdt Exp $
+
+(This patch was created in 2012 and not adequately commented.  The
+comment included "no libtool", but there is no hint of libtool
+upstream and this patch does not remove it.)
+
+Add support for MirBSD as OpenBSD and DragonFly as FreeBSD.
 
-New configure patch, no libtool
+Upstream has support to use the version with or without dots as the
+shlib suffix.  This patch forces that to be the "NODOTS" flavor
+always, perhaps for PLIST consistency.
+
+\todo Expain why PVNODOTS isn't the PACKAGE_VERSION_NO_DOTS that
+upstream configure already computes.
+
+\todo Push changes upstream, or explain why that's not appropriate.
 
 --- configure.orig     2010-09-16 20:46:47.000000000 +0000
 +++ configure