Created attachment 225086 [details] knot 3.0.6 svndiff Knot3 was upgraded to 3.0.6 recently. This patch updates the port accordingly. Tested using poudriere and portfmt.
Hello, Before I can take this PR could you take a look at: 1. Could you provide a git diff instead of svn diff? https://docs.freebsd.org/en/books/porters-handbook/upgrading/#git-diff 2. I've found several commented port options in Makefile: --- #BUILD_DEPENDS= autoconf>=2.65:devel/autoconf \ # gsed:textproc/gsed and #PLIST_FILES= etc/rc.d/${PORTNAME} Any particular reason for this comments? 3. `portclippy Makefile`: --- # PORTNAME block PORTNAME DISTVERSION CATEGORIES MASTER_SITES PKGNAMESUFFIX # Maintainer block MAINTAINER COMMENT # License block LICENSE LICENSE_FILE # Dependencies BUILD_DEPENDS LIB_DEPENDS # USES block USES +USE_LDCONFIG +USE_RC_SUBR -CONFLICTS # Configure block HAS_CONFIGURE -BINARY_ALIAS CONFIGURE_ARGS -USE_RC_SUBR # Make block +INSTALL_TARGET # Conflicts +CONFLICTS # Standard bsd.port.mk variables +BINARY_ALIAS SUB_FILES +SUB_LIST -USE_LDCONFIG -INSTALL_TARGET # Users and groups block USERS GROUPS -SUB_LIST # Packaging list block PORTDOCS # Options definitions OPTIONS_DEFINE # Options descriptions DNSTAP_DESC +FASTPARSER_DESC +MAXMINDDB_DESC # Options helpers +DNSTAP_LIB_DEPENDS DNSTAP_CONFIGURE_ENABLE DNSTAP_CONFIGURE_WITH -DNSTAP_LIB_DEPENDS DNSTAP_CPPFLAGS DOCS_CONFIGURE_OFF -FASTPARSER_DESC FASTPARSER_CONFIGURE_ENABLE -IDN_CONFIGURE_WITH IDN_CONFIGURE_OFF -MAXMINDDB_DESC +IDN_CONFIGURE_WITH MAXMINDDB_LIB_DEPENDS MAXMINDDB_CONFIGURE_ENABLE --- You use portclippy to organize Makefile blocks
Created attachment 225123 [details] knot 3.0.6 diff > git diff instead of svn diff > I seriously dislike maintaining the ridiculous amounts of dependencies for simply communicating a few lines of mutations. So by these neither, but a normal diff. > Any particular reason for this comments? > autoconf is listed as required at the instructions at knot-dns.cz, but I think it was core maintainer who removed it. The comment is a reminder of this conflicting information. For IDN_LIB_DEPENDS counts the exact same. But PLIST_FILES is likely safe to delete. Then portlint / portfmt / portclippy have not been very useful to me in the past, mainly ruining cosmetics to provide a clear overview / disclosing direct correlation with one thing and another which now need to be listed devided far away. But I admit portclippy does provide useful pointers (where portlint left you clueless). Still, clearly it's a dangerous one to "fly by the tools": It isn't always keeping in account things are within an "if" block. Also I'm not really comfortable using variables before having them defined. Is it safe to revers the order of these lines? USERS= knot GROUPS= knot SUB_LIST+= USERS="${USERS}" GROUPS="${GROUPS}" But for the rest I followed it's instructions as much as possible, and so; poudriere is happy portlint is happy portfmt is happy portclippy is almost entirely happy, and so am I.
git diff instead of svn diff I seriously dislike maintaining the ridiculous amounts of dependencies for simply communicating a few lines of mutations. So by these neither, but a normal diff. How to apply this diff? I've tried `git apply` and patch < knot-3.0.6.diff`: --- Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |--- dns/knot3.orig/Makefile 2021-05-20 13:36:55.791207000 +0000 |+++ dns/knot3/Makefile 2021-05-20 13:41:13.335190000 +0000 -------------------------- Patching file Makefile using Plan A... Hunk #1 failed at 1. Hunk #2 failed at 14. Hunk #3 failed at 45. Hunk #4 failed at 68. Hunk #5 failed at 85. 5 out of 5 hunks failed--saving rejects to Makefile.rej Hmm... The next patch looks like a unified diff to me... The text leading up to this was: -------------------------- |diff -ruN dns/knot3.orig/distinfo dns/knot3/distinfo |--- dns/knot3.orig/distinfo 2021-05-20 13:36:55.791321000 +0000 |+++ dns/knot3/distinfo 2021-05-20 13:18:01.279381000 +0000 -------------------------- Patching file dns/knot3/distinfo using Plan A... Hunk #1 failed at 1. 1 out of 1 hunks failed--saving rejects to dns/knot3/distinfo.rej Hmm... The next patch looks like a unified diff to me... The text leading up to this was: -------------------------- |diff -ruN dns/knot3.orig/pkg-plist dns/knot3/pkg-plist |--- dns/knot3.orig/pkg-plist 2021-05-20 13:36:55.791539000 +0000 |+++ dns/knot3/pkg-plist 2021-05-20 13:18:53.052281000 +0000 -------------------------- Patching file dns/knot3/pkg-plist using Plan A... Hunk #1 failed at 41. 1 out of 1 hunks failed--saving rejects to dns/knot3/pkg-plist.rej done --- Any clues? But I admit portclippy does provide useful pointers (where portlint left you clueless). Still, clearly it's a dangerous one to "fly by the tools": It isn't always keeping in account things are within an "if" block. Also I'm not really comfortable using variables before having them defined. Is it safe to revers the order of these lines? USERS= knot GROUPS= knot SUB_LIST+= USERS="${USERS}" GROUPS="${GROUPS}" Yes, you are right, I think its better maintain this order.
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=4ce74222e57947cb9134f897c971f4a7075bca65 commit 4ce74222e57947cb9134f897c971f4a7075bca65 Author: Leo Vandewoestijne <freebsd@dns.company> AuthorDate: 2021-05-22 01:08:51 +0000 Commit: Neel Chauhan <nc@FreeBSD.org> CommitDate: 2021-05-22 01:08:51 +0000 dns/knot3: Update to 3.0.6 Changes: https://www.knot-dns.cz/2021-05-12-version-306.html PR: 256001 dns/knot3/Makefile | 2 +- dns/knot3/distinfo | 6 +++--- dns/knot3/pkg-plist | 2 ++ 3 files changed, 6 insertions(+), 4 deletions(-)
I have committed the earlier diff rebased on Git.