Created attachment 234365 [details] proposed change --Fix build on -current after removal of iic_devclass global --Fix typos in help strings --Incorporate ports tree patches since 2019
Since I have a src commit bit, would you guys just prefer a phab review in the future? I'm not sure what the etiquette is for adding reviewers on the ports side of things.
Can you provide a MASTER_SITE mirror that supports http/https since ftp is being (indirectly) deprecated? Is WITHOUT_PIE still valid? Please concert options to the "new" format, see "5.13. Makefile Options" in Porters Handbook Best regards, Daniel
--I've been meaning to switch everything over to https anyway, so I'll just do that now. --I can confirm that WITHOUT_PIE is still required, linking fails if I remove it. I would assume that someone will do an exp-run to remove WITHOUT_PIE once the underlying staticlib issue is fixed. --I'm not sure what is meant by the "new" options format; the OPTIONS_DEFINE/<opt>_DESC syntax in the handbook looks identical to what I already have. Do you mean that I should use <opt>_MAKE_ARGS/<opt>_BUILD_DEPENDS instead of testing PORT_OPTIONS?
^Triage: Reporter is committer, assign accordingly. ^Triage: If there is a changelog or release notes URL available for this version, please add it to the URL field. ^Triage: Please set the maintainer-approval attachment flag (to +) on patches for ports you maintain to signify approval. -- Attachment -> Details -> maintainer-approval [+] Committers may obtain review via a Differential in Phabricator, adding the "Contributor Reviewers ($Repository)" group as a Reviewer, reaching out to other committers; directly or via mailing lists, or setting the attachment flag to: maintainer-approval ? <person-youd-like-to-review> Personally, approving directly in Bugzilla works fine :-) Thanks!
(In reply to Jason A. Harmening from comment #3) Yes, use OPT_* instead of using if statements Thanks for looking into my other questions Best regards, Daniel
Created attachment 234399 [details] proposed change
(In reply to Jason A. Harmening from comment #6) Hi Jason, The patch seems outdated. The $FreeBSD$ line the patch removes has already been removed in the repository. Would you mind adjusting the patch? Cheers.
Created attachment 234424 [details] proposed change
Looks good but doesn't compile as files/patch-client_Makefile doesn't apply
Created attachment 234428 [details] Patch for cx88 (alternative variant) Here's a variant I hacked together, feel free to adopt it if you want. - PORTVERSION --> DISTVERSION , while this isn't mandatory it's what Porters Handbook now suggests. [1] - Use http, while https works it's selfsigned which causes all sorts of warnings (you might want to have a look at something like letsencrypt) - Rephrase COMMENT to make it more compact, this is also more common with how the series are referred to - Rearrange Makefile to fit within 78 columns for readability, try to follow modern "coding" standards for ports tree - Combine all MAKE_ARGS definitions into one - Rename LINUX_COMPAT to LINUXAPI, improves overall style - Break out VERBOSE output as separate option [1] - https://docs.freebsd.org/en/books/porters-handbook/book/#porting-makefile and "Table 2. Package Naming Examples" In general I'd say that we try to avoid using underscores in variables mainly because of readability and we use it as a separator for various functions which is the main reason why I changed it. Making it short also improved (made) stile more consistent within the Makefile. I would also recommend that we try unless necessary to avoid redefining already defined variable descriptions within Ports framework for consistency which I suggested to break out VERBOSE. This is just a suggestion from my end, feel free to discard it if you don't like it.
Created attachment 234429 [details] proposed change Responding to last 2 comments: --Oops, the reason those patches don't apply is that they shouldn't be present in the update at all; I've incorporated them in the core build. I had to rewrite my patch generator to handle the ports tree switch from svn to git, and I missed a couple of things. --Regarding the proposed changes to the Makefile, I'm mostly ok with them, but with a couple of tweaks: 1) I'd rather keep WITHOUT_PIE on its own line, as that should make it easier to delete once the underlying issue is fixed. 2) I slightly prefer LINUXCOMPAT (without the _) to LINUXAPI. I've included those changes in this update.
Created attachment 234450 [details] proposed change Decided to go ahead and adopt non-self-signed certificate; updating both MASTER_SITES and wiki link to use https.
LGTM, you can drop "+" when using OPT_MARK_ARGS+= OPT_BUILD_DEPENDS+= and so on Compiles on 13.1-STABLE (amd64) Please also consider moving libtuner to https Best regards, Daniel
..and consider keep the same order for menu options (alphabetical order) through the Makefile
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=88aef5f532d02d50b3bffd191092e8887629361f commit 88aef5f532d02d50b3bffd191092e8887629361f Author: Jason A. Harmening <jah@FreeBSD.org> AuthorDate: 2022-06-10 03:26:32 +0000 Commit: Jason A. Harmening <jah@FreeBSD.org> CommitDate: 2022-06-10 03:42:55 +0000 multimedia/cx88: update to 1.5.4 --Fix build on -current after removal of iic_devclass --Fix typos in help strings --Incorporate ports tree patches since 2019 --Modernize port makefile and switch MASTER_SITES to https PR: 264381 Approved by: diizzy multimedia/cx88/Makefile | 43 +++++++++++----------- multimedia/cx88/distinfo | 6 +-- multimedia/cx88/files/cx88d.in (mode +x) | 0 multimedia/cx88/files/patch-client_Makefile (gone) | 20 ---------- .../patch-client_dvb_cx88__dvb__buffer.cpp (gone) | 11 ------ .../patch-client_dvb_cx88__dvb__capture.h (gone) | 11 ------ .../patch-client_dvb_cx88__dvb__demux.cpp (gone) | 11 ------ ...ch-client_dvb_cx88__dvb__demux__feed.cpp (gone) | 11 ------ ...patch-client_dvb_cx88__dvb__frontend.cpp (gone) | 11 ------ .../patch-client_v4l_cx88__radio__capture.h (gone) | 11 ------ multimedia/cx88/pkg-descr | 2 +- 11 files changed, 25 insertions(+), 112 deletions(-)