Update to 8.0.0.18

• Update WWW: to Makefile

• Update WWW: to Makefile

• Change versioning to handbook: Example 19. Using USE_GITHUB to Access a Commit Between Two Versions
• Update WWW: to Makefile

• reorder

rebasing this on current main updating devel/py-pyyaml PORTNAME

Move to a GH_TAGNAME from the released 8.0.0 to include some IMAP fixes
All tests pass.

Rerun script with manual update from comment 702768

Add comment about copying tests waiting for release

• Use USES=pytest

portfmt sort order

portfmt sort order

• Update USES=pytest

Yes, that works well. Thanks for the recommendation.

• update for USES=pytest

er sorry I misread that. Ok added comment.

• Add comment stating why USE_GITHUB vs sdist

They do disagree. 15.8 porters handbook recommends keeping USE_* vars together (using USE_GITHUB GH_* as example).

Upstream updated https://github.com/sethmlarson/rfc6555/pull/8 and prefer not to add test to sdist.

Update diff 14.19.3

Tested on py37:

• Added patch to fix idle folder with spaces (forget this from the init. review).
• Capped offlineimap3 to 3.7-3.9 since imaplib2 is limited to 3.9. In turn only testing offlineimap3 for py37/8/9
• Patched docs sphinx-build command to use the correct conncurrent version. Will need to fix upstream.
• Patched out tests that require a network connection and user auth. Maybe have upstream move these out of unittest?

Run unittests with python37, 38. 39. 310, and 311

Sorry in the wrong branch. Updating from correct one.

• Add comments for build depends and pypi packaging
• include CONFLICTS_INSTALL just in case someone is still using the removed py-offlineimap package

Add tests include file in files/ tests

Added tests and moved to github until tests are added

tests are not included in sdist. Requested here: https://github.com/sethmlarson/rfc6555/pull/8

You're correct the tests dir is not included in the sdist. However, those test are just to import the module. Maybe to test future absolute_import? I don't know how much value this would add. If you still think this is a good idea I can ask to include tests.

Remove :27 from USES python:27
Remove Google Analytics script

Can you upload the .vbox file?

reorder devel/Makefile

Add extra details in MOVED

2.x has been replaced by 3.x

Looks like some ports were removed between then (2/2020) and today (7/2021). Forget the exact method to update the tree at the time however this was used.

New version on main as of 7/18/21

In D30868#702305, @jrm wrote:

I'm not seeing those. http://pkg.ftfl.ca/data/13amd64-default/2021-07-16_09h32m04s/logs/bear-3.0.13.log

So I think we are ready to commit. Any objections? I did a quick runtime test of bear -- gmake on a project and all seems fine.

Added because of the following warning if they're not there. Might be related to you current issues with qa.sh?

Building with 1.38.1_1,2 and without the openssl build dep, it works.

Oh I see now. https://github.com/Kitware/CMake/blob/master/Modules/FindOpenSSL.cmake#L134

Looking at devel/grpc Makefile if you're using base SSL it removed the check for openssl from grpc's cmake CMakeLists.txt file. Should it also remove the openssl require from the .pc files too?

I also have tests working with a non-ports build however it depends on https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=257123 being fixed.

I see the staging from cmake it using ENV's DESTDIR (/wrkdirs/usr/ports/devel/bear/work/stage/) and prepending that to ${CMAKE_INSTALL_PREFIX} (wrkdirs/usr/ports/devel/bear/work/.build/stage)

The current error is grpc++ pkgconfig requiring grpc requiring openssl. Since you're not installing ports openssl and since base doesn't provide an openssl pkgconfig (/usr/libdata/pkgconfig/) openssl can't be found by pkgconfig.

https://svnweb.freebsd.org/ports?view=revision&revision=564764

In D30030#680673, @thierry wrote:

Thanks! I copied this sentence from the tt-rss forum, and with the used font did not notice the bad quote ;-)

With the right one, it goes further in single user mode, but the migration fails:

[07:50:35/51886] Starting migration to 141...
[07:50:35/51886] Failed on line: create table ttrss_user_prefs2 ( owner_uid integer not null, pref_name varchar(250), profile integer null, value longtext not null, foreign key (profile) references ttrss_settings_profiles(id) ON DELETE CASCADE, foreign key (owner_uid) references ttrss_users(id) ON DELETE CASCADE) ENGINE=InnoDB DEFAULT CHARSET=UTF8
[07:50:35/51886] Migration failed: SQLSTATE[42000]: Syntax error or access violation: 1142 REFERENCES command denied to user 'xxxx'@'localhost' for table 'ttrss_settings_profiles'

So I had to run the migration scripts manually, connected as root: this is not very user friendly…

Furthermore, the installed scripts 143.sql, 144.sql and 145.sql do not increment the schema version, and then when you connect to tt-rss it wants to re-run the migration once more!
I had to run manually:
update ttrss_version set schema_version = 145;

Are these problems fixed in a newer version? Or is it documented somewhere?

In D30030#680511, @thierry wrote:

After upgrading, I updated my config file, but I cannot run it in single user mode to update the schema.

I tried different things around
putenv(‘TTRSS_SINGLE_USER_MODE=true’);
but always without success!

Could you please put the correct one in config.php.sample?

• Add commented SINGLE_SUER_MODE example in config.php.in

With that if everything looks good I'll upload a new patch to the PR.

I've always opened them in pairs if I had a review. Open a PR in Bugzilla so it's publicly known and review so non-Phab members can follow along.

Any additional comments?

• fix typo and inline URL in UPDATING

Updating tomorrow to handle the epoch bump.

This pre maturely committed.

Had an extra non-tt-rss commit mixed up with this. Removed.

Change to 1.4.15

In D28871#649595, @decke wrote:

I think we never had a legacy port for phpvirtualbox because there are not that many users and upstream is almost dead. It should be enough for now to change the dependencies to vbox-legacy. We can update it when the dust has settled.

In D28871#646753, @madpilot wrote:

I tested now virtualbox 6 modified to force clang 11 from ports on 12.2 and it works fine. I think this fix is cleaner than forcing DEBUG.

I'm almost sure this would work for 11.x too, but it's difficult to test because I'd have to build a full set of ports using SSL from ports, due to the QT/OpenSSL issue. Maybe I'll test that too, anyway I'll remove the restriction on 11 along with the one on 12.

In D28871#646424, @madpilot wrote:

In D28871#646261, @madpilot wrote:

In D28871#646258, @dereks_lifeofadishwasher.com wrote:

I was able to build, install, run, create/run VMs with this patch on 12.2-RELEASE-p3 host. Should the ignore 12 lines be removed?

If it works yes. Tomorrow I'll test myself, but you report looks promising.

I was able to build, install, run, create/run VMs with this patch on 12.2-RELEASE-p3 host. Should the ignore 12 lines be removed?

Remove copy/paste from testing

Updated patch for py-sphinxcontrib-autoprogram using USE_GITHUB

• devel/py-python-gitlab: Update to 2.6.0
• Switch to USE_GITHUB until sdist has doc/

Looks like this could use DOCS and do-test.

When setting the docs' html convert I chose lynx since it had the most overlapping depends with default mutt and built times. At the time that was www/lynx. Also assuming you're using poudriere to build so existing system checks aren't relevant there.

Do you have any way to verify mu4e? I do not use it.

I've been using 1.4.13 for a couple days now without issues.

Turn this into an optional patch with an addition to pkg-message

In D26061#579331, @uqs wrote:

Looks good, do you want me to go ahead and commit this on your behalf?