OK, got it. Don't use optsuffix.
I didn't know it was temporary.

I re-tested it and the port looks good now.
Would you like me to commit it?

OK, this works now and looks pretty good.
Can you consider making some configuration files @sample?
https://www.freebsd.org/doc/en/books/porters-handbook/book.html#plist-config

This doesn't build for me:

/work/stage/usr/local/share/arangodb/js/server/bootstrap/coordinator-database.js
CMake Error at cmake_install.cmake:87 (file):
  file INSTALL cannot find
  "/usr/home/lifanov/src/svn/freebsd/ports/head/databases/arangodb32/work/.build/var/log/arangodb3".
gmake[1]: *** [Makefile:99: install/strip] Error 1

Looks good to me!
Thanks for taking care of it.

In D12793#269311, @dev_dudu.ro wrote:

In D12793#268412, @lifanov wrote:

This looks pretty good, but %%DATADIR%%3 and %%DOCSDIR%%3 in pkg-plist should really be directories.
Can you consider configuring these to match the port name or maybe swithing the portname to arangodb3?

I'm not sure I understand what you mean - should I replace them with the actual final value - e.g. /usr/local/share, or add @dir entries for them to pkg-plist?

I was able to test this and it works.

The fetch/package stages work. I'll check if gameplay works when I get to my desktop.
It would be neat to be able to innoextract existing gog.com distfile for Diamond in nwndata.

I own/play nwn occasionally. I'm going to review/test this update.

This looks pretty good, but %%DATADIR%%3 and %%DOCSDIR%%3 in pkg-plist should really be directories.
Can you consider configuring these to match the port name or maybe swithing the portname to arangodb3?

I can see in the build system that some components are hard to decouple.
Try decoupling what you can, like boost or libcurl.

Could you also try running "make stage-qa" and "make check-orphans" please and make sure these are clean?

How did you generate pkg-plist? The files listed in it are already relative to LOCALBASE and @dir should only specify directories that are empty.
The "make makeplist" target should generate a pkg-plist close to what you want, so then you need to review it and make any manual changes you need.

Hi and welcome to FreeBSD!

Ughh, sorry. I didn't think of that.
What do you think of adding a file in etc/newsyslog.conf.d that does something like
<include> %%LOCALBASE%%/etc/zrepl-newsyslog.conf
and install a sample for that?

pkg-descr needs to start with a capital letter.
Also, it uses "written in Go" twice in the same sentence.

I'll take one last look at it and commit it in a few minutes. I hereby implicitly approve you (woodsb02) to make any sort of change to the port in the future including grabbing maintainership.

Allan helped me find the mistake (I had primarycache=metadata set on this dataset).
With this addressed, the performance gain from switching to 1m buffer is minimal.

Woo, thank you mat!

The above check would have caught /usr/local/obscuredirectory/foo without a symlink to it.

$(find ${STAGEDIR}${PREFIX} \
    -type f -perm +111 2>/dev/null)

Can I commit this fix as-is for now so that I can commit a fix for sysutils/ansible please?

I'm not sure why we check symlinks, but exclusion of executable files pointed to by symlinks doesn't work without this check.

address antoine's feedback

address mat's feedback: move PLIST_FILES after USES/USE block

Before:

syscall                     seconds   calls  errors
cap_enter               0.000007683       1       0
cap_ioctls_limit        0.000013340       2       0
__sysctl                0.000033176       4       0
issetugid               0.000012642       2       0
write                   0.000014249       1       0
sigprocmask             0.000104417      16       0
readlink                0.000010616       1       1
read                  328.301444459   25719       0
openat                  0.000086678       6       0
open                    0.000028566       2       0
munmap                  0.000052800       5       0
mprotect                0.000022979       3       0
mmap                    0.000296068      31       0
lseek                   0.000007055       1       0
getdirentries           0.000055177       2       0
fstatfs                 0.000009918       1       0
fstatat                 0.000123692      11       0
fstat                   0.000049381       6       0
close                   0.000054059       7       0
cap_rights_limit        0.000020324       3       0
cap_getmode             0.000006426       1       0
cap_fcntls_limit        0.000012362       2       0
access                  0.000021023       2       0
                      ------------- ------- -------
                      328.302487090   25829       1

At this point, the patch works and I can commit this port and security/py-backports.ssl_match_hostname if they are ready.

Looks good! Make sure you also get maintainer's approval.

address swills' feedback:

o unhardcode some values having to do with DISTFILES
o be more precise in the LICENSE block about which license applies to what

Cool beans! Once the new version is functional enough to replace existing 1.7.0, I'll open a PR and attempt to land this.

whoops, lost a path component

address matthew's feedback: good point about just exec'ing nx

really fix for and test with Python 3

fix stage and plist issues using a post-stage and post-plist target

Oooh, there is a Gentoo mirror for this package. Hold on while I shamelessly rip off the OpenBSD port...

Approved!

Approved! Make sure you either try this port with python 3 or wait for that github issue or do USES=python:2 instead.

Looks good!

I personally prefer to see the entire change, including the category file update, but I don't feel strongly about it.

I'm good with this revision. Make sure you commit the dependencies in order or at the same time so that the tree is consistent.

address dteske's feedback: un-hardcode pid file

address mat's feedback: un-bump epoch until we get a real version
address dteske's feedback: make for list sortable by including a semicolon

address mat's feedback: clean up USE_GITHUB with default project

address dteske's feedback