In D24571#546716, @mandree wrote:

bapt, I am willing to change that, but I am not willing to put up with obstructivism and take micro change steps with re-uploading everything for new review for each and every changed two bytes for a one-line change. This is out of balance, and the discussion is, too.

Do do I get a conditional approval to commit with the test -> [ ] change, or are we going to waste more time on a trivial fix?

In D24571#546282, @mandree wrote:

same code, rediffed with "full" context - not really clear why it was needed for a one-line patch...

Could you use devel/arcanist, or at least generate a diff with full context like it does, with svn diff -x -U9999 or git diff -U9999.

Could you use devel/arcanist, or at least generate a diff with full context like it does, with svn diff -x -U9999 or git diff -U9999.

What difference is there from having either a Makefile.local with whatever you need in the port's directory, or using what everyone has been using for decades:

In any way, whatever the fix ends up to be, this one is ok for me :)

What happens when not actually building packages but simply installing a port, that is, when you don't have a PKGREPOSITORY?

How about changing the example to <le>1.9</le>, then the text is right.

In D24630#542149, @bapt wrote:

In D24630#542147, @mat wrote:

What happens if file is not a regular file, like a symlink?

The function os.remove is equivalent of remove(3) and behave the same, so it removes the symlink: https://github.com/freebsd/pkg/blob/master/libpkg/lua_scripts.c#L166

Basically the behaviour is the same as [ -s ... ] && rm -f

What happens if file is not a regular file, like a symlink?

isn't there a stat thing that you can use without having to open the file?

Then please bump the required pkg version in here when you commit this ;-)

I was wondering about ports without a distinfo file (like, say, ports-mgmt/portlint) but they could gain a distinfo file with only a TIMESTAMP in it.

I remember a review by @emaste (I think) about this, he had a default value for the TIMESTAMP.

Don't we also need to bump PKG_DEPENDS to make sure we require a pkg with lua support?

In D24231#538638, @yuri wrote:

Why do you want to save the output into a file?

Also, can you extract the awk script to is own file?

Could you use devel/arcanist, or at least generate a diff with full context like it does, with svn diff -x -U9999 or git diff -U9999.

In this case, the documentation is wrong, I will fix the documentation.

Please make the flavors not conflict.

Could you change all envcore to simply core, and make it simply say "used when building lua itself".