https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=283745

Jason E. Hale <jh...@freebsd.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |Works As Intended
             Status|Open                        |Closed
             Blocks|                            |275792

--- Comment #9 from Jason E. Hale <jh...@freebsd.org> ---
(In reply to Daniel Engberg from comment #8)
> I was under the impression it was run during build irregardless not during 
> the configure stage irregardless which breaks in our environment irregardless 
> and that is was passed to external tools if CMake wasn't the invoked 
> interpreter.

This is rather off-topic, but please drop the term "irregardless" from your
vocabulary, Daniel. Whether you're a native English speaker or not, as even
native speakers often get this wrong, "irregardless" is a non-standard word for
"regardless" and its use should be avoided in formal writing. I'm not trying to
be petty, but your use of it three times in one sentence honestly made my head
spin, distracted me from what you were trying to communicate, and prompted me
to grep the ports tree for instances of "irregardless". I found just one
instance in Mk/Uses/ninja.mk added by you in 9843444ea96e, corrected by me in
d790b4b4d0d7.

Regardless (ha!), I'm still not exactly sure what you're trying to say. CMake
is run in the do-configure target when USES=cmake is set (Mk/Uses/cmake.mk).
Setting USES=cmake:noninja basically directs CMake to generate files that
make(1) understands rather than those by ninja or samurai.

The current net-mgmt/netdata port uses autotools, so I suppose OP is trying to
update the port. Upstream has dropped autotools completely in favor of CMake in
1.45 [1]. In the end, I don't think CMake itself is the problem, but rather
environment variables being cleared by the framework. For the third time, this
shouldn't matter since ports should only download files in the fetch stage.

As I'm writing this, though, son of a...diizzy!!! (I'm reserving my 'Damn,
Daniel' card for later) You, especially, and OP could have mentioned bug
#275792. Way to bury the lead! You've been assigned to that bug for over a year
and didn't think to bring it up while I've been speculating this whole time.
This PR gave a few more details, but [2] pretty much said it all.

This is not a problem with CMake, but rather with the net-mgmt/netdata port
proposed update. I've already offered a workaround in comment #7 for the
meantime. Disabling the DASHBOARD option should also work in the proposed port
update as an even worse workaround. I don't use this software, so not sure what
functionality would be sacrificed. Regardless (ha! ha!), I'm closing this. I'll
still at least link this PR to the original issue for archival purposes.

[1] https://github.com/netdata/netdata/releases/tag/v1.45.0
[2] https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=275792#c53


Referenced Bugs:

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=275792
[Bug 275792] net-mgmt/netdata: Update to v2.1.0
-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to