[Bug 192878] comms/pr [maintainer] new source location

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=192878

--- Comment #10 from John Marino <marino at FreeBSD.org> ---
(In reply to C Hutchinson from comment #9)
> Created attachment 146131 [details]
> UPDATES [maintainer] STAGED LICENSE SVN DIFF
> 
> OK. This is as good as it gets.

Then we have a problem.

> All incantations of your proposed install loop(s) failed.

Then it wasn't implemented correctly.  I do this all the time.  And I convert
loops to single commands all the time.  It works.

> I'm not sure if you had a chance to examine the source itself.
> But it's a pure copy operation. No bin(aries), or lib's
> involved.
> Anyhow. I've made a couple of the other corrections regarding
> your "nit's" with this, and added a LICENSE, as well. I can
> imagine no reason for not letting this through. It passes with
> flying colors.
> 
> See attached svn diff, for details.
> 
> BTW ${CP} is completely valid, and provided in bsd.port.mk.
> Which I think is why no ALARM's were set off. :)

No, ${CP} is not valid.  You are supposed to use INSTALL_PROGRAM, INSTALL_MAN,
INSTALL_SCRIPT, INSTALL_DATA macros, never CP.  Those macros do more than copy,
they control permissions, ownership, and stripping.



> 
> I'm going to make a "proper" dev box, running on 11, now.
> So kindly mark this fixed. :)


I can't.
While the loops could be let go because they do the same thing more
inefficiently, the CP issue is a dealbreaker.

Attempting to create a DOCDIR directory each time a file is installed is a
dealbreaker.  Not using PORT_OPTIONS:MDOCS or equivalent is a dealbreaker.

so I can accept this is the best it's going to get.  I could fix all of that in
< 5 minutes.


> 
> The following, are for your "magnifying glass". :)

those problems aren't going to show in redports

-- 
You are receiving this mail because:
You are the assignee for the bug.