Re: Re: git: b430a140c818 - main - net-im/purple-gowhatsapp: add WhatsApp plugin for libpurple

On Sun, Jan 21, 2024 at 04:46:29AM -0500, Mikhail T. wrote:
> 21.01.24 02:04, Mathieu Arnold:
> > On Sat, Jan 20, 2024 at 07:30:55PM -0500, Mikhail T. wrote:
> > > But why would you want it backed out?!
> > Because you added a port that does not build.
> 
> That's not a reason to back it out even if it really failed due to some
> mistake on my part. I made such mistakes before.
> 
> But this time, as I pointed out already, the failure is due to some firewall
> rule on the build cluster.

Ah, no, it's not some firewall rule on the build cluster, it is how the
build has worked for the last 11 years when poudriere was created and
was made the official way we build everything.
The only time network access is authorized is during the fetch phase.
You would probably know that if you used poudriere to test your ports
like everybody else does.

> None of the thing Daniel pointed out caused the failure to build. It is a
> firewall rule. Setting DEVELOPER does not warn about it either.
> 
> > The FreeBSD ports tree is not your own ports tree where you add things to see if they build.
> It certainly is not "my own", but what's wrong with "adding things to see if
> they build"?

Well, it's not your personal sandbox, it is the official FreeBSD ports
repository.  You must test your changes before you submit them to make
sure they actually work. You obviously did not here.

> > You have to ...
> > You must use ...
> > You must also ...
> Would you mind pointing the rules and the bylaws, that you allege I
> violated? Are they spelled out somewhere? Nothing I'm able to find is worded
> anywhere as strongly as "must".

In any corporate world, when your manager tells you those things, you
say "yes" and you do it, or you go work somewhere else.
It is not the first time you have been told these things, and yet, you
still refuse to abide by the rules we all follow.
Those rules don't need to be written anywhere, especially when you've
been told those rules at one point or another along the years.

> > If you are not willing to do those basic things everyone else is doing, then please do not push changes yourself and only submit PR so that someone else can run those tests.
> 
> When a change has a potential to affect the work of others, I do that. See
> PR 276204 <https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276204> for the
> most recent example. But this was simply a new port. If it failed, no one is
> any worse off, than if it didn't exist at all. And yet, you're coming down
> on me like I ruined your day (or an even  longer time-period). Please, take
> care of your own demons privately. Thank you. Yours,

Well, your port is broken and cannot be built, yet, it is not marked
BROKEN, so it will waste time again and again on the package building
cluster.

-- 
Mathieu Arnold