Re: nvidia-driver and no update in /usr/ports/UPDATING

On Fri, Apr 08, 2022 at 03:34:57AM +0100, Steve O'Hara-Smith wrote:
>On Fri, 8 Apr 2022 00:00:34 +0100
>tech-lists <tech-lists@zyxst.net> wrote:
>
>> So *WHY* were the consequences of PR 261666 not in /usr/ports/UPDATING?
>> nvidia-driver was updated to (560) i think. 5-something.. the instance
>> before was moved to nvidia-driver-470 as it's now legacy. That's great
>> and all, but *WHY* the *F**K* isn't in *UPDATING??!??!?!?!*
>
>	You know if this was a commercial product for which you were paying
>for support and you spoke like that to a support worker you would get hung
>up on and blacklisted for verbal abuse. It is no more acceptable in text,
>especially since you are talking to users - I very much doubt that anyone
>involved will see your rant.

I disagree with you here, having supported commercial products
for over a decade. I thought my missive was mild; though that might be
cultural.

This isn't truly a rant, and I'm astonished you're discounting 
it as such. I'm frustrated because it seems some ports maintainers
will not follow freebsd policies. That costs time and money,
makes people like me less likely to trust freebsd; it's all so 
needless.

Yes, I realise freebsd is free. It markets itself as
"the power to serve". Well, it can't serve very well if port
maintainers fail to follow policy. So, maybe a better and 
more honest slogan would be "the power to serve, but if you 
use ports then, well, you're on your own". At least we'd know
where we are. As I understand it, there's no need to modify UPDATING
if the port has simply moved. But if the executable name changes 
or if the version changes majorly and *especially* if it's a kernel
module then yeah I expect it to be in UPDATING.

>	Now to answer your question - there's nothing in UPDATING because
>the port maintainer for that port didn't put anything in there when they
>made the change. 

I realised this before my post and it's the reason for it.

> So if you want to know why you could look up the
>maintainer (listed in the Makefile) and ask them *politely* about it -
>maybe they had to rush a child to hospital or attend some social event and
>forgot to do it. Whatever the reason they deserve politeness and gratitude
>for what they have done not abuse for what they haven't.

Here's a way to avoid the whole issue - update *UPDATING* before making 
the change, and make that *mandatory* in order to update a port.

Having a polite communication with the maintainer is pointless in this 
circumstance. See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=249252
which was ignored, twice. The problem is one of policy.

>	Or you could hurl anonymous abuse at the uninvolved with thinly
>disguised obscenities in a public forum to no effect whatsoever except to
>make yourself look foolish.

Foolish for articulating my frustration? You think my post is without
merit? It's the situation I'm frustrated with. There's a policy in
place for changes like this and as far as I can tell it's not being
followed in some cases. This makes like hell for me for any freebsd 
system I support.

I guess you think it's better i ditch freebsd and use something else?
Maybe I should just bite my lip, shut up and go away?
-- 
J.