Re: git: eab8b0d82c62 - main - */R-cran-*: Fixup WWW entries

> On Oct 20, 2026, at 11:11 AM, Zsolt Udvari <uzsolt@freebsd.org> wrote:
> 
> On Tue, Oct 20, 2026 at 06:37:03PM +0100, Nuno Teixeira wrote:
>> Is it an improvement having a port without WWW for human reading?
> 
> I understand your point - if you copy/select the value of WWW in Makefile and
> paste/open it into/to browser. I can't imagine other issue.
> 
> I use Makefile's WWW directly only sometimes
> becuase I've a small "server" with poudriere and I don't open WWW on it.
> If I need WWW I use "pkg query '%w'" on my laptop and open this url in
> browser (using urlview and have a tmux-hotkey to open it).
> So it isn't critical for me.
> 
> I think it's an improvement if the DRY is improvement.
> 
> Cheers,
>  Zsolt

My argument against having WWW in Makefiles was that it is that if the URL
changes (which we ourselves have no control over when or how an upstream
repository works), that we would eventually have to do things like this:

https://reviews.freebsd.org/D59714

And

When someone doesn’t get the memo, such-as-in:

https://reviews.freebsd.org/D60718

Leads to having to perform repeat drive-by, such as:

https://reviews.freebsd.org/D60822

Whereas if WWW is omitted:

1. portlint correctly identifies that the WWW has a default value (read: for
ports that do not have a default provided by USES*, portlint will complain)

2. You can use `make -V WWW` to obtain the WWW just as easily and
then click the URL that pops out from the command. That is arguably
even easier than using $EDITOR to locate the WWW amongst the noise.

3. We don’t have to hunt down old URL formats and fix them continuously

I view it as a win-win-win.

But I’m not going to force anyone to do anything they don’t want.
— 
Devin