Re: git: 0505b712df3a - main - all: Bump PORTREVISION after lang/go* update

 Hi,

On Thu, 7 Dec 2023 07:12:07 +0000
Pietro Cerutti <gahr@FreeBSD.org> wrote:

> On Dec 07 2023, 02:23 UTC, Ashish SHUKLA <ashish@FreeBSD.org> wrote:
> >The branch main has been updated by ashish:
> >
> >URL: https://cgit.FreeBSD.org/ports/commit/?id=0505b712df3ae2b4bcaba00aecc7c1e9100d8609
> >
> >commit 0505b712df3ae2b4bcaba00aecc7c1e9100d8609
> >Author:     Ashish SHUKLA <ashish@FreeBSD.org>
> >AuthorDate: 2023-12-07 02:15:46 +0000
> >Commit:     Ashish SHUKLA <ashish@FreeBSD.org>
> >CommitDate: 2023-12-07 02:22:10 +0000
> >
> >    all: Bump PORTREVISION after lang/go* update
> 
> Hi, I'm totally ignorant about the go ecosystem, but from what I get, 
> ports that USES "go" without the "run" argument do get go added to 
> RUN_DEPENDS.
> 
> For example mail/opensmtpd-filter-rspamd installs an executable written 
> in go, but doesn't care about go at run time.

 It kinda does as go static link binaries it care about go standard
libraries/functions (don't know the right name for go).

> Wouldn't bumping it (and ports in a similar situation) be equivalent to 
> bumping the version of all utilities written in C after an update of the 
> compiler? :)

 Not really, it's the same as bumping a statically linked C program
after a libc change.

> If the go update doesn't fix immediate issues with previously compiled 
> binaries, perhaps we could limit the amount of churn.. What do you 
> think?

 It's hard to know if the go compiler fixes issue without reading all
the source code of all go ports, it's easier to bump and assume that at
least all go ports will benefit from improvement of some go standard
functions.

 Cheers,

> -- 
> Pietro Cerutti
> I have pledged to give 10% of income to effective charities
> and invite you to join me - https://givingwhatwecan.org


-- 
Emmanuel Vadot <manu@bidouilliste.com> <manu@freebsd.org>