Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 14 Sep 2019 10:52:45 -0600
From:      Adam Weinberger <adamw@adamw.org>
To:        Koichiro Iwao <meta@freebsd.org>
Cc:        FreeBSD Ports <ports@freebsd.org>, ruby@freebsd.org
Subject:   Re: FLAVORS for Ruby
Message-ID:  <006FCB74-04EB-4A82-A800-6C7CA273E749@adamw.org>
In-Reply-To: <20190914042738.r3hedyqtpxsxnd5e@icepick.vmeta.jp>
References:  <20190913074519.xfu3avb4ihmfzm2o@icepick.vmeta.jp> <CAALwa8m3NJSckxSyrVTehx3LWxheSJXxUB=iAxr%2B1HG_2WFtfg@mail.gmail.com> <20190913090645.buutinhgh2pygb4h@icepick.vmeta.jp> <CAP7rwcjH6d7AiEL4XeRyQ-evT=dq%2BZi8WDTx-aYhye-zq2DKJw@mail.gmail.com> <20190914042738.r3hedyqtpxsxnd5e@icepick.vmeta.jp>

next in thread | previous in thread | raw e-mail | index | archive | help
> On Sep 13, 2019, at 22:27, Koichiro Iwao <meta@freebsd.org> wrote:
>=20
>> On Fri, Sep 13, 2019 at 09:33:43AM -0600, Adam Weinberger wrote:
>> Systems MUST be able to support concurrent installations of python2.7
>> and actual python. What is your use case for concurrent ruby?
>=20
> I know the importance of Python 2. Even if it is EoL-ed, it will be
> required over the next a few years because not a few applications don't
> migrate to Python 3. So that's true and reasonable.
>=20
> Excuse me that I'm answering your question with a question. What about
> PHP? Concurrent installation is a MUST?
>=20
> FreeBSD ports allows concurrent installations of multiple Ruby versions
> however doesn't allow concurrent installations of rubygems for multiple
> Ruby versions. This inconsistency is the issue for me.

The issue is that FLAVORS has added a substantial (and painful) complexity t=
o python ports and python.mk. It means that a number of people have had to b=
e hyper-vigilant and watch commits closely to catch errors introduced when p=
eople utilize the paradigm incorrectly. It=E2=80=99s a bitter pill, but it=E2=
=80=99s accepted because the use-case for multiple concurrent python version=
s is essential.

As Antoine said, inconsistency isn=E2=80=99t a strong enough use case. Which=
 brings us back to the original question: is there a specific use-case for c=
oncurrent ruby that makes the substantial increase in cognitive load, comple=
xity, and monitoring worth it?

# Adam


=E2=80=94
Adam Weinberger
adamw@adamw.org
https://www.adamw.org=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?006FCB74-04EB-4A82-A800-6C7CA273E749>