Re: [gentoo-user] Systemd upower

[Canek Peláez Valdés]

On Wed, Jun 4, 2014 at 5:28 AM, Greg Woodbury <redwolfe@×××××.com> wrote:
> On 06/03/2014 10:05 PM, Canek Peláez Valdés wrote:
>> On Tue, Jun 3, 2014 at 8:37 PM, Greg Woodbury <redwolfe@×××××.com> wrote:
>
>>
>>> Sure, systemd is a more elegant solution than the patchworks that have
>>> been applied several times to the original SysV concept.
>>
>> Glad to see you recognize that.
>>
>>> However, the implementors and advocates of systemd have stepped on the
>>> concerns and violated certain basic freedoms of many folks in their zeal
>>> to see their vision become predominate.
>>
>> Oh FFS. What "freedoms" have you had "violated"? The "freedom" to
>> mandate what other developers should write, or what packages they can
>> use as hard dependencies?
>>
>> You never had that "freedom". That's the developer freedom; if you
>> want some of that, become a developer.
>
> I was a developer for more years than I really care to remember.  I
> still try to contribute in ways and areas that I'm not so out-of-date with.

Good for you. Now, imagine you don't only contribute, but that you
actually *maintain* (as in, *you* are in charge) of project X. And
then you see that project X is so much easier to maintain if you
depend on project Y. So you make project Y a hard dependency of
project X.

And then a bunch of people who don't really know how to maintain code,
start yelling at you because you made project X dependant on project
Y. And they *demand* of you that you should not depend on Y, but they
don't provide the code to do it,

Will you drop dependency on project Y, even if it makes your life as a
maintainer several times easier?

> Furthermore, it is a two-way street (as I see it.) The developers write
> things they find interesting and enjoyable to work on, and users use
> things that are interesting and work well.  For many, seeing other folks
> use what they have written provides a significant measure of the
> enjoyment derived from the exercise.

That does not contradicts anything I have said.

> To see this as only freedom for the developer is part of an attitude
> shift over the years that only lessens the overall usefulness of Linux
> and FOSS. It does, in fact, push quite a few folk I know away from the
> Linux arena. It is, to use a political analogy, like the people who
> claim there "is not any real difference" between *any* opposing
> political movements -- that neglects taking into account a great deal of
> technical and historical details.

I have no idea what do you mean by the last paragraph. This is not a
political discusion (although many would like to see it that way). It
is a *technical* discusion, and therefore there is no real discusion:
the general consensus is that systemd is the technological superior
alternative.

> I occasionally think about forking projects and fixing some of the
> things I think are the most egregious fsck-ups in some of them, but then
> I really look at what I'm doing and what I enjoy doing, and realize that
> I won't get enough (emotional?) reward for giving up time in other
> significant parts of my life.

And that's your right, and it's fine. But let *other* developers
choose whatever technologies they want to choose, and (consequently)
drop support for obsolete technologies like pm-utils.

That's the reason for this whole thread: developers chose the
technological superior alternative; saying that the reason for that is
that there is cabals and conspiracies is blatant ignorance (in the
best case), or spreading FUD (in the worst case).

>> Or help Samuli to maintain upower-pm-utils; that would be *much* more
>> helpful than spreding FUD about cabals and conspiracies.
>
> There is no need for me to invent Fear, Uncertainty, and Doubt -- the
> folks involved are doing quite well on their own.

I never said you "invented" it. I say you are spreading it, and I
still think that's the case.

> Also, history (for
> those not doomed to repeat it [1]) provides all that is required to make
> calling it a "cabal"  [TINC - there is no cabal![2]]  There never was a
> Usenet Backbone Cabal in any formal sense, but there was plenty of
> semi-(un)coordinated activity -- based largely on shared ideals -- that
> gave that appearance.  {I was there when Usenet/Netnews was invented,
> closely observing, making minor and not-so-minor contributions, and was
> responsible for some of the "cabal-like" activities.}

Great; so any kind of group work "semi-(un)coordinated" can be called
a cabal, and it has no (inherent) negative connotation. Then the Linux
Kernel developers is a Cabal; the GNOME devs is a Cabal; the KDE ones
are also a Cabal; and of course the Gentoo Developers who *oppose*
systemd is a Cabal, and so are the ones that *support* systemd.

So you yourself are saying that calling out a "Cabal" of systemd
proponents have no meaning at all whatsoever, because *EVERYTHING* is
a Cabal.

> The mere coinage of terms like "Lennertware," whether or not deserved,
> show that there is a widespread awareness that some developers, in my
> opinion, have over developed egos. [3]

Yeah, please go and check out the "contributions" (when they exist) of
the people that seriously use the term "Lennartware". Doesn't matter
to what project, check out what they have contributed over the years.
Go on, I wait; it would not take you long, because they usually are
NOT developers, and the few that are actually developers haven't
contributed really that much.

Now go on and check out the contributions of the people that say that,
please, do not use the term Lennartware.

There is *NO* "widespread awareness". There is a small but really loud
group of people that are unwilling and/or unable to create good code
that all distributions wants to use, and therefore use childish
tactics like coining terms like "Lennartware".

> It is all so trite to say "become a developer and DO something instead
> of complaining"  but it is not a realistic thing to say when the
> problems are getting so large and interconnected.

And that's the root of your misunderstanding Greg. There are no
"problems"; this "interconnection" is by design, because many
developers are fed up with a Lego-like plumbing where you can
interchange any basic component like a Lego block, all of them equally
weak and fragile, which makes the testing matrices of all
distributions a nightmare to maintain.

So they are creating a single plumbing layer that we are calling
systemd. This is not a "problem", although there are developers that
see it that way; this is a technical solution to a technical problem.

Furthermore, it works GREAT, and the best proof of that is that
basically Gentoo and Slackware (I think) are the only distributions
not using it by default.

> Furthermore, it
> denigrates and devalues the "pseudo-democratic" processes that FOSS and
> Linux have worked for years to nurture.

There was *never* a "pseudo-democratic" process in FOSS or Linux.
NEVER. It would be a *terrible* mistake.

It has always been a *meritocratic* process. That's why we have
"benevolent dictators" everywhere in the community:

THOSE WHO WRITE THE CODE, MAKE THE RULES.

So if you want to change the rules, start writing some code.

Regards.
-- 
Canek Peláez Valdés
Profesor de asignatura, Facultad de Ciencias
Universidad Nacional Autónoma de México