On 11/21/14 14:00, Paige Thompson wrote:
On 11/21/14 18:14, Rich Freeman wrote:
On Fri, Nov 21, 2014 at 12:34 PM, James <[email protected]> wrote:
Here's one, very, very interesting proposals, under
serious consideration:
https://wiki.gentoo.org/wiki/Distributed_Gentoo
I'd be curious what the fine and wonderful folks at
gentoo_user think of this proposal. Old farts are
welcome to comment, even encourage to "constructively rant"....
Honestly, I think it is a bit optimistic, though something I'd love to
see. I'm more of a fan of getting the new working before dismantling
the old, and I'm not keen on proposals that start out with gutting
what we already have before there is anything new to replace it.
Burning bridges usually isn't wise.
The optimistic bit is that the proposal is that the only part of
Gentoo that would actually be developed centrally are the parts that
almost nobody is working on today. That basically amounts to just
having all the developers quit, and hoping that they all move on to
work on overlays instead of moving on to something else.
Interesting perspective that I had not considered... still, I think
there is more at play here. It sounds as if a few "chosen" old guard
are going to kick out more of the progressive and newer devs
so that these few. control the core distro. That way they can agree
and do what they want.
There are a lot of technical challenges in such an approach -
supporting overlays isn't all that unlike trying to provide kernel
internal API stability. I think it could be done better, but it would
be a big change. IMHO it makes far more sense to make those changes
and use them for our own internal benefit in the main repository, and
THEN think about whether the main repository is still needed.
Agreeded. In fact if one reads the threads on gentoo-dev that lead to
this document, one get's a bit more of the picture that is going to be
a very large undertaking, and many, more robust tools will need to be
developed, to support this (2 tier) system proposal. aka, actually
until gentoo current is mature and stablized, it's a very, very bad
idea, imho.
--
Rich
With regards to distributed gentoo;
to me this sounds like gentoo already. Overlays / layman repositories. I
even have my own overlay that is listed publicly for anybody to add:
https://github.com/paigeadele/netcrave
On the surface, yes. The way it will work out, no; hell no. There will
be no new devs moving into the core, unless the "few" bless them and
they will be clones of the existing dev mentalities that have been the
source of gentoo limitations, imho. The question is this: Is gentoo
a "boys club" or an open source evolving/morphing/multi-faceted
jaugernaut we all know and love. This porposal has good intentions,
but is "barking up the wrong tree".
Which goes to say, the obvious should be assumed since anybody can have
their own publicly listed overlay, you should know what's in it before
you add it--it's a fairly autonomous system but you have to ask to get
publicly to get your repo publicly list and If it were any easier, it
would probably get abused.
True, very true, but trivial in the deeper context of who and how the
future of gentoo is steered, imho. i.e. what you are pointing out merely
today's noise.
Technically you can even use git to manage your entire / filesystem if
you want but it would be more productive to write your .gitignore to
ignore all, then conditionally "un-ignore" things that need to be
tracked thereafter. I did this but only for /etc because I've gotten
tired of hunting down HOWTOs and more often than not a simple example is
more than what I need to get stuff done:
https://github.com/paigeadele/laptop.paigeat.info_etc/blob/master/.gitignore
Honestly though I kinda wish I had done this from the top most
directory, there are things here and there outside of /etc that I'd like
to keep for reference or whatever reason.
More truth, but it misses what is really at play here.
You can pretty much do whatever you want with gentoo, its up to you to
decide whether or not to.
TRUE, but it becomes a power struggle, just like what is occurring
presently with openrc vs systemd.
Take the JAVA example. Java a core, fundamental technology in many linux
distros, Android probably being the most visiable (my def of linux
is if you use the linux kernel, it's linux, forget everthing else), ymmv.
Java over the years has never risen to the level of respect and support
it deserves in gentoo, because of a few core gentoo_master that hate
Java. I hate oracle too, but not java. Please do not think what I'm
about to say about gentoo or Java, is in any way a slight against those
devs that support java or any other gentoo dev. It's just a common
prejudice (lack of respect) against java for a very wide variety of
reasons. So if a motivated group, comes together, we fix up java and
create many, many tool and packages that are wonderful, powerful and
popular with a very wide variety of linux users, then gentoo will
explode and the Java team will be the most visually powerful group of
devs to the masses of gentoo users. (surely under current gentoo
circumstances a slim prospect but very viable, imho).
The 'good old boys club' now being demoted to "second base", in the eyes
of the new gentoo users, are very unhappy. If you look at what
innovating on the net, JAVA is king, imho; Python is strong, but a
distant second. Haskell and Scala are the future of linux greatness, but
they have a ways to go. Java bridges to both Scala and
Haskel very, very visibly, imho.
So again, booting out all the new/lesser/younger devs from the core (of
gentoo) and making them second class (user) citizens is all a power
struggle ploy, being pitched as elevating overlays and the gentoo
citizen. Why should their work not be included into the core? Why
should Java (OpenJDK and Icedtea and such) be given a path to the core,
if not becoming a critical language to the core of gentoo?
(philisopically not presently) It is a "wolf in sheep's clothing". It is
the very poorest of ideas. I think we need 800 or 1,000 active gentoo
devs all competing for the the title of council elder or "grand poubah"
to make gentoo stronger and more attractive to the masses. I think one
day that python will be replace by another "better" language. I think
the core power_brokers do not like that possibility because they become,
second class devs and that they *might* be treated like they treat other
newer devs now.
What happens if one of the new brilliant mathematical/physics minds,
with deep admin experiences comes to gentoo and starts using haskel and
scala codes and sets the linux world on fire with innovation? Will that
person find a warm and cozy home at Gentoo? (ok quite down the
laughter). In 1990 I strated one of the first ISP in America. I
convinced SURAnet (who was the internet in 1990) to allow us to connect
to to MAE-EAST and sell TCP/IP connections to the general population.
First one to do it in American and probable in the world. It was only
possible, because we recruited a young mathematican out of England to
write the code. Many were farting around with ppp over serial links,
but this was the first stack that work reliable and was to become open
sources. (I do not want to "get into it" as many old farts have their
own perspective and I have zero desire to argue or elaborate). And no
we did not get rich. cheated is more likely the single word for "the
rest of the story." I merely point out this industry changing event, so
that you realize *ONE* mathematician can change the world. Forget
computer science, forget admins. Forget loosers like LINEART. Math is
and always shall be king, imho.
Would the gentoo elders make room for such a person, if the core is
controlled by the few? doubtless. One of the groups being moved to the
"Overlay fringe" is science/math. Clever, will robinson, clever,
deceitful and stupid all in one proposal. Wonder where Java ends up?
"out of the core" is my guess.
In fact I'd go so far as to say that "term limits" on the council is a
better starting point than enshrining the current few, imho. Please
understand, I'm a huge fan a few of the council members; but power
corrupts your vision and absolute power will certainly corrupt your
database that you draw conclusions from.
I sure hope I am wrong, but, then again, I do love a good conspiracy
theory? Am I a troll? Nope, I'm older, experience and have seen shit
you would not believe. I first proposed carbon sequestration as a young
engineer in the 1980 at Prudoe Bay Alaska. I use to sit in a pickup
truck and watch dozens of Jet engines put massive amounts of CO2 into
the air. May particular resposibility was to ensure that those dozens
of "gas injection wells" did not experience a loss in integrity, cause
the massive explosion would set Prudohe Bay on fire; necessitating that
the 2.2 millions barrels of oil we were producing a day, would have to
be "shut down for repairs". Surely that would have made headlines,
worldwide. I proposed for my graduate work to capture the turbine
emissions (primarily CO2) and inject them into those gas injections
well. Carbon sequestration was not even an issue at the time. CO2 was
and is the best source of "surfactant flooding" to enhance oil recovery.
I also propose to been the massive energy creation up to a satelitte
system, run by the military and use a diffuse beam to bounce the energy
down to chicago for usage. Funny, my professor thought that a bad idea
and wanted me to work with them on converting natural gas into methanol,
so it could be pumped with the crude oil to the Alaskan pipleline and to
market. Today, still there are laws to prevent methane being converted
to methanol; but is is ok to convert corn and other food sources to
methanol or ethanol.
I *was* a young math wize. Now, I'm mostly apathetic, but, I can sniff
out pooh with the best of them and this idea is *pooh*, at best, imho.
apologies well in advance of any pain anyone feels,
But, I have been using Gentoo, since early 2004,
thoughts?
James
