[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: A concerned user -- debian Guidelines

On Monday, 10 de November de 2014 08:57:50 Nathael Pajani escribió:
> You certainly heard about "debianfork" (http://debianfork.org/) and from a
> user point of view this is a tragedy.

A derivative and a fork are different things. A Derivative happens when a 
different project is started using the codebase of the first, with different 
goals, and both projects cross-pollinize each other. Both become stronger, if 
nough manpower is available. A fork happens when angry developers start a copy 
of the first project and abandon it, because it is unworkable. Ubuntu is a 
derivative. XOrg is a fork (and XFee86 was almost abandoned).

If Debianfork is a fork, it will be sad, as you say. But it seems to me that 
(if realized) it will become a derivative. Maybe an angry derivative with 
small cross-help, but a derivative anyway.
> >From my point of view (or from a user's point of view) what is about to
> >happen is a breach
> in Debian Social Contract.
> Debian Social Contract states as point 4 : Our priorities are our users and
> free software
> >From my point of view, users are ignored, and what we can read here and
> >there is that the
> decision is up to (put bluntly) "those who spend time in the project" (the
> Debian developers).

Agreed. But since Debian is a volunteer's project, it makes sense that the 
people who works on it (not *for* it) works on the topics they prefer. In my 
view, forcing systemd is a bad election, while choosing it as default for new 
installs is a good one. And that second decision was taken with the users as a 
priority, since it improves the user experience for our less technically 
suited users which just want "to install that Linux thing and have Word and a 
browser", since systemd makes Gnome work better. The first decision is the one 
we are actually discussing (or arguing about): whether to force init switch on 
upgrade. Ability to install a different than default init system has been 
raised as a concern as well. But systemd IS our default.
> But the project exists because thousands of other people use it and
> contribute to other free software - those 20K free software which makes
> Debian such a rich Distribution.

Not necesarily. The project exists because some people devotes time on it. 
Systemd itself, as an example at hand, was developed mainly at Red Hat 
quarters. And we use it. It does not depend on Debian in order to contribute 
to Debian. Same for tons of other projects.
> Even, some users contribute directly to Debian, by filling bug reports,
> speaking about Debian, buying goodies, .... and many many other ways.

True. Some of them will enjoy systemd, some actually hate it with no reason, 
and some want systemd-free systems for good reasons. And still some do not 
care. And the hard point is to give figures for those sets.
> When a (big ?) pool of users is not happy to the point of suggesting to
> fork because of a decision taken (or about to be taken) by the project
> developers, then (I think) that the Debian developers are not doing their
> job right.
> You'll notice that the fork has not been started yet, as (I think) many
> still hope this can end the right way, with USERS taken into account. All
> of them.

Agreed that all users must be taken into account, by our Constitution. But 
somebody must do the work.

Remember that the issue is not about the default. That wave passed time ago. 
It is about exact meaning of some words like "default" itself.

AND, if the worst thing happens here and Debianfork starts in the most hateing 
way, Debian will still be a great community (not so big, but still great) with 
the best non-company Linux distribution available. And with time hate will 
disappear and most probably Debianfork will become some kind of Debian Blend.
> Remember Ian Murdock's intention : " Ian intended Debian to be a
> distribution which would be made openly, in the spirit of Linux and GNU".

And it is. This very thread and the GRs and the CTTE decisions are proof that 
it is made openly. It just happens that some developers have excess of love 
for their init system project, and that that project is engulfing other things 
and that it has binary logs and a lot of other issues... but the decision was 
taken openly.
> Still from my point of view, this means that the user can choose between
> alternatives for almost everything when there is a choice. Let's cite a
> few to make it evident : Vim/emacs, KDE/Gnome/All/the/others/ones, Debian
> kernel/custom kernel, ....

There is option to use other init systems on Jessie. It has been shown how to 
install them. Maybe Release Notes need extra work. Do you volunteer? Surely 
Debian Installer needs a lot of extra work and testing. Do you volunteer? Sure 
as hell Debian will benefit for a new, shiny, well-designed init system in the 
UNIX principles tradition. Will you write it?

> Especially when this breaks so many of my systems. Of course I could spend
> time to learn the new init, a change all of my systems to fit the new init
> system. But this will cost me time. Time I will not spend on other free
> software projects. It will cost time to so many people.

If you do not upgrade to Jessie, nothing will break. And it seems we will make 
it work not to break anything even on upgrades.
> Yes, making an init-independent system is more work than the easy, "single
> init" solution. But I think there are people out there which will be
> willing to spend time on this. People with the ability to do it, to face
> the challenge and succeed.
> Be the first distribution to be init system independent, and propose a way
> for users to choose between different init systems. This would allow
> integration of other init systems in the future, and make Debian stronger,
> safer, more reliable, and once more, more universal. Of course, this means
> more work, and maybe delayed releases. But I though that Debian had no
> "release schedules". "It's done when it's done".

I fully agree with this. How would you contribute to make it happen?
> Other distributions may have chosen the easy single init way, but Debian is
> not other distributions. And uniformity is not an option.
> Should It be, let's forget about all Linux and GNU stuff, and all of us
> move to another system (let's not name it here), for the sake of
> uniformity.
> But GNU/Linux is NOT uniformity. It is choice. It is alternatives. It is
> options.

How many window systems can you choose among? How many C libraries? How many 
*really* different kernels? We are about choice between free software and 
propietary software, choice between freedom and slavery, but not everything on 
a Free Software system should (or can) be chosen.

I agree, anyway, that we should be able to choose init system and boot loader, 
but do not put alternatives so high in the stair.

> Please make me (us ?) happy again. Happy of using Debian, wearing my Debian
> Tee-shirt despite whatever people can tell me about it, and happy of
> "spreading the word", making others choose Debian for it's social value as
> much as it's technical one. Happy of contributing to free software, if not
> directly to the Debian project.

What do you really advocate for, that is not being worked in?


Noel Torres
er Envite
Debian Maintainer, Debian (and Mint and Red Hat and Ubuntu and Windows and 
Solaris and Android) User, Debian (and Tor) Translator, Debian (and KDE) Bug 
reporter, Debian (and Mint and Red Hat and Ubuntu and Windows and Solaris) 
Systems Administrator, Debian systemd hater, Debian lover, not a Debian 

Just open your mind and try to see a bigger picture

Attachment: signature.asc
Description: This is a digitally signed message part.

Reply to: