[gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 3910 bytes --]

Hi, everyone.

I'd like to request Infra to establish a new mailing list that would
fill in the gap between our public mailing lists and the gentoo-core
mailing list.

Name: gentoo-dev-internal

Topic: technical discussions between active Gentoo contributors

Restrictions:

- public (open subscription), initially we may optionally copy all
subscribers from gentoo-dev so that they do not miss discussion,

- archived,

- but posting restricted to opt-in member group.

Initially, the posting group would include active Gentoo developers
only. Afterwards, we will deploy a small moderator team whose purpose
would be controlling access to the list -- including both adding new
members on request and removing existing members (including developers)
if they misbehave.

I don't think we need to precisely define the rules for admitting new
members. I think the exact procedure would be at moderators' discretion
and would depend on the current 'health' of candidates -- i.e. if things
go calm they may just admit on request, and if people start abusing this
they will force explicit moderation before whitelisting.


Rationale
=========

The purpose of gentoo-dev is to allow technical discussion between
contributors to Gentoo, especially including making it possible for
developers to send RFCs and discuss their ideas.

Sadly, it is not uncommon for threads on that mailing list to turn into
trollfests, get deranged or hijacked into completely different topics.
Things are so bad that the mailing list stops serving its purpose. It
involves a number of consequences:

a. The developers lose time on the mailing lists instead of using it for
constructive purposes. Even skimming through those mails in search of
something remotely relevant is time-consuming.

b. The developers and contributors become discouraged and unsubscribe
from the mailing list. As a result, audience for reviews and RFCs
becomes smaller and even less focused on the topic.

c. The developers become discouraged and stop sending their ideas.
Either they do less, use another media or work in complete isolation
from other community members.

d. Eventually, the developers become tired of the persisting issues
and they retire (yes, it's a fact).

Other ideas on solving those issues were pretty much rejected already:

1. Bans on persisting violators were rejected as they are easily worked
around via subscribing from another e-mail address, and causing more
noise than the original issue.

2. Full-scale moderation of mail on gentoo-dev was rejected because of
technical limitations and the resulting high level of effort in handling
the moderation, plus the social effect.

3. Making gentoo-dev@ opt-in (like the suggested new list) would make it
much harder for users to contribute ideas, and would inevitably
discourage some of the users from writing.

All that considered, establishing a second mailing list with different
characteristic seems like a reasonable solution. In particular:

A. It gives a wider choice of tools for developers (and privileged
contributors) -- they can choose either the open or restricted mailing
list depending on the type of requested feedback.

B. The gentoo-dev mailing list is still open for power users
and contributors to submit their own ideas, and with no moderation
the discussion can proceed naturally.

C. The cost of moderation should be relatively low, and the methods can
be dynamically adjusted to fit the needs. In particular, good behavior
on gentoo-dev can be used to grant access to gentoo-dev-internal without
further requirements.

D. The restricted mailing list should be resilient to ban evasion since
the access is opt-in, and the moderators team can enforce direct
moderation of new members if there is a considerable risk.

Your comments?

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Rich Freeman]

On Tue, May 23, 2017 at 12:31 PM, Michał Górny <mgorny@gentoo.org> wrote:
>
> A. It gives a wider choice of tools for developers (and privileged
> contributors) -- they can choose either the open or restricted mailing
> list depending on the type of requested feedback.
>
> B. The gentoo-dev mailing list is still open for power users
> and contributors to submit their own ideas, and with no moderation
> the discussion can proceed naturally.
>

Wouldn't those inclined to do so simply crosspost threads to the
unmoderated list?  Then people will feel the need to respond to those
posts for all the reasons they reply to those posts already.

Maybe that would not happen if the unmoderated list essentially
becomes unused and has few subscribers, but in that case we are
potentially turning away contributions.

Also, given the reluctance to moderate anything around here in
general, do we think that the moderated list would actually be
moderated if it is after-the-fact?  The one advantage of requiring
moderation a priori is that it requires somebody to affirm "yes, this
post adds to the discussion" vs having to decide "do I want to be the
one to ban user xyz from the list and deal with the fallout?"

That said, it does address the sock puppet issue to a large degree,
unless somebody wants to be fairly painstaking at it.  (And if they're
willing to go to that much trouble we'd need to be screening IDs to
keep them out.  There is no reason somebody couldn't go through
recruitment as a dev 14 times today.)

-- 
Rich

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Philip Webb]

170523 Michał Górny wrote:
> Sadly, it is not uncommon for threads on that mailing list to turn into
> trollfests, get deranged or hijacked into completely different topics.
> Things are so bad that the mailing list stops serving its purpose. It
> involves a number of consequences:

As a user, I've been subscribed to this list since 2003
& can't remember any recent occasion -- and very few not recent --
when any such bad or damaging behaviour has happened.

Is this proposal itself not just a waste of valuable developer time
in moderating, censoring & deciding who is a sheep & who is a goat ?

-- 
========================,,============================================
SUPPORT     ___________//___,   Philip Webb
ELECTRIC   /] [] [] [] [] []|   Cities Centre, University of Toronto
TRANSIT    `-O----------O---'   purslowatchassdotutorontodotca

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Kent Fredric]

[-- Attachment #1: Type: text/plain, Size: 486 bytes --]

On Tue, 23 May 2017 21:31:18 +0200
Michał Górny <mgorny@gentoo.org> wrote:

> - public (open subscription), initially we may optionally copy all
> subscribers from gentoo-dev so that they do not miss discussion,

What would be the result if somebody replied to a g-dev-internal ML
without permission?

I think there should be something in place other than sending it
to /dev/null , but I can't think of any good approach that doesn't make
g-dev messier as a result.




[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Kent Fredric]

[-- Attachment #1: Type: text/plain, Size: 980 bytes --]

On Tue, 23 May 2017 15:52:52 -0400
Philip Webb <purslow@ca.inter.net> wrote:

> Is this proposal itself not just a waste of valuable developer time
> in moderating, censoring & deciding who is a sheep & who is a goat ?

Its not censorship, because censorship is the practice of preventing
something from being said in entirety.

People can still voice their opinion, and we aren't going to be
suppressing it, we're just making a decision about which channels they
can say things to.

And we're giving consumers a choice whether they want to hear
everything, or hear only a subset of things, by allowing them to
subscribe to either channel.

That's no more censorship than somebody setting their social media feed
to allow no outsiders to post on it. Individuals can still post in
their own feeds, and those who want to see it can still see it.

But when it comes to *my* feed, I'm entitled to dictate what occurs in
it, without being accused of censorship.

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Erik Närström]

[-- Attachment #1: Type: text/plain, Size: 752 bytes --]

I'snt the idea of creating a new mailing list to let gentoo-dev be mesy?,
in that case we could simply redirect all non-aproved posts from gen-int to
gen-dev.

/EKG

On 23 May 2017 22:04, "Kent Fredric" <kentnl@gentoo.org> wrote:

> On Tue, 23 May 2017 21:31:18 +0200
> Michał Górny <mgorny@gentoo.org> wrote:
>
> > - public (open subscription), initially we may optionally copy all
> > subscribers from gentoo-dev so that they do not miss discussion,
>
> What would be the result if somebody replied to a g-dev-internal ML
> without permission?
>
> I think there should be something in place other than sending it
> to /dev/null , but I can't think of any good approach that doesn't make
> g-dev messier as a result.
>
>
>
>

[-- Attachment #2: Type: text/html, Size: 1167 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Kent Fredric]

[-- Attachment #1: Type: text/plain, Size: 672 bytes --]

On Tue, 23 May 2017 20:32:03 +0000
Erik Närström <erik.narstrom@gmail.com> wrote:

> I'snt the idea of creating a new mailing list to let gentoo-dev be
> mesy?, in that case we could simply redirect all non-aproved posts
> from gen-int to gen-dev.

I mean, messy in the sense you'd have replies, but no obvious parent
for the replies, so people who wanted to read the whole thread
including external submissions would need to mentally piece together 2
independent mailing lists.

Though, I guess my email client might work as expected being subscribed
to both lists.

Just anyone who signs up to only gentoo-dev will get a lot of confusing
arguments :)

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Erik Närström]

[-- Attachment #1: Type: text/plain, Size: 1177 bytes --]

The same asnow with other words?, wouldn it be possible to repost peoples
mail in gen-dev?. I mean instead  of leting people send mail to gen-int we
kidnapped the mail and send it to gen-dev so that all using gen-dev gets
the hole conversion, and does only subscribe to gen-int only get to se
posts from pre-aproved mail-acont, wouldn't that work?, or am I missing
something fundamental?.

/EKG

On 23 May 2017 23:05, "Kent Fredric" <kentnl@gentoo.org> wrote:

> On Tue, 23 May 2017 20:32:03 +0000
> Erik Närström <erik.narstrom@gmail.com> wrote:
>
> > I'snt the idea of creating a new mailing list to let gentoo-dev be
> > mesy?, in that case we could simply redirect all non-aproved posts
> > from gen-int to gen-dev.
>
> I mean, messy in the sense you'd have replies, but no obvious parent
> for the replies, so people who wanted to read the whole thread
> including external submissions would need to mentally piece together 2
> independent mailing lists.
>
> Though, I guess my email client might work as expected being subscribed
> to both lists.
>
> Just anyone who signs up to only gentoo-dev will get a lot of confusing
> arguments :)
>

[-- Attachment #2: Type: text/html, Size: 1625 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[William Hubbs]

[-- Attachment #1: Type: text/plain, Size: 710 bytes --]

On Tue, May 23, 2017 at 09:31:18PM +0200, Michał Górny wrote:
> Hi, everyone.
> 
> I'd like to request Infra to establish a new mailing list that would
> fill in the gap between our public mailing lists and the gentoo-core
> mailing list.
> 
> Name: gentoo-dev-internal
> 
> Topic: technical discussions between active Gentoo contributors

That's what gentoo-dev is now.

I'll be honest. I'm really way too tired right now to write a long
argument, but do we really think that basically moving everyone to
another mailing list will solve these kinds of issues?

I think this is just shuffling people around and will lead to the same
thing happening on the new list over time.

William


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Tuomo Hartikainen]

Hi all,

On 2017-05-23 21:08, William Hubbs wrote:
> On Tue, May 23, 2017 at 09:31:18PM +0200, Michał Górny wrote:
> > Hi, everyone.
> > 
> > I'd like to request Infra to establish a new mailing list that would
> > fill in the gap between our public mailing lists and the gentoo-core
> > mailing list.
> > 
> > Name: gentoo-dev-internal
> > 
> > Topic: technical discussions between active Gentoo contributors
> 
> That's what gentoo-dev is now.
> 
> I'll be honest. I'm really way too tired right now to write a long
> argument, but do we really think that basically moving everyone to
> another mailing list will solve these kinds of issues?
> 
> I think this is just shuffling people around and will lead to the same
> thing happening on the new list over time.
> 
> William

I too was left wondering, what would be the role of the gentoo-dev after
the proposed change. If it is agreed that gentoo-dev has become unusable
for development work and the discussion is moved elsewhere, what would
be the intended purpose of the current list?

-- 
Tuomo Hartikainen

[gentoo-dev] Re: [RFC] Restricted version of gentoo-dev mailing list

[Benda Xu]

[-- Attachment #1: Type: text/plain, Size: 707 bytes --]

Hi Michał,

> Name: gentoo-dev-internal
>
> Topic: technical discussions between active Gentoo contributors

Basically I object to this proposal.  

1. Another layer of hierarchy is not desirable for a non-profit
   organization like us.

2. Useful discussion are diluted from 1 list into 2 lists.

3. It is really hard to whitelist/moderate in a transparent and
   objective way.


I take the intention of this proposal as that you would like to keep a
certain group of people out of your discussions.  If you personally want
to mute someone, it is straightforward to set up a blacklist in your
MTA/MUA.

I don't think a change is needed at the Gentoo infra level.

Yours,
Benda

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Ulrich Mueller]

[-- Attachment #1: Type: text/plain, Size: 934 bytes --]

>>>>> On Tue, 23 May 2017, William Hubbs wrote:

> On Tue, May 23, 2017 at 09:31:18PM +0200, Michał Górny wrote:
>> I'd like to request Infra to establish a new mailing list that would
>> fill in the gap between our public mailing lists and the gentoo-core
>> mailing list.
>> 
>> Name: gentoo-dev-internal
>> 
>> Topic: technical discussions between active Gentoo contributors

> That's what gentoo-dev is now.

> I'll be honest. I'm really way too tired right now to write a long
> argument, but do we really think that basically moving everyone to
> another mailing list will solve these kinds of issues?

> I think this is just shuffling people around and will lead to the same
> thing happening on the new list over time.

+1

IMHO another list with the same topic as gentoo-dev makes no sense.
*If* we want moderation, we should moderate the existing list, but not
create a redundant new one.

Ulrich

[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Walter Dnes]

  Let's step back a minute, and consider *WHY* non-developers feel the
need to join this list in the first place.  Do you remember the acrimony
after the decision to not officially support a separate /usr without
initramfs?  A lot of people who complained on the Gentoo-User list were
bluntly told that the devs hadn't heard much objection, and that they
should've expressed their opinion on *THIS* list *BEFORE* the final
decision was made.  This reminds me of a quote from The Hitchiker's Guide
To The Galaxy...


"But the plans were on display..."

"On display? I eventually had to go down to the cellar to find them."

"That's the display department."

"With a flashlight."

"Ah, well, the lights had probably gone."

"So had the stairs."

"But look, you found the notice, didn't you?"

"Yes," said Arthur, "yes I did. It was on display in the bottom of a
locked filing cabinet stuck in a disused lavatory with a sign on the
door saying 'Beware of the Leopard'." 


  I'd really rather not have to follow an additional list which includes
a lot of very technical items and patches that are over my head.  But
after the separate /usr fiasco, I don't feel I have any choice.  If we
could have a guarantee of proposed changes like that being posted on
Gentoo-User for comment, rather than being sprung on users by surprise,
I'd be willing to sign off this list.

-- 
Walter Dnes <waltdnes@waltdnes.org>
I don't run "desktop environments"; I run useful applications

Re: [gentoo-dev] Re: [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 1046 bytes --]

On śro, 2017-05-24 at 14:21 +0900, Benda Xu wrote:
> Hi Michał,
> 
> > Name: gentoo-dev-internal
> > 
> > Topic: technical discussions between active Gentoo contributors
> 
> Basically I object to this proposal.  
> 
> 1. Another layer of hierarchy is not desirable for a non-profit
>    organization like us.
> 
> 2. Useful discussion are diluted from 1 list into 2 lists.
> 
> 3. It is really hard to whitelist/moderate in a transparent and
>    objective way.
> 
> 
> I take the intention of this proposal as that you would like to keep a
> certain group of people out of your discussions.  If you personally want
> to mute someone, it is straightforward to set up a blacklist in your
> MTA/MUA.

Sure. And also mute the 30 random replies which might or might not be
relevant.

> I don't think a change is needed at the Gentoo infra level.
> 

I wouldn't call opening a new mailing list 'a change'. It is a regular
request made by developers who need a tool to work with.

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 1217 bytes --]

On śro, 2017-05-24 at 07:29 +0200, Ulrich Mueller wrote:
> > > > > > On Tue, 23 May 2017, William Hubbs wrote:
> > On Tue, May 23, 2017 at 09:31:18PM +0200, Michał Górny wrote:
> > > I'd like to request Infra to establish a new mailing list that would
> > > fill in the gap between our public mailing lists and the gentoo-core
> > > mailing list.
> > > 
> > > Name: gentoo-dev-internal
> > > 
> > > Topic: technical discussions between active Gentoo contributors
> > That's what gentoo-dev is now.
> > I'll be honest. I'm really way too tired right now to write a long
> > argument, but do we really think that basically moving everyone to
> > another mailing list will solve these kinds of issues?
> > I think this is just shuffling people around and will lead to the same
> > thing happening on the new list over time.
> 
> +1
> 
> IMHO another list with the same topic as gentoo-dev makes no sense.
> *If* we want moderation, we should moderate the existing list, but not
> create a redundant new one.
> 

Then please make it happen. Find a working solution and implement it.
Talking about how everything is bad is not going to solve any problem.

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 824 bytes --]

On śro, 2017-05-24 at 01:55 -0400, Walter Dnes wrote:
>   Let's step back a minute, and consider *WHY* non-developers feel the
> need to join this list in the first place.  Do you remember the acrimony
> after the decision to not officially support a separate /usr without
> initramfs?  A lot of people who complained on the Gentoo-User list were
> bluntly told that the devs hadn't heard much objection, and that they
> should've expressed their opinion on *THIS* list *BEFORE* the final
> decision was made.  This reminds me of a quote from The Hitchiker's Guide
> To The Galaxy...
> 

Next time such a thing happens, the discussion will happen on a
completely private media or not happen at all because of the state of
this mailing list. Is this what you really want?

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Kristian Fiskerstrand]

[-- Attachment #1.1: Type: text/plain, Size: 637 bytes --]

On 05/24/2017 08:41 AM, Michał Górny wrote:
> Next time such a thing happens, the discussion will happen on a
> completely private media or not happen at all because of the state of
> this mailing list. Is this what you really want?

I wouldn't agree with it being an axiom of sorts that discussion will
happen elsewhere. The noise level of the discussion of a new list or
moderation of the current dev list is greather than the noise that
spurred the discussion to begin with.

-- 
Kristian Fiskerstrand
OpenPGP keyblock reachable at hkp://pool.sks-keyservers.net
fpr:94CB AFDD 3034 5109 5618 35AA 0B7F 8B60 E3ED FAE3


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Walter Dnes]

On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
> Next time such a thing happens, the discussion will happen on a
> completely private media or not happen at all because of the state
> of this mailing list. Is this what you really want?

  Here's the part you did not quote...

>> If we could have a guarantee of proposed changes like that being
>> posted on Gentoo-User for comment, rather than being sprung on
>> users by surprise, I'd be willing to sign off this list.

  Note where I said "...posted on Gentoo-User for comment...".  What I'm
asking is for such proposed changes to be posted on Gentoo-User, and the
discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
surprise stuff seems to happen a lot in Open Source...

* Gentoo /usr
* Firefox Australis UI, and dropping ALSA and going PulseAudio-only
* GNOME getting a hard-coded dependancy on systemd
* etc, etc

-- 
Walter Dnes <waltdnes@waltdnes.org>
I don't run "desktop environments"; I run useful applications

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Rich Freeman]

On Wed, May 24, 2017 at 3:48 AM, Walter Dnes <waltdnes@waltdnes.org> wrote:
> On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
>> Next time such a thing happens, the discussion will happen on a
>> completely private media or not happen at all because of the state
>> of this mailing list. Is this what you really want?
>
>   Here's the part you did not quote...
>
>>> If we could have a guarantee of proposed changes like that being
>>> posted on Gentoo-User for comment, rather than being sprung on
>>> users by surprise, I'd be willing to sign off this list.
>
>   Note where I said "...posted on Gentoo-User for comment...".  What I'm
> asking is for such proposed changes to be posted on Gentoo-User, and the
> discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
> surprise stuff seems to happen a lot in Open Source...
>
> * Gentoo /usr
> * Firefox Australis UI, and dropping ALSA and going PulseAudio-only
> * GNOME getting a hard-coded dependancy on systemd
> * etc, etc
>

What value would be obtained by posting this stuff for user comment?
I'd also note that only one of those was posted on -dev-announce for
comment as far as I'm aware.  Two are package/project-level changes
which typically don't get wide discussion.

These sorts of changes aren't being made for the purpose of giving
users a hard time.  They're typically done because of technical
constraints.

Sure, it is valuable when somebody points out an issue nobody has
thought of.  However, dropping support for /usr not being mounted
during early boot was something that was recognized up-front as being
controversial.  It is doubtful that a bunch of additional list
contributors would have pointed out an issue that wasn't already
discussed or anticipated.  Sure, maybe we'd get 20 people posting that
they don't like the change, but that would have been unlikely to
actually change the outcome of the decision.  That basically means
that it is unhelpful.  We already knew that a lot of people weren't
going to like the change, and numerous developers said as much as
well.  The change was made because to some degree it had already
happened and it was the result of upstream forces that were becoming
increasingly difficult to work around.  For what its worth, I suspect
that a system with /usr mounted late probably isn't much more likely
to break today than it was back then - we just won't necessarily take
bug reports if it does in some corner case.

Honestly, I think the flamefests are generally not helpful.  For one
they tend to discourage contribution.  A few have already posted on
this list that Gentoo is well-known to be a community with lots of
infighting/etc.  Well, putting controversial changes out there just so
that people can complain about them isn't going to change that
reputation if we're going to make the change anyway.

Obviously there is only so much that we can do to stop people from
complaining, but keep in mind that every time somebody posts a "Gentoo
devs are out of touch" post/email/whatever it isn't exactly great for
PR.  Most of those who do contribute do so because it scratches some
personal itch and so a lot of us just ignore it (which probably wasn't
the goal of those complaining either).  However, there are probably
many who might get involved, and who might even listen to these
complaints in the future, who don't get involved because of them.  A
lot of the sentiment in these discussions is about trying to keep the
useful contributions without the noise.

My main concern with the multiple list suggestion is whether it will
actually accomplish the intended goal in the first place.  If not,
then the issue of social contract is a bit moot.

-- 
Rich

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 1684 bytes --]

On śro, 2017-05-24 at 03:48 -0400, Walter Dnes wrote:
> On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
> > Next time such a thing happens, the discussion will happen on a
> > completely private media or not happen at all because of the state
> > of this mailing list. Is this what you really want?
> 
>   Here's the part you did not quote...
> 
> > > If we could have a guarantee of proposed changes like that being
> > > posted on Gentoo-User for comment, rather than being sprung on
> > > users by surprise, I'd be willing to sign off this list.
> 
>   Note where I said "...posted on Gentoo-User for comment...".  What I'm
> asking is for such proposed changes to be posted on Gentoo-User, and the
> discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
> surprise stuff seems to happen a lot in Open Source...
> 
> * Gentoo /usr
> * Firefox Australis UI, and dropping ALSA and going PulseAudio-only
> * GNOME getting a hard-coded dependancy on systemd
> * etc, etc

And what would be the use of those 'user comments'? Do you believe it
would change anything? So what is the purpose of asking more users from
feedback *we do not want*?

Would it make you feel better if you were asked to raise your objection?
Would you feel better claiming that we did it against objections of many
users? Or do you believe we would abandon it and decide not to do
anything because of the resulting bikeshed -- which seems to be
a recurring theme lately?

Yes, I *do not want feedback* on *how to do Gentoo* from people who do
*not help me do Gentoo* but instead only complain and demand.

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] Re: [RFC] Restricted version of gentoo-dev mailing list

[Kent Fredric]

[-- Attachment #1: Type: text/plain, Size: 324 bytes --]

On Wed, 24 May 2017 14:21:13 +0900
Benda Xu <heroxbd@gentoo.org> wrote:

> 2. Useful discussion are diluted from 1 list into 2 lists.

I think dilution is the point, but the problem is not useful
discussions.

Its the useless ones that are targeted for dilution, so that the useful
ones can stop drowning in them.

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Alan McKinnon]

On 24/05/2017 12:24, Michał Górny wrote:
> On śro, 2017-05-24 at 03:48 -0400, Walter Dnes wrote:
>> On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
>>> Next time such a thing happens, the discussion will happen on a
>>> completely private media or not happen at all because of the state
>>> of this mailing list. Is this what you really want?
>>
>>   Here's the part you did not quote...
>>
>>>> If we could have a guarantee of proposed changes like that being
>>>> posted on Gentoo-User for comment, rather than being sprung on
>>>> users by surprise, I'd be willing to sign off this list.
>>
>>   Note where I said "...posted on Gentoo-User for comment...".  What I'm
>> asking is for such proposed changes to be posted on Gentoo-User, and the
>> discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
>> surprise stuff seems to happen a lot in Open Source...
>>
>> * Gentoo /usr
>> * Firefox Australis UI, and dropping ALSA and going PulseAudio-only
>> * GNOME getting a hard-coded dependancy on systemd
>> * etc, etc
> 
> And what would be the use of those 'user comments'? Do you believe it
> would change anything? So what is the purpose of asking more users from
> feedback *we do not want*?
> 
> Would it make you feel better if you were asked to raise your objection?
> Would you feel better claiming that we did it against objections of many
> users? Or do you believe we would abandon it and decide not to do
> anything because of the resulting bikeshed -- which seems to be
> a recurring theme lately?
> 
> Yes, I *do not want feedback* on *how to do Gentoo* from people who do
> *not help me do Gentoo* but instead only complain and demand.
> 


I can't let this one go.

Michał, Walter is a long-term high-quality Gentoo-user and does not
deserve to be told what you just said above.

Please get your act together and stop treating other people as worthless
nuisances.

You are annoyed by someone or something, that much has been obvious for
a long time now. I can assure you, Walter is not the correct target for
your annoyance. Please deal with the correct source.

-- 
Alan McKinnon
alan.mckinnon@gmail.com

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 2527 bytes --]

On śro, 2017-05-24 at 13:39 +0200, Alan McKinnon wrote:
> On 24/05/2017 12:24, Michał Górny wrote:
> > On śro, 2017-05-24 at 03:48 -0400, Walter Dnes wrote:
> > > On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
> > > > Next time such a thing happens, the discussion will happen on a
> > > > completely private media or not happen at all because of the state
> > > > of this mailing list. Is this what you really want?
> > > 
> > >   Here's the part you did not quote...
> > > 
> > > > > If we could have a guarantee of proposed changes like that being
> > > > > posted on Gentoo-User for comment, rather than being sprung on
> > > > > users by surprise, I'd be willing to sign off this list.
> > > 
> > >   Note where I said "...posted on Gentoo-User for comment...".  What I'm
> > > asking is for such proposed changes to be posted on Gentoo-User, and the
> > > discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
> > > surprise stuff seems to happen a lot in Open Source...
> > > 
> > > * Gentoo /usr
> > > * Firefox Australis UI, and dropping ALSA and going PulseAudio-only
> > > * GNOME getting a hard-coded dependancy on systemd
> > > * etc, etc
> > 
> > And what would be the use of those 'user comments'? Do you believe it
> > would change anything? So what is the purpose of asking more users from
> > feedback *we do not want*?
> > 
> > Would it make you feel better if you were asked to raise your objection?
> > Would you feel better claiming that we did it against objections of many
> > users? Or do you believe we would abandon it and decide not to do
> > anything because of the resulting bikeshed -- which seems to be
> > a recurring theme lately?
> > 
> > Yes, I *do not want feedback* on *how to do Gentoo* from people who do
> > *not help me do Gentoo* but instead only complain and demand.
> > 
> 
> 
> I can't let this one go.
> 
> Michał, Walter is a long-term high-quality Gentoo-user and does not
> deserve to be told what you just said above.
> 
> Please get your act together and stop treating other people as worthless
> nuisances.
> 
> You are annoyed by someone or something, that much has been obvious for
> a long time now. I can assure you, Walter is not the correct target for
> your annoyance. Please deal with the correct source.
> 

I'm sorry, I didn't mean that to apply specifically to Walter. It was
focused on the purpose of splitting the mailing lists.

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Dirkjan Ochtman]

On Tue, May 23, 2017 at 9:31 PM, Michał Górny <mgorny@gentoo.org> wrote:
> I'd like to request Infra to establish a new mailing list that would
> fill in the gap between our public mailing lists and the gentoo-core
> mailing list.

I agree with the others who've said that they don't think this is the
right solution. I've previously agreed we need moderation. I would
advocate that infra work on better moderation tools and/or mailing
list infrastructure that enables our use case.

(If we have problems with a lack of infra capacity making it
prohibitively expensive to move us forward, here's one solution that's
a little out there: just switch to hosted Discourse.)

Cheers,

Dirkjan

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Vincent-Xavier JUMEL]

[-- Attachment #1: Type: text/plain, Size: 890 bytes --]

Hi fellow Gentoo dev and user that read the list,

Le 23 mai à 21:31 Michał Górny a écrit
[…] 
> Initially, the posting group would include active Gentoo developers
> only. Afterwards, we will deploy a small moderator team whose purpose
> would be controlling access to the list -- including both adding new
> members on request and removing existing members (including developers)
> if they misbehave.
> 
[…]
This kind of idea is a terrible mistake ! It's exactly the kind of
measure that would prompt me to quit Gentoo as a user. The tone and
behavior of some members have been becoming digusting, even if those dev
do some real work on Gentoo.


-- 
Vincent-Xavier JUMEL Id: https://keybase.io/vincentxavier https://blog.thetys-retz.net

Société Libre, Logiciel Libre http://www.april.org/adherer
Parinux, logiciel libre à Paris : http://www.parinux.org

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Gregory Woodbury]

On Wed, May 24, 2017 at 9:01 AM, Dirkjan Ochtman <djc@gentoo.org> wrote:
>.....
> I agree with the others who've said that they don't think this is the
> right solution. I've previously agreed we need moderation. I would
> advocate that infra work on better moderation tools and/or mailing
> list infrastructure that enables our use case.

As I read the lists on Gentoo, I note that there are actually
only a few topics and/or contributors that make the lists
noisy.

A minimally robo-moderated list that initially filters non-
developer posters, with an ability for any developer to
clear or maintain moderation of a poster could work fairly
easily.  All developers would be initially clear of moderation
while others would need to "earn" a clearance of being
moderated, that is: new non-dev posters have to have
some developer look at the post and pass it to the list if it
seems appropriate, or they may reject it (with or without
comment.)

The robo-moderation would post a one-liner attention
message, and any cleared list member (or any of a
fairly large number of moderators) could engage the robot
off-list to examine and approve/reject the message, and
clear/maintain the posters moderation status.

Additionally, any of the chosen type of moderators (all
devs or selected moderators) could place non-dev
posters back on moderated status as necessary. If a ban
is needed, a proposal could be posted by the bot, and
any interested devs could vote to the bot (off-list) within
a given time period, and the plurality would determine
a ban or not.

Interactions with the bot would be off-list, and anyone
would get a status report from it as desired.

A full[er] specification of the moderation robot can be
developed rather quickly if desired.
-- 
G.Wolfe Woodbury
redwolfe@gmail.com

[gentoo-dev] Re: [RFC] Restricted version of gentoo-dev mailing list

[Nuno Silva]

On 2017-05-24, Michał Górny wrote:

> On śro, 2017-05-24 at 03:48 -0400, Walter Dnes wrote:
>> On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
[...]
>>   Note where I said "...posted on Gentoo-User for comment...".  What I'm
>> asking is for such proposed changes to be posted on Gentoo-User, and the
>> discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
>> surprise stuff seems to happen a lot in Open Source...
>> 
>> * Gentoo /usr
>> * Firefox Australis UI, and dropping ALSA and going PulseAudio-only
>> * GNOME getting a hard-coded dependancy on systemd
>> * etc, etc
>
> And what would be the use of those 'user comments'? Do you believe it
> would change anything? So what is the purpose of asking more users from
> feedback *we do not want*?

Is this the official policy of the Gentoo project?

-- 
Nuno Silva

Re: [gentoo-dev] Re: [RFC] Restricted version of gentoo-dev mailing list

[Rich Freeman]

On Wed, May 24, 2017 at 7:33 AM, Nuno Silva <nunojsilva@ist.utl.pt> wrote:
> On 2017-05-24, Michał Górny wrote:
>
>> On śro, 2017-05-24 at 03:48 -0400, Walter Dnes wrote:
>>> On Wed, May 24, 2017 at 08:41:25AM +0200, Micha?? Górny wrote
> [...]
>>>   Note where I said "...posted on Gentoo-User for comment...".  What I'm
>>> asking is for such proposed changes to be posted on Gentoo-User, and the
>>> discussion/feedback/flamefests/etc will be on Gentoo-User.  This type of
>>> surprise stuff seems to happen a lot in Open Source...
>>>
>>> * Gentoo /usr
>>> * Firefox Australis UI, and dropping ALSA and going PulseAudio-only
>>> * GNOME getting a hard-coded dependancy on systemd
>>> * etc, etc
>>
>> And what would be the use of those 'user comments'? Do you believe it
>> would change anything? So what is the purpose of asking more users from
>> feedback *we do not want*?
>
> Is this the official policy of the Gentoo project?
>

Unless you're reading it in an approved GLEP or some other official
summary from the Council or some other project team nothing anybody
says on this list really constitutes "official policy."

Michał put it bluntly, but basically made the same points I did in
fewer words.  The issue isn't WHO we get feedback from, but rather
what KIND of feedback we're getting.

Expressing preferences is sometimes but not often useful around here.
We've actually done forum polls and such when such preferences are
sought, but it doesn't happen often.  This isn't because we don't care
about user preference.  Rather, it is because the whole point of
Gentoo is that we don't make you express your preferences in some poll
and then you're stuck with whatever the majority votes for.  If you
have a preference you can stick it in your USE flags and you don't
have to justify it to anybody.  The whole point of Gentoo is to give
users choices.

When times come where it seems like a choice is being taken away (such
as the examples above), there is usually a driver behind it which we
find difficult to avoid.  If we don't support GNOME without systemd it
probably isn't because we're all systemd fanboys (that should be
pretty obvious).  Instead it is because the gnome team finds that with
their manpower they can't deliver the level of support to the project
while spending time working around the systemd issues.

Ultimately all this stuff is FOSS, so there is almost never a
technical issue that couldn't be worked around in theory.  The problem
is that manpower is limited and if all the people working on a project
would rather spend it on something other than fighting upstream then
that is often what ends up happening.

So, when these kinds of discussions happen it tends to be more about
searching for alternatives that maybe the team hasn't considered, or
trying to assess the impact of a change and how to make it in a way
that isn't too disruptive.  We circulate news items so that we can get
feedback so that when the change is communicated to users they have
clear instructions on how to cope.

So, if an outsider wants to point out that there is some way to
mitigate the impact of a change, or that there might be a way to avoid
the change that doesn't just involve endless patching/etc, then that
is helpful feedback to have.

On the other hand, when these kinds of changes come along there are
some who will just point out how bad the change is for them
personally, and how much work we're causing them.  While we're
sympathetic, in the end we're not employees.  If a volunteer not
spending six hours per month on patches causes your business to lose
500 hours of work, you can't really compare the two.  If there is some
way of making everybody happy I think most around here would be happy
to hear it.  However, just complaining about the impact of a proposed
change when it is clear that those proposing the change and making the
decisions already know that it will be impactful isn't helpful.

Gentoo is hardly the only community where this sort of situation
exists.  I think any non-commercial distro is going to face these
kinds of situations.  Those contributing are going to weigh the impact
of the change in terms of how much work it costs them and how much it
furthers their goals, and a lot of +1's from users aren't going to
have much impact except to the degree that they align with those
pre-existing factors.

Now, for a commercial distro (including semi-commercial ones like
Ubuntu) the situation is going to be different.  RedHat can afford to
pay developers to fork upstream projects if it comes down to it if it
makes their paying customers happy, because they actually have paying
customers.  The customers get to vote with their dollars, and those
dollars give RedHat leadership the power they need to incentivize
developers to build what will bring in those dollars.  There is no
reason this can't work on a smaller scale with Gentoo - I doubt there
is a bug or pull request sitting in limbo that would remain that way
if somebody offered a significant bounty for its resolution.  However,
absent that contributors are going to scratch their own itches whether
that is personal benefit or some kind of community-oriented goal.

-- 
Rich

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Walter Dnes]

On Wed, May 24, 2017 at 05:51:00AM -0400, Rich Freeman wrote

> What value would be obtained by posting this stuff for user comment?
> I'd also note that only one of those was posted on -dev-announce for
> comment as far as I'm aware.  Two are package/project-level changes
> which typically don't get wide discussion.

  I was using the Firefox PulseAudio event as another example of stuff
that happens in some obscure location that ordinary users don't know
about.  It was https://bugzilla.mozilla.org/show_bug.cgi?id=1247056
YES!!! It was done as a bug in the Firefox bugzilla.  How many regular
users religiously follow that?

  There were some people who "upgraded" Firefox one day, and discovered
that sound no longer worked on it.  It was done with so little
communication that several Firefox maintainers for various linux distros
(i.e. not exactly ordinary users) were unaware that they should set
PulseAudio as a hard dependancy for Firefox.

  An item I just remembered was the switchover from "/dev/hda" to
"/dev/sda".  One day I upgraded (the kernel?), and got a panic at the
next bootup.  Fortunately, I have a "Production" and an "Experimental"
kernel in my LILO menu, so I was able to reboot, and fall back to the
previous kernel.  A new kernel always goes into "Experimental".  I
"promote" the Experimental kernel to Production after it has run without
problems for a couple of weeks.

  Another one was when font packages stopped generating ISO8859-1 fonts
if you didn't add the "nls" USE flag.  No news item.  Xfreecell stopped
working because it couldn't find the font.

  The point I'm trying to make is that these kind of surprises should
not happen, be they OS-level or application-level.  Regular users should
not have to follow this list.  Even if you're not going to alter course
regardless of feedback, at least let users know so they're not caught
unawares.  If nothing else, post news items to Gentoo-user ***AHEAD OF
TIME*** warning about changes like these.

-- 
Walter Dnes <waltdnes@waltdnes.org>
I don't run "desktop environments"; I run useful applications

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 2381 bytes --]

On śro, 2017-05-24 at 11:54 -0400, Walter Dnes wrote:
> On Wed, May 24, 2017 at 05:51:00AM -0400, Rich Freeman wrote
> 
> > What value would be obtained by posting this stuff for user comment?
> > I'd also note that only one of those was posted on -dev-announce for
> > comment as far as I'm aware.  Two are package/project-level changes
> > which typically don't get wide discussion.
> 
>   I was using the Firefox PulseAudio event as another example of stuff
> that happens in some obscure location that ordinary users don't know
> about.  It was https://bugzilla.mozilla.org/show_bug.cgi?id=1247056
> YES!!! It was done as a bug in the Firefox bugzilla.  How many regular
> users religiously follow that?
> 
>   There were some people who "upgraded" Firefox one day, and discovered
> that sound no longer worked on it.  It was done with so little
> communication that several Firefox maintainers for various linux distros
> (i.e. not exactly ordinary users) were unaware that they should set
> PulseAudio as a hard dependancy for Firefox.
> 
>   An item I just remembered was the switchover from "/dev/hda" to
> "/dev/sda".  One day I upgraded (the kernel?), and got a panic at the
> next bootup.  Fortunately, I have a "Production" and an "Experimental"
> kernel in my LILO menu, so I was able to reboot, and fall back to the
> previous kernel.  A new kernel always goes into "Experimental".  I
> "promote" the Experimental kernel to Production after it has run without
> problems for a couple of weeks.
> 
>   Another one was when font packages stopped generating ISO8859-1 fonts
> if you didn't add the "nls" USE flag.  No news item.  Xfreecell stopped
> working because it couldn't find the font.
> 
>   The point I'm trying to make is that these kind of surprises should
> not happen, be they OS-level or application-level.  Regular users should
> not have to follow this list.  Even if you're not going to alter course
> regardless of feedback, at least let users know so they're not caught
> unawares.  If nothing else, post news items to Gentoo-user ***AHEAD OF
> TIME*** warning about changes like these.
> 

...and all of this is completely irrelevant to the topic. This kind of
information belongs in the news items, and for a few years now that's
where you are expected to find them.

-- 
Best regards,
Michał Górny

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Rich Freeman]

On Wed, May 24, 2017 at 9:09 AM, Michał Górny <mgorny@gentoo.org> wrote:
> On śro, 2017-05-24 at 11:54 -0400, Walter Dnes wrote:
>>
>>   I was using the Firefox PulseAudio event as another example of stuff
>> that happens in some obscure location that ordinary users don't know
>> about.  It was https://bugzilla.mozilla.org/show_bug.cgi?id=1247056
>> YES!!! It was done as a bug in the Firefox bugzilla.  How many regular
>> users religiously follow that?
>>
>
> ...and all of this is completely irrelevant to the topic. This kind of
> information belongs in the news items, and for a few years now that's
> where you are expected to find them.
>

Correct - this belongs in news.  Following any mailing list (let alone
being able to post on it) wouldn't have helped in this case.

Under-use of news items has been an issue historically for Gentoo,
though it has greatly improved.  In general devs should be thinking
about the impact of large changes and communicating it using news.
mgorny's suggestion in his new thread would probably also help in that
regard.

-- 
Rich

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Daniel Campbell]

[-- Attachment #1.1: Type: text/plain, Size: 781 bytes --]

On 05/23/2017 12:31 PM, Michał Górny wrote:
> [snip]
> Your comments?
> 

Since it's adding a list instead of warping an existing one, I say go
ahead on the condition that everything important finds its way to a more
open list. I'm subscribed to enough as it is.

I am skeptical that it will lead to an improved social experience among
Gentoo developers, but also willing to be proven wrong.

Personally, there's nothing that attracts me to the idea. I don't really
like the concept of curating a mailing list. But seeing as the point of
the list is to lessen the volume of messages, it will likely succeed.

~zlg
-- 
Daniel Campbell - Gentoo Developer
OpenPGP Key: 0x1EA055D6 @ hkp://keys.gnupg.net
fpr: AE03 9064 AE00 053C 270C  1DE4 6F7A 9091 1EA0 55D6


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Patrick Lauer]

On 05/24/2017 12:24 PM, Michał Górny wrote:

>
> Yes, I *do not want feedback* on *how to do Gentoo* from people who do
> *not help me do Gentoo* but instead only complain and demand.
>

But you do gentoo wrong, so as a user I'd like you to reconsider what 
you wrote there and maybe take a long vacation.

(Now of course this is exactly the kind of complaint you don't want to 
hear, so feel free to ignore it ...)

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Alexander Berntsen]

[-- Attachment #1.1: Type: text/plain, Size: 512 bytes --]

On 27/05/17 18:17, Patrick Lauer wrote:
> But you do gentoo wrong, so as a user I'd like you to reconsider what 
> you wrote there and maybe take a long vacation.
I too do not hate our users (in which I include myself).

Treating users as a worthless nuisance, unless they're writing ebuilds
or managing your precious GitHub crap, is a good way to lose users.

And even if I did hate our users, I agree with William & Ulrich.
-- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] Restricted version of gentoo-dev mailing list

[Roy Bamford]

[-- Attachment #1: Type: text/plain, Size: 761 bytes --]

On 2017.05.29 11:59, Alexander Berntsen wrote:
> On 27/05/17 18:17, Patrick Lauer wrote:
> > But you do gentoo wrong, so as a user I'd like you to reconsider
> what 
> > you wrote there and maybe take a long vacation.
> I too do not hate our users (in which I include myself).
> 
> Treating users as a worthless nuisance, unless they're writing ebuilds
> or managing your precious GitHub crap, is a good way to lose users.
> 
> And even if I did hate our users, I agree with William & Ulrich.
> -- 
> Alexander
> bernalex@gentoo.org
> https://secure.plaimi.net/~alexander
> 
> 

... and some of those lost users might have gone on to become devs.

-- 
Regards,

Roy Bamford
(Neddyseagoon) a member of
elections
gentoo-ops
forum-mods

[-- Attachment #2: Type: application/pgp-signature, Size: 833 bytes --]