llvm dev - Nov 2019 - [cfe-dev] RFC: Moving toward Discord and Discourse for LLVM's discussions

If I am not mistaken, there are two things that are becoming clear:
1. For email, nobody seems to be against Discourse as long as the mailing
lists are still a supported way to participate. So this seems
non-controversial.
2. For IRC, people seem to be happy with switching to a more modern
solution, but Discord is largely disliked by a significant portion of
respondents.

So perhaps we can focus the discussion on "if not Discord, what else?"

Slack appears to be problematic due to lack of moderation capabilities.
Although I don't understand that, I think it is fine - does not meet a key
goal so we can't consider it.

This Matrix thing was brought up by some as a possibly viable way forward.
Can we look into whether it meets all the goals?

Perhaps a good start would be to list the goals. So far it seems like:
- moderation capabilities
- no terms of service that give the provider ownership of content for all
eternity
- IRC integration
- preferably open source and standard protocols
- free?

And probably a bunch of other goals.

On Wed., Nov. 20, 2019, 5:27 p.m. Eric Christopher via llvm-dev, <
llvm-dev at lists.llvm.org> wrote:
> There is absolutely no reason why that would happen.
>
> On Wed, Nov 20, 2019, 2:20 PM Renato Golin via llvm-dev <
> llvm-dev at lists.llvm.org> wrote:
>
>> On Wed, 20 Nov 2019 at 21:36, Whisperity via llvm-dev
>> <llvm-dev at lists.llvm.org> wrote:
>> > there will be the other 70+ gaming-related servers constantly
nagging
>> for attention...
>>
>> I very literally cannot cope with that at all. I can't "just
filter"
>> the noise on my own, so, if that becomes the norm, I'll be forced
to
>> pass.
>> _______________________________________________
>> LLVM Developers mailing list
>> llvm-dev at lists.llvm.org
>> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20191120/6fdfbe41/attachment.html>

Hi,

On Thu, 21 Nov 2019, 10:23 Nemanja Ivanovic via cfe-dev, <
cfe-dev at lists.llvm.org> wrote:
>
> 1. For email, nobody seems to be against Discourse as long as the mailing
> lists are still a supported way to participate. So this seems
> non-controversial.
>
Does this mean that discussions will end up split into two places (Email
and Discourse) ?

Or, will Discourse push forum messages to the mailing list, and pull
replies from the mailing list to the forum ?

Cheers,
.Andrea
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20191121/48eb5f02/attachment-0001.html>

On Wed, Nov 20, 2019 at 4:52 PM Nemanja Ivanovic via cfe-dev <
cfe-dev at lists.llvm.org> wrote:
> If I am not mistaken, there are two things that are becoming clear:
> 1. For email, nobody seems to be against Discourse as long as the mailing
> lists are still a supported way to participate. So this seems
> non-controversial.
> 2. For IRC, people seem to be happy with switching to a more modern
> solution, but Discord is largely disliked by a significant portion of
> respondents.
>
> So perhaps we can focus the discussion on "if not Discord, what
else?"
>
> Slack appears to be problematic due to lack of moderation capabilities.
> Although I don't understand that, I think it is fine - does not meet a
key
> goal so we can't consider it.
>
> This Matrix thing was brought up by some as a possibly viable way forward.
> Can we look into whether it meets all the goals?
>
> Perhaps a good start would be to list the goals. So far it seems like:
> - moderation capabilities
> - no terms of service that give the provider ownership of content for all
> eternity
> - IRC integration
> - preferably open source and standard protocols
> - free?
>
I’m still not totally sold on needing extensive moderation capabilities.
As mentioned earlier, Chromium — an open source project with more
developers than LLVM — has a code of conduct similar to LLVMs and manages
to get by with a Slack server while still maintaining their code of
conduct.  It’s possible we’re fundamentally different than Chromium in some
way, but I’d like to understand what those are before we decide it’s
impossible to have a professional and welcoming environment, because there
seems to be an existence proof to the contrary.

IRC integration, as far as i can tell, is an explicit *non* goal.

FWIW, imo the best way to be welcoming to be new people and/or outsiders is
to use tools that they probably already have some exposure to.  Being open
source is a nice-to-have, but I think it’s a mistake to weigh that heavily
in comparison to usability, familiarity, and feature set
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20191120/b8b68936/attachment.html>

On Wed, Nov 20, 2019 at 6:04 PM Andrea Bocci via cfe-dev <
cfe-dev at lists.llvm.org> wrote:
> Hi,
>
> On Thu, 21 Nov 2019, 10:23 Nemanja Ivanovic via cfe-dev, <
> cfe-dev at lists.llvm.org> wrote:
>
>>
>> 1. For email, nobody seems to be against Discourse as long as the
mailing
>> lists are still a supported way to participate. So this seems
>> non-controversial.
>>
>
> Does this mean that discussions will end up split into two places (Email
> and Discourse) ?
>
> Or, will Discourse push forum messages to the mailing list, and pull
> replies from the mailing list to the forum ?
>
This is how Discourse works. Users have to enable "mailing-list mode"
in
their profile and then they get a direct email update and can answer on the
emails, their answer gets added to the thread on Discourse. I tried inline
answers and it seemed to work well.

-- 
Mehdi


>
> Cheers,
> .Andrea
>
>> _______________________________________________
> cfe-dev mailing list
> cfe-dev at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20191120/bba59694/attachment.html>

On Thu, Nov 21, 2019 at 7:11 AM Zachary Turner via cfe-dev
<cfe-dev at lists.llvm.org> wrote:
>
>
>
> On Wed, Nov 20, 2019 at 4:52 PM Nemanja Ivanovic via cfe-dev <cfe-dev at
lists.llvm.org> wrote:
>>
>> If I am not mistaken, there are two things that are becoming clear:
>> 1. For email, nobody seems to be against Discourse as long as the
mailing lists are still a supported way to participate. So this seems
non-controversial.
>> 2. For IRC, people seem to be happy with switching to a more modern
solution, but Discord is largely disliked by a significant portion of
respondents.
>>
>> So perhaps we can focus the discussion on "if not Discord, what
else?"
>>
>> Slack appears to be problematic due to lack of moderation capabilities.
Although I don't understand that, I think it is fine - does not meet a key
goal so we can't consider it.
>>
>> This Matrix thing was brought up by some as a possibly viable way
forward. Can we look into whether it meets all the goals?
>>
>> Perhaps a good start would be to list the goals. So far it seems like:
>> - moderation capabilities
>> - no terms of service that give the provider ownership of content for
all eternity
>> - IRC integration
>> - preferably open source and standard protocols
>> - free?
>
>
> I’m still not totally sold on needing extensive moderation capabilities. 
As mentioned earlier, Chromium — an open source project with more developers
than LLVM — has a code of conduct similar to LLVMs and manages to get by with a
Slack server while still maintaining their code of conduct.  It’s possible we’re
fundamentally different than Chromium in some way, but I’d like to understand
what those are before we decide it’s impossible to have a professional and
welcoming environment, because there seems to be an existence proof to the
contrary.
>
> IRC integration, as far as i can tell, is an explicit *non* goal.
Pardon me if i'm mistaken, but having a baseline non-goal of
explicitly not supporting something that is the current status quo
seems to me like the opposite of being inclusive, but more like
being "let's just shake/change things up and force everyone else
to adjust to the new reality."
> FWIW, imo the best way to be welcoming to be new people and/or outsiders is
to use tools that they probably already have some exposure to.  Being open
source is a nice-to-have, but I think it’s a mistake to weigh that heavily in
comparison to usability, familiarity, and feature set
Roman
> _______________________________________________
> cfe-dev mailing list
> cfe-dev at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev

On Thu, 21 Nov 2019 at 04:11, Zachary Turner <zturner at roblox.com>
wrote:
> I’m still not totally sold on needing extensive moderation capabilities. 
As mentioned earlier, Chromium — an open source project with more developers
than LLVM — has a code of conduct similar to LLVMs and manages to get by with a
Slack server while still maintaining their code of conduct.  It’s possible we’re
fundamentally different than Chromium in some way, but I’d like to understand
what those are before we decide it’s impossible to have a professional and
welcoming environment, because there seems to be an existence proof to the
contrary.
There's a truth to it. The problems we had in Bugzilla was spammers,
not "unreasonable people". Back in 2008, I heard stories that a couple
of people got banned from the list/IRC, but nothing since.

But forums, being a web platform (like Bugzilla), make it more
vulnerable to spammers and trolls, which may need more moderation.

We need to be aware of the functionality available, but I agree, I
wouldn't use lack of good moderation as a strong reason not to use a
particular tool.

> FWIW, imo the best way to be welcoming to be new people and/or outsiders is
to use tools that they probably already have some exposure to.  Being open
source is a nice-to-have, but I think it’s a mistake to weigh that heavily in
comparison to usability, familiarity, and feature set
If we went with popularity, we'd choose Facebook or WhatsApp. That
can't be the most important criteria. We really need to think about
accessibility, diversity, privacy and security.

With open standards and platforms, in the worst case scenario, we can
modify ourselves to fix whatever is broken and make it as
secure/accessible as we need it to be. We've done some of it with
Phabricator, for example.

We also need to make sure our data (history, threads, control) belong
to us. It would be a deal breaker if we couldn't download a dump of
our mail history, especially for llvm-dev/cfe-dev, in a format that we
can actually consume with other tools.

Finally, we really cannot force developers to sign troubling terms and
conditions or user agreements that they're not comfortable (or able)
to. Unfortunatelly, the world of closed source software nowadays is a
minefield of T&Cs and EULAs and it's almost impossible to actually
know what's going on until after you lost control of your
data/access/rights.

--renato