Category Theory
Zulip Server
Archive

You're reading the public-facing archive of the Category Theory Zulip server.
To join the server you need an invite. Anybody can get an invite by contacting Matteo Capucci at name dot surname at gmail dot com.
For all things related to this archive refer to the same person.

Stream: community: general

Topic: fab thoughts

Fabrizio Genovese (Apr 04 2020 at 11:13):

Rongmin Lu said:

T Murrills said:

anyway I think Nicholas Scheel nonetheless has a point: mathematical terms are thought of, in general, as referring to abstract objects that could not in principle be stored in a computer.

That is soooo 20th century.

I don't agree at all with this.

Fabrizio Genovese (Apr 04 2020 at 11:15):

Nicholas Scheel said:

well, you raise an interesting point: programmers expect any arbitrary object (of a certain type) to be storable as data, which basically means its countable; physicists expect an object to actually exist in the universe, somewhere (possibly inside a black hole); and mathematicians almost always work with descriptions of objects (not encodings of objects), and a large part of the trick is relating different descriptions of the same object, which requires proofs of arbitrary complexity, not just computation

I think this is false. I don't know many physicist that believe Dirac deltas and plain waves exist in the universe. Still, they are object that they use almost all the time.

Fabrizio Genovese (Apr 04 2020 at 11:16):

sarahzrf said:

mathematicians dont all say yes

For me answering "no" to this question feels like punching mathematics in the face.

Fabrizio Genovese (Apr 04 2020 at 11:21):

Ok, I managed to get to the end of this conversation. I'd like to question the most basic assumption: I don't think programmers feedback loops are better than mathematicians. I mean, maybe they are when you think about programmers as "people trying to implement research in type theory", but that's really a small subset of what's out there.

Fabrizio Genovese (Apr 04 2020 at 11:22):

Usually people dig into your code and correct it only if they really need it. Github is full of repositories no one cares about. Also, the mole of code out there is really - really - humongous. If the "virtuous feedback loop" you are talking about would really exist, we would need many times the number of programmers we have to perpetuate it.

Fabrizio Genovese (Apr 04 2020 at 11:23):

What we have, instead, is that code is often - even in industry - barely tested, or tested inadequately, and the result is that you have to turn a Boeing plane off every 50 days or so otherwise it crashes (in the literal sense of crashing and exploding, it appears). :shrug:

Fabrizio Genovese (Apr 04 2020 at 11:25):

Also, dependent types and formal verification fix this up to a point. The typechecker will check that your code is formally correct, but not that it's conceptually correct, because it doesn't know anything about what you are trying to implement. It happened to us sometimes (even if very rarely I have to say) that we wrote some code in Idris that typechecked, but didn't work as expected. What happened in that case was that the conceptual premises we started from were actually wrong, and we ended up implementing a piece of code that did the wrong thing.

Fabrizio Genovese (Apr 04 2020 at 11:27):

All in all, I believe that the idea that "programmers have a way better feedback loop" is largely a mathematician delusion. Writing code is hard, writing good tests is hard and time consuming, and everyone has already too much on their plate to really care about deeply analyzing and peer reviewing other people's code. Exactly like in maths. :slight_smile: This idea works - more or less - only if you think about stuff such as "I wanna use a typechecker to check that a mathematical proof is correct", wichi is more or less what Voevodsky wanted to do with univalent foundation. But really that's just a teeny tiny bit of what programming is.

Fabrizio Genovese (Apr 04 2020 at 11:27):

(deleted)

(=_=) (Apr 04 2020 at 14:39):

Fabrizio Genovese said:

All in all, I believe that the idea that "programmers have a way better feedback loop" is largely a mathematician delusion.

It is not a mathematician's delusion. This is speaking from extensive contact with people who work in the industry (and NOT in startups).

Writing code is hard, writing good tests is hard and time consuming, and everyone has already too much on their plate to really care about deeply analyzing and peer reviewing other people's code. Exactly like in maths. :)

That is true only in startups and places where people believe in "move fast and break stuff". Seriously.

Testers exist. These positions are well advertised. I personally know a former academic whose main role in his corporate role now is to review code. These roles exist. People get paid to do this.

Compare that to academia, where all the emphasis is on churning out new "code", and the bare minimum lip service is paid to QA. Yes, it's exactly like maths, because maths behaves like the startup world!

This idea works - more or less - only if you think about stuff such as "I wanna use a typechecker to check that a mathematical proof is correct", wichi is more or less what Voevodsky wanted to do with univalent foundation. But really that's just a teeny tiny bit of what programming is.

See, maths does not EVEN have a type checker. See where the problem is?

Fabrizio Genovese (Apr 04 2020 at 14:41):

Does Boing look as a startup to you? Obviously code testers exist. Every role can exist if you pay someone to do it. That's a very, very small part of what coding is.

Fabrizio Genovese (Apr 04 2020 at 14:42):

And yes, peer review process in academia has a lot of problems because it nearly always happens on a volountary basis, so it amounts to ask people with already a lot of work to do to do even more work which, by the way, doesn't give you any form of credit.

(=_=) (Apr 04 2020 at 14:42):

Fabrizio Genovese said:

Does Boing look as a startup to you? Obviously code testers exist. Every role can exist if you pay someone to do it. That's a very, very small part of what coding is.

Nobody pays people to "test" proofs. There's this honour system that bars people from getting paid. That's a problem, because this "testing" gets farmed off to people who're ill-equipped to do rigorous testing.

Fabrizio Genovese (Apr 04 2020 at 14:43):

Nobody pays people to test proofs because, since at least 30 years, we realized that the system of economical incentives around academic publications is totally bonkers. I don't think that's the solution.

Fabrizio Genovese (Apr 04 2020 at 14:44):

The idea of using economic incentives to get better academic publication quality is exactly the sales pitch of Elsevier and Springer, and we know how it went.

(=_=) (Apr 04 2020 at 14:44):

Also, if you're going to dismiss every single thing I've mentioned as "a very, very small part of what coding is", then we can just wrap it up here. I described a system of things.

Fabrizio Genovese (Apr 04 2020 at 14:45):

I am saying that you are making very general statements about a community that includes a few million people, by deciding to selectively see just what makes a case for your point

(=_=) (Apr 04 2020 at 14:45):

Fabrizio Genovese said:

Nobody pays people to test proofs because, since at least 30 years, we realized that the system of economical incentives around academic publications is totally bonkers. I don't think that's the solution.

It's bonkers because you don't pay a group of independent testers. The current testers we have now are not independent. They have vested interests because they themselves write original proofs.

(=_=) (Apr 04 2020 at 14:46):

Fabrizio Genovese said:

I am saying that you are making very general statements about a community that includes a few million people, by deciding to selectively see just what makes a case for your point

I am selecting the case of corporates. Startups are a different animal altogether and your description applies.

Fabrizio Genovese (Apr 04 2020 at 14:46):

post-selecting*

Fabrizio Genovese (Apr 04 2020 at 14:47):

Rongmin Lu said:

Fabrizio Genovese said:

Nobody pays people to test proofs because, since at least 30 years, we realized that the system of economical incentives around academic publications is totally bonkers. I don't think that's the solution.

It's bonkers because you don't pay a group of independent testers. The current testers we have now are not independent. They have vested interests because they themselves write original proofs.

It is way more complicated than that. Take ACT. We are more or less 500 active researchers and we basically all know each other. There is no notion of "independence" in small research community, and a big chunk of research happens in a context where you can count on your fingers how many people on the planet can actually really understand the stuff you do research on

(=_=) (Apr 04 2020 at 14:47):

Fabrizio Genovese said:

post-selecting*

Keep snarking. Of course I was talking about best practices. Why would I talk about the gunslingers in startups when they do the same awful things as academia?

Fabrizio Genovese (Apr 04 2020 at 14:49):

There are best practices in mathematics too and, to be honest, I don't think that formal correctedness statements along with theorems would help us a big deal

(=_=) (Apr 04 2020 at 14:49):

Fabrizio Genovese said:

Rongmin Lu said:

It's bonkers because you don't pay a group of independent testers. The current testers we have now are not independent. They have vested interests because they themselves write original proofs.

It is way more complicated than that. Take ACT. We are more or less 500 active researchers and we basically all know each other. There is no notion of "independence" in small research community, and a big chunk of research happens in a context where you can count on your fingers how many people on the planet can actually really understand the stuff you do research on

That's a structural and institutional problem that requires a structural and institutional solution. I'm pointing out there's such a problem and that the industry has a solution.

Fabrizio Genovese (Apr 04 2020 at 14:49):

so then again, you can't keep switching $\forall$ and $\exists$ in your statements as you please, otherwise we won't reach any conclusion whatsoever

Fabrizio Genovese (Apr 04 2020 at 14:50):

I don't think industry has a solution, because, frankly, industry has no real incentive to optimize against code quality. They optimize against profit, which coincides with code-quality up to some point

Fabrizio Genovese (Apr 04 2020 at 14:50):

I can give you _a lot_ of examples for this, but just to name a few:

(=_=) (Apr 04 2020 at 14:51):

Fabrizio Genovese said:

I don't think industry has a solution, because, frankly, industry has no real incentive to optimize against code quality. They optimize against profit, which coincides with code-quality up to some point

Better than no optimisation, which is the case for academia.

(=_=) (Apr 04 2020 at 14:52):

Fabrizio Genovese said:

so then again, you can't keep switching $\forall$ and $\exists$ in your statements as you please, otherwise we won't reach any conclusion whatsoever

Did I use quantifiers in my statements? No.

Fabrizio Genovese (Apr 04 2020 at 14:52):

Boeing has serious bugs in their codebase. Microsoft as well. Apple too. They basically don't do anything until the problems become public. Then they decide to pay someone to do audits and fix them. Or zoom. They found out a super serious security breach in their code, notified them and they did _nothing_ for months. They started fixing the problem only when the hackers, having no choice, made the zero-day exploit public.

Fabrizio Genovese (Apr 04 2020 at 14:52):

Rongmin Lu said:

Fabrizio Genovese said:

I am saying that you are making very general statements about a community that includes a few million people, by deciding to selectively see just what makes a case for your point

I am selecting the case of corporates. Startups are a different animal altogether and your description applies.

You did here

Fabrizio Genovese (Apr 04 2020 at 14:52):

Rongmin Lu said:

Fabrizio Genovese said:

post-selecting*

Keep snarking. Of course I was talking about best practices. Why would I talk about the gunslingers in startups when they do the same awful things as academia?

and here

Fabrizio Genovese (Apr 04 2020 at 14:53):

"Of course i wasn't talking about the whole thing, i was saying that there is a subset of the thing where this holds." This is exactly switching a $\forall$ for a $\exists$

Fabrizio Genovese (Apr 04 2020 at 14:54):

Which I mean, ok. But then be precise from the beginning. And it's funny that I have to say this since I cannot stand Wittgenstein. people consistently quoting Wittgenstein out of context

(=_=) (Apr 04 2020 at 14:54):

Fabrizio Genovese said:

Rongmin Lu said:

Fabrizio Genovese said:

post-selecting*

Keep snarking. Of course I was talking about best practices. Why would I talk about the gunslingers in startups when they do the same awful things as academia?

and here

Look, I want to talk about best practices, alright. That was the whole point of me talking about it. What's the point of saying, oh startups also do the same as academia? The point was that academia needed improving, and here's what I understand of what systems there are in industry to do it.

(=_=) (Apr 04 2020 at 14:55):

Fabrizio Genovese said:

Which I mean, ok. But then be precise from the beginning. And it's funny that I have to say this since I cannot stand Wittgenstein. people consistently quoting Wittgenstein out of context

I'm not here to talk about philosophy. Seriously.

Fabrizio Genovese (Apr 04 2020 at 14:55):

But I don't think that academia will improve in any way if your best practice is "pay people to review" or "force people to typeset their proofs in HoTT"

Fabrizio Genovese (Apr 04 2020 at 14:57):

We brainstormed for some months about what's the best way to set up a peer review process, and it's no easy task. But of one thing i am sure, the right incentives are not economic, nor involve imposing a super formal framework for researchers to follow

(=_=) (Apr 04 2020 at 14:57):

Fabrizio Genovese said:

But I don't think that academia will improve in any way if your best practice is "pay people to review" or "force people to typeset their proofs in HoTT"

These two options are options that I am explicitly ruling out. And let's be clear that I understand "pay people to review" to mean "pay people who currently engage in research to review".

(=_=) (Apr 04 2020 at 14:58):

I was proposing an independent cadre of reviewers. They are explicitly barred from research and can only engage in review. This is similar to the separation of concerns between auditors and accountants in Europe, as I understand it.

Fabrizio Genovese (Apr 04 2020 at 15:00):

You started by making comparisons between programmers and mathematicians, saying that programmers have a virtuous feedback loop that mathematicians do not have. And I said that I don't think this is true, and that the case of serious code testing is actually a very small minority of what happens in the programming community. Now you say that you were talking about making review process better, which I understand, but that's a completely different thing which, imho, is totally unrelated to what happens in programming. We can talk about that, for sure

(=_=) (Apr 04 2020 at 15:01):

Fabrizio Genovese said:

We brainstormed for some months about what's the best way to set up a peer review process, and it's no easy task. But of one thing i am sure, the right incentives are not economic, nor involve imposing a super formal framework for researchers to follow

I would agree that the right incentives are not just economic, but the current economic incentives are out of whack. I'd want an incentive structure that aligns with the stated goal of producing research of good quality.

Fabrizio Genovese (Apr 04 2020 at 15:01):

Rongmin Lu said:

I was proposing an independent cadre of reviewers. They are explicitly barred from research and can only engage in review. This is similar to the separation of concerns between auditors and accountants in Europe, as I understand it.

I think this is a very unfeasible idea, for a lot of reasons. For instance, why should I decide to give up on my research to just do reviews? Do I get tenure track out of it?

(=_=) (Apr 04 2020 at 15:02):

Fabrizio Genovese said:

You started by making comparisons between programmers and mathematicians, saying that programmers have a virtuous feedback loop that mathematicians do not have.

They have a slightly better feedback loop. I think the other conversationists all agreed that feedback is slightly more immediate when you have a compiler and other tools programmers have that mathematicians do not.

Fabrizio Genovese (Apr 04 2020 at 15:02):

There is a lot of good quality research already out there. For instance, I don't think that "making publishing harder with a stricter peer review process" will make any good to the community

(=_=) (Apr 04 2020 at 15:03):

Fabrizio Genovese said:

Rongmin Lu said:

I was proposing an independent cadre of reviewers. They are explicitly barred from research and can only engage in review. This is similar to the separation of concerns between auditors and accountants in Europe, as I understand it.

I think this is a very unfeasible idea, for a lot of reasons. For instance, why should I decide to give up on my research to just do reviews? Do I get tenure track out of it?

Yes you should! That's the problem innit?!

(=_=) (Apr 04 2020 at 15:04):

Fabrizio Genovese said:

There is a lot of good quality research already out there. For instance, I don't think that "making publishing harder with a stricter peer review process" will make any good to the community

There is also a lot of dodgy research out there. And it's just been announced that some allegedly dodgy research is going to be published after all. See Morgan's contribution upstream.

Fabrizio Genovese (Apr 04 2020 at 15:04):

I think that the easiest - imperfect - solution would be to give reviewers credit. You leave the system as is, but reviews are public, they are added as an appendix to a paper, and the reviewer shares partial credit with the author

Fabrizio Genovese (Apr 04 2020 at 15:04):

This incentivates the reviewes to do a good job because they really are motivated to make good quality work and, moreover, by reviewing are taking somewhat a responsibility in saying to the readers "this seems good"

(=_=) (Apr 04 2020 at 15:04):

Fabrizio Genovese said:

I think that the easiest - imperfect - solution would be to give reviewers credit. You leave the system as is, but reviews are public, they are added as an appendix to a paper, and the reviewer shares partial credit with the author

There is still a conflict of interest. That's why you need separation of concerns.

Fabrizio Genovese (Apr 04 2020 at 15:05):

Conflict of interest is unavoidable in a community with very few people, where everyone knows everyone else

(=_=) (Apr 04 2020 at 15:06):

Fabrizio Genovese said:

Conflict of interest is unavoidable in a community with very few people, where everyone knows everyone else

I'm sure auditors and accountants know each other. The point is that you don't have job functions that conflict with each other.

Fabrizio Genovese (Apr 04 2020 at 15:06):

That's a way, way bigger market, that's my point

Fabrizio Genovese (Apr 04 2020 at 15:07):

there are thousands and thousands of accountants and auditors around. Also, the process is not as perfect as you may believe

Fabrizio Genovese (Apr 04 2020 at 15:07):

Look at the glorious relationship between the SEC and Wall Street

Fabrizio Genovese (Apr 04 2020 at 15:07):

SEC is - theoretically - an independent entity, but in practice...

(=_=) (Apr 04 2020 at 15:07):

Fabrizio Genovese said:

there are thousands and thousands of accountants and auditors around. Also, the process is not as perfect as you may believe

I'm not saying it's perfect. I'm saying it's better than nothing. And nothing is what academia has right now.

Fabrizio Genovese (Apr 04 2020 at 15:08):

Again, I do not agree with this. It is not true that accademia has "nothing". There are a lot of very good peer reviewed articles out there

(=_=) (Apr 04 2020 at 15:08):

Fabrizio Genovese said:

That's a way, way bigger market, that's my point

Just because a market is smaller doesn't mean it has to be dominated by cronyism.

Fabrizio Genovese (Apr 04 2020 at 15:08):

I myself have received, sometimes, insanely detailed and insightful reviews

Fabrizio Genovese (Apr 04 2020 at 15:08):

I don't think academia is dominated by cronysm, I am sorry

(=_=) (Apr 04 2020 at 15:08):

Fabrizio Genovese said:

There are a lot of very good peer reviewed articles out there

There are also a lot of very bad peer reviewed articles out there.

Fabrizio Genovese (Apr 04 2020 at 15:09):

Yes, and unsurprisingly, there is almost often a very strong agreement about which papers are these

(=_=) (Apr 04 2020 at 15:09):

Seriously, just stop. Your arguments are predictable and they have all been rehearsed and rebutted before.

(=_=) (Apr 04 2020 at 15:10):

Fabrizio Genovese said:

Yes, and unsurprisingly, there is almost often a very strong agreement about which papers are these

That knowledge is unpublished. Which is also a problem.

Fabrizio Genovese (Apr 04 2020 at 15:10):

It is not, again science is a social process. It's totally not "unaccessible" if you just hang aroud in your scientific crowd, which I suppose it's what you do if you do research in a given field

Fabrizio Genovese (Apr 04 2020 at 15:11):

Rongmin Lu said:

Seriously, just stop. Your arguments are predictable and they have all been rehearsed and rebutted before.

I find this offensive, honestly

(=_=) (Apr 04 2020 at 15:14):

Fabrizio Genovese said:

Rongmin Lu said:

Seriously, just stop. Your arguments are predictable and they have all been rehearsed and rebutted before.

I find this offensive, honestly

I find your arguments offensive, honestly.

Fabrizio Genovese (Apr 04 2020 at 15:14):

Rongmin Lu said:

Fabrizio Genovese said:

Rongmin Lu said:

Seriously, just stop. Your arguments are predictable and they have all been rehearsed and rebutted before.

I find this offensive, honestly

I find your arguments offensive, honestly.

Ok, @Christian Williams @Jules Hedges etc. I don't think it's right for me to moderate this conversation for obvious reasons, but I guess someone should do something. Cheers.

(=_=) (Apr 04 2020 at 15:15):

Fabrizio Genovese said:

It is not, again science is a social process. It's totally not "unaccessible" if you just hang aroud in your scientific crowd, which I suppose it's what you do if you do research in a given field

Folklore is a problem, because if you don't hang out with the right crowd, you miss certain information. But hey, it's great for gatekeeping.

(=_=) (Apr 04 2020 at 15:16):

Fabrizio Genovese said:

Rongmin Lu said:

Fabrizio Genovese said:

Rongmin Lu said:

Seriously, just stop. Your arguments are predictable and they have all been rehearsed and rebutted before.

I find this offensive, honestly

I find your arguments offensive, honestly.

Ok, Christian Williams Jules Hedges etc. I don't think it's right for me to moderate this conversation for obvious reasons, but I guess someone should do something. Cheers.

Good to know I'm being "moderated". Censorship rocks, eh?

(=_=) (Apr 04 2020 at 15:22):

Fabrizio Genovese said:

I myself have received, sometimes, insanely detailed and insightful reviews

I'm sure you have. But that does not rule out the existence of bad reviews.

Christian Williams (Apr 04 2020 at 16:43):

Hi, I don't think I want to weigh in on this. What I want to say is that I think some are getting a little too comfortable on this server. There are 800 people on here that can see everything you're saying. This is not a chat room; this is a place for mathematical discussion and research. Tangents are okay, but if they don't lead somewhere productive then let them go.

Christian Williams (Apr 04 2020 at 16:43):

I've not yet been nearly proactive enough yet to help the server -- we have a nice stream structure laid out waiting to be implemented, and there are guidelines we've recognized as necessary that we haven't yet made clear to the community. I apologize for being slow with this stuff. Part of the issue is being overwhelmed -- I think many members may have have shied away from the server because there is such a flood of messages. Online communication comes easier to some than others, and we need to respect this.

Christian Williams (Apr 04 2020 at 16:43):

Everybody who has embraced this and quickly found their voice on here, that's awesome. You're having great conversations, and it's exciting to see. I only ask: When you have a long discussion on here, please pause and make sure that it is benefiting everyone. Pause and think to write condensed paragraphs rather than stream-of-consciousness, and if you can try to limit the short "chatting" back and forth like in person. These kind of things will have to be incorporated into the guidelines, but most of all they'll have to be absorbed by each member as the best practices for everyone. Otherwise, we could lose members and this will just become dominated by those that talk the most. Just take it easy... we will likely have this server for a long time. It will become a great place as we learn to use it right.

Jules Hedges (Apr 04 2020 at 16:48):

I encourage everyone to become familiar with the muting options that Zulip provides. In particular, you can mute a single topic

Fabrizio Genovese (Apr 04 2020 at 16:48):

Yes, even if you can't mute users

Morgan Rogers (he/him) (Apr 04 2020 at 17:16):

Rongmin Lu said:

Fabrizio Genovese said:

Rongmin Lu said:

Seriously, just stop. Your arguments are predictable and they have all been rehearsed and rebutted before.

I find this offensive, honestly

I find your arguments offensive, honestly.

I feel that this argument should have ended well before this point, folks. :grimacing:

Morgan Rogers (he/him) (Apr 04 2020 at 17:22):

Christian Williams said:

When you have a long discussion on here, please pause and make sure that it is benefiting everyone. Pause and think to write condensed paragraphs rather than stream-of-consciousness, and if you can try to limit the short "chatting" back and forth like in person.

There's also a private chat functionality which I haven't personally road tested. When the argument or discussion is between so few people, perhaps that might be a more suitable place for it? (although I would rather we all get along :dove: )

Morgan Rogers (he/him) (Apr 04 2020 at 17:39):

Finally, my own two cents: I went to a public lecture by Eugenia Cheng in Edinburgh last year. One of the most exciting take-aways from that was a hope that the kind of work which is crucial to the academic world which currently goes largely unrewarded, might be granted more respect (in the context of this conversation, I'm talking about the process of reviewing papers, but there was a broad range of examples). The networking we're doing here is actually one of those things, even if you can't put it on your CVs! So be respectful; if you're feeling the heat, come back later.

sarahzrf (Apr 04 2020 at 20:32):

i havent read this whole thread, i just wanted to mention that when i said something about feedback loops, i meant the experience of programming as something that has immediate feedback

sarahzrf (Apr 04 2020 at 20:32):

like, you see output from your code, you see whether it has type errors, etc

sarahzrf (Apr 04 2020 at 20:32):

its not just all in your head

John Baez (Apr 04 2020 at 20:44):

I know, I hate that. :upside_down: I've managed to get my head working pretty well after all these years, but whenever I try to program, every ten years or so, I need to learn a new language, and I spend all my time making stupid syntax bugs.

John Baez (Apr 04 2020 at 20:44):

(Just thought I'd present the perspective of the mathematician who doesn't enjoy programming. In principle I have nothing against programming: I can see that it could be great, and some of my best friends do it.)

sarahzrf (Apr 04 2020 at 20:46):

hmm, i've just learned to cope with the fact that i can expect to not be able to do anything at first :)

John Baez (Apr 04 2020 at 20:49):

Right.

John Baez (Apr 04 2020 at 22:46):

Chill out, folks.

T Murrills (Apr 04 2020 at 23:14):

perhaps this should be in #meta, but just a couple thoughts:

• it is nice at times to feel able to chat freely and just “talk” as you would in real life, with short messages and stream-of-consciousness. However, as @Christian Williams said, that’s not necessarily conducive to actual discussions about, well, math content. If people other than myself also see the appeal of having spaces to chat, though, perhaps we could have a convention where those topics would be followed by “(chat)”, and all other streams would be assumed to be as @Christian Williams described.
• Emoji reactions provide useful features by showing public perception without joining the fray; maybe we could take advantage of that by having conventions here that “bystanders” to indicate when they think discussions are getting either too heated, or off topic, or anything else server-y (perhaps with different emojis for different meanings). For example, someone could react with 🌨 or ❄️ to messages like those at the tail end of the argument above, and this would indicate to the participants of the discussion that perhaps they should step back and “chill out”. Idk if mods can search by recent emoji reaction, but if they can, it could also help mods identify potential places where they should step in if things progress. So, point is: designating specific conventions of meaning (and thereby function) for emoji reactions in the CoC or somewhere might help the public collectively “self-moderate” to an extent.

Christian Williams (Apr 05 2020 at 04:54):

These are good ideas, thank you. It's fun to start thinking of the emojis as signals.

(=_=) (Apr 05 2020 at 07:33):

My apologies to Fabrizio Genovese. This whole discussion started with a grumpy misunderstanding that got grumpier as the night wore on. As Morgan Rogers has pointed out, it should have been ended a lot earlier, and I'm sorry it didn't.

Jules Hedges (Apr 05 2020 at 09:54):

Maybe it would be a good idea to move such conversations to #off-topic either voluntarily or by mod - even if they're technically not off topic

Jules Hedges (Apr 05 2020 at 09:56):

Alternatively we could have some kind of "chat" stream where you're free to talk as much as you like. The thing about #general is it's the one thing everyone's subscribed to

Fabrizio Genovese (Apr 05 2020 at 10:31):

Now that all misunderstandings have been cleared we can also delete these messages altogether