You could have 
the post. 
1 Like
Only if I agreed with the whole thing. 
2 Likes
Some notes on discuss.python.org:
- When I sent a message, I can immediately get a permanent URL to it. No need to wait until the python-dev archive got my email
- I like the scroll bar to quickly scroll into the discussion, it’s more efficient than the one in Gmail (web UI).
- I don’t see much differences between python-dev on Gmail and discuss.python.org: the UI is mostly the same
- I like “new” and “unread” links which are more efficient than Gmail to quickly see “fresh stuff”. “Unread” is more efficient than a regular mailing list, since it only lists threads where I contributed.
- I can very easily ignore a thread if I don’t care.
About features:
- I created a wiki to compare the 7 governance PEPs: Comparison of the 7 governance PEPs I like the fact that anyone can modify it. It’s much more convenient than having a mailing list on one side and https://wiki.python.org/ on the other side. Having both at the same place helps to keep everything consistent.
- Polls avoid the very common spam issue of “+1”. Like 5 to 10 people sending “+1” email with no rationale or anything, just to like “I like what you wrote”. These emails are pure annoyance to me. Discourse also allows to “like” a post which is even closer to “+1” emails.
- I think that I like the ability to modify my own posts to fix a typo, update an URL, etc. Anyway can see the history of my post, so I don’t see this feature as a threat. The only threat is when an admin removes a message: it disappears immediately, and there is no notice that “there was a message”. See How does a suppressed message or edit message look like? discussion. Many forums allow users to modify their own post, and I don’t recall that this feature has been abused on the forums that I used. The common case is more someone who enhance their post, than trying to hide evidence of a misbehavior or something like that.
Quotes are almost a major feature of Discourse. It’s quite easy to use. In Gmail, when I would like to comment 3 lines of a PEP which takes 10 pages, I have to edit the email to remove all contents except of 3 lines… It’s super annoying. Some people give up and leave all the previous content just to add one sentence, and some email clients fail to render this properly. Example: read https://mail.python.org/pipermail/python-dev/
The UI for quotes is quite nice, like the the ability to unfold a quote, to see the content, and link to the original message. In a giant discussion, this can help a lot.
Being able to display an URL with a title and not just the raw URL is more readable. I also like to be able to use bold and code formatting. It also helps for readability.
Aha, another random note: I found the secret “dark mode” (dark background) and I really like it 
Did I mention that I like the ability to put inline pictures?
1 Like
Did you try the “mailing list mode” to get messages as emails? Sorry, I didn’t try, so I have no idea how it looks. No idea if you get “everything” or not.
Mailman 3 does this.
1 Like
It’s a little hard to discover, but once you notice that you can just highlight some text and it will quote just that text, it’s a major revelation.
1 Like
Oh wow, me too!
2 Likes
Nobody forces them to do so. Thunderbird is free software and a download away (and there are alternatives too). Pointing to the limitations of one e-mail client as the reason why mailing-lists can’t be better than DIscourse sounds like a fallacy.
OTOH if I want a threaded view for Discourse topics then I’m pretty much screwed.
1 Like
As you noted above, Discourse discourages sprawling tree-like explosion of discussion by maintaining a (rough) top to bottom flow.
With some lightweight moderation (like Brett’s split of the vote date discussion from the big Python Governance Electoral System thread) you can achieve better threading in ways that mailing lists couldn’t properly handle before (off-topic discussions about e-mail clients on the Timeline to vote for a governance PEP python-committers megathread are a great example).
In other words, you can split off messages after the fact to provide a better linear flow for everybody to enjoy. In a way then this is more a matter of proactive moderation of hot topics, rather than an intrinsic tool deficiency.
To flip this around, I think that Brett is simply saying that the benefits of threaded views in e-mail only apply to a subset of e-mail users. On the other hand the problems stemming from immutable tree-like explosive threads are universal.
In general, the next generation of Python core contributors is increasingly less likely to be e-mail first, let alone Thunderbird-first.
1 Like
Whether you are part of that subset is entirely up to you. So I frankly don’t understand your argument.
(note: Thunderbird isn’t the only client that allows threaded display, it’s just the one I’m using as an example)
I don’t know what “e-mail first” means. What I know is that e-mail as a communication tool isn’t going away, and people will still be required to master e-mail in 10 or 20 years.
I remain to be convinced that it’s gonna work on, e.g., a PEP discussion, where usually the problem is not someone veering off-topic to talk about e-mail clients. 
I still stand by my example: as a moderator, can you try to cut Python Governance Electoral System in several independent Discourse topics? If you think this is too tedious or too much work, then you’re making my point.
Ahh, that’s the point. We are deciding whether or not to make the change. So I think there should be enough reasons to motivate us to change the status quo, but not just it’s not worse. Actually I don’t really care which one to use, but the current advantages of discourse are just not that strong to change my preference.
Ah yes, I was annoyed that “my” thread on the timeline have been hijacked by yet another thread about the voting method… At the end, I kind of ignored the whole thread and so I might have missed a reply about the timeline. It’s annoying me on mailing lists. But I didn’t see the “split the discussion” feature on Discourse in practice yet, so I cannot judge the feature 
(and then people complained “what? it’s already time to vote? nobody told me!” or 
)
1 Like
I… don’t think I agree? Or rather I don’t think I agree that people are going to be required to “master” email in the way people do now. Email is increasingly becoming a notification mechanism not a communication mechanism, while the actual communication is being shifted to other platforms that just notify via email. This isn’t really related to the tech world, just in general across all areas of interest.
TBH, I think that’s highly subjective. It’s certainly both in both my corporate and personal worlds.
The nature of online communication will always evolve, but I don’t think email is going anywhere any time soon. That said, while the right email client can certainly help with rambling, branching millithreads (and NNTP such as Gmane is so much better at managing that), these kinds of discussions are difficult to manage IRL so I’m not surprised that all technologies have trouble with it. It is kind of sad that email clients have actually gotten worse with web mail and mobile clients. Maybe if we make email clients horrible enough, we’ll finally kill it off.
How’s that for an off-topic, discussion branch point?!
I’ll think you find that in most professional settings, e-mail isn’t used significantly less than it was 10 years ago. People still have discussions and exchange documents over e-mail. They aren’t even trying to switch to something else.
That’s not my experience through several large companies. I mean sure they all still have email, and things get discussed there, but there’s also probably another 5 systems in wide use, with a 6th being currently developed
OK, let me put it another way. I have heard the argument made by you and others that a reason you prefer mailing lists is because you prefer a threaded view of discussions which Discourse does not provide but your email client does. That’s fine, but I don’t have that benefit based on my choice of email client, and so this desire for a threaded view isn’t a benefit of mailing lists to me. So when I hear an argument being made for a feature that I don’t have access to based on some other choice I made it feels like I’m being told I’m “doing it wrong” when it comes to mailing lists and that if I would just switch my email client to Thunderbird or any other threaded email client then mailing lists would be better for me. So I don’t think it’s a fallacy because I never said mailing lists can’t be better, I just don’t think they are better for most people(including me), and I don’t think people will see that benefit unless they are willing to change their email client.
But as I said earlier, I don’t think there is any way to quantify this into a right/wrong/better/worse answer where people are going to suddenly agree. This part of the discussion very much comes down to “I have a way I consume emails that has these benefits that Discourse doesn’t have” versus the opposite. After that it becomes whether you think Discourse is a better baseline experience compared to how the majority of people consume email and if you do think Discourse is better in that situation whether we choose based on what some people prefer based on their email client (which I personally think typically makes them email power users).
For me personally, I do think Discourse is better than something like Gmail and the various other email clients that I suspect a large majority of people use, and I think the amount of people in that situation is large enough that I think that’s the experience we should be optimizing for; people’s email client preferences transcending Python participation and thus I don’t think it’s something we should expect others to change to more effectively participate when we have Discourse as an option. As I said I don’t think this is necessarily more right than an opposing view, it’s just my view.
Sure, but the scale is different than what we deal with here in my experience. For instance, I typically don’t have email threads at work that span 100 responses nor involve so many individuals participating. In fact, people actively stop using email and switch to a meeting at my work any time an email thread reaches double-digit responses in a short amount of time.
So on one hand it’s not ok to expect people to change their e-mail client (which there is a large choice of), but on the other hand it’s ok to force people to switch to Discourse (which there is a single implementation of)? I’m not sure how that’s supposed to be a reasonable answer to the concerns with Discourse…