+1 this is a great idea.
I’m also using NNTP. It’s fast and simple. You can always download an existing list (not possible with forums and mailing lists).
Discourse is essentially a forum with mailing list functionality.
For users it means that email is getting used more for notifications or notifications have to go to a mobile app / chrome plugin.
I don’t think that Discourse will revive the forum/mailinglists/newsgroups; part of the support has moved to StackOverflow and Slack is still used a lot for conversations and support.
Anyway, I don’t expect any other outcome than that NNTP will be dropped and that Discourse will be used.
Thanks for your feedback.
While i agree that Discourse is mainly a forum with a bit of mailing list functionality, i don’t think that people will use if for notifications only. Actually, im running another forum (not related to TYPO3 but to VW campervans) where quite a fraction of people use e-mail only to read and reply. Most of those users are not very aware of computer stuff and just reply to notifications because they’re not able to distinguish between forum notifications and actual emails. It works fine for them, so i guess it will also work for us.
Its not so much about reviving the forum/mailinglists/newsgroups, but about removing the burden to keep those services running and up to date over the next years. As you said, a lot of former discussions moved over to StackOverflow or Slack, so i think it makes more sense to invest our energy in other things.
I absolutely agree about the fast and simple part. However, “simple” is already quite subjective, as i know a lot of people wo have not heard about NNTP, but who are perfectly at home in using web based services like Discourse. Its basically the same discussion as comparing Slack to IRC.
Regarding exporting lists: Maybe we should think about a global TYPO3 content archive, where we export content from all services, wherever they’re running. I think we had a IRC channel to HTML export once, this is something i miss in Slack, and certainly something we could expand to other services like a upcoming Discourse forum.
Since a couple of other software projects either moved to Discourse from an existing forum implementation as communication platform or started with Discourse right away, I’m quite used to Discourse. As for personal opinion, using Discourse and dropping NNTP, Mailing-Lists and the Forum could be the way to go. So +1 for that.
But the thing is: It’s important to actually drop all other means of communication and try to not spread communication. Of course for lazy visitors it’s a nice thing just dropping a question wherever they are. So “as much channels as possible” might increase the number of questions asked a little bit. But to not simply drop a question anywhere but having a distinct place to search for answers before raising the same questions over and over, having as less channels as possible is imho the better approach.
My point is: Please use it as a replacement, not as an addition.
What about the guys trying to install StackExchange? Are they aware of this thread here? Is the current effort to install StackExchange an approved thing of “the TYPO3 guys” or is just one part of the community moving over there?
The same question goes to the Discourse effort: Is it somehow “approved” or just the infra structure guys trying to minimize their load?
Don’t get me wrong, I’m pretty in favor of Discourse and I’d really like to use it. But we shouldn’t take the current situation of having a limited number of complains compared to the number of actively involved people in completely different discussions of all platforms as a “the majority likes the idea”. It’s more like “the majority doesn’t care”, which is a completely different thing.
And that’s basically what happened to the StackExchange guys: They came out on 2016-09-13, basically by informing the community about having started a thing. That’s 8 months ago and they are currently at 41% of the “commitment” required to start a Public Beta.
I’d really like to see Discourse start and take over the places of all current channels of written communication. And I’d be really sad about seeing it living alongside all other channels and just pulling contribution from them.
[quote=“goli, post:9, topic:190, full:true”]
My point is: Please use it as a replacement, not as an addition.[/quote]
Years ago this was probably an issue. Today Google finds the solution, no matter if it’s posted on a blog, some mailing list or at Stack Overflow…
[quote=“goli, post:9, topic:190, full:true”]
What about the guys trying to install StackExchange? Are they aware of this thread here? Is the current effort to install StackExchange an approved thing of “the TYPO3 guys” or is just one part of the community moving over there?[/quote]
One could also say: If Stack Overflow does the job, then we can shut down the mailing lists without providing any replacement at all… But maybe it turns out that there are good reasons for both to coexist.
However, I must also say that I don’t know about the long term plan for using Stack Overflow and/or a dedicated Stack Exchange project. Communication wise, I’d say this could have been done in a much better way. So far, only a link on https://typo3.org/support/ and a mention in an Association news (https://typo3.org/news/article/diary-of-the-typo3-association-july-august-2016/) mention it.
[quote=“goli, post:9, topic:190, full:true”]
The same question goes to the Discourse effort: Is it somehow “approved” or just the infra structure guys trying to minimize their load?[/quote]
Well, the proposal you’re reading here is the formal way to approval. Waiting for feedback…
I like the plan, and decisions.typo3.org was actually planned to see how it could work at a small scale.
Still, there are some technical remarks that I’d like to see defined / in-place when removing the old legacy functionality - some you already adressed:
- Migrate the old content please. If we do that, we could just shut down the old services without any grace period for having that data “found on google” still.
- Fix LDAP / Single-Sign-On with t3o accounts.
- Let’s also dig up the current communication channel lists and properly announce “Slack”, “StackOverflow” and “Discource” as platform everywhere (e.g. introduction package, t3o etc)
- Find a way for the announce mailing list replacement - maybe we could use a newsletter system for that instead, where people could subscribe on t3o directly. So it does not have to be connected to discourse at all in the first place.
Also, it would be great if we commonly decide who is responsible for maintaining the tool.
Overall, I give a 100% yes, let’s do it, let’s just sort out the details. Maybe people could jump in on a sprint to help you?
Thanks for the efforts again - Server team!
Btw there seems to be a NNTP plugin for discourse: https://github.com/sman591/discourse-nntp-bridge
Are you able to provide me more informations here? There is nothing open, or do you know an error that I do not know?
I didn’t try Discourse but looking at the features it looks like it could replace also Slack for people that ask/give assistance.
IMHO Slack actually is not used in the right way (except for #typo3-cms-coredev). It’s good for team chat but, since the information that flows are not organized in threads (now there are threads but nobody uses them), it’s becoming a huge source of information but really hard to search and repetition will be more and more common.
Maybe #typo3-cms should be set read-only and people searching for help redirected to Discourse.
[EDIT] I add another problem of Slack: when someone ask a question, if the one person that could answer is not there reading, it’s highly probable that he will never see that question.
If we want to move to discourse, please provide an own instance under a dedicated sub-domain, otherwise the decision part will get lost in traffic.
Thanks for sharing your thoughts.
By now, my assumption was that we just keep the mailman archives (already available as HTML only) at some place. As it is HTML only, we could keep this forever without any hassle. Furthermore, we have every group available in mbox format, which we will keep as well.
However, your remark got me thinking and i just saw that there is a mbox importer for Discourse here. I will give it a try, but i cannot promise you that we’re able to import just everything 1:1.
As already pointed out by @avalarion_avarres, we’re currently not aware of any issues related to LDAP. If there are any, please let us know, im sure we’re able to fix it somehow.
I absolutely agree with those points
Tacking a quick look, it seems that this plugin syncs messages through NNTP. This means that we would still ned to run and maintain a NNTP service. Personally, i don’t think this is worth it, anyone thinking otherwise please speak up
This is not yet decided. However, there is already another point taken for keeping them apart: In here, you’re allowed to vote only if you’ve reached a certain trust level, which you can reach by participating in discussions. If we use the same installation for any communiction, you could reach this trust level also within other categories.
Please decide and commit carefully.
Don’t get me wrong, I don’t want to bash on the efforts taken here. Creating any kind of communication channel and providing services to make it work is a really important thing and I’m glad you guys do that.
My point is: The means of communication used by the community are drifting apart and to my opinion, we lack a strong decision and commitment about what the official channels are.
Just recently I stumbled across this post here:
This basically proves my point: It names seven channels to communicate TYPO3 related topics. Only one of that seven is “mailing lists, forum and slack” (which I consider as the official ones till now) and the whole “social media” is a single point of those seven, which is a whole bunch of individual channels itself.
Instead of deciding for a technology to replace existing technologies as part of the infrastructure cleanup, my strong opinion about all this is to decide for a communications strategy in the first place that covers all current channels and brings them in a distinct order.
Distinct order means: We know about all of those channels, we name all of those channels and we encourage everyone to prefer a couple of them over others.
That can be done either by a “government decision”, like asking all of but only the core team members.
Or it can be done by the whole community, asking on all currently available channels (emails, newsgroup, slack, twitter, stackoverflow) and giving it enough time to reach all of our community.
But making a technical decision, driven on infra structure changes and the manpower available for that, which was only discussed by a hand full of people is likely to be come just another option among several others when it comes to finding a place to drop an arbitrary question about TYPO3.
About integrating existing information
Whatever the platform will be, it should be a real successor of the current implementation, in all possible ways.
This include especially existing data.
Starting with a fresh installation and no content creates a feeling of something completely new that takes time to evolve.
We should avoid that feeling by filling it with all the existing threads and creating something that is mature right from the beginning.
We’re not creating a new channel, we’re replacing an existing one. With this in mind, I have little understanding for your problem. In fact, I understand that you’re unhappy with the spreading of information over various channels, like Stack Overflow, Slack or mailing lists. I agree that we can improve that by telling users where we expect information to be found.
However, besides this, our control over the Internet is quite limited. There are several TYPO3 forums which are maintained by 3rd parties, and we cannot forbid anyone to use Stack Overflow, Twitter etc. We can only decide if it should be recommended or not.
I invite you to draft such a strategy and mention the advantages you see by it. Please send it to the board of the TYPO3 association, because I assume that’s where the responsibility of such a concept would be.
Meanwhile, let’s decide if the lack of a concept is a blocker for continuing here, or not…
I am in favor of using modern tools for the communication channels.
I feel that the preferred channel to ask help for is now Slack - it is a lot since I answered or even looked at the mailing lists - because it is simple to get and offer help (too much simple?) so
my fear is that I don’t know if it will be possible to revert this “habit” and encourage people to use e.g. Discourse, instead.
I tend to agree with the remarks made by @goli
My clear expectations for a communication strategy has 3 tiers:
- casual chatting (Slack, etc…)
- support (Stackoverflow, Discourse, …)
- bug hunting (Redmine aka forge, JIRA, …)
All those products are perfectly made for their purpose.
Consequently we can shut down any forum or mailing list and make them a static lookup archive for Mr. Google.
For support Slack for me is clearly just a “first touch point”. So basically, after asking the question the answer usually should be to create either a stackoverflow question or a forge entry.
There might be only one tough point which would speak against externals services and thats the moment where those are replaced by a newer service. What happens to the TYPO3 knowledge in there?
This would indeed be a strong point for the self-hosted Discourse service instead of using Stackoverflow.
Since I do not expect Slack to be replaced with any other causal chat solution in the near future, I count that one as set. Bug tracking in Redmine works sufficiently well, so that one’s set too.
The remaining question therefore: Discourse vs Stackoverflow.
My personal tendency would indeed go into the Discourse direction as for me the chances to eventually get a dedicated stackoverflow channel for TYPO3 seem to diminish day by day due to several reasons.
Regarding that almost OT:
I really don’t get, why so many people want a dedicated StackExchange sub-site. The idea of moving there was a increased visibility, wasn’t it? Hiding in an own place on a sub-site is again a step back.
Folks will open issues directly on SO anyways. For the few communities that have their dedicated sub-site, you can see the same effect there. This just splits questions and answers once again. Sometimes, people will then vote for closing to migrate it to another site.
So this makes really no sense to me. If you (we as “programming” community) want to use SO, stay on the main site.
Is it hiding? Maybe I got that totally wrong then. I thought that still is somehow directly connected to the main SO site.
Then we should stick to the main SO of course.
When you look at the list of people proposing the dedicated TYPO3 site, most of them have <100 points reputation, IMHO a clear sign that they have no clue what they’re voting for and how StackOverflow works (in detail). Might sound harsh, sorry, but it’s IMHO really a stupid idea.
I completly agree.
I think the best way currently to try to place some quality answers on the web, if someone references the TYPO3 documentation with a small comment in StackOverflow. That is straightforward, and navigates the people to the documentation. There are a lot of example, a lot of tutorial and it is perfect to learn from it.
I understand the concerns about Slack, but I think this is just a different level of communication. You have a problem, someone tries to help, you discuss it, there is really a dialog between two or more people. Just like asking your collegue next to you.
Mailinglist was a main tool to directly notify peoples with the same intrest and maybe start a conversation about a problem. (My personal opinion: the worst thing when I find such a years old mailing thread in google and had to keep reading and paging and reading and paging… Nowdays I simply ignore them and pray not to get any of those in the first 5 results)
Forums could work just like StackOverflow with a discuss feature on it. So you can ask your question, someone can ask back, you can quotate his answer and extend your question, and so on. This results again a long thread to read and cost time to evaluate if it is relevant at all or a completly different usecase.
The main point, that it provides a quick way to evaluate a question and its accepted answer. (And the way between the solution and the question is totally irrelevant)
If the goal is to discuss something quickly, active, Slack perfect for this task.
And if something really should be discussed, maybe not quickly, maybe not so active like on a chat, because people needs opinions, this site is perfect. But these topics are not really in the “I have a problem/ support” topic.
I personally don’t see, in which way could a forum increase the current quality of the given answers or the connection to the community.
But to the original question, the idea of getting rid of these channels is a good idea. I would think about if a Discourse necessary or not.
Would be a nice experiment to shut down all of them, and see what happens.