They do, I think, by recommending Mailgun. I wish there were an idiot’s guide to SES, as it’s 10x cheaper.
I also disagree with the comment above about Meta not being a good advertisement for Discourse. The first thing you look for is to see software in action.
Perhaps the whole VPS/“droplet” requirement is a turnoff for some people. I’d used a shared server for over a decade before understanding that an unmanaged VPS would be just as easy overall.
Perhaps the whole VPS/“droplet” requirement is a turnoff for some people. I’d used a shared server for over a decade before understanding that an unmanaged VPS would be just as easy overall.
This is kind of what I was getting at. In the old days you might have just added discourse through Cpanel, but now it’s more of a headache. Integration with DO as a marketplace app or something like that would lower the bar and make it like the (good) old days when you could run a forum on a shared server for peanuts.
But we don’t want to be bound to a single provider.
I prefer Scaleway, others prefer Hetzner, some Linode …
If you want this level of simplicity perhaps what you really want is a hosting provider?
This wasn’t what I personally wanted, it was originally a suggestion to make it easier to self host. I have actually since checked the docs again and they seem better than when I first setup, but it was definitely harder than some equivalent apps. Making this even easier to package up for people would help though in getting people started.
There have been several instances of people coming to Meta asking if there was a directory of communities, forums & companies specifically using Discourse. They were directed to Discourse.com which shows major companies & communities using this along with a note that some companies/communities do not wish to be “advertised.” I have a hosted, private forum and I fall into that category. 
So people are ready to discuss about cats, politics, dogs, bakery, boxing and bakery as long as the platform is Discourse? I have hard time to believe that.
But even i was looking for Discourse forums because I liked to see what it is about before spinning on a VPs.
The frequent insistence that the install/setup process and/or UI is not complex, arguing with/about the user’s experience, or in general the request for a person to justify their experience and opinion, are symptomatic of the very problems I’ve tried to articulate here before. This is one of the (few) reasons I have at times been turned off of this community, much to my dismay.
If someone who is a Discourse admin (or user) is taking their time to come here to tell the community “my experience setting up Discourse was more difficult than I think it should be”, listen to them, accept that their experience is valid first and foremost. Yes, it’s their personal opinion, and yes it’s reasonable to want specifics of exactly how it was difficult for them. But this should not be presented as a need for the person to justify their opinion and experience vs. that of others. Even if specifics can’t be or aren’t provided, the report itself, the user’s subjective experience is important.
It is the job of software developers to try to understand user’s pain points. It is not the job of the user to know exactly why they feel the way they do and communicate it to devs. This is one of the fundamental challenges of software development, and there are countless articles and anecdotes about its difficulty. Many well-respected tech industry figures (e.g. Steve Jobs) even believe that user’s don’t even know what they want, so asking them directly isn’t even necessarily a good approach. But even if it were, most users (even admins) are not UI/UX experts, so their ability to specifically identify and describe issues they encounter can be limited.
Even if a user does have expertise to help them make a more effective issue report, it’s also asking a lot of a person’s time and energy to detail the problems they encountered. This is why user studies are done, why they are useful, and why often the method is simply observing a person performing tasks, rather than trying to have them demonstrate some specific issue, much less describe it.
I think it’s also important to recognize that people’s prior experience with other similar-seeming (for a user/admin, non-coder) tools is relevant. Many people who are building communities now have had some experience spinning up PHP applications, perhaps even another forum like PHPBB. By comparison to that, the Discourse setup is relatively complex. There are of course aspects of both PHP and the PHPBB setup itself that are suboptimal, and reasons why e.g. the email setup that it uses is not preferable for Discourse, Rails, etc. But the fact that this isn’t really well explained anywhere (not just how to setup Discourse but why it’s harder than some other systems and what those trade-offs are about) is also part of what turns off potential users.
Lastly, in regards to ease of install, documentation, and whether self-host is not a “blessed path”, it’s worth looking at the changes to the Discourse.org home page in the last few years. Here it is back in 2021: Discourse.org on web.archive.org from 2021
Notice that reference to the self-install via Github is not only much higher up the page, but also includes a direct link, whereas on the current site there is no link where it first mentions Open Source, and the word “install” is never used. So for anyone who doesn’t necessarily already understand that a link to Github is a link to install (and that there are docs there to guide the install, which are mentioned nowhere else on the Discourse website), it’s a subtle but notable thing that will discourage them from investigating self-host.
The tension between free open source and profit for the entity that develops a given OS project is a fundamental one, so I understand these changes. But it’s fairly clear to me that OS/free/self-host is being deemphasized, and that to me is an indication of it being a less “blessed path” FWIW.
fair - I think that’s a great reminder.
so fire away!
Let’s be specific about which part of the install is hard and how we can improve those instructions or improve the workflow to build a new instance.
(I’m guessing that should probably be a separate topic in installation ?)
That said - and I’m shy to say 
that back in 2017 when I took over my first Discourse forum, I had zero linux server experience (although I had had some unix training and work experience back in the day so the commands weren’t alien to me).
However, all told, it was a significant learning journey for me, but one I relished. I put the onus on me to close any gaps in my knowledge and this was of course helped by the community on this very site.
To my mind, if you make it too “easy” you will fall over too easily when problems occur later (problems that are usually self-inflicted through bad choices as the software and servers are generally incredibly robust), because you didn’t learn enough of how and why (but then again issues are most times to be relished as it’s an opportunity to learn more stuff!).
It wasn’t an explicit decision to de-emphasize this for any sort of business purpose, we still maintain that any Discourse installation is a good thing, whether we’re paid directly or not. It’s quite possible we’ll add a link there again in response to this feedback.
Maybe, but to me this smacks of a (no doubt unintentional) elitism, i.e. “I did it the hard(er) way and I think it was good for me, therefore everyone should because it’ll be good for them.” As someone who spent decades in the PHP self-hosting space before finally tackling Discourse self-host I can tell you that self-host install and maintenance of most PHP tools is easier but also has useful learning moments where you sometimes have to troubleshoot or dive deeper into config and even code (just like Discourse at times).
In fact everything I learned in my time using SMF, Wordpress, and many other PHP tools absolutely prepared me for being confident and knowledgeable enough to muddle through a Discourse Docker install, and even still I sometimes find myself banging my head against an obscure error message in the console for an hour or so before I get my instance back up and running. But the point is that my entry into Discourse self-host was softened by what I learned in PHP self-host, and that learning curve is undoubtedly gentler than the Discourse one. That’s worth keeping in mind as CDCK and the community as a whole attempt to popularize and grow Discourse as a platform. In short: I’m glad setting up Discourse helped you learn skills that are useful and applicable elsewhere, but in my view it’s still valuable to try to make that process easier and easier for people (within reasonable limits).
That’s good to hear and I hope you do! I think some subtle adjustment could strike a better balance of outlining the literally turnkey hosted options while making it clear that Discourse is also free and (somewhat) easily self-hostable, and that install files and instructions can be found at Github. As another example of this, many other OS projects have direct links to documentation in their main navigation menu, which generally includes an installation section as part of the docs. The lack of this on Discourse.org (and the fact that “getting started” link makes zero mention of self-host as far as I saw, and just assumes you already have a Discourse instance) makes it all the more important to be more clear about how to self-host if desired.
Hmmm … not at all, more like a willingness to take responsibility for my own lack of knowledge and the need to improve it, whilst recognising, humbly, that I’m benefitting, for free, from other people’s hard work.
I think the difficulty or lack thereof in terms of installing Discourse is a bit of a red herring with respect to this topic. The reality is maybe 1 in 1,000 or 10,000 Discourse users have to learn how to install it. Plus there is the paid option, so there’s no obligation to learn anything to install a forum if the price is right.
When someone wants to find or build a community, the average user doesn’t think to themselves “which platform is ideal for communicating with my community of interest?” They just use what they know. That’s why in my hobby, 99%+ of the discussion and interactions happen on a combination of Facebook/Instagram/Reddit/Discord. It’s a bit sad to me too, because typically the interactions I see on these sites are just so low quality compared to the activity I get on my Discourse forum (I might be biased 
but I really do believe this).
It’s larger and (arguably) problematic trend of all the activity on the Internet consolidating on a handful of extremely large platforms. And I can’t really blame Discourse for losing this battle since it’s a competition where the other side has multiple levels of magnitude more resources and engineers. That being said, Discourse can always do better.
People default to what they already know and people don’t people don’t know Discourse. That’s because they either have never used Discourse or have used it without knowing they have. In my opinion, the number one metric that should be optimized to solve this is converting visitors into users. The advantage that a Discourse forum has to many other platforms is that it has natural search engine discoverability. For example, I did not notice how many Discourses I visited via Google searches until I set up my own. Now I notice it all the time. And to be honest, I’ve never joined one I’ve found via Google search before (I guess except for this meta one). What I’m getting at here is you need to maximize users actually using the platform consciously. The users that use the platform and like it will ultimately be the ones recommending it.
As for the specifics of what’s holding Discourse back from this potential? It’s a hard question. I agree with the post that bumped this thread that UI is a huge factor. The UI is complex because the software is so flexible and mature, which is great. But that flexibility can also be better used to tailor the UX to match the level of experience of the user, as was mentioned above. Or development of other on-boarding tools or incentive systems to encourage a visitor of the site to interact and join. There’s also the notion that forums are outdated platforms and for boomers (I hear this often). It’s hard to overcome this despite it being more perception than reality.
I just wanted to make a side-note since it was brought up. It’s entirely my own personal opinion but I just see the entire chat system as bloatware. It’s confusingly redundant with the direct message system and the last thing anyone needed was Yet Another Way to message someone directly. It is an inferior Slack and Discord client. It’s a shame because I agree with the idea that chat and forum are complementary. But no one is leaving Discord to chat on Discourse. It’s a missed opportunity because if Discord integration was a priority instead of replacement, you’d create an easy entry-point for people familiar with Discord [a huge amount of people] to be exposed to Discourse. A very tangential point but just something that came to mind when typing all this out.
Hmm, well as the person who started this topic I can say for my part at least that this is not correct; in fact it’s almost the exact opposite of what my original intent was. The first post and discussion thread on Twitter that I referenced was about how people wanting to set up new communities (i.e. looking for a community platform) do not consider Discourse as often or as seriously as I think they should given its capabilities and benefits. The later discussion of user experience, etc. is relevant to why admins might not decide to use Discourse for a new community (e.g. “it doesn’t look modern enough”, or “people think it’s too complex”), but so is the consideration of install/setup/admin experience. As a current admin of a few communities, and past admin of many more, these were definitely primary considerations in any that I had a hand in setting up.
Maybe not, but the exact opposite (people leaving Discourse for Discord) definitely is, and that’s also a potential problem for Discourse. I definitely think that developing integrated chat was in the best interest of the platform because chat as a whole is a major communications medium these days, and a competitor to Discourse’s core interaction method. Anyone who does value chat and some of the other things Discourse has going for it (notably open source and data ownership) would not have benefitted from a Discord integration (not open, no self host, no data ownership), so a chat function aligned with the discourse model and ethos again made sense IMHO.
This is a wonderful discussion. We certainly are well aware that the complexity of the user experience (both for site admins as well as for community members) is a challenge that we have in front of us (and have had for some time).
As has been mentioned, we have a few goals that are sometimes at odds with each other - we want Discourse to be highly configurable and customizable, and by doing so, we hand over significant control to the admins of each community in defining much of the user experience for their members. At the same time, we want it to be easy for newcomers to understand, but we also have found that experienced users want a fair bit of control over personalizing their own experience.
We have our work cut out for us, but we are putting a fair bit of focus on these problems. In the last release, we put some effort into the site setup experience in particular. For this coming release we have focus areas on improvements to the admin experience more generally, as well as on the extensibility of the platform.
The latter effort is an interesting one as it relates to this conversation. On the one hand, it could be said that it could lead to further increasing complexity, in enabling more customization. But it’s also a goal to enable us to define themes that are better tailored to specific use cases, or even a simpler default experience, while maintaining backward compatibility for existing sites. For example, it’d open the door to more seriously considering ideas like this one: CDCK should develop new default Discourse themes on a regular basis to keep it looking current
I don’t want to get too in the weeds about development process in this topic, but since it’s come up, I think it’s also worth mentioning that with our growth over the past couple of years, Discourse also now includes product managers and designers that are working alongside developers on all of these efforts.
As for how we position open source on the website, it is something we have been discussing. The site has been in a bit of flux as we’ve introduced the basic plan and have been trying to be more forthcoming about what we offer on our enterprise plan. There are a number of options here and we’re trying to lead people towards the one that is likely best for them. I agree it got a bit lost in the mix and deserves to be more discoverable. It is on our pricing page right now under one of the FAQs, but that itself may change as pricing alone is likely not the main reason to choose self hosting.
Lots of stuff to do!
Again, though, I’ve really enjoyed the discussion here and am happy to see how new life can be breathed into a discussion that was started more than a year ago. That’s one of the things I personally love the most about this platform.
This is why many creators use Facebook, Patreon, or Substack. It’s like five clicks to get started. And so many people use these platforms, it’s easy for new people to get started with them.
If Discourse wants Creators to use this platform for community building, I think <1% of them want to learn coding. They just want it to work, for their users to love it, and for it accelerate their growth.
What I mean to say is of course it’s always better to put the intstallation barrier to entry as low as possible. But I really don’t think that’s the issue at play here. I would argue that a person would recommend a platform based on their experience with it as a user more than anything. Or at the very least, there will be thousands of times more people that experience Discourse as a regular user than the people that run them.
Basically I’m advocating for a bottom-up approach. It’s a better goal to make the regular users fall in love with Discourse so that they advocate for the platform rather than trying to convince a handful of decision makers that their community will love this platform they never heard of and their users are unfamiliar with. If provided with sufficient motivation from their community, there are many people willing to put in the time and effort to build Discourse-based communities.
I also don’t mean to disparage the chat feature too much. I appreciate that it’s all free for me and a lot of work and time has been poured into it. I’m just not sure what problem it solves that I can’t easily [and nore effectively] solve with Discord. I can respect the argument about data ownership, if that’s a priority to someone. But the same logic can be used to argue in favor of cloning any and all software into Discourse.
It’s not going to stop people from migrating to Discord because it’s simply an inferior product. I actually have the public chats off on my site because it is redundant with our Discord. My mod team actually exclusively uses Discord to talk about moderation issues on the forum, despite us having a staff channel on the forum. I get that this is just my experience this is all purely my opinion so I can only really speak to my experience. I much rather see the energy go to building exciting features that make the forum experience better than trying to do what many other companies are already successfully doing but be worse at it.
Lots of great points in this discussion many of which I fear will be lost. There are I think several different threads here now. I wonder if it would be useful to start some new threads with the idea of clear focus
As noted, one would really want to hear from people who chose a platform having rejected Discourse, from users who left a community because they didn’t get on with Discourse, from people who chose Discourse for their community and regretted it.
As for Product Managers and bloat, and Conway’s law, there are in my view lessons to be learned. If you reward PMs for shipping features, or if you add features because some small cohort wants them, you get complexity and a baffling array of options. If you measure build times, or install times, or performance on minimal servers, you might keep the lid on those things. So, look after what you’re measuring or incentivising.
I think a showcase of what Discourse installations for half a dozen different purposes would be an excellent idea. Meta is not a showcase, and a single instance is not a showcase.
The problem of abrasive or dismissive comments on meta from founders or senior devs is perhaps one which is better now - hopefully lessons have been learnt. But it was a negative for me.
As a case study, I see the Vintage Computer Federation was previously a vbulletin site, went through a somewhat public process of choosing between XenForo and Discourse, with membership polls and feedback, and is now on XenForo. Here’s a dead experiment. (I wasn’t there when this happened and haven’t gone looking for the discussions from that time. But I happened upon on the dead experiment by searching for the “discourse”. If I search for “forum” I get the successful successor.)
I liked Coding Horror’s approach, even when I didn’t agree with him! It was usually to avoid the feature creep that is being complained about here, as it happens.
Interesting to see that XenForo has the same header flicker problem on my phone as Discourse does on the A2HS app (though not the Hub app).
Great to hear that you’re actively looking at this. I was reminded of some blog post about how very many ways there are to shut down a Windows computer. Might be this one. The next post gives a possible (organisational) reason.
Thank you for your thoughtful and insightful comments. Overall the reignited activity here and the response from the CDCK team has been heartening. I know how difficult it can be to address all these kinds of issues, but knowing you’re very aware of them and working to solve them as best you can is greatly appreciated.
And thanks as well to all the recent contributors to the thread! Lots of interesting discussion and perspectives. Ultimately I do think one of the hardest challenges is connecting with people who either didn’t choose Discourse even though they were aware of it or those who aren’t even aware of it. The insight of these “lost sales” (so to speak, since some could be self-host) is one of the most important things for any business to have, but hard to actually get. We’ve expressed a lot of potentially interesting ideas here which may help, but the fundamental challenge of actually soliciting those user/admin experiences remains the likely biggest bottleneck.
I am more hopeful for the future of Discourse than when I began this topic, in any case. (not to say I was despairing of it, just a bit concerned 
)