I don’t know anything about the competitors you mention, but one thing I noticed in getting setup with discourse is that it works best and is easiest to show other people once there is quite a bit of content in a site. That makes onboarding hard because someone like me who knows the potential benefit or a ‘real’ community forum has to do more to show others what it’s like.
In trying to persuade my colleagues to use discourse I struggled to find other communities that were like the one we wanted to build. That partly because our use is mostly internal, but also because the existing users are quite niche. The best example I had come across before is the Stan forums but that was too technical to share with colleagues. Providing better examples of discourse in use would help (not just links… actually review them and give deep links to nice threads which exemplify why discourse is great).
Another option might be to try and poach some really big mass-market communities and offer them cheap deals as a loss leader. I almost bought the stack exchange software, for example, because so many of us were familiar with it (although qa wasn’t as good a fit as a format for all our content).
Note, meta.discourse is not a great advert for you for anyone other than open source/community nerds (for whom it’s very appealing). It’s just too detailed and nitty gritty to be attractive to normies.
It’s not the product itself though - discourse is great!
meta is not meant primarily for this purpose, it’s a product support site - that’s why there is https://www.discourse.org/ which is the “product” site.
Though I think we could do better with try.discourse.org in that regard. I already pulled out my ‘WIP’ card a few minutes ago on another topic about docs (and using it twice in quick succession brings on the stress headaches ) but I think having an interactive ‘explainer’ site with examples of lots of different features in place could really help with this.
I’m definitely interested in the potential of discourse and aware of mighty - a platform on which I’ve administrated a group, circle, haaartland and others .
For whatever reasons I’ve struggled to find all the information that I think would help trade off decisions - not just discourse but also the others
JD you’ll know that I strongly embrace Conway’s law.
Discourse exhibits manifestations of it. For example coding horrors aversion to signed posts which manifests itself in the default guides in other forums and here, The messages about sending multiple replies in a topic and not aggregating replies into one post…
On a deeper level the philosophical or cultural alignment of the dev team determines the software. For example mighty put themselves at the top of a tree where all their communities are then subs and the mighty app allows a user to search for communities of relevance / interest.
Be aware too that most people don’t have equal knowledge of multiple platforms to make a properly informed selection so technical considerations are ameliorated by the advertising and most have a ‘shiny website’, many have options that you can turn on and off without having to install and rebuild software so discourses move towards themes and things components will appeal to a developer techie software admin but not to a commercial pay the fee start using it organisation or someone who wants to gate a community behind a paywall with full PayPal stripe global taxes etc integration…
It’s not just horses for courses although that is crucial but not sufficient
right - but the first thing you want to do when evaluating something like this or demoing the possibilities to someone is to see it in use… not read a sales blurb. (at least that’s what I want). I’m not saying meta should be that, just that it doesn’t fill that niche.
Yes, I get that. It’s perhaps in part a bi-product from not being a massively popular centralised platform like whatsapp or Facebook, where you can have significant personal experience of using it already. But as @JammyDodger there is a demo site …
But the demo site makes it look like a hobby project … something like phpbb was. You can see the “features” It doesn’t show the value that all these features create.
I wonder if trying to poach a few active Reddits or FB groups on widely-accessible topics might work to give some nicer examples of how discourse enhances a community? it wouldn’t cost much to try and reach out to a couple of key mods on big fora that would benefit.
For ref, the best example I found to show colleagues was drownedinsound — but that’s perhaps only because my boss and a few other people in the dept are into new music. Perhaps target/get examples of other wide-appeal topics like parenting/sports/film?
This is my experience too. Also, many customisations of discourse are (IMO) pretty bad and the css and other layout changes people try and make actually make the experience worse both for their users and to use the sites as an example.
It get that people want to brand things, but there is definitely a tension here… most people don’t have the skills of expertise to actually improve things. The default discourse UI is good — especially for power users.
Yeah, for many it’s hobby and a learning journey - something we should also not discount.
The amount of IT knowledge people are picking up by deploying their own Discourse site is very significant and we should certainly highlight this as an extremely positive thing.
I would say it’s too much. Although there is a docker install, configuration could be a lot easier. I’m pretty experienced doing devops for various side projects, but my experience self hosting led me to think that it was being actively disincentivised. Setting up email was hard and required faffing with config files. Ditto for plugins. All this stuff could be in the admin UI. Note - I’m not complaining, because I understand if you want to keep some berries to persuade people to use hosted. It’s just that a non-slick experience doing self hosting made me worry that this wasn’t a ‘blessed’ option and that it might go away.
Would it be possible to make any $ doing referrals to a vps provider that provides an smpt-as-as-service and could setup a 1-click installer for people who don’t want hosted? That would significantly lower the bar to entry, but I guess it might cannibalise hosted sales.
I don’t think this is part of CDCK’s strategy at all.
On the contrary, I see them bending over backwards to help (together with the community).
The benefit to CDCK is the amount of eyes-on testing it provides and the additional contribution, both in code and in feature requests.
What part is ‘hard’?
I think there is a limit to the minimum required commitment and effort someone is going to need to host anything on their own platform. Also if everything is just a button press, how the heck are they going to support their own website? Clearly CDCK has to separate responsibility for that. It is reasonable to expect that someone has or is willing to gain the required knowledge for long term support either by personal involvement or the direct involvement of other SMEs via contract.
Once you’ve done it a few times, setting up a completely new, 100% independent, fully owned, site in “default” spec takes under 20 minutes - and that includes buying the domain! That’s phenomenal, no?
I personally feel your point about “surfacing examples” of the platform is a fairer comment.
I recall someone on meta who used to search something like the “This is a Civilized Place for Public Discussion” sentence as it’s a default, public string of any Discourse forum. It would work with others, and you can also search for other languages.
I think we need to be careful about what are Discourse issues and what Discourse can’t really solve. Email config is a faff, but from many years of experience of running fora that send emails I know that the biggest issue is with email and not the forum software. It’s become so locked down to stop abuse that the time and effort setting up your outbound email environment usually far exceeds the little bit of time needed to configure Discourse email.
Setting up a domain for email with DNS, SPF, DKIM, send limits, filtering, avoiding blacklists is the real pain, not the few things you need to do in Discourse to get mail to be sent out. If you can do the domain stuff then the Discourse stuff is a breeze by comparison or alternatively, if you can’t handle the bit of Discourse config then you’re probably going to have a nightmare with the non-Discourse based email config.
mmdv - it worked for me. I added one extra fairly generic word (weather) to narrow down my specific site and I came top of the list, so I assume that if I’d just searched for the string then I’d have eventually found myself somewhere in the list.
Couldn’t agree more. Email is the most complex part. Once you’ve set up a couple of sites, leveraging your existing email infrastructure makes things much easier (but you still need to jump through the domain setup hoops of course).
No. And never will be. Sorry, but my forum is not an ad for CDCK. Same applies for my WordPress. And out there is no one who is looking for community/forum using one specific engine.
I know, TINSTAFL, but every single one of my users know or should know I’m using Discourse. My WordPress is totally differen story.
There must be some other technical details to looking for. Similar as we all here know right away if a forum is builded using Discourse or by WordPress. CDCK knows I’m using theirs product
Sorry if that came across as critical. I don’t think it’s actually “hard”, it’s perhaps just that the documentation for it is pretty spread across this meta site, and if you haven’t installed a rails app before there is a fair bit of assumed knowledge. I have done lots of python (esp django) stuff before, but thought editing the config files and recompiling was an odd process for what could be runtime settings.
I guess ultimately it just felt more like a less-preferred option, which I accepted because it was free.
It’s a little akin to the problem of people here evaluating the discourse UI. We all like it and are familiar with it. But it was interesting to see my students use it for the first time last year… they didn’t always find it easy and some (not all) described it as “wordy” and confusing. Most ended up liking it, but the onboarding was more geared to people familiar with this type of community.
Anyway - would be happy to help with community documentation of the process of installing on a new digital ocean droplet, for example. I think it would also help to have a guide to the best way to setup with email. A lot of the inbound stuff is complex and not that well documented. Suggesting a good default setup would make sense.
While true that email is a pita (often for good reasons), it’s not so hard that we couldn’t offer a guide to a ‘blessed path’ (e.g. “just use sendgrid” or “setup SES like this”).
For me the issue was complicated by the fact the MS/365 just really doesn’t want you to use legacy authentication. I think that’s probably an issue for other orgs too although not sure. (As an side, we had to get a special dispensation to leave on legacy auth, and I’m still worried that I’m one policy change from everything stopping working again, and because I am in a massive org running a site for a (relatively) small group of 1500 people we aren’t always a priority).
Similarly, getting SSO setup with 365 was not that well documented as I recall, and I haven’t yet tackled that — although we really do want it. I should bit the bullet and write a guide the person coming next.
) but I think having an interactive ‘explainer’ site with examples of lots of different features in place could really help with this.