Zoom out. Said voice message could be automatically transcribed and offered in both formats. Then you meet the needs of all parties. Whenever you see an issue with a way of communicating, instead of looking at how things are, be open to accepting that how things might be could solve the problem (even if you can’t envision a solution in the moment).
What do you mean? Can you share a link to more info?
It’s not always a one size fits all. For very large communities, sometimes a custom solution might be the right fit - which isn’t necessarily a chat, or a forum.
The cost of transcription is still quite expensive. You will only see this on top tier enterprise platforms and not on Discourse, Circle, etc. Transcription is still only about 90% accurate for a clear speaker with little background noise.
That’s not accurate. Here’s a small dev shop doing it, and plenty of open source AI platforms out there enabling this as well: Navi - Transcription and live translation for FaceTime | Product Hunt.
For noise, look into Krispr.
The three pillars of communities around shared interests are anonymity (as you pointed out), language and timezone. So for anonymity, again, one needs to zoom out. You don’t need to use your face, or even your voice, in a call. You can use an avatar instead, or blur the face, or apply filters, etc. For example: https://alter.xyz.
There are always ways to remove barriers and create functionality but I still don’t see the demand for it and I still worry about diluting quality because a typed conversation is generally more concise than a spoken conversation as you lack any real editing ability.
Right, I don’t think he was suggesting that though. Think about it like this:
Anchor message [media of some kind]
- messages about said anchor message (different types, but text to it too)
Basically, just like chat.
This is true, and some argue that the photobucket catastrophe wherein millions of in-line photos were delisted was the primary “straw the broke the camel’s back” that ended forums as a mainstream platform. There’s a very lively conversation on Reddit about it now.
More broadly speaking, though, I think that - as @JoelR has better explained elsewhere - the fact that hobbyist admins were playing three roles - webmaster, forum admin, and community manager - and not doing any of those three exceptionally well, introduced the type of existential risk that made it easy for a well-run, large-scale platform-as-a-service provider to displace well-established mainstream forums. Facebook Groups and Reddit were only the latest in a long line of providers that had disrupted the hobbyist market before - like the venerable ezBoards or the proBoards that followed. The difference here, though, is the change in browsing habits and demographics shifted the overwhelming majority of consumer and casual populations to hyper-transactional mobile-first content (namely, images, video, and short-form text). Forums never really enabled these as a primary mainstay, at least, not in-line.
Fascinatingly, Slack first arose in developer communities in the early 2010s because it did a great job of pulling together all of the images, video, documents, etc., that were scattered about on share drives and workspaces, into a single shared storage and presentation facility. It was the synchronous counterpart to the asynchronous Basecamp that preceded it.
Jeff Atwood is the founder and developer of Discourse. (He also created Stack Overflow, founded Stack Exchange, and is a known quantity in the open source and developer activism space.) He is controversial purely because he expressed a very passionate perspective about how communities should be managed, and you see his philosophy expressed in the design choices he made when developing Discourse (and the other platforms he founded / co-founded).
Here are just a few qualities:
When he vowed to reinvent forums in 2013, he wanted to bridge the gap between synchronous and asynchronous content. This is controversial because most forum fans have a strong preference for the linearity of asynchronous communication, and generally view that synchronous conversations (like Chat) threaten the integrity of forums. For Jeff and Discourse, the two can and should be convergent.
On his blog, he further explained he felt that forums had basically become “a sad trombone,” and the user interface and broader client experience hadn’t changed in decades. He argued that forums needed to be redesigned to (a) be firstly for tablets, mobile devices, and hi-res browsers; (b) be super-responsive, which is only possible with a modern tech-stack, and the LAMP stack that most software - including the current legacy leaders like Xenforo and Invision Power Board - is slow and largely not fit for purpose; and, (c) needs to have built-in automated governance and moderation systems that drive self-regulation.
Also on his blog, he argues that most forum admins are incompetent. Specifically, that because most forum software is closed-source and not open source, it is inherently less secure and open to exploitation. (This is not inherently an Atwood argument, as many proponents in the OSS/FOSS movement also argue the same perspective.) But in his basic argument, he suggests that because admins don’t appreciate the importance of portability and security in their software choices (why would they? they’re just trying to set up a forum, or were when forums were alive in the hobbyist sector), they end up compromising both. So he made very specific design choices to make security and portability unavoidable and self-enforcing software choices.
There’s a lot more, and I use the term “controversial” affectionately. In my view, he is the only contributor to forums who has actually disrupted the space in a pretty big way. It is also very telling that many of his positions that he has staked on forums - from continuous-scrolling pages, to synchronous auto-loads of recent conversations - have been borrowed and influenced the product roadmap of many of the LAMP-stack providers like Invision and Xenforo. That said, he continues to rail against older technology as inherently problematic, and his perspective is that if you’re not designing for mobile first - if not mobile-only - you’re in essence dead.
Finally, and this is the most controversial thing he has stated, in his original announcement about why he wanted to create Discourse, he indicated that less conversation - not more conversation - was inherently required to arrive at a conclusion. He cited his experiences with Stack, and cited the main use case (today, especially) for forums - the Question-and-Answer or Support Discussion forum. In it, he argued:
if your goal is to have an excellent signal to noise ratio, you must suppress discussion. Stack Exchange only supports the absolute minimum amount of discussion necessary to produce great questions and great answers. That’s why answers get constantly re-ordered by votes, that’s why comments have limited formatting and length and only a few display, and so forth. Almost every design decision we made was informed by our desire to push discussion down, to inhibit it in every way we could. Spare us the long-winded diatribe, just answer the damn question already .
He then treated Discourse as a “playing for the other side,” to explore what would happen if you co-opted long-form discussions and drove more civilized discourse by discouraging walls of text and driving more actionable, real-time conversations.
He is, in my view, a visionary.
Thanks for taking the time to put out this - very useful - response. I like principled product people, and to my delight, I’m on the same page with him on several ideas you paraphrased. I need to find out more about this guy and read through his thinking.
But let me give you some very contrarian product decisions I came across, for a real-time chat product for communities, just like how some of his seem unfit:
- No DMs between members, only Tickets between members and mods.
- No ability to add new lines, which limits text to one paragraph at most.
- No input controller for text, but a canvas for apps instead.
- LiveText (which is how the 1st chat app was actually conceived, not sure when we replaced that with XXX is typing - which isn’t as useful).
And many other things. Basically, I love the fact he’s been innovating and experimenting with what a forum might be - since you can see it’s something we also experiment with, for chat.
Again, thanks for sharing, very insightful. I now have some more homework to do.
Discourse has always been very interesting to me.
I find a smalls sense of irony that in one of those articles he chastises forum software for not changing in 10 years, but the Discourse product of today is almost identical to the Discourse product 10 years ago. While there was a burst of innovation at the start, not much appears to have changed since. It is easy to innovate at the start with a fresh slate (and VC funding) but harder when you have a mature product, thousands of customers and millions of end users, most of whom strongly resist change.
I quite like some of the concepts but the UI has always felt unfinished to me. Its functional but not particularly stylish or slick.
In terms of running a forum on a LAMP stack, this is something I do agree with. As we (forum vendors) aim to compete with modern experiences, a basic PHP + MySQL approach is now simply not enough.
One must keep in mind that all forum products evolved from a downloadable product designed to be hosted elsewhere beyond the control of the software authors. To allow as much market penetration as possible, this meant keeping requirements very basic. As most forum vendors with roots in the early 2000s were bootstrapped with a team of 2-3 developers and serving smaller clients with $149 license sales, tossing out the product and starting again with a complex tech stack would have been commercial suicide.
As we move past that, I would expect to see more innovation in both the UI and feature set. We have several projects in development that would not be possible if we limited ourselves to a basic LAMP stack.
What are some ideas that come to mind, on how they could improve it? I think the Chat direction for Discourse is a bad strategy choice, so we can skip that.
I’m not deeply invested in Discourse and although I’ve used it a few times with other communities, this community here is about the most I have used it.
Most of my mental energy is focused on moving my own platform forward and navigating the tricky line between forging a new path and keeping paying customers happy.
I’m not sure that’s true. Discourse has made considerable improvements in its deployment at scale, and its interface, features, and functionality have significantly evolved over the past five years alone. You should familiarize yourself more with Discourse’s development history: it a very fascinating and exciting story.
Again, this doesn’t sound like an informed opinion to me. Have you actually seen some Discourse installations in the wild, Matt? Take a look at the World of Warcraft community, for example:
This is just template and CSS changes. The UI looks super-polished, stylish, and modern to me, while it is still indistinguishably Discourse. The rest of the Blizzard Forums are just more stock Discourse, but it follows Atwood’s design guidance and looks stunning on high-resolution devices and mobile/tablet devices of all sizes.
I undestand where you are coming from, for sure. I started playing with forums with Matt Wright’s WWWBoards and WWWThreads scripts (and later Bugdorf’s WebBBS), and it was only Madrona Park / Infopop’s Ultimate BB that represented a major sea change in UI and UX in 1997. But Atwood’s perspective is right: the table-based, linear-threaded layout in the UI hasn’t changed, and the underlying architecture hasn’t evolved much since the sea-change to PHP and MySQL in the late 1990s.
I wonder why there has been so much more incremental innovation in the synchronous chat space - from numerous client-server architectures (like the Java-based DigiChat, or the many incarnations of IM across AIM, ICQ, Yahoo! IM, and MSN Messenger); to the evolutionary pushes across voice-and-text chat with TeamSpeak, Ventrilo, and eventually, Discord. These came out of the hobbyist space, no differently than forum scripts did. Forums, by comparison, have looked glacial in innovation.
This goes back to your original post, and again, I understand the commercial realities here, but it feels like a consolation prize. Shifting to a SaaS offering for the enterprise without delivering a parallel feature set for the self-hosted market stands in sharp contrast to what others are doing in the forum space (Discourse, Flarum, NodeBB, etc.), and what other innovators are doing on the non-forum / collaboration space (like Mattermost, for instance) – in that they all offer self-hosted options with a parallel featureset to their cloud-hosted offering.
I’m not trying to pick on Invision here. Xenforo is in the same boat. vBulletin is in laughable last place given the whole Jelsoft acquisition debacle, and a myriad of other mismanagement issues. But I don’t think your earlier statement –
… is anything other than a cop-out, considering the risk-taking that many of these 1-3 people teams have engaged in without any funding. Jeff Atwood is certainly controversial and exceptional given his track record of success, but he rubs a lot of people the wrong way because he specifically calls out the culture of complacency and me-too’ism across all of the software development space. I just wish there were more risk-taking in the forum space, the way chat, collaboration, productivity, and voice/video have seen in the past five years alone.
I appreciate the thoughtful reply.
Most of the Discourse communities I’ve seen tend to all look alike. The WoW is very nice. I’m not against Atwood or Discourse and my opinion is based on the exploration I’ve accrued in passing.
I do want to underline again just how difficult it is for a bootstrapped business with a micro development team to take risks. Forum software has powered massive parts of the internet for years and this software has come from passionate individuals who wrote code after finishing their day job.
When you have just 1-3 developers, you are so busy fixing bugs, talking to customers, doing support, etc that you don’t have a lot of space or time to be make dramatic shifts. When we rewrote Invision Community a while ago it took 18 months and we didn’t ship anything new. Our entire team (of which there were about 5 of us at that time) were fully committed to supporting existing customers and refactoring the code base. License sales dried up and we almost lost the entire company. We did not have another round of funding to fall back on.
Thankfully, of our competitors of that era we are in the strongest position with a team of 9 full time developers so we are now able to take calculated risks in expanding beyond LAMP. We also have techs, sales, marketing and people responsible just for internal organisation.
We are levering cloud architecture to build new features rapidly. Again, with a limited resource pool to draw from and limited experience outside of PHP, it is the sensible option for us. We cannot open a (check|cheque) book and hire a handful of ruby/Node/whatever developers. We just do not have the resources to completely throw out our tech stack and start again with new developers.
Vision is cheap, execution is expensive.
FWIW, I’m proud to be a bootstrapped business with scars from fighting hard to keep a business moving forward for two decades. We could have sold out multiple times but I enjoy the process too much.
AOL owned AIM. Microsoft owned MSN Messenger. Huge budgets, large development teams not attempting to write complex software that runs on a $5/month hosting account.
Discord came from Jason Citron who sold his game platform for $104m which founded a new game studio. They wanted to fix VoIP for gamers and took a funding round to develop Discord.
We just go back to money. The more of it you have, the more bold and inventive you can be.
From the outside, forum developers look unimaginative. From the inside, we’ve been pulling off miracles for decades with a micro budget to allow mega-communities to fit on a tiny footprint at a low cost.
Thanks for the thoughtful answer, Matt.
In my opinion, many hobbyist forum admins are a lot like the legacy webmasters of old: they haven’t kept up with design standards. Again, why would they? The craft has evolved into a real profession, and the days of hacking up HTML to make a site look nice won’t cut it when you need to be deploying responsive frameworks (like Bootstrap) and feeling very comfortable with at least intermediate CSS3+. I don’t think anyone who sets up Discourse, without having a background in web design and modern web applications, can realistically do a nice job of making it look anything other than stock.
By contrast, I think this is one of the core strengths of scripts like phpBB and IP.Board. The administrative interface allows unbridled access to the style, so that all a hobbyist admin needs to do is just change colors, upload a logo or an image or two, and off you go. I see the foundations of the Admin Control Panel from forums of old in the WordPress sites of the past 15 years. It’s made a lot of the web wildly accessible. But it also makes about a quarter of the web look very, very similar. Is there any surprise, then, that the blogosphere and forumspace evaporated in the mid 2010s?
I think a single passionate developer can put a lot of these established incumbents out of business. You see this in Gaming (Minecraft was created in a single weekend by one Java developer in his basement!), in voice and video (Ventrilo and TeamSpeak were both after-work passion-projects), and - yes - even in disruptive forum tech. Flarum was essentially built by two people, and now has ~10 volunteers, none of whom are getting paid. And before Discourse had funding to embark on something unique, Jeff Atwood took crazy risks with Stack.
In my opinion, a lot of this boils down not to money – which is the clarion call of those of us in SaaS and PE-backed enterprises (and I am one of those such executives) – but with the total mindshift differences between scrappy entrepreneurs who live in the Free and Open Source Software space - making use of reusable and extensible code frameworks over version-exposed development collaboration spaces like GitHub - and those who run traditional brick-and-mortar-style businesses, with payrolls, employees, and - as you said it best - established legacy customers. I think it’s the last point that creates a false sense of restriction: you have an obligation to your existing customer base, and so in many ways, you feel hamstrung by having to fix bugs, communicate product roadmap, and maintain an organizational infrastructure to continue running the ‘business.’ But, unless if you cut customers loose and start a new venture to “put yourself out of business,” you run the very real certainty that someone else certainly will. In my view, this has been the prevailing reality since the early-to-mid 2010s, when PEs started to realize they had to start backing companies for the long-haul much earlier in the founder-led cycle. (This has also created a lot of financial and economic dilemmas, best saved for another conversation, but it does explain the hyper-speed of 1-person led creations vs. the glacial speed of legacy organizations that don’t “blow themselves up.”
There is no doubt in my mind that of the legacy forum developers, Invision is the leader. phpBB also continues to exist as a viable alternative, even though its feature set is much more simplified; the amateur developer ecosystem, by virtue of its open source, is extremely vibrant, as well. I tend to see Xenforo and vBulletin as two lesser-tier performers in the space, but I do lament just how much has been lost. vBulletin’s ecosystem, itself, was vast and powered countless adjacent businesses - from vBSEO to GARS and vBadvanced, and the like. All of these businesses, revenue streams, and customers are now completely gone.
Yes, but execution is a lot less expensive than you may think. Especially if you embrace the “just start creating” mindset of a lot of scrappy open-source entrepreneurs. The fact that most of the innovation is taking place outside of forums leads me to the same conclusion that forums, at least as they were originally envisioned in the 1990s through the early 2000s, are truly dead. What is being built today isn’t really the same thing: it’s web applications that approximate the forum feeling, or collaboration and communications tools that have a forum-style featureset thrown in for familiarity, or it’s cloud-based providers from the legacy set looking to penetrate enterprise customers.
The use case for forums has never been stronger as a way to bring down support costs by deflecting ticket volumes and building a self-help community. But the end-user client experience is so atrocious that, by and large, many corporations are simply going without. I worry especially about decisions that companies make - especially in gaming - to just glom on to SaaS-hosted platforms like Discord, that represent significant risk and in many cases security and scalability threats to their own end-customers. But again, that’s a discussion we’re having elsewhere here.
Thanks again for being willing to talk about this industry’s unfortunate decline. I’m eager to see the rebirth, and hope more people will follow in Jeff Atwood’s disruptive direction to address some of the core success factors and usability concerns that have made asynchronous forums so inaccessible and uncompetitive in today’s mobile-first, high-resolution, fast-paced environment.
This is the crazy thing. We have not seen a decline in terms of profitability. Our business now is stronger than it’s ever been. Most of this is down to securing enterprise customers and moving to a SaaS model where you get a recurring revenue and not a “one and done” license purchase. This has allowed our team to expand significantly over the past three or four years. It has given us the financial security to develop new tools such as sentiment analysis on content, AI image moderation, real-time trending content and presence (who is viewing/typing with you) along with a pretty neat implementation of realtime chat (in development now).
I do not believe this is simply a dead cat bounce.
Putting tech stacks to one side, most of our customers want the following:
- A faster UI (but not necessarily simpler)
- More human analytics
- Faster workflows
- Stronger discovery tools
- Robust moderation tools to combat bad actors and spam
Few are asking for a complete reinvention. I know you’re tempted to quote Henry T Ford at this point.
What would reinvention look like? For Atwood this was not a major departure from the forum layouts that preceding Discourse. There are still tables, there are still lists of posts. On the surface, it’s more of an evolution than a revolution.
Ultimately the market will win, it always does. We have reinvented ourselves multiple times and have had to make hard decisions that would scare most PR departments (if we had one).
This is a curious paradox in this industry. Forum developers have not tried to put themselves out of business, but then no-one else has either. Outside of Discourse, there has been a huge void of competition and VC money to disrupt this industry.
Using BuiltWith.com, here are stats for Invision Community.
BuiltWith shows we have 8.5k live installations, and Discourse 14k. If you focus on the paid version of Discourse (hosted) this figure is 2.6k.
I don’t think it shows a huge decline, and neither does it show that Discourse hoovered up communities from other vendors.