The future of forums?

Discussion in 'Managing an Online Community' started by Pete, Jun 15, 2018.

  1. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    So over in the GDPR thread we started talking about modularisation, e.g. the newsletter thing being more of a separated component that could be tweaked/enabled separately from the rest of it, and that going modular design allows forums to do things they currently can’t.

    Thing is, I’m not sure that’s true. And speaking as someone who is launching a niche forum platform soon, I’m definitely interested in this kind of thing.

    Sure, forums could benefit from understanding user expectations of the modern age... notifications systems, what UX conventions and so on other (not necessarily forum) platforms are using.

    And yes, there are always going to be the “noobs” who don’t know what they’re doing and make a bit of a mess, but that’s true for blogs, it’s true for online stores too. That’s not a forum specific problem.

    Then someone mentioned WordPress. I don’t think you can really compare WordPress to a forum... out of the box, it fundamentally does less. There’s more scope to add things because there’s simply less for it to interact with.

    WP has posts, and ways of getting at those posts. And ways to make “posts” into things that aren’t posts. WooCommerce is my favourite example, a product is a post, an order is a post etc. With different visibility rules and so on but it’s still just posts under the hood.

    Making this level of modular in a forum would give you a lot of trouble, wouldn’t it? Forums have topics, of which you can have comments that can be split, merged, moved around and interacted with in a lot more ways than WP... seems to me that the essential complexity of a forum is also inherently a barrier to being quite so modular.

    Ok, what if we did it at a higher level? A users module, a forum boards/topics/posts module... you still need all the moderation, banning and user management stuff inherently tied into both, making as we engineers (haha) talk about being “tightly coupled”... and for any non trivial add on, that problem scales up and outward.

    We also touched upon vB 3 and all its functionality... great that it could do many things but it made it a complex mess to work with, not to mention a stability concern. IPS could be headed that way with IPS 4 if they aren’t careful but their modules are sufficiently not quite so closely coupled that it’s ok. Especially as most of the modules are adjacent to, rather than directly integrated into, the forum.

    So... forums? I sitll think we need flavours of forums. But I don’t think that’s solvable with modularisation... partly because of the above and because I’m literally seeing this right now in StoryBB.

    StoryBB is the prime example of what I mean with flavours of forums. It started out life as SMF 2.1 a year ago and is in the process of substantial reworking. Most significantly it got roleplay characters as a feature.

    Now, this isn’t a new idea, and it’s usually done as a plugin to connect accounts together but this runs into all kinds of weird cases, not least managing PMs, bans, moderation, or the security of multiple accounts. So I ended up not doing it that way and keeping accounts as accounts, and having characters as something new underneath the accounts.

    It turns out this is such a complex process, it would be unmaintainable as a plugin, and I simply can’t envisage any platform where this would be actually viable because of playing nicely with, well, everything else.

    It turns out, though that this isn’t the only case. Consider StackExchange: isn’t that a forum but with very very specific tweaks? Sufficiently large that a conventional forum wouldn’t actually work for it? It’s still a forum but it thinks about topics and comments differently.

    Reddit is another case... highly nested replies focusing on sharing a link or a short post and seeing what shakes out.

    Here is our future, I think: forums that have increasingly specialised and bespoke features that can’t always just play nice in a conveniently encapsulated modular world.

    But I could be wrong...!
     
    • Also Wondering! Also Wondering! x 3
    • Agree Agree x 1
    • List
  2. we_are_borg

    we_are_borg Moderator

    4,628
    807
    +1,809
    This is what Alfa1Alfa1 talked about. Going to read it fully when i am done.
     
  3. we_are_borg

    we_are_borg Moderator

    4,628
    807
    +1,809
    Modular system is the way to go but too do so the tree needs to be very good to uphold the branches. Look at IPS and WBB they are building there system out the forum is not the focus anymore. I continue with IPS as example they are building out think shop, blog, gallery, downloads, cms, clubs, forums and calendar, but they can still expand even more think wiki, classified and more so building modular is good because you can add more and more. Developers need to understand that people want an one stop website, they don’t want to hop arround to multiple websites.
     
  4. zappaDPJ

    zappaDPJ Administrator

    6,370
    1,342
    +4,774
    I agree with that and I think one of the primary reasons for it is developers developing for themselves rather than for the end user. Another issue is that in general the desktop bound, elderly content providers are on the edge of extinction. Forums are unlikely to survive unless the younger, mobile consumer can be persuaded to step up to the plate and contribute.

    In my view forums need to be app driven. I've said before, we need Tapatalk that isn't Tapatalk.
     
    • Agree Agree x 1
    • Winner Winner x 1
    • List
  5. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    Tapatalk that isn’t Tapatalk is the antithesis of what I’m talking about.

    Even SMF and StoryBB could not meaningfully be supported by the same app, there are too many changes that are not compatible, and StoryBB is like 80% SMF still...

    Consider: a forum wide app essentially means creating a system that is literally the lowest common denominator across all supported platforms. Boards, topics, posts and nothing else. That erodes a lot of potentia.l away.

    But consider also: if we accept that StackExchange is a type of forum, should the same app for a forum also support StackExchange? Or is it too different?

    What about Reddit?

    I don’t think the problem is fixable at that level. Forums need to evolve what they are and provide features that a specific community needs... and that breaks the notion of having an app for all the forums.
     
  6. zappaDPJ

    zappaDPJ Administrator

    6,370
    1,342
    +4,774
    I know but I'm suggesting that in general anything that isn't app driven is unlikely to gain traction or survive long term.

    It's certainly possible to build a successful platform to fulfil a need, Discord seems to be a prime example of that but it's still (as far as I know) app driven.

    As a desktop user with poor eyesight I would like to be proved wrong but it seems to me everything has to be app driven. I recently returned to the recording studio after an absence of around 25 years and within a day the number of apps on my phone doubled. Even my pedal board needs an app to drive it :(
     
  7. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    Then forums are already dead because an app per forum is the only way that’s going to be viable, or at best an app per forum platform type and even that isn’t viable in practice.
     
  8. zappaDPJ

    zappaDPJ Administrator

    6,370
    1,342
    +4,774
    I agree, per forum, the uptake from the user base is so small it's not worth the development time. So do you see responsive as the only viable solution?
     
  9. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    How else can you do it? The forum world necessarily means having sites running different code to include different addons to support different needs. It’s sufficiently inviable doing this with a single master app, to the point where an app cannot do it.

    So responsive becomes the only option.

    Or we somehow get IPS, IB, phpBB, SMF, etc to team up and build an API that is flexible enough to cover all the different uses... which is basically impossible.
     
    • Agree Agree x 1
    • Winner Winner x 1
    • List
  10. Nev_Dull

    Nev_Dull Anachronism

    1,770
    717
    +844
    I see two problems with the idea of making flavours of forums. First, you can't possibly anticipate or meet all needs. So you end up with people forcing one flavour of forum to serve as another. Secondly, people very often don't know what they need, or their needs change. Again, they end up using the wrong product and become frustrated. That ends up reflecting badly on the software, even though it isn't its fault.

    That's why I like the idea of a more modular design. No matter how you use it, the core of every forum is the same. You have posts, threads, topic sections, members. Almost everything else could be dropped in as an optional module. Forum owners could choose what they need to tailor the experience for their purpose and audience. That includes the UI. If the near future is going to be tiny mobile screens, forums need a complete separate mobile UI for input and output, rather than a stopgap responsive design, if we ever hope to see mobile users participating in a meaningful way.
     
  11. darnoldy

    darnoldy Curmudgeon

    1,714
    782
    +460
    A forum or a blog (and many of the other forms) are, at their core, two collections of objects: members (accounts, users, etc) and posts (articles, replies, comments, etc) each having an array of properties. We tend to think of threads, forums, categories, etc as container objects--but they're really just search results.

    The primary differences between all of the forms that have been mentioned are the way user permissions are set, and the default way that search results are presented.

    Maybe what's needed is a more-robust and flexible search engine, so that site owners (and perhaps individual members) can more-easily customize how their content is presented.
     
  12. Nev_Dull

    Nev_Dull Anachronism

    1,770
    717
    +844
    I agree with all that. The problem is how to get a better search? I don't know how you get away from a simple full text search. We can't expect posters to add metadata or enter keywords for posts -- it's hard enough getting them to post at all.
     
  13. zappaDPJ

    zappaDPJ Administrator

    6,370
    1,342
    +4,774
    A predictive search would certainly help.
     
  14. cheat_master30

    cheat_master30 Moderator

    3,838
    962
    +1,001
    That's the million dollar question that basically all CMS systems need to answer now. Search features in pretty much every piece of software/on any site not made for it/using something like Google are generally rather dire, and the ones for things like WordPress are the worst. Unfortunately, it seems the way the likes of Google handle it involve bucketloads of data, machine learning and algorithms complicated enough that a whole industry sprung up about how to game them. The likes of the old fashioned 'find the keyword in the page' search systems just aren't good enough any more.

    Outside of search, I feel the future of forums is going to come with the following:

    1. Apps for forums, likely provided by the software creator. Seriously, forget Tapatalk type solutions. An app for 'forums' in general won't work out, and a bespoke one for every single separate community would be way too much work/cost far too much. Instead, I feel the answer is going to be a setup where the creator of the script (aka XenForo, IPS, vBulletin, etc) provide the basic code for an app for the site along with the script, and that's then modified when the admin changes various settings and themes in the admin panel. Third party add on creators would then take advantage of various plugin hooks to add code for the app interface, and maybe you'd have an option to either export the app settings or entire built app from there. Yeah, it's a ridiculously complicated idea, but I suspect that's the only likely way forums will be able to take advantage of the move to apps nowadays.

    2. The second one would be something akin to Reddit, except decentralised and federated. This an idea I've already wrote about in an article here, and it's one that I think may be the most viable solution for forums to gain traction on the modern internet. Remember, people don't want separate accounts for everything, and they don't want to join a ghost town. So the answer is to set up a Mastodon type solution for forums, where the sites are connected but individually managed and the user can decide which communities they want to see and which they don't from a single main point.

    But yeah, that's just my opinion on the matter.
     
  15. Nev_Dull

    Nev_Dull Anachronism

    1,770
    717
    +844
    This is a long-standing problem online. Search in a CMS, or even a blog is infinitely better than a forum because you can use metadata and keyword systems. The reason it doesn't work well in popular software like Wordpress is because very few people know how to do it well, and aren't interesting in learning about it.

    I don't think apps are the answer either. I know I don't want an app on my phone for every forum I visit. What I want is a dedicated mobile UI that I can use through my browser. One app, lots of sites.
     
  16. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    You mean other than Safari? As I already explained, this necessarily requires forums to dial down the customisation - the app can only support a lowest common denominator (or, in the fact of overwhelming support, some limited additional features).

    How often did Tapatalk successfully offer any of the additional functionality on your site(s)? How often did any customisations not work with or even actively interfere with Tapatalk? This isn't because Tapatalk is incompetent, it's because you can't possibly predict all the interactions ahead of time.

    This is also frequently a non-starter for any customisations. My example is Moodle - which has a very specific plugin architecture and their app is pretty generic, often requiring that individual plugins specifically offer extensions to be able to support the mobile app. In fact, for addons that are new types of activity to do in a course, the *only* way you're getting that in the app is to explicitly provide support for it, often duplicating all your PHP logic into JavaScript. You're writing large chunks of your plugin twice to make that work, and even then you're still pigeonholed by the types of plugin supported; not all plugin types are supported and not all plugins work even in the official Moodle app, nor can be made to work.

    And that's for an organisation that has millions in the bank from external backers, a thriving network of partner companies who all fund Moodle's development, a dev team considerably larger than any of the forum softwares currently have. This is an organisation that can run a multi-day coder-focused workshop event to teach coders how to implement mobile and GDPR features in their plugins.

    The forum manufacturers don't have those resources (except possibly IB and they won't want to commit to that) - and the customisations that will be exposed in any such app will necessarily be limited.

    Now this is an interesting concept - though I think things like Mastodon are surprisingly non-viable... outside of the people I work with who are all techy, almost no-one I know knows anything about it other than 'isn't that a dinosaur' - and these are people who would be quite happy on a forum. Getting from one forum to another seamlessly is a huge deal, and one of the perks Tapatalk had, though many forum users were concerned about people going to competitor forums.

    Search in general is fairly poor in forums, but incidentally why is that? Is it because the search algorithm is poor? Is it more likely that content is just written by regular people who don't focus on making *good* titles or keyword-rich posts or any of the usual things that help search engines (even internal ones) get to your content? People don't use metadata unless there's a reason to do so, and 'because it'll be easier for other people to find posts' is sufficiently abstract a concept that people don't do it.

    I'm not entirely sure I agree with that. Yes, on a purely technical level that's true - but on a different level, there are contexts which separate these. A blog is a person or a group of people writing content specifically engaging for replies. It is a few-to-many broadcast system, whereas a forum is much more a free-for-all, much more around the notion that everyone can have a voice. The downside is that everyone has a voice, and hearing things in a crowd is harder.

    Additionally you have the fact that in a blog, the controls around publication and metadata are stronger and often employed better. But these don't have to be solely on blogs; I've run forums as blogs before now, making a forum with specific permissions and a custom theme into a blog for one-to-many publication (because I hate WP that much :p)

    But as much as conceptually it's the same thing, it's conceptually not the same thing. I also don't necessarily see them as just search results - if I'm coming from Google, maybe, but once I'm already at a site, it's not so much just a search result, it's a view on part of a thing, that I can choose to interact with, and whose interaction potentials are formed primarily by the publication context.

    Now this is always going to be a problem - but it's not one we don't already have. You can make a roleplay site on any of the major forum platforms, but it'll have some major shortcomings. You can make a Q&A site too, but compared to StackExchange (which is really Q&A focused in how it does everything), it will likely have some major shortcomings.

    These shortcomings we overcome with addons today, and that wouldn't happen in a more flavoured setup. Heck we even already have that in, say, directing people to IPS to have forums+pages rather than a base forum with some kind of weird bridge to some CMS. Same functionality, but one of these feels better to use and we encourage people to do that.

    I'm just suggesting that we start looking at platforms that actively embrace some use cases that were traditionally an existing platform with a raft of barely-usable-together plugins.
     
  17. zappaDPJ

    zappaDPJ Administrator

    6,370
    1,342
    +4,774
    It's probably a bit of both. Google often finds content that a forum search fails on so that suggests forum algorithms are not as good as they could be. I also think the search function in general is somewhat underspecified. Why can't I search by date range or search for images by user?
     
  18. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    Searching by date range should be doable (SMF has it albeit with a lazy interface).

    Searching for images by user depends on what you're trying to do - if you know who it is, maybe go to their profile. If it's an image in a post... how do you want to search it?

    The problem for images is that forums cannot have the resources Google can throw at machine learning to teach it about images...
     
  19. Nev_Dull

    Nev_Dull Anachronism

    1,770
    717
    +844
    Can you explain a bit more about how flavoured forums would work? I'm just not seeing it.

    A forum developer wants to build new software. Presumably, they're going to want to make at least a couple of different flavours to make sales from a variety of client types. Are you saying they would build each one, independently from scratch? I don't see that. I expect they would share the same codebase and simply build each one with the features to serve a particular need. Wouldn't it be smarter to build a generic base and then build separate modules to add those feature sets, so buyers could build the forum they wanted? I'm not a developer, so I don't know.
     
  20. Pete

    Pete Flavours of Forums Forever

    1,773
    227
    +602
    OK, so I'm biased because I'm literally developing one right now. But my use case demonstrates my points pretty firmly.

    So I started from SMF and started building a ton of new functionality on top. My particular use case is roleplay forums - where you need subaccounts, where you can have an account and various 'accounts' that you can post under in different places, where each of the accounts has its own avatar. Where you're looking at the site as one of your personas and see a theme that fits the character - one minute you're looking at the site as Harry Potter, with a gold/red theme, and access to Gryffindor only areas, next you're Draco Malfoy, tinged in silver and green, and seeing only the Slytherin areas, but still on the same account.

    Now, the usual approach is to make people create new accounts and glue them together, but this raises some hefty considerations in terms of management, namely that you don't actually want multiple accounts.

    If you have multiple physical accounts, you have to deal with the fact that each account requires its own email and password. Each account can then send/receive PMs. Each account may have its own permissions - and being able to see all that in one place is a nightmare if you're trying to do it as an addon.

    And that's before we get to things like moderation where you want to track moderation actions under the main account rather than any of the 'characters' or personas. Where warnings apply to the account consistently, where you don't have to remember to hit all the things. The more you look at it, the more you realise just how much you actually have to change to properly support it rather than taking the seemingly-easier-but-fundamentally-flawed approach.

    This is a major change of functionality and while I originally developed it as an addon for SMF (it's even on GitHub), it's not even close to the full depth of functionality you truly need to actually make this work in general rather than for a single site, and then maintaining that when a platform inevitably makes changes. Consider, for example, the requirements of supporting GDPR in a situation like that.

    This is also the point at which the much vaunted modularisation stops working - you get to a point where you're constrained by it, not liberated by it, and compatibility across addons becomes an unmaintainable nightmare because you have such huge changes that you're notionally trying to keep as plugins when they're at the scale they should become core features in a piece of software rather than bits bolted onto another piece of software.

    Flavours of software dedicated to purpose is not a new idea - Ubuntu does it, where you have the core Ubuntu, KDE-based (Kubuntu), XFCE-based (Xubuntu), education-based (Edubuntu), MythTV/DVR (Mythbuntu)... heck, every Linux distro is a flavour of Linux as an overarching ecosystem. Same concept, different scale.

    Even XF and IPS are different flavours of the forum concept when you think about it - XF is a flavour aimed at the core forum experience, IPS is clearly aimed at larger corporate endeavours where intermixing the different apps is more the priority than focusing on the core forum.
     
    • Informative! Informative! x 1
    • List
Verification:
Draft saved Draft deleted
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.