I think what changed over the years in the "product discussions" space/is that the sentiment has shifted from "users will come to us" to "we must go to our users". Platforms like Twitter (surprising as it is), Discord, Reddit etc. where a user already is dissuades a lot of products from hosting their own forums.
It's just quicker for a user to join a Discord community or tweet/DM a question where they already have an account. Even with things like OAuth2, the mental burden of creating a new account is there. Not to mention learning how to use the new platform. And the total gain after all the effort is trivial.
I think at this point in time, it's not reasonable to ask users to sign up to your own private platform unless you have a huge and loyal following. Users have become lazy due to the abundance of alternatives.
When I get an issue/bug/question, the first thing I check is if they have an issue tracker. Preferably GitHub. After that comes their various social accounts; Discord, Twitter, Reddit etc. If none of those work out I would most probably ditch the software. Reaching out via custom channels is cumbersome.
With that being said, this looks really cool although "delightfully simple" won't be an objective feeling among your users. Good job none the less! I wish you the best.
It's funny I don't use Discord much and consider it more of a pain in the a** than doing a quick forum signup (but I use a password manager so less friction than the average user).
The Hard Problem of forum software is how to prevent spam bots and drive-by exploits. It's what killed the previous generation of forums. Reddit usually gets the blame, but that is only true in that they had the economies of scale to survive the increasingly hostile internet ecosystem.
Aside from bundling an Akismet extension, as a development team we constantly adapt to changing spam runs and publishing these preventions for the public to use. Additionally, being extremely extensible, third party developers have contributed extensions that provide spam prevention and cleanup too, like stopforumspam and others.
About time Flarum is noticed!
I have been using Flarum through unitedbsd.com. It's nice and snappy.
Might be a lil too modern for some communities. Especially OSS ones. But is fine IMO.
My problem with discourse has been that Firefox shows FB tracker warning in irregular intervals on the top-left corner. It's says FB tracker. It makes me uneasy. But somehow it is the standard for all OSS communities.
I noticed this before and upon looking at the source, it only seems to apply to Discourse sites that allow Facebook for account creation and login. There's no built-in Facebook tracking (that would be scary indeed).
Am on a similar boat, I simply find the boards that combine different categories of threads into one page and have an infinite scrollbar on the side to jump to different points in time in the thread instead of pages (like discourse) very confusing to navigate.
It may be nostalgia but I spent a decade perfecting my navigation skills through such forums, and I don't see an improvement with the new ones.
I don't think it's just nostalgia. The bizarre, not quite phone, not quite PC monitor infinite scrolling design of these neoforums is awful, and far worse for effectively finding and cataloguing info.
My thoughts as well, when I think "forum" I think hierarchical subforums full of threads listed in last-posted order. I do not like the modern tagged jumble of threads style like Discourse and this, and I hate the fact that so many companies in particular are removing long-standing classic forums in favor of this junk. Especially because as usual they're half-assing the migration so years or even decades of history are just being lost.
Looks nice. Though my mindshare goes to Lemmy at this point. It's slightly different, but a fediverse project and pretty feature compete at this point:
I'm happier than ever with Discord since I started using it for casual screen sharing/voice chat with my gaming buddy. It really is a stretch for everything except gaming/social communities.
I've seen quite a few Flarum forums pop up recently. I wonder how it stacks up against Discourse in resource requirements? Discourse isn't a lightweight app (not a criticism, just an observation).
Although Flarum scales very well, it is not losing sight of shared hosting environments. The only requirement is composer/ssh; hosting providers usually offer these nowadays.
Flarum only released its first stable last year and so has some features missing, yet https://extiverse.com already lists over 500 extensions compatible with Flarum v1.3. Relying on much used technology allows for fairly easy adoption by developers to contribute.
Why do entrepreneurs NOT use ready-made forum softwares like this to start startups? Clever use of general tools like this (maybe not specifically Flarum) can help them launch communities or marketplaces in less than a day. Is it as optimal as a in-house built forum? No. Should they be saving time on reinventing the undifferentiated code part so that they can test the core business concept and build out their differentiated networks of users? Yes.
- you're competing with hobbyists that have the same look & feel
- hobbyists are online when the community is (mostly outside of 9-5)
- members of the community demand custom functionality specific to their interest.
Also:
- building communities is very hard
- they're not easy to monetise
I've built two successful communities in related niches and I'm pretty sure that succeeded because I was young and early. It's not easy to replicate now.
Most, let's call it, web software is a handrolled system with its own framework, design patterns, etc. Flarum is a great example of web software that rolled it's own, WordPress is another far more popular piece of software that does it's own thing.
The point being, that unless you're already familiar in the "system language" of the software you want to hack upon, or unless the software is very light on abstraction, it's gonna be pretty inconvenient to MVP on top of it.
I've tried in the past to take that approach with few small projects and gave up pretty quickly. In my view it's the same as getting onboarded on a new project - it's slow. The only difference is that with OSS software you might have a bit more documentation available
* PHP 7.3+ with the following extensions: curl, dom, fileinfo, gd, json, mbstring, openssl, pdo_mysql, tokenizer, zip
MySQL 5.6+/8.0.23+ or MariaDB 10.0.5+
* SSH (command-line) access to run Composer
Unfortunately the dependencies are not as simple as I hoped (other than SSH). Are there single binary forum software (perhaps written in Go/Rust) that use SQLite, that people have experience with?
I guess I am looking for the Caddy of forum software.
This is a bog standard LAMP setup, they just specified the specific php modules which might not be enabled by default (but usually are). Not sure what's complex about it, I'm sure you can drop it into a default XAMPP install and run.
You can ;) We are not losing sight of the lowly shared hosting users. The only tough requirement - which many shared hosting companies already provide - is commandline/ssh access so you can use/install composer.
Would be more viable if they had a file upload/download module that didn't allocate a files worth of space per download per user. The FoF module is not usable.
A new version of FoF Upload is being worked on to truncate unused files and map uploads to posts. I think it would be a great addition to add user restrictions too. What an excellent idea, it's the first time I've see someone mention it.
I think the right word would be "empowerment". Software opens up the ability of communities to grow. Security, usability, accessibility, performance and many more factors play an important role to empower communities in growing a great community.
can the replies be configured contained in the related thread that is indented too for easy reading? I found put all discussions and replies at the same level is hard to follow.
There was one extension that managed to do that with one nested level, but it's been abandoned. I also know of a commercial party whom are currently working on a threaded view extension as well.
I know it feels harmless, but this kind of low-effort drive-by PR is loathed by maintainers. It's not contributing anything significant to the project, but requires some of their time and gives you visibility in the changelog.
Personally, I have sent these kinds of PRs out and received only approvals and thanks from maintainers. When in an approver/maintainer role myself, these kinds of PRs are trivial to approve and I appreciate them. The PRs that change a bunch of stuff, are large, hard to review, poorly described, poorly justified, poorly tested, etc., are the annoying ones.
I agree that spelling mistakes in a readme / website / docs should be fixed. Comments, on the other hand, are irrelevant, and not worth pinging the maintainers for unless they are affecting comprehension (not the case here). On a project with 0 PRs open it may be ok, but in more active projects that have dozens/hundreds of PRs at any given time, it's just adding noise.
It's also been a trend in the past few years for people to make this kind of PR, which is very low hanging fruit, just to increase the number of projects they "contributed to" and their profile activity. The author may have not had this intention, but it sure plays into the stereotype.
I mainly do C/Assembly for a living and haven't touched webdev stuff since I was ~14, and I haven't used Windows as my main OS since ~2008 so I'm quite a bit out of the loop here on both issues, and I'm sorry if I make some silly/outdated assumptions here, but that statement about PHP on Windows still comes a bit as a surprise to me.
Back in the day, we simply downloaded a WAMP/XAMPP package, which was a typical windows installer. The thing came with a simple management UI where you could e.g. point the document root to the root directory of your development tree and everything just worked. IIRC similar packages were floating around with other languages or Postgres instead of MySQL.
So what the hell happened since then? Why did those things get so complicated? Does the answer involve complexity introduced by container/scaling/management/IDE/framework/... software? Did installing Software on Windows become more difficult than simply running an installer? Doesn't Windows even have package management nowadays (chocolatey/some App store/...)? If it's so complicated, don't Windows devs nowadays have their trusty WSL anyway?
Things like Laragon (https://laragon.org/) have taken up the XAMPP mantle in recent years. Not sure about production (I moved away from Windows for dev stuff similarly), but developing with PHP on Windows has definitely got easier, imo.
Ah, there're some services that provide cli tool, which uses PHP.
First step to install that tool is to install php.
I don't want a web environment here, just a php with version in my system to install that tool, but failed. I don't know how to begin, try and failed. Hopelessly.
Flarum is a sizable open source application that uses Mithril.js as the front-end framework. Flarum devs: I'm wondering, how well did Mithril scale and are you happy with the choice?
Just before releasing stable we revisited this choice and looked at alternatives like vuejs, react and svelte. But mithril came out as a winner, it's very thin, extremely flexible and extensible. Especially extensibility - for us - has been a huge factor, because that is Flarums superpower.
Now that we've added typescript, things only became easier.
It's just quicker for a user to join a Discord community or tweet/DM a question where they already have an account. Even with things like OAuth2, the mental burden of creating a new account is there. Not to mention learning how to use the new platform. And the total gain after all the effort is trivial.
I think at this point in time, it's not reasonable to ask users to sign up to your own private platform unless you have a huge and loyal following. Users have become lazy due to the abundance of alternatives.
When I get an issue/bug/question, the first thing I check is if they have an issue tracker. Preferably GitHub. After that comes their various social accounts; Discord, Twitter, Reddit etc. If none of those work out I would most probably ditch the software. Reaching out via custom channels is cumbersome.
With that being said, this looks really cool although "delightfully simple" won't be an objective feeling among your users. Good job none the less! I wish you the best.