Best Forum Software in your opinion

[emanuele]

That's what I said "If only it were possible to 'pick 'n' mix' I reckon I could get fairly close to a 'best forum' software" note the 'I reckon I' bit lol.

I guess my eyes were caught by the 'pick 'n' mix' and skipped the following word. lol

What I find interesting about this registration argument is that virtually everyone who says that registering on a forum is hard has no problem registering on Facebook where they ask for more information. That said, I think there is something to be done about limiting the amount of information required on registration - and I think there's scope for making it easier for people to register by default on platforms. Login with FB, Twitter, G+, etc. is something that would simplify that problem.

In my opinion (that is the one of a person not using social media apart from twitter to post the releases of his software and that's all), the problem is on a completely different scale.

Registering on facebook means registering once and access "everything" (i.e. any people registered to facebook and (kind of) discuss about any thing). Registering on a forum just means have registered to exchange comments with an handful of people on a certain topic, if you want to discuss another topic with a different set of people you have to register somewhere else, and so on.
I tried to suggest different approaches to forum owners (I was not saying I was able to build the solution, but I proposed the idea of "sharing" (I know there are several implications) accounts among forums), but the reaction has been *NEVER*, because, of course, forum owners are jealous of their users and fear the possibility that one of them will have easy access to a "competitor" website.

This one is easy, the guest can post but it doesn't appear until they register. The intent being that they've gone this far, so why not complete the process. Personally I think this is a great idea - whether it works or not will be proof in the doing. I suppose a time limit can be included, if the guest walks away without registering the post is automatically deleted, so no need for admins to get their hands dirty with mountains of moderation.

This is indeed interesting, so the approach would be:
* box for reply always visible
* guest posts
* forum takes the post and sends back "hey, if you want to see your post fast, why don't you complete the registration clicking the link in the email we have sent you?"
* guest can decide to leave or click.
Nice.

But that's exactly where the devs have the problem. They have to balance out whether they think more or less people will benefit from a feature, plus whether it will have consequences, plus whether it will be seen as bloat.

It's the same problem Word has at this point; it has a million features, each person probably only uses 2-5%, but they all use a slightly different 2-5% of the features. Short of including all the features - which does have a tangible cost and problem (which is why Word is as sluggish to use now on a high end machine as it was 15 years ago on what was then a high end machine - the additional power is eaten by the additional complexity)

I get where you're coming from but in reality the only way to really get the pick 'n' mix deal is for third party devs to make it happen, because first party devs cannot - and should not - attempt to satisfy all the permutations of all the things that all the people want. Unfortunately this holds true for all platforms in all environments.

I don't know how good it will turn, but the approach I started implementing in elk 1.1 is to extract anything that doesn't belong to a certain controller into it's "module" and each "module" can be turned on/off. This makes almost any "auxiliary feature" (like attachments for example) optional. It's not an addon, it's "core", but still if the feature is off, the "core" doesn't even know the feature exists (except maybe for few translatable strings hanging around in a common file.

 

[Maddox]

But that's exactly where the devs have the problem. They have to balance out whether they think more or less people will benefit from a feature, plus whether it will have consequences, plus whether it will be seen as bloat.

I understand this, but it's back to what 'they' think, they rarely if ever ask their clients; it's based on an assumption taken from a pool of their opinions. I'm not suggesting that 'every' idea be fed back to the client base, that would be unworkable, but for ideas that present a major code investment they should not be dismissed because the developer thinks it's not wanted. In a roundabout way, it's like we (in the UK) vote for a Government and we expect them to govern the general day-to-day running of the country, but when it comes to major life changing issues they ask the people, much as they did with the referendum for staying or leaving the EU. We have to place 'some' trust in developers to keep the general day-to-day wheels turning and they input some nice touches here and there, but they don't ask their clients what they think of an idea, especially if it's a major change or feature; they either just go ahead and do it and wait for the praise or fallout, or they bin it. If developers and clients could work together more closely then we might see some changes, both small and monumental.

It's a bit like if you were offering your services as a developer and a client comes to you with an idea, you would (hopefully) work closely with the client to deliver what is being asked for. Along the way there may be some issues, but you work with the client to resolve these. Naturally, we're talking one-2-one here, which is different when you involve many people, but the principle is the same and end results could be massive.

It's the same problem Word has at this point; it has a million features, each person probably only uses 2-5%, but they all use a slightly different 2-5% of the features. Short of including all the features - which does have a tangible cost and problem (which is why Word is as sluggish to use now on a high end machine as it was 15 years ago on what was then a high end machine - the additional power is eaten by the additional complexity)

I must admit that I've been using Word since it more or less came out and I use it on a daily basis and I've never found it sluggish (but that's me) and I agree with you that I only use a fraction of the features, but those I do use work, those I don't just never get in my way, but at some point in the future if I want to use one of those features I know that it's there to be used. I'm using what I consider to be a high end machine, though there are many out there that would make mine look like a snail by comparison lol.

I get where you're coming from but in reality the only way to really get the pick 'n' mix deal is for third party devs to make it happen, because first party devs cannot - and should not - attempt to satisfy all the permutations of all the things that all the people want. Unfortunately this holds true for all platforms in all environments.

I get this wholeheartedly and I have stated that it's impossible to cater for every whim, but the major stuff and add-ons that people want to make their sites more that just a forum would greatly appreciate the ability to do this pick 'n' mix - ISP have done it with all their major add-ons and are now including some of the more sought after features into the core, such as groups and reactions; not exactly ground breaking stuff, but it beats the hell out of searching for a 3rd party add-on that may or may not stick around, or even play nice with others. It's actually very refreshing to see this - sure, not everyone wants everything, but it can be switched on and off, and as I said earlier (with Word) if I need one of those switched off features further down the line, I know that all I need do is switch it on and configure it - job done.

It's never easy to move out of a comfort zone, to step outside the box and take a fresh look, but sometimes doing so can reinvigorate and give incentive to do things differently from all the others - really, it's often like watching sheep following each other but none taking the lead and being more adventurous. They do it, so will we and the wheel goes around and around, but never gets anywhere - like being on a treadmill lol.

 

[Pete]

but the major stuff and add-ons that people want to make their sites more that just a forum would greatly appreciate the ability to do this pick 'n' mix

There's the simple matters of logistics that make this unworkable.

On the one side, none of the forum teams has the manpower that Word has. None of them are even close, and aren't likely to be - Microsoft has tens of thousands of developers, and at any one time hundreds of those can be working on the flagship products. If you got the core dev teams of all the major forum platforms together, it would make a few dozen at most.

On the other, there is the small matter of testability. For each piece of code it's possible to write tests to prove that things aren't broken, and that you can add new features without breaking them.

Every new on/off switch doubles the testing workload because you have to prove that for all the code, all permutations work. So adding a new checkbox doubles the number of permutations you have to deal with at any given time because you have to test everything else with it on, and everything else with it off. 10 checkboxes = 1024 permutations. 20 checkboxes = 1048576 permutations.

That's already too many permutations to meaningfully test all the combinations. And the pick 'n' mix approach makes it worse with each modular item.

It's a neat idea theory but in practice but there's nothing that makes first party devs any better than third party devs at making things work together, and what ends up happening is that core devs frequently cull less used features to keep the testable surface testable.

I sometimes think the way to actually solve this is for all the dev teams of all the forums to be bought out and have one larger team producing the software.

 

[emanuele]

On the other, there is the small matter of testability. For each piece of code it's possible to write tests to prove that things aren't broken, and that you can add new features without breaking them.

Every new on/off switch doubles the testing workload because you have to prove that for all the code, all permutations work. So adding a new checkbox doubles the number of permutations you have to deal with at any given time because you have to test everything else with it on, and everything else with it off. 10 checkboxes = 1024 permutations. 20 checkboxes = 1048576 permutations.

That's already too many permutations to meaningfully test all the combinations. And the pick 'n' mix approach makes it worse with each modular item.

It's a neat idea theory but in practice but there's nothing that makes first party devs any better than third party devs at making things work together, and what ends up happening is that core devs frequently cull less used features to keep the testable surface testable.

Anyway, delegating the "checkbox" to an "extension writer" simply means deny the need of cross-testing, because no extension writer will ever test all the possible permutations with all the possible other extension ending in just test its own work and calling it good.

I sometimes think the way to actually solve this is for all the dev teams of all the forums to be bought out and have one larger team producing the software.

And I guess you already know that it would not solve the problem because complexity grows exponentially, while the team will never be able to do the same.
j/k anyway.

 

[Maddox]

If it's in-house code (the first party developers) then the testing would be good because they know their product, or ought to. I understand the problematic symptoms that come with all the permutations, we see it now when one 3rd party add-on plays havoc with another and there ensues an argument as to whose add-on is the culprit. This is one of the reasons I try very hard to steer clear of 3rd party add-ons unless I know the developers are on the ball with customer service and 'want' to clear up any misgivings; I've mentioned that there are a 'few' exceptional developers who go the extra mile to appease the client who becomes the victim of said incompatibilities. Which then gives way to why I believe that first party developers should up their game and either add features to the core or create their own add-ons. I also appreciate that this can be difficult (though not impossible) for small teams of developers as opposed to the likes of Microsoft who have factories full of people working away at any given issue or feature upgrade.

If a developer knows their trade, and they should, then they ought to know what will work - at least in theory. I know that the real world scenario is different, but there is nearly always a work-around. I believe developers would call this 'bug busting' lol.

I know when I develop websites for clients they want the latest and greatest, I also know what will and won't work together and where conflicts can arise. I also know that keeping your finger on the pulse and staying up-to-date with the latest innovations are important. I also have to contend with the different browsers that people use, which is why I have to become familiar with all their quirks and differences, which, thankfully, are becoming less as time progresses. I believe that the same applies to developers in the PHP world, as opposed to the HTML world that I work in.

It can be tough, but hey this is the industry that we have chosen to work within and clients becoming more demanding as time goes by and we have to try our best to placate them if we want to retain their business.

 

[Pete]

then the testing would be good because they know their product

On the contrary. Firstly, knowing the product is actually more likely to be contemptuous - you 'know' it works, therefore you don't need to test it so thoroughly.

Secondly, my experience of first party developers doesn't magically make them better than third party developers. Just more invested in theory in getting it right, but as we've seen that's not always true.

 

[Maddox]

On the contrary. Firstly, knowing the product is actually more likely to be contemptuous - you 'know' it works, therefore you don't need to test it so thoroughly.

I would be wholeheartedly dismayed if this were truly the case - one of the most important aspects of trading is 'trust' and if what you say is true then that trust is going to be hit hard and make people think whether they're buying quality or crap. What also derails this comment is the amount of Beta's that are released, if what you say is true then their would be no Beta's.

Secondly, my experience of first party developers doesn't magically make them better than third party developers. Just more invested in theory in getting it right, but as we've seen that's not always true.

I would agree with this other than if it's first party it's more likely to receive better support (I know there are excellent 3rd party developers out there that give great support) in the long-term and, providing the applications in question remain of value, they are more or less guaranteed to receive continued support and future upgrades. Take your pick from amongst first party apps and they all offer this.

 

[Pete]

I would be wholeheartedly dismayed if this were truly the case - one of the most important aspects of trading is 'trust'

I may have made it sound more callous than intended, but it's absolutely true in every single software development team under the sun. There's a reason you try to have separate people who aren't the devs doing QA, it's because devs have a habit (and it's normal, they're human) to assume things. They assume certain behaviours that worked before still work as they remember. They assume that just because they plug in 2 and 2 to a routine and get 4 that it works to add numbers in all cases. It may happen that it generates 4 as a by-product.

It's also why automated tests are important - but as has been observed too many times, devs that write tests are still relying on their assumptions, and I can remember having arguments with developers about how amazing unit tests are right up until the point where something goes wrong and they claim unit tests would have solved it... except if you assume something can't ever happen, why test for it? You don't test for things that can't happen, right?

Like all things, it's a spectrum and some devs are better at catching it than others, but no-one's infallible. If there were infallible devs, there wouldn't be any bugs.

I would agree with this other than if it's first party it's more likely to receive better support

In theory, but this assertion is not held up by the software industry as a whole. Forum software actually seems to be something of an outlier there where the first party devs are generally - but not unilaterally - better at supporting than third party devs.

 

[Maddox]

I understand the 'assumption' part and I do agree that third party QA is a good idea, providing this third party don't also make assumptions lol.

For the most part I would give ground to forum devs as being more supportive than other areas of software development, though I do believe that this is improving (there are still exceptions) as competition ramps up.

Whilst my area is different from PHP development (though I do use it at times) anything I produce for a client is 'always' put into a test environment so that others can play with it on a variety of devices to ensure that it works as intended, and I 'never' make a site live until the client is 100% happy and I am 100% happy, if there is the slightest problem it is scrutinised and corrected and if it can't be corrected (mainly due to conflicts, which are rare) it is replaced. I would like to believe that if any dev values their reputation and work, they would at least try to do the same with their products before unleashing them into the public domain.

I understand that bugs do creep in, especially if there are thousands of lines of code, which is what the Beta's and RC's are supposed to iron out and I would rather see 50 Beta's in order to get as clean a code out as possible than have it rushed and then fall over.

 

[Pete]

I do agree that third party QA is a good idea

I wasn't even talking about third party QA. I literally meant just someone who isn't on the dev team - but could still equally be first party. At my place we have a dev team and we have a couple of people whose job it is to do a bunch of stuff, but certainly QA before it goes to the client.

especially if there are thousands of lines of code

A typical forum platform runs into the couple of hundred thousand lines of code and invariably not all of it is first party code either. XF 1 for example is built on bits of Zend Framework but also has components like SabreDAV. Haven't looked at the others to get a feel for integrated third party code though.

 

[MagicalAzareal]

Gosora because I created it :d

Well, there is a reason I like it.

I can soak up large numbers of requests without paying more than I really want to, I can have it precisely how I want it, trim the fat (forums have like a million and one features no one cares about), keep things simple, little dashboards showing me how much CPU / RAM I'm using, etc.

There's more to come there too, it's very MVP-ish.

For other software... Don't tell rafalp I said this, but I kinda like Misago.
For traditional things, there's MyBB, I guess.

A typical forum platform runs into the couple of hundred thousand lines of code and invariably not all of it is first party code either.

Off the top of my head, Gosora is 20k lines of code and around 10k lines of generated code.
I should probably get around to chopping this down tbh.

 

[overcast]

For me SMF and IPS are two good software. I recently checked out Flarum which is from the fluxbb developers. It's a good forum script too.

 

[MagicalAzareal]

For me SMF and IPS are two good software. I recently checked out Flarum which is from the fluxbb developers. It's a good forum script too.

Flarum is a merger of two projects. EsoTalk by Toby and FluxBB which joined in way later.
That said, I don't think EsoTalk was ever that high profile lol

 

[Doug Heffernan]

For me SMF and IPS are two good software. I recently checked out Flarum which is from the fluxbb developers. It's a good forum script too.

For me too. Smf and Ipb are the best forums out there till now imho.

 

[overcast]

Flarum is a merger of two projects. EsoTalk by Toby and FluxBB which joined in way later.
That said, I don't think EsoTalk was ever that high profile lol

What that has to do with end product?

 

[R0binHood]

There's a big one missing off the list these days.

Not because it's the best, but because it's one of the most prevalent, and that's facebook groups.

Love it or hate it, it's the forum and community software of choice for a massive amount of people in 2019, 5 years after this thread started

 

[MagicalAzareal]

There's a big one missing off the list these days.

Not because it's the best, but because it's one of the most prevalent, and that's facebook groups.

Love it or hate it, it's the forum and community software of choice for a massive amount of people in 2019, 5 years after this thread started

We'll see how long that lasts with the crazy AI and moderators who are constantly trying to commit suicide, let alone make reasonable decisions.

 

[R0binHood]

I’ve been trying to use it more recently and it’s lack of organisation, insistence on sub threaded threads and reliance on AI to drive content to me other than purely chronologically and via categorisation is driving me nuts. They have the users and media contribution and consumption bandwidth and capability, but the usability game in terms of discoverability is leaving it all up for grabs at the moment. The question is whether classic forums will seize the opportunity to grab communities back or not given the global forum climate.

 

[MagicalAzareal]

I’ve been trying to use it more recently and it’s lack of organisation, sub threaded threads and reliance on AI to drive content to me other than purely chronologically and via categorisation is driving me nuts. They have the users and media contribution and consumption bandwidth and capability, but the usability game in terms of discoverability is leaving it all up for grabs at the moment. The question is whether classic forums will seize the opportunity to grab communities back or not given the global forum climate.

I'm not surprised, Silicon Valley is obsessed with AI right now, even when it would be better to use something far more normal.

 

[Bluefish]

I've read these boards over the years and finally joined. In fact, this is my very first post... but I don't feel like I'm new here. I've been lurking around for a very long time. I suppose what pushed me to join was the fact that a forum I have had for about 20 years is pretty much dead. I can't bring myself to bury it for some reason. At its peak, I had over 600 active users and averaged about 150 posts a day. Now I'm down to a dozen active users and some days we get no posts at all. On average, probably about 4 a day. Most of my users gave my forum up for Facebook. I design and manage websites for clients, and Facebook has become part of my job description. More and more my clients want me to know how and be able to utilize Facebook for their benefit. For someone who despises what it is and what it has done (and is doing) I still find myself spending more and more time there. I hope there will be a turn of popularity back to this type of social interaction, but I'm not seeing it.

As to the OP's question -
I've experimented with Burning Board, but never used a live version of it.
I've used IPB and it's superb. Couldn't ask for a better forum.
I've used vBulletin for a very long time. It's a very reliable forum.
I used MyBB for a short period and found it to be quite acceptable.
I've use SMF the longest and have extensive experience with it. Maybe it's because I'm more used to it than the others, but it always fit best for me.