Compare Forums- Part 1
SMF ReviewIn this section we compare website forum applications, together with installation choices regarding multiple applications on the same webspace - especially of course regarding CMS issues.
Part 1 - SMF review; the start page for the forum software comparison section - Introduction - Installing Forums - [this page]
Part 2 - phpBB review
Part 3 - Kunena and Fireboard for Joomla review
Part 4 - vBulletin review
In the spirit of the CMS comparison pages elsewhere on the site, we thought we'd write down our experiences with some forum apps. However, this is not an exhaustive set of forum reviews, by any means: we only report on five applications here. As usual, our forum reviews are honest and revealing, and come from an expert user's point of view. When reading our forum comparisons, you should carefully bear in mind our standard comparison and rating methods, which you can find at the link here: compare website software - our judging criteria.
You should also note that the reviews are only accurate at the review date - software is constantly updated. >> PLEASE NOTE THE STATEMENTS IN THE DISCLAIMER AT THE END <<
SMF Reviewreview date: 2008-02-15 SMF overview
The Simple Machines Forum is something of a benchmark in this area: many have used it and know it well. It is not the best choice for a giant forum but is recognised as ideal for most new forum websites and forum owners, as it is very suitable for small and medium-size sites, easy to manage, does most tasks effortlessly, and has many good features.
Installation
SMF is free, and is of the usual PHP - MySQL format for popular dynamic webapps. That means you must install it on a LAMP server, or any other server type that has PHP scripting and MySQL databases on it. Not a problem, since that's around 95% of servers. The only servers that don't always accord with this specification are exotic specialist servers and Microsoft hosting resellers. In these cases you should certainly vote with your feet, and move to a proper web host. All web hosts offer PHP and MySQL on their servers; if they don't, they aren't web hosts.
SMF installs easily and smoothly; you just need your database name, username, and password. There is one weird glitch we found though: using the Firefox browser, there is a problem regarding password allocations. With luck you won't experience this issue.
Documentation
On looking back at this section of my notes, there's a double question mark, like this: ?? It means I don't know (at that time anyway). It is the finest compliment you can make to an app, when you don't need the docs for anything. It all just worked.
Looking at the docs now, I would describe them as basic. There are a bunch of separate web pages at SMF central, but no PDF download. It would be nice to have a proper PDF, but of course getting the docs right is a tough job and they no doubt have enough on their plate. The score is 3 out of 10 for docs since they are not in an optimal format and are fairly sketchy - but as we stated, there are good reasons not to consider this a major sin; most of the time you won't need them. The only negative is that if you do need to consult them, we found that it is unlikely that the topic you need to research is likely to be covered, since advanced issues are not usually discussed.
In OSS, it is normally the case that only the very largest projects, or those with a core contingent of enterprise-level commercial users, manage to get the documentation organised and half-right. The most obvious clue to this is the existence of one or more books on the software; the existence of a couple of books invariably points to a mature and organised project. Examples are Plone, DotNetNuke, Joomla and OpenCMS.
Templates
Called 'themes' in this project. Very easy to change: just an upload via the application backend, no messing around with FTP or anything like that. A couple of clicks and it's done - couldn't really be easier.
There are some very good templates available; the commercial ones are well worth buying and some are fairly sophisticated. To qualify as 'excellent' they would need to include - as standard - features that expert users require: such as additional menus for other site links outside of the forum; module positions for additional content displays; module positions for banner adverts; module positions for PPC adverts; and so on. We didn't find anything like that in standard templates, though some advanced commercially modded sites had them; but no doubt as the SMF community becomes more sophisticated, these facilities will appear. As a basic requirement, a forum needs an additional menu to be available for external or additional requirements; and this sort of feature is probably not hard to either add via the PHP template or to have collapse if it is not used.
Plugins
It is very easy to add plugins: a smooth procedure that can't really be bettered. There is a useful 'test' mode that runs first, before you actually commit - this can find problems before it's too late. We really liked this plugin system and can't think of another app that does it better, even in CMS. The SMF plugin system serves as a benchmark in forums to show others how it should be done.
SEO
A poor SEO score for the basic app; which rises to merely 'not good' with plugins.
There has to be some sort of concession to the fact that search engines exist - and that people need search traffic on their sites. Here, the basic app just doesn't cut it, though of course that doesn't really matter in the world of dynamic webapps - as long as there are plugins to fix it.
URLs: there is a choice of two SEF URL plugins, and we tried one of them. The resulting URLs are OK, but not perfect (they look a lot like Joomla core SEF URLs, or ASP CMS SEF URLs for instance). However, the bonus with the SEF solution we used (PrettyURLs) is that it's all done within the app - there is no htaccess involved. That's really handy, as, with other dynamic apps in the same webspace, there could otherwise be htaccess conflicts.
Metadata: aarghhh! Boilerplate metadata - identical sitewide. The standard stuff can be replaced with a meta plugin, but that only means you get all your meta identical on every page, instead of the default meta. In fact you could do that yourself by hacking the files, all the plugin does is put a text box in the admin backend to make it easier. A module job if ever I saw one. And of course it's sending a boy to do a man's job: this needs a component-level plugin, to fix things at a higher level so that per-page meta can be input.
One day (perhaps when a core dev needs to make some money out of one of his own sites), they'll realise that search engines are important; and that search success demands unique per-page metadata. Four variables are vital here: title, description, keywords and follow/nofollow, index/noindex. (And please don't try and tell me that this meta or that meta is no longer necessary: this is my daily work and I can assure you that they are all needed and I'd bet you a million bucks on it if I had it - I sure couldn't lose. Any testing procedure will show you that 4 meta fields as a minimum are needed.)
Here, the title is changed automatically to be the post title; the rest are missing or boilerplate.
The single thing that stands out in all our forum reviews is that the software authors, universally, have no idea how important search engine traffic is to website owners.
Session ID problem
There are also worrying signs that this forum issues SIDs under some circumstances. Session IDs were used extensively in the past for ecommerce applications, in order to identify customers through the checkout procedure. This is no longer the case as they are now identified by cookies / IP / UA string (except in applications that still use the old technology). Since SIDs are not even necessary in ecommerce, why would a forum need them? Good question. It looks as if they are sometimes issued, and an effort has been made to avoid giving them to searchbots by identifying them as such; SIDs are then not issued to user-agents identified as genuine spiders. Looking in the database, there are a lot of spider user-agent identifiers. This approach is doomed to failure as it won't work all the time. Our own site forum is now issuing SIDs to certain bots, which demonstrates the fallibility of this approach; so we will probably attempt to disable PHP SIDs via htaccess or php.ini and see what happens.
Other problems
Not much really. This app does its job smoothly, and if you are reasonably lucky you won't even need to look at the docs. That's what we call praise around here - the real sinners have us poring over the PDFs like there was no tomorrow.
Of course, there are some niggles: there is far too much footer linkspam, for instance. Four to six footer links on every page is basically linkspam, no other way to describe it. One link per page is surely enough - it means the authors will get several hundred backlinks after all. Any more is crazy and doesn't make any sense. I suppose this situation will continue, though, until Google gets around to penalising the vast numbers of pointless sitewide footer links out there.
Something we really didn't like is the bot worm probes from SMF, or someone closely associated, trying to get private files off the server, soon after SMF was installed and various plugins and templates were being tested. This is a nasty thing to get involved with, by SMF or someone close. These may have originated from a template author, not the application authors. Hack attempts are not something we would expect OSS people to be doing. Yes, we monitor our server and see these things.
Admin usability
Not bad at all. A simple text-based admin panel that works as well as any text-based panel. If they changed to an icon-based structure it would of course be better by an order of magnitude (partly because many other things get improved as part of that process); but this is OK.
ACL
Basic but OK; nowhere near the levels provided by phpBB, though. Of course, that also means that the problems inherent in phpBB3 are not present, which is a tremendous advantage.
The visitor experience
A nice-looking forum that's easy to use by reasonably experienced visitors. Any visitor usability issues revolve entirely around the templates (or themes) here, so fixing any problem is a template issue. In practice, the app looks good and works well. As with any app there are new-user issues, but this is a universal problem. For instance, in our testing, we found that a new user unfamiliar with forums did not know from the outset how to post in the forums. This could be fixed by having a very prominent Help menu link. As stated, this would most likely be a universal problem.
Unfortunately, it isn't possible to add menu links on the popular templates except by hacking the code; and in any case:
(a) Such a menu link has to be extremely prominent in order to catch a new user's eye;
(b) There is a desperate need for a full menu bar that can have links added via the admin backend. Hacking the code to add menu links is just not on.
We sometimes include a 'How To Use This Site' page in a prominent position, on many of our clients' sites. There will always be brand-new users; and the more of them, the better of course. The trouble is that usability is hardly ever a priority; here, we would be inclined to think that a prominent menu bar, left free for user requirements, would solve this and other issues: a site howto page, forum rules, forum info and so on. Why is it that devs can't see this?
Conclusion
An excellent simple forum solution, as it says on the box. In fact it does many complex jobs well, and the clever thing about it is that it does the tricky stuff so well you think it's simple. The easy extensibility is an excellent feature. SEO is still in the dark ages but that is the case for many webapps, especially where the developers do not see commercial success or social popularity as a key requirement.
This is a good quick forum solution with some first-class features, and it will justifiably be the number one choice for many users. Usability is good for both owner and visitor, and if the SEO score were to be improved, there would be even more reasons to put this one at the top of the list. There aren't many ways this forum could be substantially improved - which says it all. The SEO is certainly one aspect - but if your site has other more important pages than the forum, then it's not such a big hit. The admin backend could learn some lessons from Joomla CMS - but what webapp couldn't?
Security: it seems that phpBB3 is now a little stronger in this department, but it isn't our field so we shouldn't comment here; phpBB2 was vulnerable to serious exploits (documented by Google), but this does not seem to have ever been the case with SMF.
Plugins: well, every dynamic web application needs more plugins.
Menu and module positions needed: undoubtedly.
SEO: as you can see, SMF certainly needs a lot more in the SEO department, for sure. Basically, though, this is a great forum that can only get better.
Forum Application: Tech Spec
cost: free
type: PHP - MySQL
version reviewed: SMF v1.1.4
zip installer size: 2.13MB
choice of templates: 100's
number of plugins (estimate): 500
central site: www.simplemachines.org
example sites: start from the central site's forum members
tags: SMF review, SMF comparison, compare SMF, forum software reviews
Comparing ForumsTo choose the best forum software for your needs, you should write down a list of the features you will need, and those you might want. If you compare that list against the advertised features of several forums, you can see how the land lies.
Then, install the software on your local LAN and play with it. This stage should show the good and bad points.
The basic features of most forums are very, very similar. Mostly, they are PHP apps that run on a MySQL database. The big differences are in SEO, ACL, and usability for admins. Unfortunately, there is little difference in SEO ratings: they are all lousy, though some are even worse than others. Some have slightly better prospects due to availability of plugins.
Some forums don't even have a plugin system, and they should be avoided like the plague. That was OK 10 years ago but times have changed slightly.
A critical factor is the ACL provision. There are a lot of things to weigh up here if you need very good ACL; most of us, luckily, will be more than happy with the basic ACL provision of virtually all forums. Trying to deliver a high level of granular ACL has failed miserably in the case of one forum application, with the result that admin usability is a total nightmare.
For these reasons, relative ACL provision would be one of the most important factors to check out on a forum feature matrix - there are sites that offer this facility. If the matrix didn't show detailed ACL or any of the half-dozen or so critical SEO features, then it would give the impression the comparisons there are pretty much a waste of time.
Real-world user reports are always a better proposition than feature matrices, especially where the users have no axe to grind.
Forums: The Visitor ExperienceThere are two classes of user for a webapp - the end-user (who is the person who owns it, lives with it, and works with it every day) - and the visitor (who might see it for 8 seconds in their life, total); a visitor-user if you like. Therefore the user (ie the owner) comes first in any discussion of usability issues etc. We can't forget the visitor, of course, since they pay the bills in the end. We already dissected the forums from the point of view of the owners - now is the turn of the visitors.
All the forums we tested had a very similar visitor experience - good. We couldn't spot any major issues here except those introduced by individual templates. It would be hard - if not impossible - for visitors to tell which forum they were using, if the critical labels were covered over. This is good as it means they all work and there aren't any notable problems. In fact, any problems at all are going to be introduced by the templates (or 'themes', 'styles' etc according to which app you are talking about).
The only issue of any consequence, which applies to all forums for all visitors, is that forum applications are not as advanced as CMS ones. Actually, that is too polite a way to put it; compared to modern CMS applications, forums are in the Stone Age with regard to SEO, admin usability, on-page assets, and just about everything else. As an example, it is fairly obvious (to CMS users) that there is a crying need for menu and module positions on a forum front page.
As these are only available in commercially-modded forums, there is a general lack of sophistication and modern website capability. Update Q2 2011 Finally, after years of hoping this would get fixed, forum software authors are beginning to see the problem. The capability to add modules and menus to the index page has been introduced in most of the apps, usually via a plugin. At last...
The average new forum user may have usability issues, since nowhere is there a prominent menu link to till them what is going on. We proved that, as our novice testers (office staff who had never used a forum before) could not work out how to post; they hadn't realised that you must register first, then the appropriate buttons become visible. And note that these testers use the Internet every day in their work: they are not total noobs. There are always a number of usability problems that only become visible when you run tests; and otherwise you will remain in complete ignorance of these issues.
Of course, how to use a forum is obvious to experienced users; but you will be turning away new users if you don't resolve these issues. New users are important and we see no reason on earth to tell them to go away. That's what you are doing if you don't carry out usability testing. In fact there is often a tiny help button on forum index pages - but usability tests show this is useless as the new users, who are the ones who need it, cannot see it.
Usability, of course, is ignored and despised by 99% of developers. If most of them never even validate their code, how likely is it they will take usability seriously?
What do we think forums need most?It isn't hard to answer that. Forum software needs to be dragged kicking and screaming into the 21st Century. All these and more need fixing:
- Realisation that website success depends on acceptability to search engines.
- Fixing the glaring errors such as session IDs, no SEF URLs, and no per-page metadata.
- A smooth plugin method, for any apps that don't have that.
- More on-page assets. This refers to all those things you can find on modern dynamic websites now that forums don't have (unless heavily modded by pros): additional menu positions (here, for links out to other site pages); extra module positions (for content, news, menus, adverts or whatever); banner ad positions; PPC ad positions; and so on.
- A default working forum directly after installation. This is easy to arrange on any database-driven application: sample data can be loaded. You could have an option to load it or not, for new or experienced users, as is given with enlightened applications. No forums we have worked with do the obvious, which is to have a working board on install. It is the new user's biggest hurdle - and completely unnecessary as it would be so easy to fix. And what you certainly don't want (except for LAN testing) is a ton of useless sample data that is always hard to remove totally. All you need is one single working forum - PLEASE. It can then be cloned for the next one.
- Visitor usability tests show us that prominent menus are desparately needed.
- Admin usability doesn't need testing: just look at the steam coming out of the ears of admins using some of these applications. Grim is not a strong enough word. One forum app even takes the award here for Worst Application Ever Tested, as regards admin usability. Because it's very hard to criticise open-source applications too much as it's all done for free, we didn't put this particular label on the offending application's review page here; but you can probably guess which one as there are plenty of hints.
SEO for forumsAs a rule, forum SEO is lousy. Really, really terrible. The developers just don't seem to get it: people need traffic on their websites - and it mainly comes from search engines.
In general forums are many years behind other web applications here. There is a brand new forum series that doesn't even have a plugin system. None of them have additional menus that can be added, or any of the other obvious on-page assets that are needed. It's hard to work out what the developers have been doing for the last 8 years - and this applies to almost all of them.
The list of things wrong with forums for search compliance, accessibility, usability, marketing or any other component of SEO is seemingly endless. It's the single biggest reason why the Kunena forum looks an attractive option; as a plugin for a good CMS, all those problems have already been solved by developers who had a vague idea what the meaning of the word quality is.
Forums have a long, long way to go before they are worthy of existence in the 21st century. Part of that process will have to be some re-education of the developers, who mostly seem to have been living in a cave for the past eight years.
Forum installation options and issues
Forum as pluginUsing a forum as a plugin for a CMS has pros and cons.
The disadvantages are fairly short and simple:
1. The apps are then joined at the hip. Any problem (like an exploit) will affect both. Again, database issues will affect both - since they are on the same one.
2. A plugin forum is probably not a good idea for sites with over about 10k visits per day. It's just too much extra load, unless your hosts are A1 on load balancing. Even so, you'll have to start with load balancing long before the point where this would have been necessary otherwise - forums place a lot more load on a server. It depends of course if the forum is a very busy part of the site or not; if it gets less than 1k visits per day then there should be no problem. 3. Plugin forums are very limited in functionality, compared to discrete forums. 4. These forums are good for basic use where they are simply an addition to the main website and get maybe 25% or less of the site traffic.
The advantages:
1. It's all in one box, there is only one base webapp to worry about.
2. The benefits of your CMS are often transferred to the forum. Since forums are still in the Stone Age as regards SEO etc, this can be a big plus. It's likely that the SEF URLs and metadata situation will be vastly improved.
3. You can use your CMS templates in the forum. This could be useful.
4. Single sign-on, just one username for login.
5. There won't be any membership or htaccess conflicts as it's all under one roof.
6. You can use a good forum like Kunena for Joomla, that doesn't work standalone.
7. It's likely that the forum admin will be much simpler, for many reasons.
The bridged forumAnother way to use one of the big-name forums is to bridge it to a CMS. This has some of the advantages of both the independent set-up and the integrated one. A bridge normally employs a small middleware application that sits between the two main apps.
The pros:
1. You can use a well-known forum within your CMS, as long as there is a bridgeavailable. 2. It's all under one roof, so admin will be easier.
3. The DB and membership lists are common - you get a single sign-on.
4. Some of the CMS advantages might be transferable to the forum, depending on the bridge type and functionality.
The cons:
1. A problem with one sometimes affects both.
2. The DB may be common, this can be both a pro and a con.
3. A heavily-loaded site may not find this the best arrangement. Bridged programs are the slowest of all, and cause the highest server load, of any of the choices. That's because you are running two applications at the same time, plus a bridge, and running traffic across all three. But for this to be a negative, traffic on the forum needs to be higher than the CMS traffic. Low-traffic forums are not a problem.
4. In the case of a site that experiences a high level of attacks and probes, any kind of integrated forum may not be the best solution.
As you can see, essentially this racks up the same as a plugin forum, as the issues are the same. The main difference is that you can use a big-name app that will be far more capable than a native CMS plugin.
As a working example, you can bridge SMF to Joomla and get the best of both worlds. There are then some superb RocketTheme J templates that put other forum templates to shame.
Another well-known bridged forum is vBulletin with Drupal - a quality solution by any measurement. There are even several choices of bridges to use here, but we would recommend the simplest of all. The most complex one effectively makes Drupal a plugin to vBulletin and that may not be what you would want.
Separate forumA co-installed forum - one installed on a website where there is another web application such as a CMS - has pros and cons, like any other method. A standalone forum like this has the following advantages:
1. It is the best solution if the site has a high level of probes and attacks.
2. It is best for very high traffic sites.
3. As it will use a separate database, this means it is totally discrete and problems are isolated within each application. They are firewalled off from each other.
4. It will also be easier to change out the application for another, as it is completely separate.
5. You will need to carefully consider the directories and the htaccess situation. The index.php and htaccess files will certainly conflict, so the CMS should reside in the webroot and the forum (and its htaccess file) should go in a directory.
And of course there are disadvantages:
1. The SEO factors: as forums' SEO is usually terrible, a native CMS plugin is best - but has the least capability. 2. Management will be more complex as there are two separate apps to deal with.
Standalone v integrated: pros and consCMS owners may therefore ask: should I use a plugin forum, or a bridge to integrate two separate apps; or should I co-install them (ie install each separately in the same webspace) and thus keep them separate? The answer depends on your reaction to the questions above. Probably, with a very busy site, it would be best to keep them separate.
To minimise problems with one affecting the other, keep them separate.
For the smoothest possible solution, and the easiest to manage: integrate them. Without a doubt - since individual forum SEO capability is so appallingly grim (at least, for the ones we tested) - this is the best for SEO. This will certainly be the best way if you will run fairly simple set-ups with minimal ACL and so on. Using a CMS plugin forum is only viable if you need the most basic functionality, though.
Sub-domain forums
Don't install a forum to a subdomain unless advised to so by expert hosts in communication with expert SEO advisors. There are a bunch of reasons why this is now a bad move (though it was used in the past). Essentially, this is a completely different website and unrelated to your main domain. If you do this then any reputation each section has, and all other benefits, will not transfer to the other section. It results in an URL like this: forum.a3webtech.com
Instead, install your forum so the URL is in a main site directory like this: a3webtech.com/forum/
The forum can profitably be named for its subject, like for instance /art-forum/ and so on.
Further considerations
There are some other considerations that may or may not affect you, such as:
Statistics
Will the installation type affect your web analytics? Usually no - the stats will work the same for an integrated or discrete forum. However, this does not apply when the stats app has a problem identifying some of the CMS data. This happens when it cannot reliably 'see' some of the CMS pages correctly, and either misses or misinterprets some of the data. This can happen when the pages are of an unusual fileformat, or when the CMS is of a complex type rather than a straightforward PHP application. So in this case, a separate forum will work better as far as stats reliability goes.
Marketing and image
It may be convenient to keep the forum reasonably distant from the main site's business; or perhaps the two should be seen as inseparable. A sports club's website would probably benefit from a fully-integrated forum; an online bank would no doubt keep the customers' forum at a little distance. ___________________________________________________
Please note:
1. The reviews can only be accurate at the review date. Software is constantly being updated (or should be), so one month later some details may have changed.
2. These reviews are the personal opinion of the reviewer. They are an honest appraisal from that person's point of view - but you might disagree entirely. Different people have different expectations and viewpoints. Where we have criticised, you may disagree completely. Where we approve, you might not. We only give a viewpoint, which may or may not be factually correct.
3. There will inevitably be those who completely disagree with our interpretation - and you should consider this fact. Please be aware that there are users of any given application for which we provide reviews who would probably disagree with comments made in the reviews.
YOU MUST TRIAL THE SOFTWARE YOURSELF IN ORDER TO COME TO YOUR OWN CONCLUSION.
|
SMF Review
