Is Jeff Atwood really a visionary?

He is best known for being co-founder of the StackExchange group of websites. His reason for creating it was to make a Q&A forum and to unseat the then popular ExpertsExchange. ExpertsExchange is still around but no longer ranks as high as it once did in Google page rank.

Jeff is also a prominent blogger on where he is best known for saying bad things about PHP. He is also a champion of the Ruby language. On one of his posts he said he wanted to make Ruby projects as easy to install as PHP on shared hosts citing PHP’s famous 5 minute installs. You can even see that comment on one of his new projects on page. Discourse is Jeff’s new forum project which claims to bring forums into the 21st century and be a game changer. I tried to download and install discourse and after 3 weeks I still haven’t achieved it. (Admittedly I am no Ruby on Rails expert, I have much more experience with Python and more recently PHP. I tried Rails years ago but that was back in version 2 and I’m completely lost now.) However I installed Vanilla forums in 30 minutes with no prior experience with Vanilla and Vanilla is written in PHP!! So I would say that, in my opinion, that Discourse forums is not an example of Ruby 5 minute installs being achieved.

Also I’m not sure that Discourse is that visionary. For instance we all know forums are often beset by the wrong type of people such as spammers, trolls, etc but are we really able to deal with people problems with technology. Do we really want StackExchange style robot moderators for our forums? Is he really creating software that can’t be written in a language, such as PHP, that wouldn’t need 1 Gb of memory to work. Yes, you heard right, Discourse needs 1 Gb of RAM and no one on the Discourse discussion board can tell me (or wants me to know) why the high memory requirement. Could this just be code bloat? I wonder.

So here is my question: Is Jeff Atwood really a forum software visionary or is he simply riding on the fame and high page rank of stackexchange?

Ha, very interesting post @Kiwiheretic; and somewhat timely from where I’m sitting. I looked at Discourse for another forum that I am starting up and as soon as I read this on the about page:

The state of forums has been unchanged for so long that forums are considered unworkable and undesirable; few sites want forums any more because the software is so poor.
… I was put off. It comes across as arrogant to those of us that don’t consider forums unworkable or undesirable.

I then looked at it again more recently with a view to moving SItePoint over. I have tested much more thoroughly in a sandbox and I actually really like the software. It has a few holes if you’re looking for a really good community tool, but it has a lot of other benefits and I’m really impressed.

I think it’s a bit of both. I have a lot of respect for Jeff as a developer, but I feel like there are some fundamental holes in his software when it comes to the psychology of a community. So the jury is still out.

I’m interested to hear the perspective of others.

Yep, we’re discussing Discourse a lot since @HAWK; brought it to our attention.

The software is very dynamic and it has many features that I would like to see in this forum. Some other, I don’t see them that exciting. The product is still unfinished but it has a lot of potential.

Who knows!

I think Discourse is a great example of someone thinking slightly differently about forums, trying to be innovative and trying new ways of things. That’s a good thing IMHO, if people don’t keep innovating the web will start to stand still and not evolve.

Well, perhaps Discourse does have some new ideas but now that J. Atwood has popularised those ideas I predict that they will also start to appear in other forum software if they are not already.

I dont think Discourse will “take the world by storm” for the following reasons:

  1. There is little in the way by documentation and lack of documentation is always a deal breaker for me. (OK, its still in beta so maybe that’s still coming.)

  2. Its extremely opinionated software. They only officially support 64 bit Ubuntu with Postgresql. I guess if one thinks that J. Atwood’s design decisions are wonderful simply because he made them that may be no problem for you. However, there are those of us who still actually like Centos and MySql.

  3. Its huge memory requirements mean that you are not likely to be able to install it on your hobby vps or shared hosting plan at Hostgator or GoDaddy. In one thread, over on, Mr Atwood said he saw no reason why 32Gb shouldn’t be required as ram always gets cheaper over time and he is working to a 10 year plan so he is not concerned about not having bloated code!!

My belief is that the “forum of the future” will be built on good open source javascript libraries where most of the presentation work is done on the browser. (This is the vision the developers of the MeteorJS framework have.) This would keep network traffic to the web server down to a minimum and making the backend server application simply a session manager attached to a database.

The only problem with this is that mobile devices may not all be powerful enough to do that kind of processing on the front end, yet! However I think thats the way its heading and I personally think Mr Atwood’s crystal ball is broken when it comes to Discourse.

I should add that as far as I can tell Discourse doesn’t support Tapatalk Wheras I did find one for Vanilla forums.

Also if you intend to write your own plugins then I feel that would be easier with Vanilla.

The big reservation I have about personally maintaining a forum in rails is that it encourages monkey-patching. (I run a small forum currently using MyBB but I dislike how they don’t adhere to 3NF for database tables.) I’m looking at other forum software options myself. I dont really want to see a rails version of phpBB which also suffers from monkey-patching plugins.

I can just see it happening with discourse, there could be one plugin that totally wipes out the reliability of your whole forum in one foul swoop.

I wouldn’t think it really needs to, though. Like StackOverflow, it looks pretty nice as is on a small screen device.

I still think Tapatalk support is an advantage even though it may one day become irrelevant I dont believe that time has already come. Tapatalk uses less bandwidth and caches content offline which is very useful for mobile devices that are temporarily out of network range.

Maybe one day there will be internet everywhere including the amazon jungle. However I believe that time is still a long way off.

So its not just about how it looks.

Speaking personally, I’ve found Discourse difficult to use and difficult to navigate - which makes me wonder how easy it would be for somebody using AT. (I have asked, but haven’t had a real answer.) Keyboard support is virtually non-existent, although to be fair, I believe they’re now working on that.

It feels to me like yet another example of the software/app becoming an end in itself, and to pot with the users. It only works with modern browsers, so anybody using an older system is automatically excluded. When I learned web design, I understood the basic principal to be Tim Berners-Lee’s idea of universal access. Make sure the basic site is, as far as possible, accessible on all devices, and then enhance the experience for those on more capable systems. The current thinking of many sites seems to be that it’s fine to exclude large sections of the population, based solely on their access to technology or ability to use it. :frowning:

A point that I seem to find difficult to deal with is the lack of familiar symbols. The “unintuitive” blank page look. I would imagine that people who think abstractly, such as artists, programmers, etc, find it easy to “connect the dots” and work with. I must tend to think more concretely, relying on more familiar symbols or associations that make sense, because I find myself having difficulty “connecting those dots” in the Discourse environment. Discourse strikes me as too “sterile” (like StackOverflow) to foster a sense of community. It’s not a social environment. As such, it is an exclusive, not an inclusive, piece of software. While SO wins hands down for those quick in-and-out impersonal but usually technically competent posts, I prefer the feeling of a community, but maybe the social trend today is otherwise.

Yes, I think that is a good point. I wonder if they have become so “tunnel visioned” with their 10 year plan that they’ve lost touch with the present.

I also think its important to realise that Discourse is not a community driven project. Its a venture capitalist driven project with “community” clothes on. See this thread:

This thread above was a real eye opener for me.

codinghorror co-founder
Thanks for the effort! Unfortunately this (and by this I mean per-user language toggle, on top of the per-site language toggles we already have) is such a large, invasive, breaking feature change across the surface of the server and client that we would have to effectively stop all other feature work for a period of weeks or months to implement it. I don’t feel that would be a wise use of the limited venture capital funds we have at this time.

I agree, the “venture capital funds” is rather alarming, especially since it’s open source.
If I submit code and it’s accepted, do I get a share?

That’s an interesting question.

I think that Jeff Atwood does have a vision. Whether that vision and mine have something in common is yet to see :lol:

The software has great potential. If that potential leads somewhere is yet to be known. As per today, it is not very intuitive for me (but maybe because it is kind of SO-ish and I rarely visit SO, only when my Google search drives me there and I may not as familiar with how things work there)

I does have some features that I like such as posting youtube videos. I can simply copy the youtube url and the paste in my post and that’s it. Those are silly things but that make a bit of difference. It is not that it is very hard to do in a classic forum like this one but every little helps.

In the same way, there a little things that I find hard to find, if I find them at all… and that also annoys me.

But I think is good that someone brings a fresh approach to the same problem (venture or not). If he is able to listen to feedback, his product can be even revolutionary.

So is vBulletin :wink:

Well, the code is open & free, so its not as if they’re making money selling it. What they’ll be making money off is by offering hosting services.

Another example is WordPress which is an open source project and many contribute to it including Automattic & Audrey Capital folks (both companies by Matt Mullenweg). But Automattic provides a PaaS using WordPress and makes money off it. Atwood has mentioned the same on Discourse website as well. But he did say that they’ll know best since they wrote it, he’s forgetting that Zend is the major force behind PHP and Zend Framework isn’t the best out there, eh! :wink:

Is it against the open source use agreement to create a business model around offering a service to easily deploy forums using discourse using paid hosting? In theory I see something like this lending itself very well to replacing Proboards or something similar. That is if it is so *easy to deploy and customize which I’m a little skeptical off since marketing mumbo gumbo always seems to say something like that for every technology based product.

Not if my experiences are anything to go by and I have posted questions on a number of forums including the official discourse forum but generally I get nil replies. See my post I made on sitepoint linked below:

I suspect its not in their (Jeff Atwood & Co) best interest to make it easy to install because its more hosting $$$ for them.

I often find installing software to be a Bear. Trying to install Discourse to localhost is no exception, I didn’t manage to succeed my first attempt and I’m not looking forward to my next effort to do so. I’m stubborn persistent and I will keep trying, but I would definately not call it an “easy install”.

Of course there is always

There prices seem to be competitive. If sitepoint is determined to go down the route its going then I think a hosted solution is the best for them. At least until the product matures.

If you, as a rails expert, is having this much trouble installing it then I don’t feel so bad as a Django/python guy not having it installed in a couple of hours.

The other proposition is to get a DigitalOcean account for $10 per month, which gives you about 1 Gb of ram, and install it via Docker on Ubuntu and nginx. I haven’t tried it myself but the replies on the thread sounded promising.

Unless sitepoint has a good SLA (service level agreement) with Jeff Atwood’s I think they’re asking for trouble. I would at least wait until the product was out of Beta before pushing it to a production environment.

It sure will get interesting when their cellphones start going off at midnight because their site has gone down, hacked or both. Not a job I would like. I hope their staff are paid well for overtime.