Great feedback guys, please keep it coming! We're pretty happy with how our new publishing initiative is going so far, but we do of course want to keep improving it so that we can provide the best possible resources for everyone! Any and all feedback is welcome and encouraged, even if you're just throwing out an idea that you think is too "out there" for us to use -- we'll consider everything, and even if we don't end up using a particular suggestion it might lead to a similar idea we do use.
For anyone not already aware of it, this is all part of our Game Developer Library Project to try to collect as many great resources on as many (relevant) topics as possible. As the collection grows again we then want to do our best to filter that content for the best quality and that which is most relevant to any given user. All articles have a very basic editorial review before being approved for the site, but this is just to look for obvious formatting errors or typos and to filter out blatant spam or completely irrelevant topics before they reach the community.
We then currently provide three methods of providing feedback so that we can filter by (or improve) quality:
- The current peer-review system allows selected users (currently Crossbones, staff, and moderators) to either mark an article as peer reviewed, with the following prompt: "Does this article meet the quality standards to be presented as a peer reviewed article to fellow GameDev.net members?". It takes three votes in favour for an article to receive peer review status, which suggests at least a minimum level of quality and confers a reputation bonus to reward the submitting author. Perhaps more importantly, those with peer review status can also vote that an article does not meet those minimum quality standards, choosing a main general reason from a brief list of options. An article which reviews three of these votes will be removed from public view into a special category so that the author may try to improve and re-submit the article if they wish.
- We allow any registered user simple up/down voting on articles, in the same way you can vote on posts in the forums, and the total is displayed at the top of the article. A significant number of up-votes generally indicates lots of people found an article helpful or at least interesting, whilst a significant number of down-votes potentially indicates some problem.
- You can comment on articles to offer your own commentary, suggestions or corrections. In a number of cases already this has resulted in edits to the original article to correct mistakes or add additional helpful information or links.
Obviously this still isn't a perfect system, but it's very simple to use and seems to provide a good starting point for filtering out the worst articles, identify the best articles, and to provide a feedback loop for improvement of any article. We think it's a great starting point to build from. You can read a little more about the process in "how to publish on GameDev.net".
Second, has the deluge brought any (serious) talk of an Article Tribunal (for lack of a better phrase)?
The site used to have an "editorial review board" that examined every article before approval, and whilst this did result in a higher minimum level of quality it a) created a lot of work for quite a few people, and b) introduced some problems of it's own, including:
- It was a much lengthier process for an article to be posted to the site, with authors sometimes having to wait weeks (or in a few cases even a month or more) before their articles were published, and we weren't able to accurately judge this lead time in advance. Along with other factors, this contributed to author frustration and discouraged future submissions.
- We didn't always have an expert on a given article topic. This meant either holding the article while we found someone, or having someone do a non-expert review.
- A lot of articles were just swallowed up by the process and never saw the light of day, as the ERB would suggest some small changes but the author would sometimes never bother to re-submit.
I'm a bit afraid anything I would write would be too obscure
If it's relevant to game development we're happy to have it -- you just might find that over time a significant number of people find a more obscure article helpful, and for more unusual topics there tend to be less already written materials available.
I wouldn't suggest slowing it down (or speeding it up) any more than it already is
We would actually like to speed it up, but only if we'll a) still be able to provide a steady stream of content rather than bursts of faster articles, and b) able to better filter the content so you only see the best and/or most relevant stuff.
Ideally I'd like to add a second layer on top off the current system for peer review. The current system would still be enough to approve articles onto the site, but we'd add an optional tagging system, where, for example, one of our DirectX MVPs could come along and tag a give GPU-related article as being "validated" by a subject-expert...
This is something that's been suggested a few times and that we've discussed quite a bit, and we're considering implementing some variation on it. The biggest challenge is to ensure that we can accurately identify exactly which subjects someone qualifies as an expert for and only allow them to tag those articles. Ideally we would also try to find a way where we don't have a constant stream of potential "experts" to be vetted and to remove barriers to the vetting process, as this would be yet more work someone would have to do.