OpenWetWare talk:Blogs

Should we have a 'real' blog available for each OWW user, using wordpress for instance?

 * Reshma 14:47, 29 August 2007 (EDT): Is there such a thing as importing blogs? For people who want to move their blog from their existing site to OWW?  Or vice versa?
 * Ilya 15:15, 30 August 2007 (EDT): Yes, WordPress MU (the blog software we are using now) supports importing blogs in many different formats, including Blogger, LiveJournal and even just RSS feeds. It can also export data in WordPress eXtended RSS (WXR) XML format which would contain posts, comments, custom fields, and categories.

Previous discussion

 * Jason R. Kelly 09:45, 7 June 2007 (EDT): Austin and I have been discussing this via email, I'm moving it here to get more input.
 * Jason R. Kelly::how hard do you think it would it be to give users blogs through OWW? you think it will be a disaster? Or should I just sign up for blogger?
 * Austin:It depends on what kind of a blog you are envisioning. These are the possible solutions I can think of:
 * each user signs up for a different blogging service. no integration at all.
 * run wordpress or similar and do some kind of integration with oww. You get whatever bells and whistles of 'real' blogging software but it becomes a bit disconnected from the wiki
 * blog on the wiki itself. good integration. The only question is what features of a blog you really want and whether it's easy to implement on the wiki. The basics like showing the most recent posts is easy (e.g Blogs) but Mac says it's not a blog because "it doesn't look like a blog." I don't know what that really means but even if we dewikified it and put some skin on to it e.g. it may still not be a 'real' blog. RSS feeds should be doable using the filtered recent changes. User comments are easily doable if we want to install an extension. Other features I don't know about.
 * Jason R. Kelly:Yeah, I think you need to be an actual "blog", e.g. calendar on the side, place for comments, etc. Plus people will want to use stuff like feed burner to keep track of who subscribes, how many hits they get, etc.  All the exciting stuff about self-publishing!  If you don't have that, you lose some of the incentive. So i like this option:
 * run wordpress or similar and do some kind of integration with oww. You get whatever bells and whistles of 'real' blogging software but it becomes a bit disconnected from the wiki
 * Austin: Like I said, these features are just code. For example, if we run our own wordpress, we'd also have to use a plugin to get feed burner to work. Getting it to work with the wiki almost certainly isn't that much more difficult. In the end, almost any feature could be implemented on the wiki, so feature list shouldn't be an issue. The primary difference is a psychological one. On a wiki, you are implicitly saying that anyone else is welcome to come and edit your post.
 * Jason R. Kelly:Open question is would users actually use this rather than blogger? http://scienceblogs.com/ is an example -- the bloggers can only customize the header graphic, o/w all the blogs look the same. Don't know if something like this would be popular w OWW users, but we could find out.
 * Austin: I see no reason to run our own wordpress farm. Why not just have everyone blog on scienceblogs? What's the benefit of connecting it with OWW in this case? The only benefit of us hosting it if we're going to implement closer connections with the wiki (e.g. perhaps can use wikilinks) but in that case, it seems easier to implement on the wiki.
 * Jason R. Kelly: ScienceBlogs handpicked bloggers to be in their group, I can't get a scienceblogs blog. The benefit for OWW is that we will be encouraging scientists to communicate freely.  Some people would be more comfortable with a blog than the wiki since it's "their voice".  It seems possible that a scientist who wouldn't go sign up for blogger might click a blog link from their OWW userpage and start blogging.  This may not be true, but I'd like to run the experiment.
 * Austin:  I see. So we'd have to make an extension to the wiki no matter what to at least add a blog link.
 * Jason R. Kelly: I don't know whether it would be more work to make the wiki look like a blog, and have all the features of a blog, rather than running a wordpress farm. Any idea which would be harder?
 * Austin:I have no idea how hard it is to run a wordpress farm but in the end it just means an extra piece of software to manage, separation of the information out there, and in general it means people need to go to multiple sites to get their information.
 * Wordpress is about 51,000 lines of code (as a comparison, mediawiki is 234,000 lines of code). The benefit(?) of wordpress is you get all those lines of code and they get updated for you. The disadvantage is the complexity it adds and what we have to maintain. Most of those lines of code probably overlap with the wiki's functionality.
 * I think if we wanted to, it wouldn't be hard for a tech to implement most of the essential features of a blog in a wiki in a very short time (could probably copy lots of code from such as things like trackback). I can't imagine the wordpress farm being that difficult either. But in my mind, using the wiki is a cleaner long term solution.
 * Jason R. Kelly:Why is it important to integrate it with the wiki? To have it show up in recentChanges?
 * Austin: Just like you said that scientists may not sign up for another blogging service, having it actually on the wiki means they don't have to learn another interface and other things that go with it. I don't know how complicated wordpress is to set up and administer, but do users really want to mess with different blog settings, skins, etc (if they do, then we should definitely go with wordpress).
 * Austin: The other thing is if we're looking at blogs, forums, etc., aren't we looking at just running a CMS like drupal that can do them all together?
 * Reshma 10:32, 7 June 2007 (EDT): I think that regardless of the implementation details we want some integration with OpenWetWare to promote a more uniform experience and to highlight that your blog posts are still part of your OWW contributions.
 * Same login (ideally, logging into the wiki logs you into the blog)
 * The OWW blogs should have a similar look and feel to the wiki (i.e. similar default graphics and color scheme ... we can let users customize this but the default should be the same)
 * The blogs should promote a similar mission as the wiki: making content open and available via the same copyright licenses (so content can be moved back and forth from the blogs to the wiki) and encourage folks to "sign" their work (i.e. no anonymous posts).
 * The URL for people's blogs should be in accordance with their username. (Perhaps we also need to enable lab blogs or group blogs.  Not sure how to enforce naming conventions on this).
 * Jason R. Kelly 18:55, 7 June 2007 (EDT):The first post of a new blog could be guaranteed to be linked to from the OWW front page, as an incentive to blog on OWW. basically I look at this as a hook to get scientists more aware of the OWW ecosystem, some people are probably more inclined to blog (or read a blog) than to jump on the wiki right away.
 * Ilya 17:17, 14 June 2007 (EDT): I don't think we should outsource any of the services we provide because this way we lose flexibility and control over them. Also, I don't think making a blog out of a wiki is a good idea.
 * Worpress is easy to install (already running at http://openwetware.org/blog) but it doesn't provide multi-blog functionality out of the box. There are hacks, similar to the wiki farm we are using but they are not scalable enough to provide every OWW user with a blog.  However, there's WordpressMU which is a separate project (based on Wordpress) designed for hosting multiple blogs.
 * It may be more useful to begin using a CMS or a social networking software instead since it would provide additional functionality we are interested in (e.g., forums). However, free social networking software doesn't seem to be mature enough at the moment, so it may be useful to focus only on CMS.
 * I think to answer this question, we need to figure out what kind of functionality that we'd like to have (as a prioritized list) and then I could do some research to find out what application would be the best fit for our needs.