Lloyd wrote:Charles, would you want your site to be such a platform? Would you like to have hundreds or thousands of people using it for mostly science?
Absolutely! I have put a lot of effort into facilitating collaboration, because I think that there is a huge amount of potential in it.
Lloyd wrote:If we get one science topic started for discussion on the site, we could invite CNPS members and others to join the discussion.
I think that we need to get a workgroup going, with a handful of people at first, to flesh out some of the folders, and to load test the features. Many of the folders simply have place-holder content, that I put in there just so that I could visualize working folders. Is there any chance that we could get Brant and/or Michael to get involved, to help flesh out the solar models folders? All kinds of data can be added to the Observations folders, and more of the arguments for/against can be added to the Hypotheses folders. zyxzevn is interested in the paranormal, and in particle simulations, so maybe he'd like to try fleshing out those folders. Then there are the team folders, and related functionality. So I think that a small workgroup should test drive the functionality, and the protocol that we develop, before putting out a press release saying that we have invented something.
Zyxzevn wrote:What I do miss in that site is that when 10000 people vote, you only see the idea that the majority has. So I introduced viewpoints. But that is not necessary on a small site.
I don't know what you mean by "viewpoints", but I had a related idea that I'll run past you. I understand the problem -- votes aren't necessarily useful, if the majority of people think differently from you, or if a circle-jerk network up-votes everything within their circle, and down-votes everything else. To get past this, I was thinking that votes could be weighted. So if you're reading somebody's blog, and you develop a high opinion of the person, you can give the person a high ranking. If somebody does a list of blogs, and they sort it by rankings, your vote will influence the sort order. But more importantly, every time that person ranks something else, that person's vote will have more weight for you. So if you have ranked "Fred" highly, and if Fred has ranked an item highly, in a folder that is sorted by rank, it will be like Fred casted 5 votes, instead of 1, thereby weighting his vote stronger, and this could affect the sort order. If you have ranked all of the rankers, then you'll see results that will be much more like your world view. In other words, the stuff that interests you the most, and/or that you consider to be high in quality, will bubble up to the top for you. Meanwhile, somebody else who has ranked the rankers differently would see a different sorting order, because the weightings will work out differently. It will be a pain to implement this, but I think that there could be an enormous amount of value in it.
Zyxzevn wrote:The idea to have "theory"/"experiment"/"observation"/"conclusion" is interesting, but it is also restricting
things that are hard to theorize.
I agree. That's why there are also the Documents and the Discussions folders, for free-form material. We could expect a lot more activity in those folders, than in the more structured Observations ~ Hypotheses ~ Predictions ~ Conclusions folders.
Zyxzevn wrote:It forces everything to have a observation basis.
So instead of having "big bang" as a section, we have "redshift" & "CBR".
Yes, redshift is an observation, and big bang is an hypothesis. Separating the observations from the hypotheses allows everybody to scrutinize the observations, to make sure that they're being presented in a theory-independent way. And then they can all take a shot at their own explanations. The downfall of an hypothesis is generally that the modeler wasn't aware of some relevant data. But since everybody only presents the data that support their own hypotheses, it's hard to find a complete inventory of the data. So this is what the Observations folder is about -- if everybody contributes the stuff that they know about, it will make it easier for modelers to check their models against the data. It also stresses the point that we're all trying to explain the same things, so we shouldn't all have to duplicate the labor of presenting the data too -- that should be factored out into its own folder.