I suppose it also depends on what software. GoogleDocs are not Open Source but being freely accessible takes the sting out of that. - Jean-Claude Bradley

I think the point Ian was making, getting to the data, and having access to data is perhaps more important that how the software works. Doesn't mean open source isn't important, but in a service based world, that open data gets even more important - Deepak Singh

Well without the data there is nothing for the software to crunch :) - Jean-Claude Bradley

And without the software, there where to put the data :) Seriously... it's because of software that we now actually see why OpenData matters... if we had not the tools to process it, we just as well redo the original experiment and see for ourselves... - Egon Willighagen

Well if we had the data, we would be able to and develop our own toos. I think the key is that they're all important, and depending on the specific usecase, one might be more important than the other - Deepak Singh

Deepak, that's how I ended up in cheminformatics... I had data, but no tools... - Egon Willighagen

I think openness is equally important for both of them. In the end it is all about "standing on the shoulders of giants" instead of wasting time for unnecessary redundant work (creating software or data sets from scratch). Transparency is another advantage of it. So no, I don't share the opinion of this statement. - Konrad Förstner

Konrad, absolutely! I've seen several people reimplement subgraph isomorphism /MCSS (which takes half a PhD year to get it right) which could have been spend on actual data analysis... - Egon Willighagen

I think all of this is both true and wrong. Sure open data and open source are great. But the progress of science is made from knowledge and ideas. Those are captured mostly in papers not data or programs. - Jonathan Eisen

It also depends on the data. Some data should never be freely exchanged. - alphaxion

Jonathan, knowledge is nothing more than models that fits the data. If people only publish the models, without the data, then making alternative models, or just prove the published model wrong, is made unneededly difficult. And one major obstacle why published papers do not contain data, is the like of good tools. - Egon Willighagen

Egon - Access to data is fundamentally important. Access to tools is too. But if we cannot read the papers (or indirectly get access to the knowledge contained in the papers), all we have is an overwhelming amount of information and some toys to play with it. Standing on the shoulders of giants is not about using their data or their tools per se. It is about building upon their discoveries and their knowledge. - Jonathan Eisen

Jonathan, OD and OS are equally important to me. Knowledge is a derived good. Data is more important to me. Gigantic knowledge is nice, but showing that that knowledge is actually gigantic or crap, often still requires peer review of the data. Getting cutting edge data, requires expensive analytical devices, or many year organic synthesis... access to the raw data is much better than Open Access to some PDF hamburger, as the peer review will be much more independent and allows much more efficient ... - Egon Willighagen

... establishment if the acchievement is indeed gigantic or nonsense (or fraud for that matter)... - Egon Willighagen

It's easy to stand on the shoulders of Darwin, or Newton... it's more difficult to stand on the shoulders of Nobel prize winners, and it's impossible to stand on the shoulders of future Nobel prize winners and smart scientists otherwise... that's all. - Egon Willighagen

Egon - again, I support OD, OS completely and have been pushing it for years. But I guess we will have to disagree on the OA part. In general I think one is able to use the data and understand how the software works only because directly or indirectly one has had access to ideas, conclusions, discussions, etc. Production and release of OD and OS is wonderful. But as someone who has produced reams and reams of genomic data, I know that I am primarily judged by my papers and not releasing data. - Jonathan Eisen