Linkfest

Linkfest 2018-08-06: Now with more science!

Barrow Steelworks

By unknown - 1877 or earlier, republished by University of Strathclyde project - http://victoria.cdlr.strath.ac.uk/browseTimeline.php?group=&year1=&year2=, Public Domain, https://commons.wikimedia.org/w/index.php?curid=14652342

Somehow I had never really captured the term, Second Industrial Revolution. This is the far more interesting one that came in the late-nineteenth, early-twentieth century. This is where we got electricity and steel and mass production.

A long journey to reproducible results

Reproducibility is often an afterthought in science, which means it is often quite hard to *actually* reproduce someone's results from their method section. Sometimes it is hard even if you call the scientist and ask them how they did it. True standardization is one of the fruits of the second industrial revolution, but we have forgotten how to use it.

Plan to replicate 50 high-impact cancer papers shrinks to just 18

A high profile project runs into trouble because of a lack of attention to standardization and reproducibility when experiments were first run. If you have experience doing this, it can be easy to help the next experimenter down the line. But you only get that experience by doing it....

Not a problem limited to the sciences either. One of the ways in which you can enable replication is to make all of the intermediate products of your research available, which I think ought to be a wider practice, especially for publicly funded research. With the raw data, and the analysis script(s), you can then run the numbers yourself and see what happens. With online appendicies, this could be easy.

A fine thread on the implications of the ability to make guns at a craft scale instead of the factory scale. 3D printing isn't the real issue, it is about machining know-how and a ready market in non-gun parts that can be turned into truly functional modern firearms.