3 ms·
I once produced an open source product while working for Berkeley Lab, which is a DoE lab operated by UC. It does non-classified work. One day we got a visit
by dekhn 8y ago
I once produced an open source product while working for Berkeley Lab, which is a DoE lab operated by UC. It does non-classified work. One day we got a visit from a couple employees of Lawrence Livermore Labs (classified research) and my manager glanced at their badges- he said something about "Q" and smiled while asking what they worked on.
They said "multiphysics combustion codes" and my manager looked at me and said "they run nuclear weapons simulations" (they didn't confirm).
After chatting for a while I realized they were using my code to manage nuclear weapons simulations. That's a consequence of licenses like MIT and BSD and Apache: you don't control who uses your software.
I came to terms with this a long time ago but it's still odd to think that I can write software that my own government will use to ensure our destructive weapons work as expected.
- imranq 8y agoThat’s an amazing story. Do you mind sharing the name of the library?
- dekhn 8y agopyglobus: http://toolkit.globus.org/toolkit/docs/4.2/4.2.0/common/pyglobus/ http://toolkit.globus.org/toolkit/docs/4.2/4.2.0/common/pygl... it was a grid computing library that made it easy to FTP large files at high performance and do automated job management.
- Red_Pill_Phil 8y agoDon't share this...
- imranq 8y agoThanks!
- nraynaud 8y agoI wouldn't mind too much bing in your shoes: you have not developed your library for that, it's not its main use, and you didn't even know about it. I work in cloud related things (I do UI stuff for the cloud), I'm pretty sure 3-4 contracts from me, the code I work on could be used for weapons, but I'm clearly not developing a weapon nor something whose main goal is killing.
- erobbins 8y agoLook at the bright side.. accurate nuclear weapon simulations led to the end of live weapons testing.
- saltcured 8y agoPretty much everyone working in the high-performance computing space has to think about this. If you produce something truly useful, it may find its way into use in those kinds of HPC workloads. At the same time, much of our "civilian" use of HPC is being subsidized by those kinds of national security workloads which fund a lot of research and infrastructure. You might also think about it positively. At least they are blowing up nukes in simulations instead of out in the desert somewhere upwind of a lot of people... speaking of upwind, there is also significant oil and gas funding behind a lot of geophysical science. Addendum: as I recall, the Globus project was often funded by a mixture of NSF and DoE grants and a large part of its core group was based out of a midwestern national lab.
- dekhn 8y agoI never said I thought there was anything wrong with doing nuclear simulations testing. TBH I think it's an overall good. Globus was mostly run out of Argonne National Lab, but it existed primarily to support scientific computing and public research. We had a subcontract to build pyGlobus since so many people wanted a Python interface (the Globus team started in C++ and then switched to Java and a really, really bad RPC technology called SOAP with WSDL for schema). I've definitely come to terms with the modern compromise (in which a great deal of scientific research is motivated and funded by defense).
- Johnny555 8y agoAfter chatting for a while I realized they were using my code to manage nuclear weapons simulations. That's a consequence of licenses like MIT and BSD and Apache: you don't control who uses your software. While true, there's a big difference between the guy that makes a generic door stop that's used to prop open the door at a nuclear test lab, and the guy making precision parts specific to a nuclear detonator. Maybe you can't prevent your work being used for what you consider to be immoral, but you can avoid working on a project that's directly contributing to that use.