I just wanted to say thank You for the answers. As of 2015_11_28 it seems to me that one of the main contributions of the Nix project is a systematic and proper description of the problem that arises from the fact that different software components depend on each other and the fact that due to data transfer bottleneck between CPU and RAM, CPU and HDD, etc. it is not possible to include all dependencies recursively to every project, not to mention that RAM is a limited resource.
It also seems to me that there are multiple approaches for trying to solve the problem, a bit like the 2 Cold War era space programs (Soviets and USA), where information was "exchanged" with the help of spies, except that in the open source world people can just read the code or ask questions from the developers of a "competing" project. However, again, by pulling examples from the Cold War era, in Soviet Union a de facto, not necessarily de jure, prerequisite of a success of a technical design was that drunk and sloppy workers could get the thing to work/withstand-nature even with their "working" style. I see a parallel here with the packaging system design and sloppy approach of package maintainers, who do not take their time for testing things before publishing. In the industry of mechanical things the meticulous and hardworking people can write the tests and the "drunk proletariat" just reads the test results from the screen or from some "test machine indicator lamp", but in the software industry the "drunk proletariat" needs to write the tests and everyone can see, where that has led us. :-D I know that it sounds like a cliche and is used a little bit out of context now, but we are living in interesting times. :-D Thank You for reading my letter and thank You for the aswers. [email protected] :-)
signature.asc
Description: OpenPGP digital signature
_______________________________________________ nix-dev mailing list [email protected] http://lists.science.uu.nl/mailman/listinfo/nix-dev
