6,094 karma · joined December 13, 2010
On the other hand, if you can hide nukes in small packages in Mars orbit, maybe you don't need to launch all the way from Earth.
Terraform isn't a wrong decision, just that CNCF now has a more stable API/CLIs you can work with, with which Terraform also needs to integrate.
Last checked at least 10% of NY apartments sat empty (landlords holding out for rents to go up).
https://www.wsj.com/articles/manhattans-apartment-vacancy-ra...
Some things tried: totally custom solution (most flexible), GitLab (unlike Jenkins, you can define a pipeline as a graph), BPMN (pretty heavy Java solution, you still need to build a frontend, there are a lot of tooling but kind of big overhead and low flexibility), Airflow (good visualisation, good for short duration workflows, there is some infrastructure overhead).
It may help more to define more precisely a few scenarios where you experience problems, so people can be more specific with suggestions.
Lacking more specific information, you might consider putting automatic self destruct / deprovisioning for created cloud resources, e.g. 7 days, so you don't worry as much about what is created. What also helps is idempotency in these scripts, to just tell developers to retry things if there are transient errors.
Supercritical CO2 or helium also perform better when the heat source is hotter, which is suitable for the Brayton cycle. It's also why they are frequently mentioned when talking about high temperature next-gen nuclear, or concentrated solar.
[1] - https://sites.science.oregonstate.edu/~gablek/CH334/Chapter5...
Also a possible problem if you grow mirror algae, which can kill regular zooplankton and so out-compete regular algae.
Also if your application is shipped as orchestrated containers (like docker-compose), or as multiple containers in a 'pod' (e.g. sidecars), you may want the ability to run containers from containers as part of CI.
>so if more components of the plant, or even the entire plant, could be built offsite under controlled factory conditions, such extra costs could be substantially cut
From the paper (locked behind paywall) there is also a mention of rising labor costs.
I wonder if this has more to do with the size of reactors, which went up from 600-800MWhe up to the 1500MWhe range, i.e. if containment buildings (safety systems etc) at those power levels are just a massive dis-economy of scale.
There is also a cron that checks authorized_users vs git and sends email when something is out of sync.
Creates pretty bad incentives if implemented (people will seek out the virus). Also relies on accuracy of individualized antibody tests.