TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

The Hierarchy of Hazard Controls

40 pointsby surprisetalk3 months ago

3 comments

AlotOfReading3 months ago
I haven&#x27;t encountered the hierarchy in years, but I&#x27;ve always felt it was a bit of a weird idea.<p>The first and third tiers (elimination and controls) are the same thing: separating the hazard from potential injury conditions. This can be spatial separation, temporal separation, or a dozen other kinds. A real-world example is how xray techs have to physically leave the exposure area and stand behind a lead panel to turn on the beam. The hazard (radiation exposure) remains, but the system is safe because the tech can&#x27;t be exposed to it. We use this all the time in software. Process boundaries are a form of spatial separation, mutexes enforce temporal separation. test&#x2F;prod systems, principle of least privilege, separation of concerns, etc.
评论 #43304603 未加载
评论 #43313458 未加载
评论 #43304391 未加载
nonrandomstring3 months ago
Love this synthesis of safety and software thinking that seems apropos Ross Anderson&#x27;s musings on software safety engineering.
theamk3 months ago
If the hazard is &quot;a production shell with unrestricted privileges&quot;, then won&#x27;t &quot;elimination&quot; mean &quot;don&#x27;t give people production shells&quot;? This can be achieved by many ways, such as pre-made queries (with new queries that could only be added vi a PR&#x2F;review process), better observability, using IaaC like terraform, etc...<p>Instead, the author says &quot;we could eliminate the production environment or we could eliminate the database&quot; which really makes no sense.