2 ms·
It's worth asking, I think, how much energy should be spent against guarding against an attack like this in the future. I think every organization's risk model
by shadowgovt 5y ago
It's worth asking, I think, how much energy should be spent against guarding against an attack like this in the future. I think every organization's risk model is different.
In this case, the Net interpreted Marak's actions as damage and seems to have effectively routed around them. There was disruption, but we already have systems in place for flagging broken or malicious versions (`npm audit` being one example).
For people willing to have the system work 99% of the time with the occasional Marak-screw, we already have a working solution in what currently exists. For people who cannot tolerate that level of risk, you have to firewall your package sources and only allow use of vetted code dependencies in your software.
Open source relies, in the end, on trust. Marak broke trust and that's all. The question isn't "How do we prevent people breaking trust" (that's impossible; humans have free will) but instead "How do we mitigate damage / protect that which cannot be risked if someone chooses to break trust?"