Sandwich bill of materials

36 points by asteroid a day ago on tildes | 9 comments

zod000 | a day ago

Requiring a sandwich bill of materials for sandwiches will likely end up with not getting many sandwiches, but this was a fun (to me) link. Thanks for sharing.

[OP] asteroid | a day ago

I laughed. I thought you might, too!

chocobean | 20 hours ago

Post-incident analysis revealed that 94% of affected sandwiches had no lockfile and were resolving eggs to latest at assembly time.

I want to eat inspect these 6% of sandwiches with a lockfile

This was my favourite bit:

CVE-2019-SPROUT: Alfalfa sprouts were found to be executing arbitrary bacteria in an unsandboxed environment. Severity: High. The vendor disputes this classification.

moocow1452 | 16 hours ago

Ugh, sandwich licenses. Why do I need to open up the sandwich and check third party sources so I can put Sriracha sauce on myself when I can just pay for a sandwich that has brand name Sriracha at the proprietary shop down the road? And I know about OpenHot and Yet Another Spicy Sauce, they just don't fit my use case. /j

pete_the_paper_boat | 10 hours ago

YASS would probably do well on brand recognition alone

HiddenTig | 16 hours ago

Stupid things broke - I tried to compile a hotdog and it crashed despite meeting spec requirements.

moocow1452 | 9 hours ago

Works on my end, try a previously unused kitchen?

sorkceror | 10 hours ago

This is great, thanks for sharing!

tanglisha | 4 hours ago

I love this!

SBOMs are such a great idea, but always felt hopeless to me because of dependency chains.

Add that to vulnerability scanners that complain about everything right down to the language you're using and defenestration seems like the best option.