redcalcium
Pdf has a mind-bogging array of features, which make it so entrenched in the corporate world with no viable replacements at the moment. Things like forms where users can fill them out and submit (surprisingly a popular feature), cryptographic signing to prevent tampering, DRM, etc. Heck, I think you can even add JavaScript code to a pdf.
Every once in a while security researchers would discover sophisticated exploits that would allow malwares to take over your computer via multimedia files, but those are actually rarely exploited in the wild by run off the mill malwares.
Unless you’re an important person being targeted by hackers and three letter agencies, your biggest source of threat is running infected programs from untrusted sources, e.g. cracks downloaded from random torrents or warez sites, shady sites serving ads that trick you to run some executables, etc.
How do you sanitize ai prompts? With more prompts?
I may not agree with the devs political view, but I think their work developing lemmy is excellent and made me subscribe to monthly donation on opencollective. Lemmy is an open source project where the devs have absolutely no say over how the software being used, as evidenced by so many lemmy instances defederating from lemmygrad and lemmy.ml. Their political belief won’t affect other instance.
If we fire all developers and allow AIs to program themselves, the AIs are going to commit virtual seppuku after a few days.
the tests are now larger than the thing itself
The purpose of the code is to make the tests pass.
I’m truly torn with this. The first one seems sensible (action -> target) and easier to read and reason about (especially with long names), while the other one looks more organized, naturally sortable and works great with any autocompletion system.
