Melllvar

From a national security standpoint of the government, it absolutely does matter who has the data.

TestDisk and PhotoRec. TestDisk can recover broken drive partitions, PhotoRec can recover deleted files even if the partition table is borked.

Shut up and keep looking apologized to.

I just use Everything desktop search and let the files fall where they may.

Just lemmy in a browser for me. Never used facebag or twatter or others besides reddit.

Looks like compatibility hacks for various websites.

Interventions - are deeper modifications to make sites compatible. Firefox may modify certain code used on these sites to enforce compatibility. Each compatibility modification links to the bug on Bugzilla@Mozilla; click on the link to look up information about the underlying issue.

User Agent Override - change the user agent of Firefox when connections to certain sites are made.

https://wiki.mozilla.org/Compatibility/UA_Override_&_Interventions_Testing

If social media companies exist to collect massive troves of personal info from users–and they do–then there is a valid national security concern over social media controlled by an adversary. This is distinct from the individual privacy concerns towards domestically-controlled social media.

No it’s not. There’s no bail, for example, and no plea bargaining in civil cases; jail time isn’t on the table, the district attorney isn’t involved, the standard of evidence is lower, and the rules of procedure are different.

First comes the discovery phase where both sides exchange evidence and the court settles any evidentiary questions. This phase can frequently take longer than the trial itself.

This is a civil case.

Today, the US is withholding billions of dollars in Afghan assets, cutting off a crucial economic lifeline for the people of Afghanistan.

What assets are they talking about here?

Annual commemorative pastry observance

Lisa needs braces!

Super Mario Brothers 2

Stop, drop, and roll

Florence Foster Jenkins singing the Queen of the Night’s aria.

The output of a LLM is analogous to re-saving an image as a lo res JPEG. Data is being processed and altered using statistics, but nothing “new” is being created, only lower quality derivatives. That’s why you can’t train a LLM on the output of a LLM.

35

https://www.citizensforethics.org/reports-investigations/crew-investigations/election-certification-under-threat/

I’m not upgrading because I don’t trust Windows 11. Not that 10 has my confidence, of course, but 11 seems worse.

Windows 95 had to change the time zone selection map because of disputed borders.

https://devblogs.microsoft.com/oldnewthing/20030822-00/?p=42823