@0xAlbus@0xzeiv The other big gap and why I don’t believe it is that he claims the phone following him gave up the list of wifi networks IT saw to a random unpaired/bonded device?
Just checking in to make sure you understand why working from home hurts collaboration, but outsourcing your entire department overseas, with time-zone gaps of 8 to 12 hrs, apparently doesn’t.
What are you talking about? My harness searches only using my searxng instance dozens of times a day. I’ve designed additional engines for searxng to search several local data stores on top of the defaults searxng has. No rust is needed, the search engine part isn’t the slow part of web search.
@Alezander9 I’ve been testing mnemonic-proxy with local models. It’s still early but I’ve practically one-shotted a TUI Pac-Man, and two local data store mcp servers with semantic search with a model that I’ve intentionally held the context at 64k in order to test the proxy.
ah yes, but I think this is where you'd have to tell the model about the specific device and toolchain, and perhaps even go farther to spell out things like "in order to maximize parallelism, use 80% of the multipliers through this module" or "generate a multiplier module" that it could then use later. Good question.
@fuzziphy I had a WD 2TB nvme give up out of nowhere. Just reports a generic bootloader mfg/version. This drive and the other with it were absolutely babied. I am *not* a fan of SSDs.
@LinkedIn tell me please why your job search filters do not filter out positions I've already applied to by default, and have *no* way to filter these out at all?
Seems kind of a no-brainer...