deleted by creator

Just tell them unlock their phone so you can take a look of his browser history. Works quite a few time for me.

I have simplex notification service running 24x7. while rarely open, i never missed a message when it arrive (i use it as a message bridge between my devices). Nor I feel it uses more battery that it can’t hold a day of use despite it running constantly in the background. I’m using S21FE btw.

OpenBao, fork from vault

https://github.com/openbao/openbao

Define your criteria for an ideal messenger. What do you need actually? What’s your security requirement?

Addition. Use a Makefile to push and deploy in one make command.

GH action is all what you need. If you really worried open ssh over the internet, use Tailscale or Cloudflare Tunnel. Or use a firewall rule to block off traffic except from GH IP ranges. TBH, I have VPSes that have SSH open to the whole world. Yes, it got many hits everyday but they doesn’t do anything beyond that.

Backup mate. Either local or something over the network. When comes to data loss, it will come find you eventually.

What FS you’re on? I’m using BTRFS and have the same problem. Simply because disk analyzer doesn’t read snapshots.

Any UK registrar recommended?

No wonder why some reported my mail fall into spam dispite I rarely sent any. God. I had it for almost 10 years already, and migrating would be painful.

You sure you can train a model deterministically down to each bits? Like feeding them into sha256sum will yield the same hash?

Not just LLMs but all kinds of models are equivlant to freeware, aka the model itself and other essential bits for it to work. I won’t even call it source avaliable as there is no source.

Take redis as example. I can still go grab the source and compile a binary that works. This doesn’t applies on ML models.

Of course one can argue the training process isn’t determistic thus even with the exact training corpus, it can’t create the same model in terms of bits on mulitple runs. However, I would argue the same corpus provide the chance to train a model of similar or equivalent performance. Hence the openness of the training corpus is an absolute requirement to qualify a model being FOSS.

So you’re including free models like freeware, not FOSS only, by non big tech.

Your choice of models will be quite limited as the compute resource and training corpus needed to make a viable base model isn’t anyone can do.

What’s FOSS-AI? A model everyone can download and use for free? Or in the OSS spirit that everything need to be open and without discrimination of use, aka OSS training data corpus and no AUP attached?

Or you mean the inference engine running those models?

Shit. I have my peraonal domain hosted on .xyz for email. Guess time to migrate. Any TLD suggestions?

deleted by creator

On features ZeroTier > Tailscale, but the existance of headscale and ionscale brings me back all because ZeroTierOne is BSL, aka not FOSS.

Decide what good for me

It is quite a bloat. Llama3 7B is 4.7GB by itself, not counting all the dependencies and drivers. This can easily take 10+ GB of the drive. My Ollama setup takes about 30GB already. Given a single application (except games like COD that takes up 300GB), this is huge, almost the size of a clean OS install.