Should be standard operating procedure anyway…
Should be standard operating procedure anyway…
Thanks for that, worth knowing.
True, but OP refers to ‘some cherished items’
I assumed the context here was torrenting rather than streaming, download blu-ray remuxes and encode to your liking.
A 16Tb manufacturer recertified drive is USD160 which should sort storage (and later get a second for offline backup). I’m actually holding out until I get GPU encode (apparently CPU is somewhat better, but power considerations, maybe next gen). Do wish the scene would get on with switching, though, are we dinosaurs?
Absitively, use case here IMO is set and forget autoupdate to stay current and SELinux (which actually reduces surface)
For a media server speed matters little (5400rpm is plenty), if you’ve only got one drive, warranty is king. Thing is you shouldn’t only have one drive, drives will fail, and warranty doesn’t get your data back, so you plan for it. At the very least, you should look at getting an offline backup as soon as possible, now you don’t care if your drive fails and can get the cheapest ones. Ideally, you also set up a RAID5 (or Unraid, or mergerfs+SnapRAID) on your server, now you just get a replacement drive and rebuild. Remember RAID is not a backup, it doesn’t protect against accidental deletions for example, so you still want the offline backup.
Also, don’t sleep on manufacturer recertified drives, as long as you have a backup they’re significantly more cost-effective.
TLDR: set yourself up so that a drive failing is not a problem.
Nah, I’m suggesting you actually use some human agency. I’ve found it pleasing, a couple of hours can often net thousands of hours of listening time. You’ve got a list to start from, why not give it a go ?
While I love the *arrs for video, I found Lidarr pretty damn awful, perhaps it’s OK with usenet but not torrent without specialised private trackers. Anywho, I found the Soulseek network, spiritual successor to Napster, which has most of everything, usually at high bitrate, and take pleasure hand curating a personal library. I like the Nicotine client. There’s so much more music than video that it makes sense to be choosy, I’m my own personal DJ…
I was here to say the same as pezhore, separating storage and compute is almost as important as separating church and state. Muck around, break things, have fun, all the while your data is safe (don’t forget offline backups though). The MS-01 is a fine looking box, but any old NUC / SFF will do for your purposes (modern AMD cpu or a graphics card if you need / want plex transcode).
Edit to add, old laptops are great compute nodes (maybe moreso from my ex corporate thinkpad laptop bias, but still)…
So, by way of answering my own question, I found this map, which lets you work it out for yourself. Interesting stuff.
Which brings me to the question of how to benchmark now that I can make good guesses. Speedtest to US and Europe?
ETA: For those that follow, what I did was spin up multiple gluetun instances, Singapore, Hong Kong, Japan, hook them to firefox with FoxyProxy (different ports, duh) and compare speeds on https://librespeed.org/. Bottom line was the bottleneck was my ISP (surprise!!)
Still, I suspect there are ways to leverage crap internet to useful torrenting.
I’d be going through Singapore to get to Japan, so I doubt that’d be better… Hmm maybe, after looking at the map, it might actually work out better, depends on how saturated the direct cable is I guess.
Agreed, and the chance of it backfiring on them is indeed pleasingly high. If the compute moat for initial training gets lower (e.g. trinary/binary models) or distributed training (Hivemind etc) takes off, or both, or something new, all bets are off.
Oh hell no to subscribing, thought didn’t cross my mind. Save us from MBAs with revenue streams on their mind, wannabe rentiers.
They’re talking about a Trackball mouse e.g., not the pre laser mice (or they’re nuts). Good ergonomics, useful for carpal tunnel etc.
FWIW I’ve found their high-end mice pretty robust, my MX Master 2 is still going strong 5+ years in, if cosmetically challenged. Amortised over time, the price is not so bad.
Please let local, open equivalents be available (see LocalAI for an example of not being far off) before this. The sheer scale of data harvesting this will enable boggles the mind.
Ditto on the hate, technical, but important distinction here, they support open-weight ML. They do not release training source code or data sets to actually make your own (granted you’d need millions in video cards to do it, but still). Open-source gets thrown around a lot in AI, presumably virtue signalling, but precious few walk the walk.
Never underestimate the value of getting hordes of unpaid workers to refine your product. (See also React, others)
does rclone not work?
Hmm, Church and State… I much prefer having a separate RSS reader (FreshRSS in my case) for news, as I see it, and lemmy for more frivolous purposes. YMMV.
And a few people will get extremely rich by siphoning off a percent or two, circle of life baby /s