As usual it depends (and TDPs are highly misleading). First of all the 6700k is a 14nm chip, Vs. 32nm for the E5-2620. And the 6700k is a Skylake generation chip, compared to Sandy Bridge for the Xeon, which brings significantly better power-states. But on the other hand the 6700k is much higher clocked and has turbo-boost, with the latter being notoriously power hungry (can be disabled in the bios though).

In my educated guess the 6700k will use significantly less power if it mostly idles or does only burst tasks, which is actually what most self-hosters have as as task-loads. But if you serve websites to thousands of users which results in a consistently high CPU load, the Xeon is probably overall the better chip, including power-consumption under load.

Edit: I realized now that it is a E5-2620v2, which is Ivy Bridge and 22nm. So the difference is probably less, but overall the same considerations apply.

Sharkey is great. There is even a nice way to integrate it with a Prosody XMPP server.

Really depends on the specific workload.

Another thing to keep in mind is that the 6700k is significantly more power efficient, especially when it isn’t consistently under high load.

Also if you do any sort of media processing the 6700k has a gpu and quicksync built in that can speed these things up significantly.

You can say what ever you want in your own place. This place isn’t yours and we reserve the right to show you the door.

They are misrepresenting the facts and arguing in bad faith. Very typical concern trolling.

Sorry but these age old troll tactics of baiting me into a response to these completely unfounded accusations will not work 😜

We already explained in detail why we supported their moderation decision and you bringing it up again without providing the necessary context in which this happened is just bad faith shit slinging hoping some of it sticks.

This is not the first time this is discussed and bringing it up out of context is just bad-faith arguing.

I think you will see that these accusations have little substance, are taken out of context and argued in bad faith 🤷‍♂️

There are some credible rumors that Durov had financial issues during the pandemic and went back to Russia to strike some sort of deal with the regime there, which would also explain why since then there have been no further attempts to block Telegram in Russia.

I think you need to be more careful with your use of “scaling up”. You mean vertical scaling, while ActivityPub is good in horizontal scaling.

Indeed, with ActivityPub you will probably never have a single point that has a global view of the network, which (at least for now) is the way ATProto presents the network. Arguably that is not a desirable feature anyways, but lets just say that people have different opinions on that.

This is what you get if you use silly generic terms as names for your chat protocol 🙄

You are overlooking the idle power-consumption. A 7600 basically idles at 5W or so (and turns off all fans). Now think about how many hours you surf the web etc. with your PC Vs. actual gaming and you can see that this makes a big difference.

Depends… its a very capable 1080p card with extremely good power efficiency.

There is the overclocked “xt” version that is generally considerd bad value as it loses the power efficiency aspect and costs significantly more for only little extra performance.

Dunno, realisticly speaking it is a slightly cheaper 7600, hardly a market shake-up.

No, that divide isn’t political. Its about utilitarianism.

The side you are opposed to cares more about getting many people “good enough” privacy, than getting the best possible privacy for a few while alienating most other people.

I ran it some years ago, and over all it wasn’t bad.

The UI is a bit old fashioned and some of the built in plugins seem pretty useless, but the webdav/carddav/caldav integration is nice.

Nomadic identity also works fine, but if you are selfhosting it there will be probably not much use for it.

Due to checksum based auto-correction ZFS and btrfs (in raid1) are actually less sensitive to data-corruption due to non-ECC ram.

This is a common setup for WiFi routers, where the idea is that most traffic will be on WiFi.

Nice find indeed.