Viewing profile — mkesper
mkesper
HN member- Joined
- Fri, Aug 02, 2013, 6:35 AM UTC
- HN karma
- 3,601
- Public activity
- 1,386 items
- HN profile
- View on Hacker News ↗
About mkesper
No profile information was provided.
Recent public activity
-
comment
Comment #49174548
Related, pencils for more accurately drawing skin tones: https://hautfarben.org/en/products/skin-tones-pencils
-
comment
Comment #49082269
Disk failures don't neatly wait for idle times so investing in a special vdev (which can be added to an existing pool as the article points out) should be well worth it.
-
comment
Comment #49057715
How do these compare to the fish builtin history search with dir awareness and completion menu? I found it quite an improvement to bash history handling but fish isn't POSIX compat…
-
comment
Comment #49046986
These feel quite different. IRGC attacked Amazon compute. Ukraine attacked several warehouses. Comparing Wildberries to Amazon is also only half the truth. Wildberries is directly …
-
comment
Comment #48956414
Do not run Ubuntu on these old machines, use plain Debian instead (no fat desktop needed, no overhead due to braindead snap packaging etc.). Xfce4 should be fine. SSDs help and als…
-
comment
Comment #48934423
Be sure not to buy any 'branded' variant, though (e.g. from Verizon etc.)!
-
comment
Comment #48923173
Something like that would also be great for scripting, not only agents.
- story
-
comment
Comment #48902475
Have a look at Donato Capitella's video: https://m.youtube.com/watch?v=zp8j4vO-wz0 He provides also toolboxes and benchmark tables as text (in the video description)
-
comment
Comment #48898979
Regarding M5, this has similar size but runs Linux and is cheaper: https://www.kickstarter.com/projects/m5stack/cardputerzero
-
comment
Comment #48730742
DGX Spark does not have high memory bandwith. M3 Max (Mac Studio) features more memory bandwith than that one. See https://aimultiple.com/dgx-spark-alternatives
-
comment
Comment #48716266
You do not have to use all resources that are under the systemd umbrella. That's just BS, sorry. And prod deployments breaking... Well I guess that never happened with those best-m…
-
comment
Comment #48705907
Currently, llama.cpp clusters don't support tensor parallelism, have a look at Donato Capitella's detailed report: https://m.youtube.com/watch?v=PkKXm_mKCCM He also provides rocm t…
-
comment
Comment #48662962
But existing resources were not affected.
-
comment
Comment #48617513
Tangentially, in theory, k3s + kubevirt + microvms sounds like the optimal combination for lightweight but isolated deployment. Does anyone have experience with that?
-
comment
Comment #48551209
Please show sources. Maybe no one appreciates it because it's only rumors spread by enterprises massively profiting from such rumors? And about your toilet paper analogy: Every pro…
-
comment
Comment #48543668
Server auction still features cheap servers with ECC: https://www.hetzner.com/sb/#ecc=true
-
comment
Comment #48487508
Outdated! Should have linked directly to https://github.com/philipl/inferencefs/ obviously.
-
comment
Comment #48474742
It was a giant mistake to keep 'classic' (umnaintained, unpatched) yarn around. Switched to pnpm in part due to that reason.
-
comment
Comment #48461286
Hank Green also explains 400 billions are really very much cash: https://www.youtube.com/watch?v=AcjnLc4TH4M
-
comment
Comment #48459731
Not at all. If you're lucky you can (after some fiddling) install a real coreboot, this site is incredibly helpful for that: https://docs.mrchromebox.tech/docs/getting-started.html…
-
comment
Comment #48396190
Kudos for the idea and being fully open to the state of this project (AI code, expect breakage)!
-
comment
Comment #48371650
This is a GNU extension so not portable.
-
comment
Comment #48370997
There's another big feature: You're not relying on the time zone to which the server was set (like with cron) but can explicitly specify a time zone: https://www.freedesktop.org/so…
-
comment
Comment #48369894
The Pi is really a GPU with a CPU added on so the whole system needs to be brought up by the (binary blobs for running the) GPU. Hopefully not the case in these more PC-like system…