Szpadel

7 ms·

61.
by Szpadel 1y ago
I just want to mention that ZIL is just to speed up sync writes, as it ends syscall when data are written to ZIL, but might be still in progress on slower storage. ZIL is also basically write only storage, therefore sad without very signifi
62.
by Szpadel 1y ago
something that people forget with raid1 is that this only protect from catastrophic disk failure. this means your your drive need to be dead for raid to do it's protection and this is usually the case. the problem is when starts corrup
63.
by Szpadel 1y ago
there is also other scenario, field might obviously looks like it should have value, and in check it might event always have it, but it might be "lazy" value. eg. you might have some bug CSV uploaded and your have number of rows i
64.
by Szpadel 1y ago
In my case it stayed the same, but I tested it on android Firefox, maybe it works better on desktop, but I do not have access to it right now to verify
65.
by Szpadel 1y ago
there are some more or less legit causes for fingerprinting. like bot protection or to identifying scammers that just create another account when previous is banned. whether this is justified is of course subjective
66.
by Szpadel 1y ago
other downsides, cloudflare, PayPal and all kinds of finance related sites will assign high threat level for you and you will make your life miserable for causes ranging from captcha through rejecting your purchases to even blocking you acc
67.
by Szpadel 1y ago
I use litellm as my personal AI gateway, and from user point of view there is no difference if proxy uses official SDK or not, this might be benefit for proxy developers. but I can give you one example: litellm recently had issue with handl
68.
by Szpadel 1y ago
I personally use unifi doorbell, mostly because I already had dream machine, but AFAIK they have also less expensive options.
69.
by Szpadel 1y ago
> quite impressed with the smoothness of the UX on this relatively cheap machine. I recently tried installing cachyos kernel on fedora I tried sched_ext with it (scx_lavd specifically) and to my suprise changing CPU scheduler resolved al
70.
Show HN: Mixture-of-Models Gateway(github.com)
2 points by Szpadel 1y ago|0 comments
71.
by Szpadel 1y ago
true, but they still borrowed patches from clear linux
72.
by Szpadel 1y ago
that's might also be false positive, I use languagetool for grammar/spelling correction and one of the corrections it to replace dashes with em-dashes
73.
by Szpadel 1y ago
why it required forking vscode? what prevents this from being just extension? from article I do not see any obvious functionality that would prevent such implementation
74.
by Szpadel 1y ago
in case you also have home assistant, you can enable location reporting too. you can request precise lock remotely and use it for buch of automations. I found my wife phone that way
75.
by Szpadel 1y ago
> It would be interesting to see if you could feed the file into an LLM and get it to produce the feedback. I did something like that already and It works ok. I'm just electonic amateur and it happened to me to incorrectly wire up M
76.
by Szpadel 1y ago
with RL it's hard to define score function in many categories. rhis is especially visible in current coding capabilities. LLM will very often create sloppy solutions because they work well in RL. hardcoding API keys? ignoring errors? d
77.
by Szpadel 1y ago
In our case this was very heavy specialized endpoint and because each request used different set of parameters could not benefit from caching (actually in this case it thrashed caches with useless entries). This resulted in upscale. When ha
78.
by Szpadel 1y ago
the worst offender I saw is meta. they have facebookexternalhit bot (they sometimes use default python request user agent) that (as they documented) explicitly ignores robots.txt it's (as they say) used to validate links if they contai
79.
by Szpadel 1y ago
Try combining it with --long My use cases are usually source code, SQL dumps and log files. Sometimes xz gave marginally better results, but difference was well below 1%
80.
by Szpadel 1y ago
description from openrouter: > Grok 4 is xAI's latest reasoning model with a 256k context window. It supports parallel tool calling, structured outputs, and both image and text inputs. Note that reasoning is not exposed, reasoning c
81.
by Szpadel 1y ago
in my experience using zstd --long --ultra -22 gives marginally better compression ratio than xz -9 while being significantly faster
82.
by Szpadel 1y ago
I never wrote single line of zig code, but I follow it as an interesting language, so I might be missing something fundamental. I agree that this looks extremely breaking, and feel like providing no compatibility layer fragments community,
83.
by Szpadel 1y ago
php is also jited nowadays. currently I believe the main advantage is that hack is async, you can fire multiple SQL/http requests in parallel and cut some wall time.
84.
by Szpadel 1y ago
> Listen in on what people talk about, run it through a STT engine and a ML model to condense it down this is something I was talking when LLM boom started. it's now possible to spy on everyone on every conversation. you just need
85.
by Szpadel 1y ago
isn't also SOAP self described?
86.
by Szpadel 1y ago
You can file a GDPR complaint in responsible institution in your country.
87.
by Szpadel 1y ago
I tried to use one LLM based tool to rewrite sentence in more official corporate form, and it rewrote something like "we are having issues with xyz" into "please provide more information and I'll do my best to help"
88.
by Szpadel 1y ago
what concerns me is that it's wild laser beam in potentially public space. anyone that would accidentally cross that potentially invisible line could get killed or severely hurt. imagine something reflective entering this beam, it coul
89.
by Szpadel 1y ago
for sure they are no longer clear winners, but they try to be just barely on top of others. right now new Gemini surpassed their o3 (barely) in benchmarks for significantly less money so they cut pricing to be still competitive. I bet they
90.
by Szpadel 1y ago
how do you sample it behind the scenes? usually best of X means you generate X outputs and you choose best result. if you could do this automatically, it would be game changer as you could run top 5 best models in parallel and select best

More ›