Kissaki

joined 7 months ago
[–] Kissaki@beehaw.org 4 points 14 hours ago* (last edited 14 hours ago)

lol; Thanks Dave Kinne! /s

Note: This is from August. So it's been a while since then.

[–] Kissaki@beehaw.org 3 points 15 hours ago* (last edited 15 hours ago)

I will use AVC and HEVC terminology for h264/x264 and h265/x265 respectively.

For video, you get jumps of compression quality from AVC to HEVC, from 8-bit to 10-bit, and from HEVC to AV1.

Depending on your source material, changing compression settings, like target quality, variable bitrate, etc, can also have significant gains. It will depend on your source and target though, and may need some testing to get "right". If you're looking for the best compression, that may be on a file by file basis, because different kinds of video have significantly different compression behavior or concerns. That's likely not feasible for a mass of files though.

Playback compatibility should also be considered. AVC mp4 is the most compatible, right now, if you consider all kinds of and older mobile and embedded devices. If you're fine with modern or desktop, you can go for the best compression codecs.


To get an idea of encoding time investment and quality, you can use ffmpeg with default quality settings, and target the different encoding targets.

AV1 10-bit, Opus audio:

ffmpeg -i input.mp4 -c:a libopus -c:v libsvtav1 -pix_fmt yuv420p10le out_av1-10bit-opus.mkv

AVC mp4 (when targeting mp4 these codec settings are the default, so in fact don't have to be specified):

ffmpeg -i input.mp4 -c:a aac -c:v libx264 -pix_fmt yuv420p out_avc-aac.mp4

HEVC 10-bit, opus:

ffmpeg -i input.mp4 -c:a libopus -c:v libx265 -pix_fmt yuv420p10le out_hevc-10bit-opus.mkv
[–] Kissaki@beehaw.org 4 points 15 hours ago

Let me check how many subscriptions I have.

uuuh, 1,499 🫠

[–] Kissaki@beehaw.org 1 points 3 days ago

so the issue is not digital, but DRM

[–] Kissaki@beehaw.org 5 points 4 days ago

even smaller than it already is (~100MB)

cries in 1 GB apt GitLab update

[–] Kissaki@beehaw.org 5 points 6 days ago (2 children)

any non-video source?

[–] Kissaki@beehaw.org 4 points 6 days ago (2 children)

But it's still digital when pirated…

[–] Kissaki@beehaw.org 4 points 6 days ago (1 children)

Will we call it WeirdPress?

[–] Kissaki@beehaw.org 2 points 6 days ago

Is that what the pilot calls "streaming through the cloud"?

[–] Kissaki@beehaw.org 1 points 6 days ago

Simulating data in flight. Makes sense.

[–] Kissaki@beehaw.org 4 points 1 week ago* (last edited 1 week ago)

The company has reported net losses every quarter since becoming a public company, with last twelve months (LTM) losses totaling $1.07 billion.

it's crazy that that's feasible and not already an issue. And for such a super popular platform. If it's not profitable now, when will it be?

When did it go public?

Roblox went public on the NYSE via a direct listing on March 10th, 2021, and has a current market capitalization of ~$27 billion.


Interviews reveal Roblox effectively has two sets of books for counting users: one for internal business decisions, in which multiple accounts are ‘de-alted’, and one used by the finance team that reports higher metrics to investors.


To better understand the company’s reported engagement, we hired a technical consultant that monitored the top ~7,200 Roblox games across ~2.1 million Roblox servers, collecting 297.7 million rows of real-time player data.

They did extensive, founded analysis.

[–] Kissaki@beehaw.org 1 points 1 week ago

Looks like they publish source on release as a source zip https://sourceforge.net/projects/hedgecam2/files/

23
submitted 2 weeks ago* (last edited 2 weeks ago) by Kissaki@beehaw.org to c/technology@beehaw.org
 

This GitHub repository has the technical details.

 

Abstract (added emphasis and paragraphing):

Anthropogenic methane (CH4) emissions increases from the period 1850–1900 until 2019 are responsible for around 65% as much warming as carbon dioxide (CO2) has caused to date, and large reductions in methane emissions are required to limit global warming to 1.5°C or 2°C.

However, methane emissions have been increasing rapidly since ~2006. This study shows that emissions are expected to continue to increase over the remainder of the 2020s if no greater action is taken and that increases in atmospheric methane are thus far outpacing projected growth rates.

This increase has important implications for reaching net zero CO2 targets: every 50 Mt CH4 of the sustained large cuts envisioned under low-warming scenarios that are not realized would eliminate about 150 Gt of the remaining CO2 budget. Targeted methane reductions are therefore a critical component alongside decarbonization to minimize global warming.

We describe additional linkages between methane mitigation options and CO2, especially via land use, as well as their respective climate impacts and associated metrics. We explain why a net zero target specifically for methane is neither necessary nor plausible. Analyses show where reductions are most feasible at the national and sectoral levels given limited resources, for example, to meet the Global Methane Pledge target, but they also reveal large uncertainties.

Despite these uncertainties, many mitigation costs are clearly low relative to real-world financial instruments and very low compared with methane damage estimates, but legally binding regulations and methane pricing are needed to meet climate goals.

121
Steam Families is here - Steam News (store.steampowered.com)
submitted 1 month ago* (last edited 1 month ago) by Kissaki@beehaw.org to c/gaming@beehaw.org
 

Up to 6, sharing your shareable games library

Adult and child accounts, limit child accounts, approve and pay for child buy requests,

Intended for close household family; can't join a different one until one year after joining

If a family member gets banned for cheating while playing your copy of a game, you (the game owner) will also be banned in that game. Other family members are not impacted.

haha

 

researchers conducted experimental surveys with more than 1,000 adults in the U.S. to evaluate the relationship between AI disclosure and consumer behavior

The findings consistently showed products described as using artificial intelligence were less popular

“When AI is mentioned, it tends to lower emotional trust, which in turn decreases purchase intentions,”

 

Reading the post and comments on Softbank plans to cancel out angry customer voices using AI made me think it could be an interesting topic to chat about.

I think the best support I received was in the chat application and service Slack. A competent, friendly human responds. I had two or three support inquiries with them.

The last issue I had in Slack was when I opened via try icon click my clipboard content was being pasted. I was surprised they were able to identify the issue which was due to a third-party application that had only just released with the issue a day earlier. Slack support was responsive with a first message before the solution, and fast to respond with the second message with the identified cause.

I'm not sure any stand out as particularly awful for me. [Kinda] Bad seems to be the norm. Sometimes bots sit in front of being able to write a message (my bank, I have to write the same inquiry a second time), sometimes the first response is automated or templated, sometimes the first response is automated and immediately but a human will follow up, sometimes you call and can hardly understand them because of accent or even awful intonation. Often you receive incompetent answers that don't respond to your message or issue. Sometimes they're unwilling or incapable of resolution or agreeable conclusions.

 

I stumbled upon their videos and watched three. It's absurd and often hilarious how bad most of the games are.

Jauwn shows us through the games and their gameplay, but also checks further into the mechanisms trying to bait people and the publishers and developers at times linking them to previous scams.

 

Today, we had European elections in Germany.

We have the Wahl-O-Mat, a state-funded service, where you can answer 38 questions, and then match your positions against a selection of or all political parties that could be elected. It then shows you how much overlap (a percentage) you have with the various parties and their answers to those questions.

I find this to be a very important and useful tool for citizen information.
Campaign adverts are shallow and colorful PR. Broad slogans.
Individuals are not necessarily what the broader party policies are and how they vote. Personal sympathy can even be misleading in that a sympathetic person may not hold the values and positions you do.
Voting for a party, I think their program and stances should be the primary decision factor. (Alongside assessment of whether you can trust them of course.)
It obviously and drastically shows you misconceptions about parties and your alignment, and shows you parties relevant to you that you may not have known about before.

Do other countries have something/things like that too? A tool to match personal stance against political parties' stances? [In a concrete and up-to-date way.]

view more: next ›