ubercow13

Totally disagree, I have a wrapper I wrote myself for converting things, often for sharing the odd little clip online or such. It produces a complex command that is not easy to just type out, that does multiple things to maximise compatibility like

- making sure pixel are square while resizing if the video resolution is too large

    ("scale=w=if(gt(iw*sar\\,ih)\\,min(ceil(iw*sar/2)*2\\,{})\\,ceil(iw*sar*min(ih\\,{})/ih/2)*2):h=if(gt(ih\\,iw*sar)\\,min(ceil(ih/2)*2\\,{})\\,ceil(ih*min(iw*sar\\,{})/iw/sar/2)\*2):out_range=limited,zscale,setsar=1")

- dealing with some HDR or high gamut thing I can't really remember that can result from screen recording on macos using some method I was using at some point

- setting this one tag on hevc files that macos needs for them to be recognised as hevc but isn't set by default

- calculating the target bitrate if I need a specific filesize and verifying the encode actually hit that size and retrying if not (doesn't always work first time with certain hardware encoders even if they have a target or max bitrate parameter)

- dealing with 2-pass encoding which is fiddly and requires two separate commands and the parameters are codec specific

- correctly activating hardware encoding for various codecs

- etc

And this is just for the basic task of "make this into a simple mp4"

Then why wasn't someone else employed at -10% compensation instead of you before LLMs?

Well, I don't know but many LLMs are multimodal and understand pictures and images. You can upload videos to Gemini and they're tokenised and fed into the LLM. If some programming blog post has a screenshot with the result of some UI code, why would that not be scraped and used for training? Is there some reason that wouldn't be possible?

Why wouldn't they be?

What?

Am I right in reading "Figure. Estimation of All-Cause Mortality at 4 Years in Vaccinated Compared With Unvaccinated Individuals Using Weighted Cox Models: Main and Stratified Analyses" to show that vaccines were more beneficial to the 18-29 age group than any older group? Isn't that unexpected?

Your whole comment is predicated on the idea that the software isn't actually good. However it is, so the rest of it doesn't make sense.

Alacritty didn't even have scrollback for years.

Alacritty's search is less useful than Ghostty's implementation as it you have to exit search mode to do anything else.

You can set a keyboard shortcut for that GUI menu entry (and most others) in macOS system settings.

Why do you think it'd be more reliable? That's one of the main advantages of a NAS

I found it probably the worst of anywhere I've ever been, you can taste it and just being outside slightly burns the back of your throat. I still really like visiting though.

He is not claiming being blocked from one IP proves the geoblock is complete, he is claiming it shows that the geoblock in place is also active for the mirror site.

>What appears to have happened is that SaSu had a site mirror and that someone figured out a way to hit the mirror – which was also subject to the geoblock, something which took me under a minute to personally confirm – without using a VPN.

Seems quite hard to avoid

- file management such as running "ls" where any filename has any non-ascii character such as a CJK or accented letter

- editing any text file in a terminal editor that isn't 100% ascii

- viewing/printing any data from any source, such as a log file/the web/'curl'ing something, where any language other than English or non-ascii character is used

- using various modern command line tools that insist on printing emojis in their output

You can build dolphin on macos using their Craft build tool. It works ok-ish.

I do because the quality of av1 on youtube is often significantly better than vp9 and especially h264, even though the filesize is usually lower than both. And the quality of the video at 1080p when only the worse formats are available is noticeably bad.

I think mailbox doesn't really feel smooth because the age of each frame presented is fluctuating randomly with every frame, so smooth animations will appear jittery.

More like, almost 3 decades.

It seems instant to me?

Are you sure it's not you, a parent, whose opinion might be irrelevant? I mean out of you and non-parents, you are the one who has an extremely deep and instinctive emotional attachment that might cloud your judgement and affect your ability to think rationally or objectively on such a topic.

Does that argument work in other cases? "Sure it's insert bad thing, but if you were a parent, you'd understand. I'd do _anything_ to protect my lil one"

No, being a parent doesn't make your opinion more relevant really.

Claude doesn't seem very good at translation to me compared to GPT*. It also can't understand video like Gemini.

I'm not sure how that's relevant. You're saying that below the event horizon you can't lift your hand above your head faster than your head is moving towards the singularity. But that's only true of your hand and your head in relation to their position and 'velocity' in global coordinates, which is not 1m/s.

Of course cash is expensive, you have to handle it, count it, transport it. Haven't you ever seen those heavily armoured cash delivery vehicles? I mean just think how inefficient cash obviously is in every aspect of how it works compared to modern tech.

That doesn't sound right. If you're on the event horizon you're not going at very slow speeds in that sense, the space around you is already falling into the black hole faster than light.

If you're "travelling at 1m/s so you can only raise your hand above your head at 1m/s by expending infinite energy" then you're already travelling at c-1m/s away from the black hole through local space just to 'stay still' at 1m/s 'velocity'. No wonder you need infinite energy to accelerate your arm 1m/s further and things get weird - you're travelling at relativistic velocities.

Supposedly 39% of phone thefts in Europe happen in the UK https://www.theguardian.com/technology/2025/jul/26/mobile-ph...

Previously most Porsche development tools were available to everyone though, such as GCC.

Even the largest SMBHs mass is a minute fraction of their host galaxies' total mass so it is not the case that everything is just orbiting the SMBH.

Because there aren't more efficient and precise tools capable of the same things?

"AI is impressive because it solves problems better than the previous solutions"? Yes?

Is this true? If the upstream changes some lines of code once, but your branch changes them 5 times in separate commits, iterating on the correct change to those lines, you will have to resolve conflicts 5 times. And 4 of those times you'll be expending effort to merge the correct code from upstream with code you know isn't correct/final on your branch.

Likewise if multiple iterations on those lines were made upstream while your feature branch exists.

Fundamentally expending any effort to find a coherent combination of two pieces of code where one of them you know isn't going to remain is a waste of time.

The only advantage would be if those incremental changes actually make it easier to arrive at the final merge of the last changes on each branch but I'm not sure your above tips really guarantee that. IME it rarely happens.

You could try just exporting it as one text or XML file and seeing if it fits in Genini's context