Least of all, a browser publisher.

>I will no longer be helping writers/journalists with a clear anti-AI bias.

Take out the 'anti-AI' qualification, and you've got a decent general principle.

I still think it's superior to '!='.

You start regedit.

cursor up/down, page up/down, or F3 to search, hitting enter to follow directories, until you reach the key.

Hit enter on the key. The values will be displayed.

Hit tab until you are on the key you want to edit.

Hit enter. A popup with the value will appear.

Type the new value, hit enter.

A much more painless process than I anticipated, honestly, and it refreshingly uses the same conventions that used to be used everywhere in windows.

Yeah, that wasn't an attempt to remove keyboard navigation, that was fixing a painful error in keyboard navigation.

Screwing with a keyboard combo in Excel... like Shift-Space to highlight a row... is like an IDE deciding to swap the key-bindings for '{' and '['. You'd have entire accounting and marketing departments in rebellion.

Devs consistently underestimate how many power users of Excel there are, and often don't understand how they operate.

I tend to feel the same way. The medication is obviously useful in some cases and for some conditions, but there's also a pushback against the notion that it isn't the best solution for other cases and conditions. There are cases where other solutions might be healthier, especially in the long term, but the availability of a potential pharmaceutical fix interferes.

And look at the new batch of weight-loss drugs. It's a similar circumstance. It's a wonder drug in many ways, helps so many people that wouldn't have been helped otherwise. Yet, also being used in some cases as a shortcut to avoid a behavioral change that would be advantageous for a person to do.

It is important to remember that there are people for whom these drugs are literal lifesavers. And at the same time, there are those for whom the drugs enable avoidance of fixing a root cause.

Then, why are you in this thread?

I found it bc I started getting new traffic to the site from HN and turns out someone else was actually plugging it in another comment here

I bet half the devs who say 'we don't use it', probably have it on their site and don't realize it.

Conversion from jpg to bmp to jpg is a lossy process, so it may obscure the watermarked pixels.

Sounds better to do postprocess filtering explicitly? For example, adding random noise with amplitude 1/255 (which should be as invisible as the watermark), followed by a smart blur that blurs more in directions where the colors are more similar (making the blur less obvious to humans).

But this is all moot really, if MS Paint is watermarking shit, it’s better to just use something else. Nothing from Microsoft is trustworthy.

or maybe add 3 lsb noise, then let local diffusion model denoise the image. (is there something like convolutional diffusion denoiser?)

If it works anything like synthid, it is more resilient than that.