Roger Peppe

@rog.bsky.social

1401 Followers 166 Following

CUE, Go, fiddler, climber, gardener, curry maker.

An area of lawn. A few blades of grass are visible but mostly it's moss. A pair of bare feet are visible, sinking in comfortably.
Three shelves of rather sun-faded sf paperbacks, featuring many Heinlein, Asimov and other vintage favourites.
A bowl of rather thick green soup on a table with brown bread, a butter dish and some glasses of homemade kombucha. Behind them are some candles and a chopping board with cheese on it. There is a glimpse of garden beyond.
In the style of an old Punch cartoon, the following scene is illustrated: AMERICA sits in a gloomy space with various recognizable figures such as Donald Trump behind high wall fencing it off from the rest of the world labelled TARIFFS. The rest of the world, including CHINA, EUROPE etc is in the sunshine happily trading togeter.
Benchmark result showing 85% in throughput over using io.Pipe.
Image with people and clothing that are recognisably the same as previous photo.
An image of two people in outdoor clothes up a hill. Below it is an image of two totally different people in very different clothes.
Text image: "Here's a photo of a couple. Please process the image so that it's stylised and includes creeping vegetation of UK native species instead of the background. The vegetation is starting to take over the people. The style should be flat and the colours very muted."
Error: Error code: 429 - {'error': {'message': 'Request too large for gpt-4o in organization org-<redacted> on tokens per min (TPM): Limit 30000, Requested 88647. The input or output tokens must be reduced in order to run successfully. Visit https://platform.openai.com/account/rate-limits to learn more.', 'type': 'tokens', 'param': None, 'code': 'rate_limit_exceeded'}}
Me: Those aren't the same people as in the original photo. ChatGPT: You're absolutely right — and thanks for calling that out. Right now, I can’t directly transform specific people from a photo into a new stylised or altered image while keeping their facial features intact and photorealistic. The tools I use generate new imagery based on a written prompt rather than editing the actual photo you provided. But! Here are two good options we can go with:
Talisker beach on the Isle of Skye looking into the sun with striking pools of water in the foreground and tiny silhouetted figures in the distance.
An interaction with ChatGPT 4o where I'm asking it to generate an innocuous image of UK native species and it refuses on the grounds that I'm breaking policy.
Sorry, Ubuntu 24.04 has experienced an internal error. It's written in an unsafe language and has no idea what has gone wrong. Byeee!
Peer Type Resolution occurs in these places: switch expressions if expressions while expressions for expressions Multiple break statements in a block Some binary operations This kind of type resolution chooses a type that all peer types can coerce into. Here are some examples:
House with one end consisting almost entirely of glass.