James Cridland on Nostr: Being absolutely smashed currently by some hungry bot in Singapore (within an Amazon ...
Being absolutely smashed currently by some hungry bot in Singapore (within an Amazon IP range).
But, very oddly, it's requesting "fallback header hero" images like this example for The Daily, which are only ever built when we write a press release about a podcast where we don't have an image. There is nothing that links to these images programmatically in my entire codebase, except if you write a press release and link to one podcast in it.
The file image name isn't just an ID number, but a hash of the original image file URL. So, there's no way to crawl them without deliberately rewriting these URLs in order to get this image. And none will be cached, so they need making (and uploading to Amazon S3) every single time.
I'm currently blocking the entire 47.128.xxx.xxx IP range, in an effort to tell them to bugger off.
Published at
2026-04-08 13:40:50 CESTEvent JSON
{
"id": "94204b9904d2e740308873df593dcfba03841d9686d3230b639a1166fe0dfe4a",
"pubkey": "8e10c71d95b93c41fb37c3ef057b75c43bf22965355da78642065f84df3065d2",
"created_at": 1775648450,
"kind": 1,
"tags": [
[
"imeta",
"url https://assets.bne.social/media_attachments/files/116/368/896/758/688/492/original/795729994c88a768.png",
"m image/png"
],
[
"proxy",
"https://bne.social/@james/116368896854140360",
"web"
],
[
"imeta",
"url https://assets.bne.social/media_attachments/files/116/368/896/729/335/621/original/cdc0e2c06cfce4db.jpg",
"m image/jpeg"
],
[
"proxy",
"https://bne.social/users/james/statuses/116368896854140360",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://bne.social/users/james/statuses/116368896854140360",
"pink.momostr"
],
[
"-"
]
],
"content": "Being absolutely smashed currently by some hungry bot in Singapore (within an Amazon IP range).\n\nBut, very oddly, it's requesting \"fallback header hero\" images like this example for The Daily, which are only ever built when we write a press release about a podcast where we don't have an image. There is nothing that links to these images programmatically in my entire codebase, except if you write a press release and link to one podcast in it.\n\nThe file image name isn't just an ID number, but a hash of the original image file URL. So, there's no way to crawl them without deliberately rewriting these URLs in order to get this image. And none will be cached, so they need making (and uploading to Amazon S3) every single time.\n\nI'm currently blocking the entire 47.128.xxx.xxx IP range, in an effort to tell them to bugger off.\nhttps://assets.bne.social/media_attachments/files/116/368/896/729/335/621/original/cdc0e2c06cfce4db.jpg\nhttps://assets.bne.social/media_attachments/files/116/368/896/758/688/492/original/795729994c88a768.png\n",
"sig": "75848eb57affbaa0662c3dc1d0a2447cbc18d83783cacb4d6cc3fdfc90af484c655597d4282256f8ab4b05aea258e8613611ed888b753b694cdc911792eaa189"
}
