I know. But if it's there for longer than 2 seconds my head sets on fire. 😂

@James that code you sent just rewrites www to root.

Very cool though. Must admit.

It sends 404's back to https://rydercragie.com though.

This works though: https://discord.com/channels/595317990191398933/909458221419356210/1095770378556674201 Just seems to spit out occasional timeout errors.

Never mind. ChatGPT's fixed it by adding this: const response = await fetch(request, { cf: { cacheEverything: true, cacheTtl: 14400 }, timeout: 10000 });

Does anyone know what happened to the darkflare GitHub repo?

Told it to optimise it in addition to that to make it more efficient (as it suggested I should do) and it seems to have done something to it...

Umm

Hi, folks I'm getting this warning

md
Total Upload: 4178.69 KiB / gzip: 1457.82 KiB
▲ [WARNING] We recommend keeping your script less than 1MiB (1024 KiB) after gzip. Exceeding past this can affect cold start time

Earlier the size was

Total Upload: 928.13 KiB / gzip: 187.40 KiB

but it increased a lot after I installed

langchain

Could someone please tell me approx how bad this could be?

You mean Denoflare? Not heard of Darkflare before

> gzip: 187.40 KiB you're still good right now, the warning is thrown because your uncompressed size is almost 1 MiB > Total Upload: 928.13 KiB But your compressed size (the one that matters) is still only 1/5th of the way there

After I installed

langchain

the gzip size is

1457.82 KiB

Is this too bad?

Current size is this

Total Upload: 4178.69 KiB / gzip: 1457.82 KiB
▲ [WARNING] We recommend keeping your script less than 1MiB (1024 KiB) after gzip. Exceeding past this can affect cold start time

Oh I see, I misread

Ok so if your script isn't cached, you'll have a slower cold start. It won't be very fast for the first request but after that it'll be cached and just as quick as normal. Unless you don't have a request for a while, you won't cold start. So I wouldn't worry too much about it

Got it thanks 🫡

Ah right, yeah seems they deleted it. You can ping the author in the thread and ask if they've abandoned it

Looking at the averages on the CPU time here, would I be generally OK within the Workers Bundled tier (50ms CPU time limit)

Gotcha.

Yeah that looks fine to me. Only a few over, you need to be averaging over for it to be a problem

Okay

I guess as the D1 DB grows, the more CPU time will be used to fetch it though 😅

Nah it shouldn't impact your CPU time

Oh? What is it that's using CPU time then?

(and I could probably Promise.all() and bulk fetch and all that as well to improve things)

Hmm that's weird

Yeah definitely shouldn't be using anywhere near ~40ms for this

Am I doing something bad here?