aha, add the all important important "if..." in there

admin, domain is lost for whatever reason, that's the end of this world. Anon will see it done.

nope, utf8 mangles the data :/

i'll need to rewrite my entire nntp lib for this, since it treats the data as strings, and it clearly cannot, oh well

If it helps, you can probably do a UTF-8 decode before sending it to the decoder, as long as the data doesn't otherwise get mangled. You'll lose some performance with doing a round trip through UTF-8, so ideally you shouldn't, but it'll likely work.

yeah i assumed that was the issue, that's very annoying to say the least, since that means i cant treat the data stream as text :/

I suspect the issue is your usage of TextEncoder. yEnc is not text - it's binary, so encoding it to UTF-8 is incorrect.
You need to be passing the raw bytes received from the network to the yEnc decoder. Note that the body will contain yEnc headers/footers, so you'll need to either handle these yourself or use functions that do so (yencode.decode doesn't take care of the headers).

notably i'm using tosho's own nzbs to and tosho's own yencode parser and i can get it to work, does the lib not support multipart? from_post doesnt work at all, and manually removing the yencode metadata like =yend and =start and parsing the data via yencode.decode returns 1306159 bytes instead of 716800

yeah I have no idea how to use ur node-yencode lib, i have an article body and i cannot find a way to parse it into a buffer or uint8

nvm figured it out, just gotta figure out why the output data length is different
```js
const [, text] = await this.nntp.body(`<${segment.messageId}>`)
const textEncoder = new TextEncoder()
const data = yencode.decode(textEncoder.encode(text))
console.log(data.length, segment.size)
```

and i get `1306341 739519` which is annoying

I also noticed a lot of white space, idk if i should be stripping that
```
"♫�◄,�裪����¶�∟�        �o\牓�K�a� ]��♣�ٔ�������‼↕j��}�����1,G�↔&Vb0��↨�`)�U��>�`�
N�a�H�L֞♠F1�_ת�"��s��♦aI�ۥ=@���.�=@↨�O♣ح'+c�+���↔eļ%�q^&d�B�!�G6SΆ=J�C� �B��:jѣPZ�
��5↑�:}gka_ps�,H'‼♀@5�,=yend size=716800 part=1 pcrc32=d401c3cf
```

yeah, i'm struggling a bit with the nzb structure, i have no clue how to get the article id for the body request of a segment from an nzb

I'm guessing you're looking for yEnc - you can find the spec here.
Best of luck with your project!

I don't plan to do such. If this domain is lost for whatever reason, that's the end of this site.

I guess more of an accurate question is what encoding does the article body use

hello señor tosho man, I have a question regarding how nntp and NZB works, I'm trying to make my own lib for streaming nzbs without needing to download them, and I got everything except decoding the data, when I grab an NZB segment with the BODY command, how do I decode the text into binary data? I realize this is partially off topic, but I know you're informed enough to know as I couldn't find much info on this myselfm

Admin-san, if the badmen ever block the tosho url, I hope you can do the thing that other anime sites do where the url changes to one of the gazillion available domain extensions. It's a pain but at least it seems that when sites change to .is or .pe or whatever it can last a while before needing changed again.