Live data from Hacker News

Ask HN: Decode Youtube 500 page

news.ycombinator.com

11–20 of 21 posts

Re: Ask HN: Decode Youtube 500 page

#12
It is surely an encrypted stack trace, encoded in base64, which is probably being spit out by the load balancer proxy. I would not be surprised if Youtube developers used their own Chrome extension for making that binary meat readable.

I have once set out to write a module for Nodejs for the same purpose, but never finished it. Can't really see a downside in this way of reporting errors.

Re: Ask HN: Decode Youtube 500 page

#14
The code is indeed very useful to debug issues. When asked for it by YouTube, please provide the actual text rather than a screenshot...I get sent the screenshot a lot and then I have to use OCR software in order to decode it.

Re: Ask HN: Decode Youtube 500 page

#15

The code is indeed very useful to debug issues. When asked for it by YouTube, please provide the actual text rather than a screenshot...I get sent the screenshot a lot and then I have to use OCR software in order to decode it.

Can you divulge what kinds of information is stored in it. Or the encoding it uses?

Re: Ask HN: Decode Youtube 500 page

#17

Perhaps questions like this one should be asked somewhere else, possibly StackOverflow: http://stackoverflow.com/questions/21681084/decoding-youtube...

Isn't that OT?

Every interesting question is OT over there. Or has ultiple answers. Or doesn't belong there for some other reason.

( I used to like SO very much and I hope someone will come up a similar site with more useful guidelines. : )

Re: Ask HN: Decode Youtube 500 page

#18
1. Non-standard base64 and 2. Definitely Compressed ... by guess would be snappy compression 3. Possibility of serialization using protocol buffer 4. not sure if such information would be encrypted after such a server failure

Re: Ask HN: Decode Youtube 500 page

#19
post #8

Earlier quoted context omitted.

CBC mode AES?

More likely a header. It's 3081 byte blob after base64 decoding. If we consider it's encrypted with some block cipher with 128 or 256-bit blocks (IV and/or MAC would be likely to be 128-bit, too), there are 9 bytes for some header and/or padding.

Good catch.
Post reply on HN