Missing/incorrect documentation

166 views
Skip to first unread message

Povilas Balciunas

unread,
Aug 6, 2016, 7:52:13 PM8/6/16
to QUIC Prototype Protocol Discussion group
Hi, I'm building QUIC client library in python (https://github.com/povilasb/pyquic) using wire layout and crypto specs.

I'm testing my implementation against the demo quic server from https://github.com/google/proto-quic.

While building library I've discovered some missing or incorrect information.

1. Message 12 byte FNV-1A hash is still undocumented. Not sure if every message requires that because I'm still only playing with CHLO message.

2. According to wire spec public header has an optional 32 byte diversification nonce. The problem is that demo quic client sets the flag that diversification nonce is set, but does not actually send it:


Diversification nonce is supposed to go between connection ID and packet number.
I've also attached pcap for this request.

3. This one is not really a documentation issue, but I think it is related.
    When the CHLO message hash is wrong, server responds with "No recent network activity".
    At first server logs the "Unable to decrypt payload" message, then it timeouts handling the request and sends "No recent network activity".
    But that is all client sees, which is counterintuitive.

If I missed some documents and that's the reason I could not find the information, please let me know.

Also, do you guys have a process how to contribute to the docs?

chlo.pcap

Alexis La Goutte

unread,
Aug 7, 2016, 2:22:41 PM8/7/16
to QUIC Prototype Protocol Discussion group
Hi Povilas,


On Sun, Aug 7, 2016 at 1:52 AM, Povilas Balciunas <balci...@gmail.com> wrote:
Hi, I'm building QUIC client library in python (https://github.com/povilasb/pyquic) using wire layout and crypto specs.

I'm testing my implementation against the demo quic server from https://github.com/google/proto-quic.

While building library I've discovered some missing or incorrect information.

1. Message 12 byte FNV-1A hash is still undocumented. Not sure if every message requires that because I'm still only playing with CHLO message.

2. According to wire spec public header has an optional 32 byte diversification nonce. The problem is that demo quic client sets the flag that diversification nonce is set, but does not actually send it:


Diversification nonce is supposed to go between connection ID and packet number.
I've also attached pcap for this request.
What release (of Wireshark) do you a using ? (because missing the information in the screenshot)
About the diversification nonce, it is only on cert to client flow
 

3. This one is not really a documentation issue, but I think it is related.
    When the CHLO message hash is wrong, server responds with "No recent network activity".
    At first server logs the "Unable to decrypt payload" message, then it timeouts handling the request and sends "No recent network activity".
    But that is all client sees, which is counterintuitive.
it look like a bug on quic_server...

If I missed some documents and that's the reason I could not find the information, please let me know.

Also, do you guys have a process how to contribute to the docs?
for the moment, only Google guy can update documentation.
it will be nice to have a public wiki/repo for push update/contribute to spec...

Cheers

--
You received this message because you are subscribed to the Google Groups "QUIC Prototype Protocol Discussion group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to proto-quic+unsubscribe@chromium.org.
To post to this group, send email to proto...@chromium.org.
For more options, visit https://groups.google.com/a/chromium.org/d/optout.

Povilas Balciunas

unread,
Aug 8, 2016, 12:25:20 PM8/8/16
to proto...@chromium.org, alexis....@gmail.com
Alexis, I used Wireshark 2.1.1-477-g7083839 - it's compiled from sources.

Guys, any plans to move docs to github or something so that we could contribute too? :)
Hi Povilas,


To unsubscribe from this group and stop receiving emails from it, send an email to proto-quic+...@chromium.org.
Reply all
Reply to author
Forward
0 new messages