wget-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Wget-dev] [GSOC 2019] Interested in adding QUIC support


From: Tim Rühsen
Subject: Re: [Wget-dev] [GSOC 2019] Interested in adding QUIC support
Date: Mon, 1 Apr 2019 12:29:02 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.6.1

Hi Edward,

On 4/1/19 7:12 AM, Edward Hui wrote:
> Hi community,
> 
> This is Edward, an Information Engineering student in Hong Kong. I have
> read the documentation of QUIC and I hope to participate with implementing
> it in wget2, I have experience on socket programming in C++ with TCP&UDP,
> together with rich internship and competition experience.
> 
> Besides, I feel really interested in the idea of QUIC and I would like to
> seek professional advice from developer team. Is it possible to set
> planning like:
> 1. Implement simple client and server connection with QUIC first
> 2. Migrate the code with wget2
> 
> Or all the coding and testing is recommended to perform together with the
> source code of wget2?
> Thank you very much for the time spent on reading my email and I look
> forward to hearing from you soon.

thanks for considering to work on GNU Wget2 !

Bad news: I just removed QUIC for GSOC (better: tagged QUIC as not being
part of GSOC any more).

I am bit sorry about it, but my mind changed after some very detailed
technical talks and discussions at #Curlup2019. Just came back from
there this night.

Long story short: it's very complicated stuff, implementations are still
all 'experimental'. The spec hasn't been finalized and may still change
a lot. Application layer integration will be hard, not as straight
forward as HTTP/2 (it really wasn't easy, though). The needed TLS
functionality for the QUIC layer hasn't been implemented by upstream TLS
libraries (except boringSSL and  NSS which we don't use). There is
already a patched version of OpenSSL, but they are also waiting for the
final spec before merging into the release branch.

That's the main reason why I took QUIC from the list of GSOC projects.
It's far too heavy and anything we do now might have to be redone when
the spec is going to be finalized.

Regards, Tim

Attachment: signature.asc
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]