On 12/4/2021 2:12 AM, Ryan Schmitt wrote:
Oleg's not wrong about the value of tests, but here's a functional
improvement I've proposed for dual-stack connections:
https://issues.apache.org/jira/browse/HTTPCLIENT-2178
This is a very good option!
I can also propose a few others
* Rewrite of the Kerberos auth scheme
This is a single most needed improvement with a major impact on the
project, but it is an extremely tough one. One would need to invest a
considerable amount of time and effort into studying the protocol and
setting up a development infrastructure.
Another problem is that only Michael would be able to help and review
the changes.
* Content compression / decompression for the async transport
That would be a nice and neatly isolated feature with a major visibility
for our users. It would also be a good introduction into the internal
working of the async transport and its APIs, and a stepping stone toward
other improvements in that area.
* Support for 'Upgrade' handshake (RFC 2817)
Not the most wanted feature but reasonably well isolated and requiring a
good understanding of the advanced HTTP/1.1 features. The next step
would be adding support for the TLS protocol upgrade.
* HTTP cache protocol conformance (RFC 7234)
The HTTP caching code has not been receiving enough attention for some
time now and is currently not well maintained. Someone needs to step in
and do a thorough review of the existing code base and ensure its
conformance to the RFC 7234.
* HTTP/2 protocol improvements
This is a tough one but if you want to learn the HTTP/2 protocol and
need some exposure to it, I could think of a few improvements you could
work on.
* WebSocket protocol support (RFC 6455)
That would start off a whole new development stream within the project.
I would not recommend it but it is an option worth mentioning.
My advise to you would be to choose between the dual-stack connections
proposed by Ryan, the async content compression / decompression and the
RFC 7234 conformance.
Cheers
Oleg
On Fri, Dec 3, 2021 at 1:33 PM Arturo Bernal
<[email protected]> wrote:
HI Oleg,
I haven't a specific area of interest, i find it fascinating all and each
modules related with the httpcomponents. I'll start with the test and see
from there what improvement can we implement,
Kind regards,
Arturo Bernal
[email protected]
On 3 Dec 2021, at 13:35, Oleg Kalnichevski <[email protected]> wrote:
On 12/3/2021 1:11 PM, Arturo Bernal wrote:
HI @all,
I have been collaborating with the project for some time, trying to get
involved not only technically but also functionally. I would like to
continue on this path in order to continue expanding my technical and
business logic knowledge of httpcomponents. I know there is a long way to
go to get a decent full-featured domain offered by the httpcomponents
project, but I'm working on it.
So, I’m open and free to any help that the project need it.
Looking forward to hear from yours,
Kind regards,
Hi Arturo
Please let us know if you have any particular areas of interest, such
as, say, network programming, security, CI, docker, and so on?
HC is a mature project and there is presently no major feature requests
in the pipeline. There is however a dire need of help with writing unit
tests, integration tests and making our CI better. Writing tests is
non-glorious and underappreciated work but extremely valuable for us as a
project.
As far as functional improvements are concerned, I have a number of
things in mind that I can propose, but I would like to see what kind of
ideas other people may come up with.
Cheers
Oleg
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
