On 28/01/15 20:18, Andrea Faulds wrote:
> Hi Michael,
>
>> On 28 Jan 2015, at 19:13, Michael Wallner <[email protected]> wrote:
>>
>> - Client
>>
>> The http stream wrapper is a hack and the existing libcurl binding
>> is subpar. They could be improved separately, but that is not
>> subject of this RFC.
>>
>> Currently only libcurl is implemented as a provider for the client
>> functionality. Provides most of the functionality of current
>> libcurl. Representation of the request and response how a client
>> sees them. Support for parallel requests and optional libev{,ent}
>> support.
>
> If the client is merely a wrapper around cURL, what benefit does it
> offer over ext/curl except a better API?
>
> Personally, I’ve never liked that PHP requires cURL for doing HTTP
> requests. It’s a language made for the web, it should have built-in
> HTTP, and it should share code between its server-side HTTP and
> client-side HTTP stuff. I don’t think that the HTTP stream wrappers
> are a “hack” - they’re what PHP should have had all along. I think we
> should focus on improving them (so there’s no need to use cURL)
> rather than adding yet another HTTP client.
Sounds a bit like NIH.
We could implement a http_client driver with the code from the stream
wrapper and see how they compare in feature set and performance.
--
Regards,
Mike