Re: HTTP 2 client API

2015-07-31 Thread Michael McMahon
Hi Wenbo, The latest version of the docs is available at: http://cr.openjdk.java.net/~michaelm/8087112/2/ I am hoping to finalize this very soon. Thanks, Michael. On 31/07/15 00:26, Wenbo Zhu wrote: Has the API docs been moved? Is there an updated version of it? Thanks, Wenbo On Mon, Mar 9,

Re: HTTP 2 client API

2015-07-31 Thread Simone Bordet
Hi, On Fri, Jul 31, 2015 at 9:00 AM, Michael McMahon wrote: > Hi Wenbo, > > The latest version of the docs is available at: > http://cr.openjdk.java.net/~michaelm/8087112/2/ > I am hoping to finalize this very soon. Comments below. == HttpRequest.create(URI) and related. Using a URI is not corr

Re: HTTP 2 client API

2015-07-31 Thread Michael McMahon
Thanks for the quick review, Simone. Some answers below. Michael On 31/07/15 14:10, Simone Bordet wrote: Hi, On Fri, Jul 31, 2015 at 9:00 AM, Michael McMahon wrote: Hi Wenbo, The latest version of the docs is available at: http://cr.openjdk.java.net/~michaelm/8087112/2/ I am hoping to final

Re: HTTP 2 client API

2015-07-31 Thread Michael McMahon
On 31/07/15 14:56, Michael McMahon wrote: T == HttpRequest.create(URI) and related. Using a URI is not correct since requests may be made for targets that are not URIs, see https://tools.ietf.org/html/rfc7230#section-5.3. Examples: OPTIONS * HTTP/1.1, CONNECT foo:9090 HTTP/1.1 where "*" and "fo

Re: HTTP 2 client API

2015-07-31 Thread Simone Bordet
Hi, On Fri, Jul 31, 2015 at 3:56 PM, Michael McMahon wrote: > Ok. Now, on the whole flow-control issue, here is my thinking. > First that isFast() method is a relic from a previous version and should not > have been there. > > My thinking is that the body processing methods of the response > body

Re: HTTP 2 client API

2015-07-31 Thread Simone Bordet
Hi, On Fri, Jul 31, 2015 at 4:19 PM, Michael McMahon wrote: > Sorry, just on that point again. It was never intended to support > anything other than URIs here, whether as a String or URI object. > We don't want to allow CONNECTs directly and since it is possible > to create a request without a U

Re: HTTP 2 client API

2015-07-31 Thread Michael McMahon
On 31/07/15 16:06, Simone Bordet wrote: Hi, On Fri, Jul 31, 2015 at 3:56 PM, Michael McMahon wrote: Ok. Now, on the whole flow-control issue, here is my thinking. First that isFast() method is a relic from a previous version and should not have been there. My thinking is that the body process

Re: HTTP 2 client API

2015-07-31 Thread Simone Bordet
Hi, On Fri, Jul 31, 2015 at 6:33 PM, Michael McMahon wrote: > Well, it needs co-operation between the producer and the consumer > obviously. But, the onResponseBodyChunk() method could write to a queue > and block if it reaches a certain size. The onRequestBodyChunk() would > read off the data an

Re: HTTP 2 client API

2015-07-31 Thread Simone Bordet
Hi, On Fri, Jul 31, 2015 at 7:54 PM, Wenbo Zhu wrote: > Thanks for the update. > > === > > Is WebSocket out of the scope now? > > == async streams > > I.e. how bodies are to be read/written asynchronously, with flow-control > (aka back pressures). > > There are many different styles or abstractio

Re: HTTP 2 client API

2015-07-31 Thread Wenbo Zhu
On Fri, Jul 31, 2015 at 11:32 AM, Simone Bordet wrote: > Hi, > > On Fri, Jul 31, 2015 at 7:54 PM, Wenbo Zhu wrote: > > Thanks for the update. > > > > === > > > > Is WebSocket out of the scope now? > > > > == async streams > > > > I.e. how bodies are to be read/written asynchronously, with flow-c

Re: HTTP 2 client API

2015-07-31 Thread Simone Bordet
Hi, On Fri, Jul 31, 2015 at 8:37 PM, Wenbo Zhu wrote: >> > Or we follow the NIO2 model (readiness), >> >> Please no ! :) > > Ignoring the epoll part, is the issue in the API styles or the actual model? https://webtide.com/on-jdk-7-asynchronous-io/ and http://www.slideshare.net/SimoneBordet/ser