Re: drop experimental from http2 for 2.4.next?

2017-05-17 Thread Stefan Eissing 
I was talking about the way ahead for HTTP/2 support in httpd.

Regarding it as no longer "experimental" helps, I hope, to make
better progress on this. There have been two aspects to "experimental"
in my mind:

 1. The "we do not guarantee that this is how it will stay". Anything
may change in the next release to make it better and better.
 2. We try this out. It's an experiment. It may go away again.

Removing "experimental" makes it clear to everyone that the project
has been working on 1 and that http2 has become a core part of
httpd, similar to ssl. And as more than one person is working on
ssl, I hope that we will have more people working on http2.

-Stefan

> Am 16.05.2017 um 20:50 schrieb Eric Covener :
> 
> On Mon, Apr 17, 2017 at 4:24 AM, Stefan Eissing
>  wrote:
>> What needs to be done? From what I saw in the last two years, these
>> are key areas to improve:
>> 
>>  1. separation of semantics and serialisation
>>  2. connections with >1 requests simultaneously
>> 
>> mod_http need to spin off a mod_http1 with the parts that read
>> and write headers, handle chunked encoding in requests
>> and responses. etc.
>> 
>> mpm needs facilities for processing slave connections and assign
>> its resources to slave/master connections in fair and performant
>> ways.
>> 
>> As much as I like to work on it, I am certainly not able to do
>> that by myself. So, yes, I welcome getting rid of experimental.
> 
> 
> I'm sorry, but can you clarify the relationship between the initial
> bit and the welcoming of it not experimental?
> 
> -- 
> Eric Covener
> cove...@gmail.com

Re: drop experimental from http2 for 2.4.next?

2017-05-16 Thread Eric Covener 
On Mon, Apr 17, 2017 at 4:24 AM, Stefan Eissing
 wrote:
> What needs to be done? From what I saw in the last two years, these
> are key areas to improve:
>
>   1. separation of semantics and serialisation
>   2. connections with >1 requests simultaneously
>
> mod_http need to spin off a mod_http1 with the parts that read
> and write headers, handle chunked encoding in requests
> and responses. etc.
>
> mpm needs facilities for processing slave connections and assign
> its resources to slave/master connections in fair and performant
> ways.
>
> As much as I like to work on it, I am certainly not able to do
> that by myself. So, yes, I welcome getting rid of experimental.


I'm sorry, but can you clarify the relationship between the initial
bit and the welcoming of it not experimental?

-- 
Eric Covener
cove...@gmail.com

Re: drop experimental from http2 for 2.4.next?

2017-04-19 Thread Jim Jagielski 
Do we foresee an issue w/ it moving to RTC? I don't...
I think we could get the required 3 +1s quite easily.

> On Apr 18, 2017, at 2:35 PM, Eric Covener  wrote:
> 
> On Sat, Apr 15, 2017 at 8:15 PM, Nick Kew  wrote:
>> Why would it be good for a stable (i.e. non-experimental)
>> component of httpd to have an entirely different commit
>> policy to the project as a whole?  Surely the CTR is in
>> recognition of its experimental status, to lubricate the
>> process of hacking it into shape.
> 
> For the record, I have no defense for keeping it CTR. I primarily
> wanted to highlight that it was a factor.

Re: drop experimental from http2 for 2.4.next?

2017-04-18 Thread Eric Covener 
On Sat, Apr 15, 2017 at 8:15 PM, Nick Kew  wrote:
> Why would it be good for a stable (i.e. non-experimental)
> component of httpd to have an entirely different commit
> policy to the project as a whole?  Surely the CTR is in
> recognition of its experimental status, to lubricate the
> process of hacking it into shape.

For the record, I have no defense for keeping it CTR. I primarily
wanted to highlight that it was a factor.

Re: drop experimental from http2 for 2.4.next?

2017-04-18 Thread Graham Leggett 
On 17 Apr 2017, at 10:24 AM, Stefan Eissing  
wrote:

> These modules, they grow up so fast...
> 
> For the project, it would be good to drop that "experimental" and 
> treat HTTP/2 as an integral part of httpd. Not only for political
> posturing (which is important), but also for very technical reasons.
> 
> Looking at https://w3techs.com/technologies/details/ce-http2/all/all
> one can see that HTTP/2 is used by 13% of all sites, which is almost
> double from 1 year ago. Firefox telemetry reports HTTP/2.0 now 
> on 35% of all responses received.
> 
> What needs to be done?

I would say what needs to be done is make it a solid and viable HTTP2 
implementation, declare it non-experimental and let it fly.

> From what I saw in the last two years, these 
> are key areas to improve:
> 
>  1. separation of semantics and serialisation
>  2. connections with >1 requests simultaneously
> 
> mod_http need to spin off a mod_http1 with the parts that read
> and write headers, handle chunked encoding in requests
> and responses. etc.
> 
> mpm needs facilities for processing slave connections and assign
> its resources to slave/master connections in fair and performant
> ways.

These are great to have for httpd v2.6 - let’s develop these above there.

Regards,
Graham
—



smime.p7s
Description: S/MIME cryptographic signature

Re: drop experimental from http2 for 2.4.next?

2017-04-18 Thread Graham Leggett 
On 15 Apr 2017, at 11:02 PM, Eric Covener  wrote:

> Hi everyone, shall we drop experimental from mod_http2 for 2.4.next?

+1.

> We could drop it and keep CTR.

It can’t be not-experimental and CRT at the same time.

Regards,
Graham
—



smime.p7s
Description: S/MIME cryptographic signature

Re: drop experimental from http2 for 2.4.next?

2017-04-18 Thread Ruediger Pluem 


On 04/16/2017 02:15 AM, Nick Kew wrote:
> On Sat, 2017-04-15 at 17:02 -0400, Eric Covener wrote:
>> Hi everyone, shall we drop experimental from mod_http2 for 2.4.next?
>>
>> We could drop it and keep CTR.
>>
> Why would it be good for a stable (i.e. non-experimental)
> component of httpd to have an entirely different commit
> policy to the project as a whole?  Surely the CTR is in
> recognition of its experimental status, to lubricate the
> process of hacking it into shape.
> 

I am with Nick here. It is either or. If we remove the experimental tag it 
should be RTC.
Given that I would like to keep it experimental for some time.

Regards

Rüdiger

Re: drop experimental from http2 for 2.4.next?

2017-04-17 Thread Jacob Champion 

On 04/17/2017 01:24 AM, Stefan Eissing wrote:

As much as I like to work on it, I am certainly not able to do
that by myself. So, yes, I welcome getting rid of experimental.


Sounds good to me.

I agree with Nick that we should probably switch to RTC at the same 
time, for backports to 2.4.x.


--Jacob

Re: drop experimental from http2 for 2.4.next?

2017-04-17 Thread Stefan Eissing 
These modules, they grow up so fast...

For the project, it would be good to drop that "experimental" and 
treat HTTP/2 as an integral part of httpd. Not only for political
posturing (which is important), but also for very technical reasons.

Looking at https://w3techs.com/technologies/details/ce-http2/all/all
one can see that HTTP/2 is used by 13% of all sites, which is almost
double from 1 year ago. Firefox telemetry reports HTTP/2.0 now 
on 35% of all responses received.

What needs to be done? From what I saw in the last two years, these 
are key areas to improve:

  1. separation of semantics and serialisation
  2. connections with >1 requests simultaneously

mod_http need to spin off a mod_http1 with the parts that read
and write headers, handle chunked encoding in requests
and responses. etc.

mpm needs facilities for processing slave connections and assign
its resources to slave/master connections in fair and performant
ways.

As much as I like to work on it, I am certainly not able to do
that by myself. So, yes, I welcome getting rid of experimental.

Cheers,

-Stefan


> Am 16.04.2017 um 15:00 schrieb Jim Jagielski :
> 
> Agreed. The "tag" is being used to keep it out of distros
> as well as to continue the FUD that httpd doesn't "really"
> support http/2.
> 
>> On Apr 15, 2017, at 5:02 PM, Eric Covener  wrote:
>> 
>> Hi everyone, shall we drop experimental from mod_http2 for 2.4.next?
>> 
>> We could drop it and keep CTR.
>> 
>> -- 
>> Eric Covener
>> cove...@gmail.com
>

Re: drop experimental from http2 for 2.4.next?

2017-04-16 Thread Jim Jagielski 
Agreed. The "tag" is being used to keep it out of distros
as well as to continue the FUD that httpd doesn't "really"
support http/2.

> On Apr 15, 2017, at 5:02 PM, Eric Covener  wrote:
> 
> Hi everyone, shall we drop experimental from mod_http2 for 2.4.next?
> 
> We could drop it and keep CTR.
> 
> -- 
> Eric Covener
> cove...@gmail.com

Re: drop experimental from http2 for 2.4.next?

2017-04-15 Thread Nick Kew 
On Sat, 2017-04-15 at 17:02 -0400, Eric Covener wrote:
> Hi everyone, shall we drop experimental from mod_http2 for 2.4.next?
> 
> We could drop it and keep CTR.
> 
Why would it be good for a stable (i.e. non-experimental)
component of httpd to have an entirely different commit
policy to the project as a whole?  Surely the CTR is in
recognition of its experimental status, to lubricate the
process of hacking it into shape.

-- 
Nick Kew

drop experimental from http2 for 2.4.next?

2017-04-15 Thread Eric Covener 
Hi everyone, shall we drop experimental from mod_http2 for 2.4.next?

We could drop it and keep CTR.

-- 
Eric Covener
cove...@gmail.com