4:36 a.m.
Hi guys,
I generated the EDR1 document for a final review feedback before making it official:
https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html
<https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html>
Last changes concerns Request and Application context in SE which should be temporarily
specified for EDR1 (CDI-530 will bring the true spec to them):
https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#request_context_se
<https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#request_context...
They are both active during all method invocation and are destroyed when container is shut
down. The difference is that application context is hared between all methods invocation
Don’t miss the last changes in Events chapter and more precisely in observer method
invocation context:
https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#observer_method...
<https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#observer_method...
Remember that EDR1 is not a final proposition and that it could be changed (even during
its review). This said if you find errors or if you have serious issue with the proposal,
your feedback is welcome.
If some of you have the courage to do a review of API split, it would be great.
Antoine
5:03 a.m.
On 06/23/2015 11:36 AM, Antoine Sabot-Durand wrote:
Last changes concerns Request and Application context in SE which
should be temporarily specified for EDR1 (CDI-530 will bring the true
spec to them):
https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#request_context_se
Is
there a diff/pull request anywhere on github?
They are both active during all method invocation and are destroyed
when container is shut down. The difference is that application
context is hared between all methods invocation
So request context isn't shared between method invocations? I really
cannot tell from the proposal as the wording is very vague. Has any
thought been given on how this is supposed to be implemented?
5:07 a.m.
Yes, you’ll get everything is PR 251 that I rebased to have all the modifications at
hand:
https://github.com/cdi-spec/cdi/pull/251
Le 23 juin 2015 à 12:03, Jozef Hartinger <jharting(a)redhat.com>
a écrit :
On 06/23/2015 11:36 AM, Antoine Sabot-Durand wrote:
> Last changes concerns Request and Application context in SE which should be
temporarily specified for EDR1 (CDI-530 will bring the true spec to them):
> https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#request_context_se
Is there a diff/pull request anywhere on github?
>
> They are both active during all method invocation and are destroyed when container is
shut down. The difference is that application context is hared between all methods
invocation
So request context isn't shared between method invocations? I really cannot tell from
the proposal as the wording is very vague. Has any thought been given on how this is
supposed to be implemented?
5:11 a.m.
and if you want to only focus on async event PR 250 has also been updated, but there won’t
be modification linked to contexts and observer invocation context in it. Quite hard to
live with API split PR and others ;.
Antoine
Le 23 juin 2015 à 12:07, Antoine Sabot-Durand
<antoine(a)sabot-durand.net> a écrit :
Yes, you’ll get everything is PR 251 that I rebased to have all the modifications at
hand:
https://github.com/cdi-spec/cdi/pull/251
> Le 23 juin 2015 à 12:03, Jozef Hartinger <jharting(a)redhat.com> a écrit :
>
> On 06/23/2015 11:36 AM, Antoine Sabot-Durand wrote:
>> Last changes concerns Request and Application context in SE which should be
temporarily specified for EDR1 (CDI-530 will bring the true spec to them):
>> https://dl.dropboxusercontent.com/u/2898173/cdi-spec.html#request_context_se
> Is there a diff/pull request anywhere on github?
>>
>> They are both active during all method invocation and are destroyed when
container is shut down. The difference is that application context is hared between all
methods invocation
>
> So request context isn't shared between method invocations? I really cannot tell
from the proposal as the wording is very vague. Has any thought been given on how this is
supposed to be implemented?
>
>
3475
days inactive
3475
days old
3 comments
2 participants
participants (2)
-
Antoine Sabot-Durand -
Jozef Hartinger