Sorry I suggested to enable this parameter as default
You only need to know this parameter if you need the old behavoir

Regards

Bernd

On Thu, Sep 22, 2011 at 1:30 PM, Michael Kurz <[email protected]> wrote:
> OK, good. But: This would then be parameter 94 to consider?
>
> Best regards
> Michi
>
> Am 22.09.2011 13:17, schrieb Mark Struberg:
>>
>> of course there is:
>>
>> http://myfaces.apache.org/core20/myfaces-impl/webconfig.html
>>
>> LieGrue,
>> strub
>>
>>
>>
>> ----- Original Message -----
>>>
>>> From: Michael Kurz<[email protected]>
>>> To: MyFaces Development<[email protected]>
>>> Cc:
>>> Sent: Thursday, September 22, 2011 1:14 PM
>>> Subject: Re: [VOTE] fix MYFACES-2552 for 2.0.x and 2.1.x branches
>>>
>>> I agree with Jakob, nobody will know about a parameter like this (are
>>> they documented anywhere btw?). The effect (without setting it): for
>>> application developers JSF seems to be broken.
>>>
>>> Best regards
>>> Michi
>>>
>>> Am 22.09.2011 12:16, schrieb Jakob Korherr:
>>>>
>>>>  Hi,
>>>>
>>>>  Both suggestions for a name are reasonable, because they suggest
>>>>  different things!
>>>>
>>>>  What Leo suggested (STRICT_JSF_2_COMPATIBILITY), is the flag we
>>>>  (internally) have been discussing (and postponing) for a long time.
>>>>  This config parameter would enable us to include non-spec behavior in
>>>>  MyFaces core, which would make us fail the TCK, if we included it by
>>>>  default. However, by enabling it via config parameter only
>>>>  (STRICT_JSF_2_COMPATIBILITY=false), we would be able to fix obvious
>>>>  spec issues already in MyFaces core 2.0 (e.g. MYFACES-2552) and still
>>>>  passing the TCK. Thus we would not need to wait for the next spec
>>>>  release (which may not even contain the fix....), to fix critical
>>>>  issues.
>>>>
>>>>  What Mark and Bernd suggested is to include the fix for MYFACES-2552
>>>>  and make a config parameter just for this behavior.
>>>>
>>>>
>>>>  Actually, I think introducing yet another config parameter for (in
>>>>  this case) very specific behavior is not the best idea, b/c no-one
>>>>  outside of the MyFaces developers will use it. Introducing a more
>>>>  generic config param, which would allow us to fix a lot more issues
>>>>  which are just like this one, is IMO a far better idea.
>>>>
>>>>  Thus, my vote is +1 for STRICT_JSF_2_COMPATIBILITY and +0 for Bernd's
>>>>  suggestion.
>>>>
>>>>  Regards,
>>>>  Jakob
>>>>
>>>>  2011/9/22 Martin Marinschek<[email protected]>:
>>>>>
>>>>>  +1 in general, +1 to Bernd's suggestion.
>>>>>
>>>>>  best regards,
>>>>>
>>>>>  Martin
>>>>>
>>>>>  On Thu, Sep 22, 2011 at 9:59 AM, Gerhard Petracek
>>>>>  <[email protected]>   wrote:
>>>>>>
>>>>>>  @bernd: +1
>>>>>>  regards,
>>>>>>  gerhard
>>>>>>
>>>>>>  http://www.irian.at
>>>>>>
>>>>>>  Your JSF powerhouse -
>>>>>>  JSF Consulting, Development and
>>>>>>  Courses in English and German
>>>>>>
>>>>>>  Professional Support for Apache MyFaces
>>>>>>
>>>>>>
>>>>>>
>>>>>>  2011/9/22 Bernd Bohmann<[email protected]>
>>>>>>>
>>>>>>>  I agree STRICT_JSF_2_COMPATIBILITY is too generic. I think the
>>>
>>> param
>>>>>>>
>>>>>>>  should contain EL, TYPE, MAP, RESOLVER, NULL. And it should
>>>
>>> enabled by
>>>>>>>
>>>>>>>  default.
>>>>>>>
>>>>>>>  Regards
>>>>>>>
>>>>>>>  Bernd
>>>>>>>
>>>>>>>  On Wed, Sep 21, 2011 at 11:50 PM, Mark
>>>
>>> Struberg<[email protected]>   wrote:
>>>>>>>>
>>>>>>>>  Shouldnt the config contain the text EL_TYPE or so?
>>>>>>>>  We have far too many strict JSF spec flags already ;)
>>>>>>>>
>>>>>>>>  LieGrue,
>>>>>>>>  strub
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>  ----- Original Message -----
>>>>>>>>>
>>>>>>>>>  From: Jakob Korherr<[email protected]>
>>>>>>>>>  To: MyFaces Development<[email protected]>
>>>>>>>>>  Cc: [email protected]
>>>>>>>>>  Sent: Wednesday, September 21, 2011 10:20 PM
>>>>>>>>>  Subject: Re: [VOTE] fix MYFACES-2552 for 2.0.x and
>>>
>>> 2.1.x branches
>>>>>>>>>
>>>>>>>>>  +1 for including the fix in 2.0.x and 2.1.x + adding
>>>>>>>>>  org.apache.myfaces.STRICT_JSF_2_COMPATIBILITY
>>>>>>>>>
>>>>>>>>>  Regards,
>>>>>>>>>  Jakob
>>>>>>>>>
>>>>>>>>>  2011/9/21 Leonardo Uribe<[email protected]>:
>>>>>>>>>>
>>>>>>>>>>    Hi
>>>>>>>>>>
>>>>>>>>>>    @Matt Benson: Could you attach at least the
>>>
>>> fragment with the
>>>>>>>>>>
>>>>>>>>>>  solution
>>>>>>>>>>    for MYFACES-2552 ? so I can check it, create a
>>>
>>> patch for myfaces and
>>>>>>>>>>
>>>>>>>>>>    write a page on:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>> https://cwiki.apache.org/confluence/display/MYFACES/Composite+Components
>>>>>>>>>>
>>>>>>>>>>    with the explanation and the solution using a
>>>
>>> custom EL resolver.
>>>>>>>>>>
>>>>>>>>>>  That
>>>>>>>>>>    would be very helpful.
>>>>>>>>>>
>>>>>>>>>>    regards,
>>>>>>>>>>
>>>>>>>>>>    Leonardo Uribe
>>>>>>>>>>
>>>>>>>>>>    2011/9/21 Leonardo Uribe<[email protected]>:
>>>>>>>>>>>
>>>>>>>>>>>    Hi
>>>>>>>>>>>
>>>>>>>>>>>    It should be a param starting with
>>>
>>> org.apache.myfaces, like
>>>>>>>>>>>
>>>>>>>>>>>    org.apache.myfaces.STRICT_JSF_2_COMPATIBILITY
>>>>>>>>>>>
>>>>>>>>>>>    The important part is that by default it
>>>
>>> should be disabled, to
>>>>>>>>>>>
>>>>>>>>>>>    prevent receive over and over the same
>>>
>>> report.
>>>>>>>>>>>
>>>>>>>>>>>    In theory, it is possible to write a custom
>>>
>>> EL resolver that check
>>>>>>>>>>>
>>>>>>>>>>>  if
>>>>>>>>>>>    the base object implements
>>>>>>>>>>>
>>> javax.faces.el.CompositeComponentExpressionHolder and if that so, do
>>>>>>>>>>>
>>>>>>>>>>>    the required stuff only on getType(). So, if
>>>
>>> somebody is writing a
>>>>>>>>>>>
>>>>>>>>>>>    composite component that relies on this
>>>
>>> behavior, it is possible to
>>>>>>>>>>>
>>>>>>>>>>>    write the fix in a portable way to both
>>>
>>> Mojarra and MyFaces (thanks
>>>>>>>>>>>
>>>>>>>>>>>  to
>>>>>>>>>>>    CompositeComponentExpressionHolder).
>>>>>>>>>>>
>>>>>>>>>>>    Note the change does not cause any side
>>>
>>> effects, because nobody
>>>>>>>>>>>
>>>>>>>>>>>  relies
>>>>>>>>>>>    on the "wrong" behavior, and what
>>>
>>> user wants is components
>>>>>>>>>
>>>>>>>>>  work as
>>>>>>>>>>>
>>>>>>>>>>>    expected.
>>>>>>>>>>>
>>>>>>>>>>>    regards,
>>>>>>>>>>>
>>>>>>>>>>>    Leonardo Uribe
>>>>>>>>>>>
>>>>>>>>>>>    2011/9/21 Mark
>>>
>>> Struberg<[email protected]>:
>>>>>>>>>>>>
>>>>>>>>>>>>    Not sure about that.
>>>>>>>>>>>>    Does the param start with javax.faces? In
>>>
>>> this case we should
>>>>>>>>>
>>>>>>>>>  rather use an own internal one.
>>>>>>>>>>>>
>>>>>>>>>>>>    Btw, if it's not in the spec even
>>>
>>> Mojarra would not be allowed
>>>>>>>>>
>>>>>>>>>  to use a proprietary parameter with
>>>
>>> "javax...."
>>>>>>>>>>>>
>>>>>>>>>>>>    LieGrue,
>>>>>>>>>>>>    strub
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>    ----- Original Message -----
>>>>>>>>>>>>>
>>>>>>>>>>>>>    From: Matt
>>>
>>> Benson<[email protected]>
>>>>>>>>>>>>>
>>>>>>>>>>>>>    To: MyFaces
>>>
>>> Development<[email protected]>
>>>>>>>>>>>>>
>>>>>>>>>>>>>    Cc:
>>>>>>>>>>>>>    Sent: Wednesday, September 21, 2011
>>>
>>> 6:29 PM
>>>>>>>>>>>>>
>>>>>>>>>>>>>    Subject: Re: [VOTE] fix MYFACES-2552
>>>
>>> for 2.0.x and 2.1.x
>>>>>>>>>
>>>>>>>>>  branches
>>>>>>>>>>>>>
>>>>>>>>>>>>>    +1
>>>>>>>>>>>>>
>>>>>>>>>>>>>    However, let's simplify the
>>>
>>> context parameter by giving it
>>>>>>>>>
>>>>>>>>>  a name
>>>>>>>>>>>>>
>>>>>>>>>>>>>    relating to JSF 2.2 compatibility.  I
>>>
>>> submitted the final
>>>>>>>>>>>>>
>>>>>>>>>>>>>    implementation for Mojarra, so have
>>>
>>> every right to add the same
>>>>>>>>>
>>>>>>>>>  to
>>>>>>>>>>>>>
>>>>>>>>>>>>>    MyFaces.
>>>>>>>>>>>>>
>>>>>>>>>>>>>    Matt
>>>>>>>>>>>>>
>>>>>>>>>>>>>    On Wed, Sep 21, 2011 at 11:19 AM,
>>>
>>> Gerhard Petracek
>>>>>>>>>>>>>
>>>>>>>>>>>>>    <[email protected]>
>>>
>>> wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>     +1 for it in combination with
>>>
>>> the context parameter
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>     regards,
>>>>>>>>>>>>>>     gerhard
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>     http://www.irian.at
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>     Your JSF powerhouse -
>>>>>>>>>>>>>>     JSF Consulting, Development and
>>>>>>>>>>>>>>     Courses in English and German
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>     Professional Support for Apache
>>>
>>> MyFaces
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>     2011/9/21 Rudy De
>>>
>>> Busscher<[email protected]>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>     +1
>>>>>>>>>>>>>>>     And if we create a context
>>>
>>> parameter, it should behave
>>>>>>>>>
>>>>>>>>>  by default as in
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>     the JSF 2.2 Spec.  If users
>>>
>>> want strict spec
>>>>>>>>>
>>>>>>>>>  (2.0/2.1)behaviour they
>>>>>>>>>>>>>
>>>>>>>>>>>>>    have to
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>     set the parameter value.
>>>>>>>>>>>>>>>     regards
>>>>>>>>>>>>>>>     Rudy
>>>>>>>>>>>>>>>     On 21 September 2011 17:08,
>>>
>>> Grant Smith
>>>>>>>>>
>>>>>>>>>  <[email protected]>
>>>>>>>>>>>>>
>>>>>>>>>>>>>    wrote:
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>     +1 if it's
>>>
>>> configurable in a
>>>>>>>>>
>>>>>>>>>  <context-param>. How about
>>>>>>>>>>>>>>>>
>>>>>>>>>    org.apache.myfaces.EL_RESOLVER_GETTYPE_RETURNS_NULL ?
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>     On Wed, Sep 21, 2011 at
>>>
>>> 5:35 AM, Michael Kurz
>>>>>>>>>>>>>
>>>>>>>>>>>>>    <[email protected]>   wrote:
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     +1
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     Am 21.09.2011 um
>>>
>>> 14:20 schrieb Leonardo Uribe:
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >   +1
>>>>>>>>>>>>>>>>>     >
>>>>>>>>>>>>>>>>>     >   2011/9/21
>>>
>>> Leonardo Uribe
>>>>>>>>>
>>>>>>>>>  <[email protected]>:
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   Hi
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   More than
>>>
>>> a year ago, it was found
>>>>>>>>>
>>>>>>>>>  that EL expressions
>>>>>>>>>>>>>
>>>>>>>>>>>>>    like
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>
>>> #{cc.attrs.test} does not resolve its
>>>>>>>>>
>>>>>>>>>  type correctly,
>>>>>>>>>>>>>
>>>>>>>>>>>>>    because the
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   composite
>>>
>>> component EL resolver is
>>>>>>>>>
>>>>>>>>>  not able to find
>>>>>>>>>>>>>
>>>>>>>>>>>>>    the right type.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   Instead,
>>>
>>> MapELResolver always return
>>>>>>>>>
>>>>>>>>>  Object.class as
>>>>>>>>>>>>>
>>>>>>>>>>>>>    type, breaking
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   composite
>>>
>>> components that use
>>>>>>>>>
>>>>>>>>>  h:selectOneXXX into its
>>>>>>>>>>>>>
>>>>>>>>>>>>>    internals. See
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>
>>>>>>>>>  https://issues.apache.org/jira/browse/MYFACES-2552
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   The
>>>
>>> problem with this issue is we
>>>>>>>>>
>>>>>>>>>  need to change the
>>>>>>>>>>>>>
>>>>>>>>>>>>>    way how
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>
>>>>>>>>>
>>> org.apache.myfaces.el.unified.resolver.CompositeComponentELResolver
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   works. JSF
>>>
>>> 2.0 spec clearly says in
>>>>>>>>>
>>>>>>>>>  its section
>>>>>>>>>>>>>
>>>>>>>>>>>>>    5.6.2.2 that
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   getType()
>>>>>>>>>>>>>>>>>     >>   for that
>>>
>>> EL resolver should return
>>>>>>>>>
>>>>>>>>>  null.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   The issue
>>>
>>> was reported to the EG and
>>>>>>>>>
>>>>>>>>>  a fix was
>>>>>>>>>>>>>
>>>>>>>>>>>>>    included in JSF 2.2.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   spec, see:
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>
>>> http://java.net/jira/browse/JAVASERVERFACES_SPEC_PUBLIC-745
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   but we
>>>
>>> still receive reports about
>>>>>>>>>
>>>>>>>>>  the same issue
>>>>>>>>>>>>>
>>>>>>>>>>>>>    (MYFACES-3311 and
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   others
>>>
>>> (last comment on MYFACES-1890)
>>>>>>>>>
>>>>>>>>>  ).
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   So, the
>>>
>>> current behavior even if is
>>>>>>>>>
>>>>>>>>>  described by the
>>>>>>>>>>>>>
>>>>>>>>>>>>>    spec is too
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>
>>> inconvenient. Note we already have
>>>>>>>>>
>>>>>>>>>  some places in our
>>>>>>>>>>>>>
>>>>>>>>>>>>>    implementation
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   that does
>>>
>>> not follow strictly the
>>>>>>>>>
>>>>>>>>>  spec, to keep things
>>>>>>>>>>>>>
>>>>>>>>>>>>>    working as
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   users
>>>
>>> expect. To follow the protocol
>>>>>>>>>
>>>>>>>>>  in these cases,
>>>>>>>>>>>>>
>>>>>>>>>>>>>    we need an
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   official
>>>
>>> community decision about
>>>>>>>>>
>>>>>>>>>  include it in 2.0.x
>>>>>>>>>>>>>
>>>>>>>>>>>>>    and 2.1.x
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   branches.
>>>
>>> Please vote:
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   +1 if you
>>>
>>> want this fix included in
>>>>>>>>>
>>>>>>>>>  2.0.x and 2.1.x.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>   +0
>>>>>>>>>>>>>>>>>     >>   -1 and the
>>>
>>> reason why if you see this
>>>>>>>>>
>>>>>>>>>  could cause any
>>>>>>>>>>>>>
>>>>>>>>>>>>>    problem.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   regards,
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   Leonardo
>>>
>>> Uribe
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>     >>   [1]
>>>>>>>>>>>>>
>>> http://www.apache.org/foundation/voting.html#ReleaseVotes
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>     >>
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>     --
>>>>>>>>>>>>>>>>     Grant Smith - V.P.
>>>
>>> Information Technology
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>     Marathon Computer
>>>
>>> Systems, LLC.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>     --
>>>>>>>>>>>>>>>     Rudy De Busscher
>>>>>>>>>>>>>>>     http://www.c4j.be
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>  --
>>>>>>>>>  Jakob Korherr
>>>>>>>>>
>>>>>>>>>  blog: http://www.jakobk.com
>>>>>>>>>  twitter: http://twitter.com/jakobkorherr
>>>>>>>>>  work: http://www.irian.at
>>>>>>>>>
>>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>>  --
>>>>>
>>>>>  http://www.irian.at
>>>>>
>>>>>  Your JSF powerhouse -
>>>>>  JSF Consulting, Development and
>>>>>  Courses in English and German
>>>>>
>>>>>  Professional Support for Apache MyFaces
>>>>>
>>>>
>>>>
>>>>
>>>
>

Reply via email to