Home » Modeling » EMF » OCL Support In EMF?
| OCL Support In EMF? [message #380102] |
Thu, 23 October 2003 18:26  |
Eclipse User |
|
|
|
Hi All:
Thanks in advance for your response.
EMF seems to be a very valuable toolkit. I posted something similar a few
months back asking about plans for adding validation support to EMF. Please
correct me if what I state below is overlooking current capabilities
available in EMF.
We recently exposed what seems to be a deficiency with this framework and
almost all frameworks of this type. We downloaded the WSDL 1.2 XSD, created
an Ecore model from that, and generated model, edit, and editor source code.
Very powerful! Unfortunately, in the process, we lost validation described
in the XSD during the generation process. I realize we can reactively
invoke validation checking on a populated resource. I'm interested in
proactive validation. I would like my generated EMF API to disallow the
creation of content that will ultimately be found to be invalid. I also
realize XSDs are only one source from which an Ecore model can be created. I
am curious, however, if anyone has given thought and/or if there are plans
to acknowledge the value of and address the need for this capability in EMF?
It seems to me there would be tremendous value leveraged by transferring
relevant validation content expressed within an XSD into the Ecore model.
For annotated Java interfaces adding annotations related to validation would
be very useful. For UML models capturing OCL constraints would be very
useful. I notice there is a meta-model in the works for UML Object
Constraint Language (OCL). Any thoughts on extending Ecore by adding the
OCL meta-model or some useful subset therein? Once the information is
captured in the Ecore model code generation can produce validation-oriented
methods that can be invoked proactively via the notification framework.
Proactive vs. reactive validation could be configurable.
I think EMF has nailed the meta-model for content and generation technology
that produces predictable and polished content management and editing APIs.
It seems one of the next steps is supporting the capture of behavioral
constraints and characteristics associated with modeled content in a
technology platform independent fashion and subsequently leveraging that
meta-data during technology platform generation activities.
Thanks for a great tool!
Chris Potter
|
|
|
| Re: OCL Support In EMF? [message #380108 is a reply to message #380102] |
Fri, 24 October 2003 11:13  |
| Eclipse User |
|
|
|
Chris,
We are interested in adding the kind of validation support you describe,
although we have no committed plan stating how and when. We're imagining some
kind of ecore model annotations (EAnnotation), that would affect the code
generation - similar to what I believe you're suggesting. These annotations
could be produced from XML schema, Java annotations, or UML model constrains.
For constrains on simple types (e.g., the value of an int attribute is in the
range of 1..10), we can imagine generation of proactive checking, but for things
like the multiplicity of an association, we believe that the checking can only
be done when the user indicates that the model is consistent. Is that not how
you see it?
At any rate, we imagine EMF providing the ability to generate validation code,
based on constraints defined in the original model, at some point in time. A
prereq for this support is the more general requirement to make the EMF code
generator more customizable/extensible, which we will be working on soon.
Frank.
Chris Potter wrote:
> Hi All:
>
> Thanks in advance for your response.
>
> EMF seems to be a very valuable toolkit. I posted something similar a few
> months back asking about plans for adding validation support to EMF. Please
> correct me if what I state below is overlooking current capabilities
> available in EMF.
>
> We recently exposed what seems to be a deficiency with this framework and
> almost all frameworks of this type. We downloaded the WSDL 1.2 XSD, created
> an Ecore model from that, and generated model, edit, and editor source code.
> Very powerful! Unfortunately, in the process, we lost validation described
> in the XSD during the generation process. I realize we can reactively
> invoke validation checking on a populated resource. I'm interested in
> proactive validation. I would like my generated EMF API to disallow the
> creation of content that will ultimately be found to be invalid. I also
> realize XSDs are only one source from which an Ecore model can be created. I
> am curious, however, if anyone has given thought and/or if there are plans
> to acknowledge the value of and address the need for this capability in EMF?
> It seems to me there would be tremendous value leveraged by transferring
> relevant validation content expressed within an XSD into the Ecore model.
> For annotated Java interfaces adding annotations related to validation would
> be very useful. For UML models capturing OCL constraints would be very
> useful. I notice there is a meta-model in the works for UML Object
> Constraint Language (OCL). Any thoughts on extending Ecore by adding the
> OCL meta-model or some useful subset therein? Once the information is
> captured in the Ecore model code generation can produce validation-oriented
> methods that can be invoked proactively via the notification framework.
> Proactive vs. reactive validation could be configurable.
>
> I think EMF has nailed the meta-model for content and generation technology
> that produces predictable and polished content management and editing APIs.
> It seems one of the next steps is supporting the capture of behavioral
> constraints and characteristics associated with modeled content in a
> technology platform independent fashion and subsequently leveraging that
> meta-data during technology platform generation activities.
>
> Thanks for a great tool!
>
> Chris Potter
|
|
|
| Re: OCL Support In EMF? [message #380137 is a reply to message #380108] |
Mon, 27 October 2003 10:46 |
| Eclipse User |
|
|
|
Thanks for your response Frank. I completely agree with the approach you
describe below. The one item I'm uncertain of, regarding implementation, is
the sufficiency of EAnnotations as the container for this information. It
seems there may be enough content/complexity to describing constraints that
a content model would be beneficial.
Chris Potter
"Frank Budinsky" <frankb@ca.ibm.com> wrote in message
news:3F99418E.A4F2A017@ca.ibm.com...
> Chris,
>
> We are interested in adding the kind of validation support you describe,
> although we have no committed plan stating how and when. We're imagining
some
> kind of ecore model annotations (EAnnotation), that would affect the code
> generation - similar to what I believe you're suggesting. These
annotations
> could be produced from XML schema, Java annotations, or UML model
constrains.
>
> For constrains on simple types (e.g., the value of an int attribute is in
the
> range of 1..10), we can imagine generation of proactive checking, but for
things
> like the multiplicity of an association, we believe that the checking can
only
> be done when the user indicates that the model is consistent. Is that not
how
> you see it?
>
> At any rate, we imagine EMF providing the ability to generate validation
code,
> based on constraints defined in the original model, at some point in time.
A
> prereq for this support is the more general requirement to make the EMF
code
> generator more customizable/extensible, which we will be working on soon.
>
> Frank.
>
>
> Chris Potter wrote:
>
> > Hi All:
> >
> > Thanks in advance for your response.
> >
> > EMF seems to be a very valuable toolkit. I posted something similar a
few
> > months back asking about plans for adding validation support to EMF.
Please
> > correct me if what I state below is overlooking current capabilities
> > available in EMF.
> >
> > We recently exposed what seems to be a deficiency with this framework
and
> > almost all frameworks of this type. We downloaded the WSDL 1.2 XSD,
created
> > an Ecore model from that, and generated model, edit, and editor source
code.
> > Very powerful! Unfortunately, in the process, we lost validation
described
> > in the XSD during the generation process. I realize we can reactively
> > invoke validation checking on a populated resource. I'm interested in
> > proactive validation. I would like my generated EMF API to disallow the
> > creation of content that will ultimately be found to be invalid. I also
> > realize XSDs are only one source from which an Ecore model can be
created. I
> > am curious, however, if anyone has given thought and/or if there are
plans
> > to acknowledge the value of and address the need for this capability in
EMF?
> > It seems to me there would be tremendous value leveraged by transferring
> > relevant validation content expressed within an XSD into the Ecore
model.
> > For annotated Java interfaces adding annotations related to validation
would
> > be very useful. For UML models capturing OCL constraints would be very
> > useful. I notice there is a meta-model in the works for UML Object
> > Constraint Language (OCL). Any thoughts on extending Ecore by adding
the
> > OCL meta-model or some useful subset therein? Once the information is
> > captured in the Ecore model code generation can produce
validation-oriented
> > methods that can be invoked proactively via the notification framework.
> > Proactive vs. reactive validation could be configurable.
> >
> > I think EMF has nailed the meta-model for content and generation
technology
> > that produces predictable and polished content management and editing
APIs.
> > It seems one of the next steps is supporting the capture of behavioral
> > constraints and characteristics associated with modeled content in a
> > technology platform independent fashion and subsequently leveraging that
> > meta-data during technology platform generation activities.
> >
> > Thanks for a great tool!
> >
> > Chris Potter
>
|
|
|
| Re: OCL Support In EMF? [message #380138 is a reply to message #380137] |
Mon, 27 October 2003 11:10 |
| Eclipse User |
|
|
|
Chris,
Maybe you are interpreting Frank's comment as saying that we would use
EAnnotations to "model" the constraints. The point is more that
EAnnotation.contents can hold any EObject so using an annotation any other model
can be embedded directly into the Ecore model.
Chris Potter wrote:
> Thanks for your response Frank. I completely agree with the approach you
> describe below. The one item I'm uncertain of, regarding implementation, is
> the sufficiency of EAnnotations as the container for this information. It
> seems there may be enough content/complexity to describing constraints that
> a content model would be beneficial.
>
> Chris Potter
>
> "Frank Budinsky" <frankb@ca.ibm.com> wrote in message
> news:3F99418E.A4F2A017@ca.ibm.com...
> > Chris,
> >
> > We are interested in adding the kind of validation support you describe,
> > although we have no committed plan stating how and when. We're imagining
> some
> > kind of ecore model annotations (EAnnotation), that would affect the code
> > generation - similar to what I believe you're suggesting. These
> annotations
> > could be produced from XML schema, Java annotations, or UML model
> constrains.
> >
> > For constrains on simple types (e.g., the value of an int attribute is in
> the
> > range of 1..10), we can imagine generation of proactive checking, but for
> things
> > like the multiplicity of an association, we believe that the checking can
> only
> > be done when the user indicates that the model is consistent. Is that not
> how
> > you see it?
> >
> > At any rate, we imagine EMF providing the ability to generate validation
> code,
> > based on constraints defined in the original model, at some point in time.
> A
> > prereq for this support is the more general requirement to make the EMF
> code
> > generator more customizable/extensible, which we will be working on soon.
> >
> > Frank.
> >
> >
> > Chris Potter wrote:
> >
> > > Hi All:
> > >
> > > Thanks in advance for your response.
> > >
> > > EMF seems to be a very valuable toolkit. I posted something similar a
> few
> > > months back asking about plans for adding validation support to EMF.
> Please
> > > correct me if what I state below is overlooking current capabilities
> > > available in EMF.
> > >
> > > We recently exposed what seems to be a deficiency with this framework
> and
> > > almost all frameworks of this type. We downloaded the WSDL 1.2 XSD,
> created
> > > an Ecore model from that, and generated model, edit, and editor source
> code.
> > > Very powerful! Unfortunately, in the process, we lost validation
> described
> > > in the XSD during the generation process. I realize we can reactively
> > > invoke validation checking on a populated resource. I'm interested in
> > > proactive validation. I would like my generated EMF API to disallow the
> > > creation of content that will ultimately be found to be invalid. I also
> > > realize XSDs are only one source from which an Ecore model can be
> created. I
> > > am curious, however, if anyone has given thought and/or if there are
> plans
> > > to acknowledge the value of and address the need for this capability in
> EMF?
> > > It seems to me there would be tremendous value leveraged by transferring
> > > relevant validation content expressed within an XSD into the Ecore
> model.
> > > For annotated Java interfaces adding annotations related to validation
> would
> > > be very useful. For UML models capturing OCL constraints would be very
> > > useful. I notice there is a meta-model in the works for UML Object
> > > Constraint Language (OCL). Any thoughts on extending Ecore by adding
> the
> > > OCL meta-model or some useful subset therein? Once the information is
> > > captured in the Ecore model code generation can produce
> validation-oriented
> > > methods that can be invoked proactively via the notification framework.
> > > Proactive vs. reactive validation could be configurable.
> > >
> > > I think EMF has nailed the meta-model for content and generation
> technology
> > > that produces predictable and polished content management and editing
> APIs.
> > > It seems one of the next steps is supporting the capture of behavioral
> > > constraints and characteristics associated with modeled content in a
> > > technology platform independent fashion and subsequently leveraging that
> > > meta-data during technology platform generation activities.
> > >
> > > Thanks for a great tool!
> > >
> > > Chris Potter
> >
|
|
| |
| Re: Annotations in EMF (was: OCL Support In EMF?) [message #380156 is a reply to message #380146] |
Mon, 27 October 2003 15:33 |
| Eclipse User |
|
|
|
Dave,
I kind of doubt that the EMF model for the XHTML schema would be usable, but it
does seem feasible.
In EMF 2.0 we will be able to handle the XML Schema "anyType", which has mixed
content, and lax element and attribute wildcards. As such we will be able to
read in (and write back out) the equivalent of arbitrary unvalidated DOM...
Dave Carlson wrote:
> Ed,
>
> In a giant leap of faith... I'm thinking back to an earlier thread about
> generalizing the model for representing XSD annotations in the EMF model
> derived from a schema.
>
> If you take the XML Schema for XHTML (probably the "basic" subset) and
> generate an EMF model for it, could you use this to represent any XHTML
> content as explicitly modeling EMF objects in an annotation? The same might
> be useful for models other than XSD, where rich-formatted annotations are
> represented as XHTML objects.
>
> Is this feasible? Easy to do?
>
> Cheers,
> Dave Carlson
>
> "Ed Merks" <merks@ca.ibm.com> wrote in message
> news:3F9D4392.5E5F1204@ca.ibm.com...
> > Chris,
> >
> > Maybe you are interpreting Frank's comment as saying that we would use
> > EAnnotations to "model" the constraints. The point is more that
> > EAnnotation.contents can hold any EObject so using an annotation any other
> model
> > can be embedded directly into the Ecore model.
> >
> >
|
|
|
| Re: OCL Support In EMF? [message #380179 is a reply to message #380138] |
Tue, 28 October 2003 00:19 |
| Eclipse User |
|
|
|
Hi Ed:
Thanks for contributing to this thread. I think I'm following you and do
see how EAnnotations could be used as the vehicle for adding validation
"hooks" into Ecore models. It seems to me, however, validation would best
be represented within the EMF kernel or core meta-model. I think the type
of validation constructs supported within an Ecore meta-model would be
different for EAttributes, EReferences, and EClasses. Adding support for
validation by creating EAnnotation instances within an Ecore model wouldn't
constrain validation constructs to what would be appropriate for each of
these specialized EModelElements. I view EAnnotations as an excellent means
by which to annotate an Ecore model with functionality specific to a
particular EMF consumer (or in the case of XML Schema, model enrichment
unique to that meta-model source). I view the need to express validation
content within EAttributes, EReferences, and EClasses as important as
expressing that an EClass can have an association of EReferences.
I hope I'm not missing your point.
Thanks,
Chris Potter
"Ed Merks" <merks@ca.ibm.com> wrote in message
news:3F9D4392.5E5F1204@ca.ibm.com...
> Chris,
>
> Maybe you are interpreting Frank's comment as saying that we would use
> EAnnotations to "model" the constraints. The point is more that
> EAnnotation.contents can hold any EObject so using an annotation any other
model
> can be embedded directly into the Ecore model.
>
>
> Chris Potter wrote:
>
> > Thanks for your response Frank. I completely agree with the approach
you
> > describe below. The one item I'm uncertain of, regarding
implementation, is
> > the sufficiency of EAnnotations as the container for this information.
It
> > seems there may be enough content/complexity to describing constraints
that
> > a content model would be beneficial.
> >
> > Chris Potter
> >
> > "Frank Budinsky" <frankb@ca.ibm.com> wrote in message
> > news:3F99418E.A4F2A017@ca.ibm.com...
> > > Chris,
> > >
> > > We are interested in adding the kind of validation support you
describe,
> > > although we have no committed plan stating how and when. We're
imagining
> > some
> > > kind of ecore model annotations (EAnnotation), that would affect the
code
> > > generation - similar to what I believe you're suggesting. These
> > annotations
> > > could be produced from XML schema, Java annotations, or UML model
> > constrains.
> > >
> > > For constrains on simple types (e.g., the value of an int attribute is
in
> > the
> > > range of 1..10), we can imagine generation of proactive checking, but
for
> > things
> > > like the multiplicity of an association, we believe that the checking
can
> > only
> > > be done when the user indicates that the model is consistent. Is that
not
> > how
> > > you see it?
> > >
> > > At any rate, we imagine EMF providing the ability to generate
validation
> > code,
> > > based on constraints defined in the original model, at some point in
time.
> > A
> > > prereq for this support is the more general requirement to make the
EMF
> > code
> > > generator more customizable/extensible, which we will be working on
soon.
> > >
> > > Frank.
> > >
> > >
> > > Chris Potter wrote:
> > >
> > > > Hi All:
> > > >
> > > > Thanks in advance for your response.
> > > >
> > > > EMF seems to be a very valuable toolkit. I posted something similar
a
> > few
> > > > months back asking about plans for adding validation support to EMF.
> > Please
> > > > correct me if what I state below is overlooking current capabilities
> > > > available in EMF.
> > > >
> > > > We recently exposed what seems to be a deficiency with this
framework
> > and
> > > > almost all frameworks of this type. We downloaded the WSDL 1.2 XSD,
> > created
> > > > an Ecore model from that, and generated model, edit, and editor
source
> > code.
> > > > Very powerful! Unfortunately, in the process, we lost validation
> > described
> > > > in the XSD during the generation process. I realize we can
reactively
> > > > invoke validation checking on a populated resource. I'm interested
in
> > > > proactive validation. I would like my generated EMF API to disallow
the
> > > > creation of content that will ultimately be found to be invalid. I
also
> > > > realize XSDs are only one source from which an Ecore model can be
> > created. I
> > > > am curious, however, if anyone has given thought and/or if there are
> > plans
> > > > to acknowledge the value of and address the need for this capability
in
> > EMF?
> > > > It seems to me there would be tremendous value leveraged by
transferring
> > > > relevant validation content expressed within an XSD into the Ecore
> > model.
> > > > For annotated Java interfaces adding annotations related to
validation
> > would
> > > > be very useful. For UML models capturing OCL constraints would be
very
> > > > useful. I notice there is a meta-model in the works for UML Object
> > > > Constraint Language (OCL). Any thoughts on extending Ecore by
adding
> > the
> > > > OCL meta-model or some useful subset therein? Once the information
is
> > > > captured in the Ecore model code generation can produce
> > validation-oriented
> > > > methods that can be invoked proactively via the notification
framework.
> > > > Proactive vs. reactive validation could be configurable.
> > > >
> > > > I think EMF has nailed the meta-model for content and generation
> > technology
> > > > that produces predictable and polished content management and
editing
> > APIs.
> > > > It seems one of the next steps is supporting the capture of
behavioral
> > > > constraints and characteristics associated with modeled content in a
> > > > technology platform independent fashion and subsequently leveraging
that
> > > > meta-data during technology platform generation activities.
> > > >
> > > > Thanks for a great tool!
> > > >
> > > > Chris Potter
> > >
>
|
|
|
| Re: OCL Support In EMF? [message #380180 is a reply to message #380179] |
Tue, 28 October 2003 07:44 |
| Eclipse User |
|
|
|
Chris,
My concern with extending the Ecore model directly to express constraints is the
possibility or even likelihood that a constraint model would quickly dwarf the
core model itself; one need only look at XML Schema to get a sense for how
quickly complexity can grow beyond the grasp of mere mortals. So my initial
preference would be to ensure that the core can standalone and that additional
features act like layers on the onion that can be peeled away by those who only
need or want the core. I can imagine that embedding an EReferenceConstraint in
an EReference's annotation, and an EAttributeConstraint in a EAttribute's
annotation, would support the kind of specialized constraints you allude to
below.
Chris Potter wrote:
> Hi Ed:
>
> Thanks for contributing to this thread. I think I'm following you and do
> see how EAnnotations could be used as the vehicle for adding validation
> "hooks" into Ecore models. It seems to me, however, validation would best
> be represented within the EMF kernel or core meta-model. I think the type
> of validation constructs supported within an Ecore meta-model would be
> different for EAttributes, EReferences, and EClasses. Adding support for
> validation by creating EAnnotation instances within an Ecore model wouldn't
> constrain validation constructs to what would be appropriate for each of
> these specialized EModelElements. I view EAnnotations as an excellent means
> by which to annotate an Ecore model with functionality specific to a
> particular EMF consumer (or in the case of XML Schema, model enrichment
> unique to that meta-model source). I view the need to express validation
> content within EAttributes, EReferences, and EClasses as important as
> expressing that an EClass can have an association of EReferences.
>
> I hope I'm not missing your point.
>
> Thanks,
> Chris Potter
>
> "Ed Merks" <merks@ca.ibm.com> wrote in message
> news:3F9D4392.5E5F1204@ca.ibm.com...
> > Chris,
> >
> > Maybe you are interpreting Frank's comment as saying that we would use
> > EAnnotations to "model" the constraints. The point is more that
> > EAnnotation.contents can hold any EObject so using an annotation any other
> model
> > can be embedded directly into the Ecore model.
> >
> >
> > Chris Potter wrote:
> >
> > > Thanks for your response Frank. I completely agree with the approach
> you
> > > describe below. The one item I'm uncertain of, regarding
> implementation, is
> > > the sufficiency of EAnnotations as the container for this information.
> It
> > > seems there may be enough content/complexity to describing constraints
> that
> > > a content model would be beneficial.
> > >
> > > Chris Potter
> > >
> > > "Frank Budinsky" <frankb@ca.ibm.com> wrote in message
> > > news:3F99418E.A4F2A017@ca.ibm.com...
> > > > Chris,
> > > >
> > > > We are interested in adding the kind of validation support you
> describe,
> > > > although we have no committed plan stating how and when. We're
> imagining
> > > some
> > > > kind of ecore model annotations (EAnnotation), that would affect the
> code
> > > > generation - similar to what I believe you're suggesting. These
> > > annotations
> > > > could be produced from XML schema, Java annotations, or UML model
> > > constrains.
> > > >
> > > > For constrains on simple types (e.g., the value of an int attribute is
> in
> > > the
> > > > range of 1..10), we can imagine generation of proactive checking, but
> for
> > > things
> > > > like the multiplicity of an association, we believe that the checking
> can
> > > only
> > > > be done when the user indicates that the model is consistent. Is that
> not
> > > how
> > > > you see it?
> > > >
> > > > At any rate, we imagine EMF providing the ability to generate
> validation
> > > code,
> > > > based on constraints defined in the original model, at some point in
> time.
> > > A
> > > > prereq for this support is the more general requirement to make the
> EMF
> > > code
> > > > generator more customizable/extensible, which we will be working on
> soon.
> > > >
> > > > Frank.
> > > >
> > > >
> > > > Chris Potter wrote:
> > > >
> > > > > Hi All:
> > > > >
> > > > > Thanks in advance for your response.
> > > > >
> > > > > EMF seems to be a very valuable toolkit. I posted something similar
> a
> > > few
> > > > > months back asking about plans for adding validation support to EMF.
> > > Please
> > > > > correct me if what I state below is overlooking current capabilities
> > > > > available in EMF.
> > > > >
> > > > > We recently exposed what seems to be a deficiency with this
> framework
> > > and
> > > > > almost all frameworks of this type. We downloaded the WSDL 1.2 XSD,
> > > created
> > > > > an Ecore model from that, and generated model, edit, and editor
> source
> > > code.
> > > > > Very powerful! Unfortunately, in the process, we lost validation
> > > described
> > > > > in the XSD during the generation process. I realize we can
> reactively
> > > > > invoke validation checking on a populated resource. I'm interested
> in
> > > > > proactive validation. I would like my generated EMF API to disallow
> the
> > > > > creation of content that will ultimately be found to be invalid. I
> also
> > > > > realize XSDs are only one source from which an Ecore model can be
> > > created. I
> > > > > am curious, however, if anyone has given thought and/or if there are
> > > plans
> > > > > to acknowledge the value of and address the need for this capability
> in
> > > EMF?
> > > > > It seems to me there would be tremendous value leveraged by
> transferring
> > > > > relevant validation content expressed within an XSD into the Ecore
> > > model.
> > > > > For annotated Java interfaces adding annotations related to
> validation
> > > would
> > > > > be very useful. For UML models capturing OCL constraints would be
> very
> > > > > useful. I notice there is a meta-model in the works for UML Object
> > > > > Constraint Language (OCL). Any thoughts on extending Ecore by
> adding
> > > the
> > > > > OCL meta-model or some useful subset therein? Once the information
> is
> > > > > captured in the Ecore model code generation can produce
> > > validation-oriented
> > > > > methods that can be invoked proactively via the notification
> framework.
> > > > > Proactive vs. reactive validation could be configurable.
> > > > >
> > > > > I think EMF has nailed the meta-model for content and generation
> > > technology
> > > > > that produces predictable and polished content management and
> editing
> > > APIs.
> > > > > It seems one of the next steps is supporting the capture of
> behavioral
> > > > > constraints and characteristics associated with modeled content in a
> > > > > technology platform independent fashion and subsequently leveraging
> that
> > > > > meta-data during technology platform generation activities.
> > > > >
> > > > > Thanks for a great tool!
> > > > >
> > > > > Chris Potter
> > > >
> >
|
|
|
Goto Forum:
Current Time: Thu Sep 04 19:51:08 EDT 2025
Powered by FUDForum. Page generated in 0.03690 seconds |
Home