Home » Modeling » EMF "Technology" (Ecore Tools, EMFatic, etc) » EEF Toolchain
| EEF Toolchain [message #134636] |
Fri, 03 April 2009 15:05  |
Eclipse User |
|
|
|
I was just taking a look at the EEF creation review, congratulations
and I'm looking forward to employing it. I was wondering about plans
for the toolchain and how that might evolve toward the planned December
(i.e. not ported) release. I'm not super familiar with Accelo proper
but it looks similar to JET approach and there appears to be a
generation workflow model. I'm wondering if you have looked at XPand
for future iterations? This is something that I'd mentioned over on the
EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
is that the template language itself is much more modular / extendible
and it fits into a toolchain that works very well for user-driven code
generation. This might not matter as much for this usage especially as
you have a lot of plans for meta-model driven generation and API
extensibility. Wondering the same thing about MWE, is there a
possiblity of having a hook into that?
I realize that this a bit of a can of worms as there are two very good
(and pretty complimantary) M2T approaches within Eclipse Modeling --
I'm just wondering if there has been any consideration of the issue and
wether folks have any more thoughts on it.
|
|
|
| Re: EEF Toolchain [message #134651 is a reply to message #134636] |
Mon, 06 April 2009 05:04  |
| Eclipse User |
|
|
|
Hi,
M2T/Acceleo is just so much better than XPand ;)
More seriously, EEF is based on M2T/Acceleo which is the Eclipse implementation of the OMG Standard "Mof Model To Text". It's the "next gen" implementation of the Acceleo you are talking about (JSP like syntax). It provides, among other things, better extensibility, modularity and workflow support and, as a side note, standard compliance.
Moreover we are keeping the "pragmatism and simplicity" which made Acceleo.org a success.
You should definitely have a closer look on the M2T/Acceleo project :
http://www.eclipse.org/modeling/m2t/?project=acceleo#acceleo
For the MWE, that's not something which is planned but the team would be happy to integrate a contribution. As Acceleo is standalone and already provides ANT tasks using it with the MWE is quite straightforward.
Cédric
Miles Parker wrote:
>
> I was just taking a look at the EEF creation review, congratulations
> and I'm looking forward to employing it. I was wondering about plans
> for the toolchain and how that might evolve toward the planned December
> (i.e. not ported) release. I'm not super familiar with Accelo proper
> but it looks similar to JET approach and there appears to be a
> generation workflow model. I'm wondering if you have looked at XPand
> for future iterations? This is something that I'd mentioned over on the
> EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
> is that the template language itself is much more modular / extendible
> and it fits into a toolchain that works very well for user-driven code
> generation. This might not matter as much for this usage especially as
> you have a lot of plans for meta-model driven generation and API
> extensibility. Wondering the same thing about MWE, is there a
> possiblity of having a hook into that?
>
> I realize that this a bit of a can of worms as there are two very good
> (and pretty complimantary) M2T approaches within Eclipse Modeling --
> I'm just wondering if there has been any consideration of the issue and
> wether folks have any more thoughts on it.
|
|
|
| Re: EEF Toolchain [message #134690 is a reply to message #134651] |
Mon, 06 April 2009 11:58 |
| Eclipse User |
|
|
|
This is a multi-part message in MIME format.
--------------020708050901030102060607
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Hi,
see my comments below.
Cédric Brun a écrit :
> Hi,
>
> M2T/Acceleo is just so much better than XPand ;)
>
> More seriously, EEF is based on M2T/Acceleo which is the Eclipse implementation of the OMG Standard "Mof Model To Text". It's the "next gen" implementation of the Acceleo you are talking about (JSP like syntax). It provides, among other things, better extensibility, modularity and workflow support and, as a side note, standard compliance.
> Moreover we are keeping the "pragmatism and simplicity" which made Acceleo.org a success.
>
> You should definitely have a closer look on the M2T/Acceleo project :
> http://www.eclipse.org/modeling/m2t/?project=acceleo#acceleo
Like said cédric, the EEf project uses the eclipse's acceleo project (
formerly named MTL ) as genreration framework. it perfectly fits our
needs, and we already used the dynamic/extensible capabilities of it to
generates specific code ( for the papyrus project ).
Regards,
Stéphane Bouchet.
>
> For the MWE, that's not something which is planned but the team would be happy to integrate a contribution. As Acceleo is standalone and already provides ANT tasks using it with the MWE is quite straightforward.
>
>
> Cédric
>
> Miles Parker wrote:
>
>> I was just taking a look at the EEF creation review, congratulations
>> and I'm looking forward to employing it. I was wondering about plans
>> for the toolchain and how that might evolve toward the planned December
>> (i.e. not ported) release. I'm not super familiar with Accelo proper
>> but it looks similar to JET approach and there appears to be a
>> generation workflow model. I'm wondering if you have looked at XPand
>> for future iterations? This is something that I'd mentioned over on the
>> EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
>> is that the template language itself is much more modular / extendible
>> and it fits into a toolchain that works very well for user-driven code
>> generation. This might not matter as much for this usage especially as
>> you have a lot of plans for meta-model driven generation and API
>> extensibility. Wondering the same thing about MWE, is there a
>> possiblity of having a hook into that?
>>
>> I realize that this a bit of a can of worms as there are two very good
>> (and pretty complimantary) M2T approaches within Eclipse Modeling --
>> I'm just wondering if there has been any consideration of the issue and
>> wether folks have any more thoughts on it.
>
--------------020708050901030102060607
Content-Type: text/x-vcard; charset=utf-8;
name="stephane_bouchet.vcf"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename="stephane_bouchet.vcf"
YmVnaW46dmNhcmQNCmZuO3F1b3RlZC1wcmludGFibGU6U3Q9QzM9QTlwaGFu ZSBCb3VjaGV0
DQpuO3F1b3RlZC1wcmludGFibGU6Qm91Y2hldDtTdD1DMz1BOXBoYW5lDQpv cmc6T2Jlbw0K
YWRyO3F1b3RlZC1wcmludGFibGU6Ozs7UmV6PUMzPUE5Ozs0NDQwMDtGcmFu Y2UNCmVtYWls
O2ludGVybmV0OnN0ZXBoYW5lLmJvdWNoZXRAb2Jlby5mcg0KeC1tb3ppbGxh LWh0bWw6RkFM
U0UNCnZlcnNpb246Mi4xDQplbmQ6dmNhcmQNCg0K
--------------020708050901030102060607--
|
|
|
| Re: EEF Toolchain [message #134702 is a reply to message #134651] |
Mon, 06 April 2009 17:18 |
| Eclipse User |
|
|
|
On 2009-04-06 02:04:47 -0700, Cédric Brun <cedric.brun@obeo.fr> said:
> Hi,
>
> M2T/Acceleo is just so much better than XPand ;)
OK, I should have known better! :D Actually, semanitcally it looks a
lot closer to XPand now then JET. ;) And it looks like the JSP-like
syntax :P is gone. Extensiblity and modularity aside from syntax
preference were the areas that have been frustrating for me w/ JET, so
I'll look forward to taking a closer a look at the MTL stuff. Is there
an aspect like construct? (maybe thats "before, after"..)
It looks like there will be two main M2T language threads for Modeling
M2T so we can't hope for one answer fits all -- which as we all know is
impossible anyway.
>
> More seriously, EEF is based on M2T/Acceleo which is the Eclipse
> implementation of the OMG Standard "Mof Model To Text". It's the "next
> gen" implementation of the Acceleo you are talking about (JSP like
> syntax). It provides, among other things, better extensibility,
> modularity and workflow support and, as a side note, standard
> compliance.
> Moreover we are keeping the "pragmatism and simplicity" which made
> Acceleo.org a success.
>
> You should definitely have a closer look on the M2T/Acceleo project :
> http://www.eclipse.org/modeling/m2t/?project=acceleo#acceleo
>
> For the MWE, that's not something which is planned but the team would
> be happy to integrate a contribution. As Acceleo is standalone and
> already provides ANT tasks using it with the MWE is quite
> straightforward.
>
>
> Cédric
>
> Miles Parker wrote:
>
>>
>> I was just taking a look at the EEF creation review, congratulations
>> and I'm looking forward to employing it. I was wondering about plans
>> for the toolchain and how that might evolve toward the planned December
>> (i.e. not ported) release. I'm not super familiar with Accelo proper
>> but it looks similar to JET approach and there appears to be a
>> generation workflow model. I'm wondering if you have looked at XPand
>> for future iterations? This is something that I'd mentioned over on the
>> EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
>> is that the template language itself is much more modular / extendible
>> and it fits into a toolchain that works very well for user-driven code
>> generation. This might not matter as much for this usage especially as
>> you have a lot of plans for meta-model driven generation and API
>> extensibility. Wondering the same thing about MWE, is there a
>> possiblity of having a hook into that?
>>
>> I realize that this a bit of a can of worms as there are two very good
>> (and pretty complimantary) M2T approaches within Eclipse Modeling --
>> I'm just wondering if there has been any consideration of the issue and
>> wether folks have any more thoughts on it.
|
|
| | |
| Re: EEF Toolchain [message #134755 is a reply to message #134702] |
Tue, 07 April 2009 05:44 |
| Eclipse User |
|
|
|
> OK, I should have known better! :D Actually, semanitcally it looks a
> lot closer to XPand now then JET. ;) And it looks like the JSP-like
> syntax :P is gone. Extensiblity and modularity aside from syntax
> preference were the areas that have been frustrating for me w/ JET, so
> I'll look forward to taking a closer a look at the MTL stuff. Is there
> an aspect like construct? (maybe thats "before, after"..)
>
Yes, you can override a template and call the "original" one from your overriding template, so you can basically have
Before stuff
call the original template
After stuff.
And use conditionals in your template to manage your extension.
Cédric
|
|
|
| Re: EEF Toolchain [message #620777 is a reply to message #134636] |
Mon, 06 April 2009 05:04 |
| Eclipse User |
|
|
|
Hi,
M2T/Acceleo is just so much better than XPand ;)
More seriously, EEF is based on M2T/Acceleo which is the Eclipse implementation of the OMG Standard "Mof Model To Text". It's the "next gen" implementation of the Acceleo you are talking about (JSP like syntax). It provides, among other things, better extensibility, modularity and workflow support and, as a side note, standard compliance.
Moreover we are keeping the "pragmatism and simplicity" which made Acceleo.org a success.
You should definitely have a closer look on the M2T/Acceleo project :
http://www.eclipse.org/modeling/m2t/?project=acceleo#acceleo
For the MWE, that's not something which is planned but the team would be happy to integrate a contribution. As Acceleo is standalone and already provides ANT tasks using it with the MWE is quite straightforward.
Cédric
Miles Parker wrote:
>
> I was just taking a look at the EEF creation review, congratulations
> and I'm looking forward to employing it. I was wondering about plans
> for the toolchain and how that might evolve toward the planned December
> (i.e. not ported) release. I'm not super familiar with Accelo proper
> but it looks similar to JET approach and there appears to be a
> generation workflow model. I'm wondering if you have looked at XPand
> for future iterations? This is something that I'd mentioned over on the
> EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
> is that the template language itself is much more modular / extendible
> and it fits into a toolchain that works very well for user-driven code
> generation. This might not matter as much for this usage especially as
> you have a lot of plans for meta-model driven generation and API
> extensibility. Wondering the same thing about MWE, is there a
> possiblity of having a hook into that?
>
> I realize that this a bit of a can of worms as there are two very good
> (and pretty complimantary) M2T approaches within Eclipse Modeling --
> I'm just wondering if there has been any consideration of the issue and
> wether folks have any more thoughts on it.
|
|
|
| Re: EEF Toolchain [message #620780 is a reply to message #134651] |
Mon, 06 April 2009 11:58 |
| Eclipse User |
|
|
|
This is a multi-part message in MIME format.
--------------020708050901030102060607
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Hi,
see my comments below.
Cédric Brun a écrit :
> Hi,
>
> M2T/Acceleo is just so much better than XPand ;)
>
> More seriously, EEF is based on M2T/Acceleo which is the Eclipse implementation of the OMG Standard "Mof Model To Text". It's the "next gen" implementation of the Acceleo you are talking about (JSP like syntax). It provides, among other things, better extensibility, modularity and workflow support and, as a side note, standard compliance.
> Moreover we are keeping the "pragmatism and simplicity" which made Acceleo.org a success.
>
> You should definitely have a closer look on the M2T/Acceleo project :
> http://www.eclipse.org/modeling/m2t/?project=acceleo#acceleo
Like said cédric, the EEf project uses the eclipse's acceleo project (
formerly named MTL ) as genreration framework. it perfectly fits our
needs, and we already used the dynamic/extensible capabilities of it to
generates specific code ( for the papyrus project ).
Regards,
Stéphane Bouchet.
>
> For the MWE, that's not something which is planned but the team would be happy to integrate a contribution. As Acceleo is standalone and already provides ANT tasks using it with the MWE is quite straightforward.
>
>
> Cédric
>
> Miles Parker wrote:
>
>> I was just taking a look at the EEF creation review, congratulations
>> and I'm looking forward to employing it. I was wondering about plans
>> for the toolchain and how that might evolve toward the planned December
>> (i.e. not ported) release. I'm not super familiar with Accelo proper
>> but it looks similar to JET approach and there appears to be a
>> generation workflow model. I'm wondering if you have looked at XPand
>> for future iterations? This is something that I'd mentioned over on the
>> EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
>> is that the template language itself is much more modular / extendible
>> and it fits into a toolchain that works very well for user-driven code
>> generation. This might not matter as much for this usage especially as
>> you have a lot of plans for meta-model driven generation and API
>> extensibility. Wondering the same thing about MWE, is there a
>> possiblity of having a hook into that?
>>
>> I realize that this a bit of a can of worms as there are two very good
>> (and pretty complimantary) M2T approaches within Eclipse Modeling --
>> I'm just wondering if there has been any consideration of the issue and
>> wether folks have any more thoughts on it.
>
--------------020708050901030102060607
Content-Type: text/x-vcard; charset=utf-8;
name="stephane_bouchet.vcf"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
filename="stephane_bouchet.vcf"
YmVnaW46dmNhcmQNCmZuO3F1b3RlZC1wcmludGFibGU6U3Q9QzM9QTlwaGFu ZSBCb3VjaGV0
DQpuO3F1b3RlZC1wcmludGFibGU6Qm91Y2hldDtTdD1DMz1BOXBoYW5lDQpv cmc6T2Jlbw0K
YWRyO3F1b3RlZC1wcmludGFibGU6Ozs7UmV6PUMzPUE5Ozs0NDQwMDtGcmFu Y2UNCmVtYWls
O2ludGVybmV0OnN0ZXBoYW5lLmJvdWNoZXRAb2Jlby5mcg0KeC1tb3ppbGxh LWh0bWw6RkFM
U0UNCnZlcnNpb246Mi4xDQplbmQ6dmNhcmQNCg0K
--------------020708050901030102060607--
|
|
|
| Re: EEF Toolchain [message #620781 is a reply to message #134651] |
Mon, 06 April 2009 17:18 |
| Eclipse User |
|
|
|
On 2009-04-06 02:04:47 -0700, Cédric Brun <cedric.brun@obeo.fr> said:
> Hi,
>
> M2T/Acceleo is just so much better than XPand ;)
OK, I should have known better! :D Actually, semanitcally it looks a
lot closer to XPand now then JET. ;) And it looks like the JSP-like
syntax :P is gone. Extensiblity and modularity aside from syntax
preference were the areas that have been frustrating for me w/ JET, so
I'll look forward to taking a closer a look at the MTL stuff. Is there
an aspect like construct? (maybe thats "before, after"..)
It looks like there will be two main M2T language threads for Modeling
M2T so we can't hope for one answer fits all -- which as we all know is
impossible anyway.
>
> More seriously, EEF is based on M2T/Acceleo which is the Eclipse
> implementation of the OMG Standard "Mof Model To Text". It's the "next
> gen" implementation of the Acceleo you are talking about (JSP like
> syntax). It provides, among other things, better extensibility,
> modularity and workflow support and, as a side note, standard
> compliance.
> Moreover we are keeping the "pragmatism and simplicity" which made
> Acceleo.org a success.
>
> You should definitely have a closer look on the M2T/Acceleo project :
> http://www.eclipse.org/modeling/m2t/?project=acceleo#acceleo
>
> For the MWE, that's not something which is planned but the team would
> be happy to integrate a contribution. As Acceleo is standalone and
> already provides ANT tasks using it with the MWE is quite
> straightforward.
>
>
> Cédric
>
> Miles Parker wrote:
>
>>
>> I was just taking a look at the EEF creation review, congratulations
>> and I'm looking forward to employing it. I was wondering about plans
>> for the toolchain and how that might evolve toward the planned December
>> (i.e. not ported) release. I'm not super familiar with Accelo proper
>> but it looks similar to JET approach and there appears to be a
>> generation workflow model. I'm wondering if you have looked at XPand
>> for future iterations? This is something that I'd mentioned over on the
>> EMF newgroups WRT to EMF.Edit. A principal advantage from XPand for me
>> is that the template language itself is much more modular / extendible
>> and it fits into a toolchain that works very well for user-driven code
>> generation. This might not matter as much for this usage especially as
>> you have a lot of plans for meta-model driven generation and API
>> extensibility. Wondering the same thing about MWE, is there a
>> possiblity of having a hook into that?
>>
>> I realize that this a bit of a can of worms as there are two very good
>> (and pretty complimantary) M2T approaches within Eclipse Modeling --
>> I'm just wondering if there has been any consideration of the issue and
>> wether folks have any more thoughts on it.
|
|
|
| Re: EEF Toolchain [message #620782 is a reply to message #134702] |
Tue, 07 April 2009 04:26 |
| Eclipse User |
|
|
|
Miles Parker schrieb:
> It looks like there will be two main M2T language threads for Modeling
> M2T so we can't hope for one answer fits all -- which as we all know is
> impossible anyway.
>
Yes, although Xpand and Acceleo are somewhat similar the important
difference is, that
- Acceleo is based on an OMG standard
- where Xpand is not
It's good and important to have both, because standards should evolve
from practical experience. If we only have standards implementations and
not the possibility to break them and try new things, we won't be able
to evolve them in a practice proven way.
Cheers,
Sven
|
|
|
| Re: EEF Toolchain [message #620783 is a reply to message #134715] |
Tue, 07 April 2009 05:28 |
| Eclipse User |
|
|
|
I would not have been able to say it in a better way :) Both are important for the community.
Cédric
Sven Efftinge wrote:
> Miles Parker schrieb:
>> It looks like there will be two main M2T language threads for Modeling
>> M2T so we can't hope for one answer fits all -- which as we all know is
>> impossible anyway.
>>
>
> Yes, although Xpand and Acceleo are somewhat similar the important
> difference is, that
> - Acceleo is based on an OMG standard
> - where Xpand is not
>
> It's good and important to have both, because standards should evolve
> from practical experience. If we only have standards implementations and
> not the possibility to break them and try new things, we won't be able
> to evolve them in a practice proven way.
>
> Cheers,
> Sven
|
|
|
| Re: EEF Toolchain [message #620785 is a reply to message #134702] |
Tue, 07 April 2009 05:44 |
| Eclipse User |
|
|
|
> OK, I should have known better! :D Actually, semanitcally it looks a
> lot closer to XPand now then JET. ;) And it looks like the JSP-like
> syntax :P is gone. Extensiblity and modularity aside from syntax
> preference were the areas that have been frustrating for me w/ JET, so
> I'll look forward to taking a closer a look at the MTL stuff. Is there
> an aspect like construct? (maybe thats "before, after"..)
>
Yes, you can override a template and call the "original" one from your overriding template, so you can basically have
Before stuff
call the original template
After stuff.
And use conditionals in your template to manage your extension.
Cédric
|
|
|
Goto Forum:
Current Time: Wed Jul 16 04:46:09 EDT 2025
Powered by FUDForum. Page generated in 0.59099 seconds |
Home