|
|
|
Re: Reading (parsing) EMF models without EMF runtime classes [message #414235 is a reply to message #414233] |
Fri, 26 October 2007 16:50 |
Ed Merks Messages: 33140 Registered: July 2009 |
Senior Member |
|
|
This is a multi-part message in MIME format.
--------------050909090500080807000509
Content-Type: text/plain; charset=ISO-8859-15; format=flowed
Content-Transfer-Encoding: 7bit
Bogdan,
I can imagine it might take time, but geesh, if all these companies with
ultra conservative lawyers can use EPL in their products surely anyone
can use it.
http://www.eclipse.org/membership/showMembersWithTag.php?Tag ID=strategic
After all, Eclipse has a very effective process for IP reviews and
pedigree tracking that mitigates the risk of using EPL code.
I certainly don't have non-EMF alternatives so I would suggest that it
will likely take you longer to solve the technical problems than it will
to solve your internal legal problems. Even "reverse engineering" EMF's
solutions into a different solution strikes me as more legally risky
that simply using EMF.
As I finish writing this note I notice your @sap.com address which makes
your comments even more ironic to me given that SAP is a strategic
member at Eclipse and given that I spent a whole day with folks from SAP
discussing their EMF needs:
http://ed-merks.blogspot.com/2007/10/eclipse-summit-europe-d ay-two-big.html
I was hoping the end result of this interaction would be SAP's direct
involvement as committers in the EMF project, so perhaps you should have
a chat with Axel Uhl and Simon Helsen.
Bogdan wrote:
> Ed,
>
> It is not about technical inability to use EMF runtime and it is not
> about EPL compliance for commercial products but it is rather right in
> the opposite direction - the commercial license does not easily allow
> for EPL. It is not that it completely forbids it but it could take me
> several months to get an approval for EPL usage within the product and
> I would like to have a solution sooner. I do not really want to write
> JETs for my own impl and I am checking if there is somebody that
> already did that. So to say the solution should consist of completely
> generated code and reusable runtime components. I know it is strange
> and I know that from architectural point of view it is better to have
> the EMF Runtime but that's the reality :).
>
> Best regards,
> Bogdan
>
--------------050909090500080807000509
Content-Type: text/html; charset=ISO-8859-15
Content-Transfer-Encoding: 8bit
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-15"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Bogdan,<br>
<br>
I can imagine it might take time, but geesh, if all these companies
with ultra conservative lawyers can use EPL in their products surely
anyone can use it.<br>
<blockquote><a
href=" http://www.eclipse.org/membership/showMembersWithTag.php?Tag ID=strategic"> http://www.eclipse.org/membership/showMembersWithTag.php?Tag ID=strategic</a><br>
</blockquote>
After all, Eclipse has a very effective process for IP reviews and
pedigree tracking that mitigates the risk of using EPL code.<br>
<br>
I certainly don't have non-EMF alternatives so I would suggest that it
will likely take you longer to solve the technical problems than it
will to solve your internal legal problems.
Ed Merks
Professional Support: https://www.macromodeling.com/
|
|
|
Re: Reading (parsing) EMF models without EMF runtime classes [message #414239 is a reply to message #414150] |
Fri, 26 October 2007 17:41 |
Miles Parker Messages: 1341 Registered: July 2009 |
Senior Member |
|
|
Hi Bogdan,
I don't know if this is exactly what you are looking for, but I think I
have a solution that works for wat you describe. I export XSD (not
XMI/XSD) and then as long as you supply the resulting xsd files and
regular editors will support basic editing and valdiation. (But not
OCL,Check, EDIT validation, etc.. just basic relational strucuture.)
hth..
On 2007-10-26 04:54:26 -0700, bogdan.vatkov@sap.com (Bogdan Vatkov) said:
> Hi Modelers,
>
> I have the following usecase:
> I am using Eclipse based plugin to create some EMF model (using self
> defined meta-model based on ecore) and I would like to be able to later
> read the model file (XMI) in an environment that does not allow for EMF
> Runtime.
>
> So I guess something like an XML parser, that would instantiate model
> elements(the Concrete Syntax Tree) that implement the EMF Model API but
> the implementation to be a different one - not the default one that is
> generated by EMF.
>
> Could you recommend me some solution?
> I would defintely prefer a parser generator based solution - since I
> would like to quickly adapt changes in the meta-model definition.
>
> Thanks in advance.
>
> Best regards,
> Bogdan
|
|
|
Re: Reading (parsing) EMF models without EMF runtime classes [message #414258 is a reply to message #414235] |
Sun, 28 October 2007 14:50 |
Bogdan Vatkov Messages: 16 Registered: July 2009 |
Junior Member |
|
|
Hi Ed,
Just to avoid any misunderstandings - we are intensively using EMF in our
Eclipse based solutions. I was talking about a solution that would need to
consume EMF models but is not based on Eclipse itself.
As I said the commercial license could allow for EPL, but there is a
general standard process for review and approval of any usages of any 3rd
party components (no matter of the license). The review checks tons of
assets like licensing, stability, support, extensibility, standards
compliance and so on. And I am not saying that there is a chance that EMF
will not meet the criteria, I believe it will, but still we have formal
approval process that might span several release cycles. What I am getting
now from you is that there is no such generated only version of EMF
Model API and it is not so big issue - I will simply wait for the review
process approval at my side.
I know there were some discussions between SAP colleagues and you wrt. EMF
on the ESE2007 but my question has nothing to do with those discussions.
I just wanted to run a prototype based on EMF models and wanted to avoid a
formal step. It is not a blocker, it would just speed up my development a
little bit. I will continue with standard EMF (no tweaks) and will run the
formal approval process in parallel.
Best regards,
Bogdan
|
|
|
Re: Reading (parsing) EMF models without EMF runtime classes [message #414259 is a reply to message #414258] |
Sun, 28 October 2007 15:34 |
Ed Merks Messages: 33140 Registered: July 2009 |
Senior Member |
|
|
Bogdan,
I think Martin Taal (the guy working on the Teneo EMFT project) has
experimented with something that generates POJOs. But of course that
solution doesn't provide for XML serialization or anything else you
might need...
Bogdan wrote:
> Hi Ed,
>
> Just to avoid any misunderstandings - we are intensively using EMF in
> our Eclipse based solutions. I was talking about a solution that would
> need to consume EMF models but is not based on Eclipse itself.
>
> As I said the commercial license could allow for EPL, but there is a
> general standard process for review and approval of any usages of any
> 3rd party components (no matter of the license). The review checks
> tons of assets like licensing, stability, support, extensibility,
> standards compliance and so on. And I am not saying that there is a
> chance that EMF will not meet the criteria, I believe it will, but
> still we have formal approval process that might span several release
> cycles. What I am getting now from you is that there is no such
> �generated only� version of EMF Model API and it is not so big issue -
> I will simply wait for the review process approval at my side.
>
> I know there were some discussions between SAP colleagues and you wrt.
> EMF on the ESE2007 but my question has nothing to do with those
> discussions. I just wanted to run a prototype based on EMF models and
> wanted to avoid a formal step. It is not a blocker, it would just
> speed up my development a little bit. I will continue with standard
> EMF (no tweaks) and will run the formal approval process in parallel.
>
> Best regards,
> Bogdan
>
>
Ed Merks
Professional Support: https://www.macromodeling.com/
|
|
|
Powered by
FUDForum. Page generated in 0.03652 seconds