|
|
|
|
|
Re: tooling workflow [message #381766 is a reply to message #381383] |
Thu, 15 November 2007 13:30 |
Paul Elder Messages: 849 Registered: July 2009 |
Senior Member |
|
|
Christian:
One thing I find with hot code replacement is that you have to be really
careful when modifying code that is on the stack. So, when I'm doing this
with JET templates, I try to make the modifications after the JET
transformation has completed.
Also, a third way. Copy the JET Transformation project into your runtime
workbench, and make sure it is NOT one of the plug-ins loaded in the runtime
workbench. (This is, it is NOT selected on the Plug-ins tab of the runtime
workbenches launch configuration. This way, when you run the JET
transformation, you can take advantage of JETs runtime loading capabilities.
Paul
"Christian Hall" <christian.hall@equifax.com> wrote in message
news:db47fc767a1f19a2f330d7e7909ad03f$1@www.eclipse.org...
>I actually tried both before I posted. I think I know why one didn't
>work...here are the details:
>
> - I installed the semi-stable model plug-in into the workbench, but I
> forgot to close the plug-in project...I suspect that is why that didn't
> work.
> - I tried running the runtime workbench in debug. I got one of two
> behaviors. Either I'd get an error that said runtime code modification was
> not allowed or I'd make the change and it just would not stick in the
> runtime workbench. Given this is supposed to work, I'll try to see if I
> can narrow down why things were not working.
>
> Thanks for the response.
>
|
|
|
Re: tooling workflow [message #601602 is a reply to message #381374] |
Wed, 07 November 2007 19:56 |
Ed Merks Messages: 33218 Registered: July 2009 |
Senior Member |
|
|
Christian,
It's best to ask on the M2T newsgroup, which I've added to the "to" list
of the reply. It doesn't sound quite right to me though...
Christian Hall wrote:
> Just to be sure I've got this right. Consider I have three projects:
>
> EMF model project (plugin)
> JET transformation project (plugin)
> target project
>
> Can't get the transformation to run without loading up the runtime
> workspace because of the emf dependency and because I want to have the
> artifact placed into the target project...or maybe just because of the
> EMF model.
> If I am working on getting the JET template the way I want it, I find
> I am having to change the template (e.g. fix a line of template code),
> restart runtime workbench, repeat. Each workbench startup takes about
> 30 seconds or so which is feeling like a pretty slow iteration.
> Is this just the way it is?
>
Ed Merks
Professional Support: https://www.macromodeling.com/
|
|
|
|
|
|
Re: tooling workflow [message #602375 is a reply to message #381383] |
Thu, 15 November 2007 13:30 |
Paul Elder Messages: 849 Registered: July 2009 |
Senior Member |
|
|
Christian:
One thing I find with hot code replacement is that you have to be really
careful when modifying code that is on the stack. So, when I'm doing this
with JET templates, I try to make the modifications after the JET
transformation has completed.
Also, a third way. Copy the JET Transformation project into your runtime
workbench, and make sure it is NOT one of the plug-ins loaded in the runtime
workbench. (This is, it is NOT selected on the Plug-ins tab of the runtime
workbenches launch configuration. This way, when you run the JET
transformation, you can take advantage of JETs runtime loading capabilities.
Paul
"Christian Hall" <christian.hall@equifax.com> wrote in message
news:db47fc767a1f19a2f330d7e7909ad03f$1@www.eclipse.org...
>I actually tried both before I posted. I think I know why one didn't
>work...here are the details:
>
> - I installed the semi-stable model plug-in into the workbench, but I
> forgot to close the plug-in project...I suspect that is why that didn't
> work.
> - I tried running the runtime workbench in debug. I got one of two
> behaviors. Either I'd get an error that said runtime code modification was
> not allowed or I'd make the change and it just would not stick in the
> runtime workbench. Given this is supposed to work, I'll try to see if I
> can narrow down why things were not working.
>
> Thanks for the response.
>
|
|
|
Powered by
FUDForum. Page generated in 0.04812 seconds