I haven't figured it out either. My guess is that we will pass in a
structpath and a
nodepath as parameters to each build. We create those off the required project
declarations. Each generator will have to pull from all the structs in
each path.
Alternatively we could pass in full filespecs of all nodes and structs, but
that seems
less readable. Time for some experimentation.
I'll get the outline checked in.
cheers,
-ep
At 11:21 AM 2/26/2002 -0500, Steve Magoun wrote:
>Yah, that might be a good idea. Getting all of the structs into all of the
>right places isn't something I've figured out yet...
>
>Steve
>
>
>
>On Tuesday, February 26, 2002, at 10:58 AM, Eric Parker wrote:
>
>>My first thought was to just get the minimum struct/node declarations we
>>need to compile, but I'm wondering if it's better to start with something
>>that is more representative. My guess is that it might be good to do this
>>sooner rather than later, so that we have a relatively complete example.
>>
>>Here's the structure I'm thinking:
>>sand
>> apps
>> basics
>> structs:
>> - as specified, leaving LoopScriptStruct out of it for now
>> nodes:
>> - MessageDriver
>> - DataManager
>> security:
>> structs:
>> - AuthWrapper
>> nodes:
>> - Gateway
>> deploy
>> demo
>> - requires basics and security
>> structs:
>> - DemoUser extends BaseUser
>> - Widget
>> nodes:
>> - WidgetTransactor
>>
>>For this to work, we need to pass all the relevant structs and nodes into
>>each project build. I guess that's the main reason to put the full
>>structure in place now, rather than later.
>>
>>Thoughts?
>>
>>
>>_______________________________________________
>>Sandboss-devel mailing list
>>Sandboss-devel@...
>>https://lists.sourceforge.net/lists/listinfo/sandboss-devel
>
>
>_______________________________________________
>Sandboss-devel mailing list
>Sandboss-devel@...
>https://lists.sourceforge.net/lists/listinfo/sandboss-devel
|