[news.eclipse.technology.cosmos] Re: How and what to check out from CVS?

["Ali Mehregani" <amehrega@xxxxxxxxxx>]

Hi Michael,

The CVS is structured based on COSMOS 4 subprojects:
1) data-collection
2) data-reporting
3) management-enablement
4) resource-modeling

The plug-ins under the subprojects are what gets regularly build.
In addition to the 4 root folders listed above, there are a few other 
folders that are used for various reasons:

* doc - Used for documentations
* examples - Contains samples that illustrate the usage of the COSMOS 
framework
* releng - Contains release engineering resources
* sandbox - Place holder for experimental code that is not yet ready to be 
included as part of the build
* tests - Contains testing resources for plug-ins that belong to each 
subproject

What you need really depends on what you intend to do with COSMOS.  I 
recommend reviewing the consumer documentations available on COSMOS website:

http://www.eclipse.org/cosmos/home/documents/

Good luck,

Ali Mehregani

"Michael Moser" <mmo@xxxxxxxxxxxxxx> wrote in message 
news:fj3ij1$oen$1@xxxxxxxxxxxxxxxxxxxx
>I am a bit lost re. the cosmos CVS organization and how to convert (parts 
>of) it into eclipse projects again.
>
> When I look into the CVS then the content seems to be a load of *nested* 
> Eclipse projects (some actually doubly nested)!
> But - at least to my knowledge - Eclipse doesn't support "nested" 
> rojects  - or am I missing something here?
>
> So, if I check out the top-node (org.eclipse.cosmos) I end up with a 
> strange mess of things. There is ONE big project, that contains everything 
> but nothing is built in it, because it's .project contains no "buildSpecs" 
> and no "natures". Then there is a list of maybe 20 projects which are 
> partly put into my \JavaDevt-folder (where I usually save things), partly 
> in my .workspace folder (no idea, who/what controls what ends up where). 
> Most of these projects show build and/or compile errors because they can't 
> find libraries or references to other projects etc.
>
> The same holds true, if I check out the the next "layer" of nodes (i.e. 
> only data-collection or data-reporting individually).
>
> Is one really supposed to traverse the CVS and check out all the 
> leaf-projects individually? There are, I would roughly estimate, about 50 
> of them, so this would take quite some effort and at least an hour just to 
> check out the files from CVS. Tell me, that this is NOT what is expected, 
> please!
>
> So, what is the correct approach to get all or a subset (like 
> "data-collection") out of the CVS and building in optimal time?
>
> Michael
>
>
>
>