Jack Schwartz wrote: >> I am confused because I think >> that even media-specific processing is unique based on the end goals of >> projects that uses the DC. >> So, I don't see how DC can provide "generic" media-specific modifications. >> Thinking about all the special processing we did for the Slim CD in the >> Oct preview, I really can't separate >> which part is the distro-specific processing and which part is the >> media-specific processing. Can you >> give an example? >> >> > Distro-specific processing can include media processing as well as > anything else specific to the distro. The generic processing was to be > provided by us as a starting point, done in the spirit of "Why make all > projects reinvent the wheel?" In the simple cases, that would be all > that is needed. If projects required more processing, they can add what > they need. > > The finalizer can be set up to call any processing script, whether > generic or distro-specific, in any order. If tweeks need to be make > after the generic processing, call a distro-specific script after the > generic one. This wasn't made clear in the diagram; the accompanying > text in the design doc needs to explain this. > > Does this answer your question? > > Thanks, > Jack > I think I understand the intention now. We are trying to provide what we think are the common steps in generating a certain type of image. However, I think that the "generic processing" for the different image type should be provided as a set of libraries by the DC. The distro-specific processing can call one or more of these library utilities if they choose to. I don't think it is right to first call generic processing and then, distro-specific processing for any type of image, because only the people creating the image would know exactly what they want, the DC won't know.
--Karen
