I agree it makes sense to extract something like Pier-Pillar, the way you describe this.
I do not agree to make this a separate project, as this is only one package. But that is
On Oct 9, 2013, at 10:06 AM, Damien Cassou wrote:
On Tue, Oct 8, 2013 at 3:29 PM, Diego Lont
Pier-Model depends on Magritte Core.
only a subset of Pier-Model depends on Magritte-Core. Look at what I
put on the Pillar project:
Gofer new smalltalkhubUser: 'Pier' project: 'Pillar'; configurationOf:
(ConfigurationOfPillar project version: #bleedingEdge) load
(note how I put Pillar in the Pier team because I think this is were
it belongs right now. I'm not trying to steal anything :-)).
Pillar contains the document model of Pier (PRDocumentItem subclasses,
the parser, the abstract visitor, and the text and wiki writers).
These parts do not depend on anything (neither Seaside, nor Magritte).
Pillar has 2 additional packages to export static HTML (what is
currently done in Pier-Book) and LaTeX. The HTML exporter depends on
Seaside-Core (but just for the brushes which I plan to rewrite for
Pillar because I don't need most of the things in Seaside-Core. Pillar
can then be used to write books and class documentation just like
Markdown and other wiki syntaxes.
Pillar does not contain any CMS-related feature. These features are
kept in Pier: the structures (PRPage, PRStructure,...) the context
handling, most of the visitors that update the document model and
query it... Pillar does not contain any view (so, no Pier-Seaside or
Pier-Morphic that are parts of the CMS).
- Pillar is a syntax, its parser, a document model, and exporters
- Pier is a CMS that uses the Pillar syntax and document model to
describe its content, but that could use any other wiki syntax
provided there is a parser which creates a pillar document model for
And Pier-Seaisde depends on Magritte Seaside. And
other display-layers then Pier-Seaside are possible as well and when they are stable
enough we should add them to Pier, and not the other way around. At least, that is what I
think. But I have the feeling this is not about the display of Pier, since you name
No, I'm not talking about the display of Pier, you are right. Pier
keeps its views and I'm not interested in them because I want to write
static books (in HTML, LaTeX, markdown, ePub...).
But maybe you want to isolate the part that Pier
uses to parse its CMS?
Currently Pier-Model has a PRDocument parser
integrated into Pier-Model-Document itself. It might make sense to extra this part, so you
can exchange this for Markdown and also the other way around.
we agree :-)
"Success is the ability to go from one failure to another without
Magritte, Pier and Related Tools ...