![]() |
Boost Docs : |
From: Matias Capeletto (matias.capeletto_at_[hidden])
Date: 2007-06-26 21:06:51
On 6/26/07, Matias Capeletto <matias.capeletto_at_[hidden]> wrote:
> On 6/26/07, Rene Rivera <grafikrobot_at_[hidden]> wrote:
> > Matias Capeletto wrote:
> If I have understand your proposal, you want this layout:
>
> * doc/src/ ( Sources needed to generate the doc )
>
> * doc/src/lib_name/lib_name.xml ( dockbook docs )
> * doc/src/lib_name/lib_name.xsl ( stylesheet )
> * doc/src/lib_name/Jamfile.v2 ( generate the docs inside
> "doc/html/lib_name" )
>
> * doc/src/lib_name/*.qbk ( quickbook source )
> * doc/src/lib_name/example/*.cpp ( used code samples )
Having other things (html, pdf, style, etc) as my last mail, I want to
know what are the pros of this layout for sources instead of the
actual one:
* libs/lib_name/doc/
* doc/lib_name/doc/lib_name.xml ( dockbook docs )
* doc/lib_name/doc/lib_name.xsl ( stylesheet )
* doc/lib_name/doc/Jamfile.v2 ( generate docs inside
"doc/html/lib_name" )
* doc/lib_name/doc/*.qbk ( quickbook source )
* doc/lib_name/example/doc/*.cpp ( or doc/example )
The thing I like about this layout is that everything a library author
maintainer has to modify lives in "libs/lib_name". ( Not counting the
files inside the boost headers directory, because that is a different
business )
I fail to see why the other scheme better separate the sources from
the output (html, pdf, etc ).
Can you comment on this?
Best regards
Matias