[slf4j-dev] W3C process and parallels to slf4j

[Curt Arnold]

Since SAX was brought up, I thought I might pass along some links to a 
more formal process that might be applicable to the type of situation 
that confronts SLF4J.

http://www.w3.org/2004/02/Process-20040205/activities.html#Activities	
http://www.w3.org/2004/02/Process-20040205/tr.html#Reports

Here is a crude overview:

A W3C Activity is roughly equivalent to something like the ASF Logging 
Services project.

An interest Group (IG) within the activity is community of members with 
an interest in some area, kind of like log4j-users at logging.apache.org 
representing a community of users with an interest in logging services 
for Java.

If the interest group decides that something needs work, it charters a 
working group.  This would be like the log4j-user's group deciding that 
there needs to be a log4j 1.3 with certain features and it charters the 
log4j 1.3 developer group.

The Working Group (WG) creates a Requirements Document, Working Drafts, 
Proposed Recommendations, et al.  The requirements for transition 
between the levels are described in the #Reports link.  One of the 
crucial points is that recommendations require multiple independent 
interoperable implementations before progressing to recommendation to 
prove that the spec is implementable and sufficiently complete.  
Working Group may provide software for conformance testing.  
Implementations are usually done by teams within the member companies.

The W3C doesn't create standards, its most binding document is a 
recommendation which is basically making a statement like "If you want 
to display hyperlinked documents, we recommend using HTML and here is 
its definition but you are free to ignore our recommendation and use 
anything else you'd prefer"

Members can submit Notes of existing implementations that could be used 
as starting points.  For example, Adobe submitted a PGML (portable 
graphics markup language) and Microsoft submitted Vector Markup 
Language as Technical Notes to the W3C Graphics Activity and the SVG 
Working Group used them as references to create SVG Requirements and 
Working Drafts.  The XML Schema WG collected 5 or so previous efforts 
before starting the XML Schema work.  Basically, the idea is that W3C 
development (as opposed to research) should be reserved for 
technologies that are sufficiently important that more than one company 
has worked on the problem independently and mature enough that they 
have been able to get something to work.

SLF4J.org seems to be a combination activity (Logging), Interest Group 
(Unified API) and Working Group (SLF4J 1.0 WG).  In a W3C world, these 
would each have their own charter:  See http://www.w3.org/XML/Activity 
for an example of an Activity Statement and discussion of the XML 
Working Groups.

JCL and UGLI would be similar to Member Notes containing Jakarta 
Commons and Log4j's efforts on the same problem domain (similar to 
Adobe's PGML and Microsoft's VML) that could serve as references for 
the SLF4J WG.  The SLF4J use-case/requirements documents might contain 
a review off how JCL or UGLI does or does not fulfill the requirements 
in the document.  The requirements docs would likely go through several 
iterations before being accepted by the WG or IG.  The WG would then 
start working to create something that fulfilled the agreed 
requirements and produce Working Drafts at a periodic interval for 
review and finally work through the requirements to reach a 
Recommendation status.