[cdr-discuss] About the CDR process

[Pascal Costanza]

Hi everybody,

On 12 Nov 2008, at 16:45, Tobias C. Rittweiler wrote:

>
> Hi,
>
> 1. quoting CDR 4, ``Common Lisp Document Repository (revised)'':
>
>      ``There will be an initial period of six weeks in which you can
>        send us updated versions, for example to correct typos, etc.,
>        which will replace any older version with the same CDR
>        number. You can negotiate with us a longer initial period, but
>        the maximum length is one year.''
>
>  I think a default initial period of six week is way too short. I'd
>  suggest a period of six _months_ instead.
>
>  Otherwise, there's the danger that CDRs will be finalized which  
> haven't
>  received any real feedback. Feedback, that is public scrutination, is
>  really crucial as I've found out myself while writing the
>  With-Readtable-Iterator proposal.
>
>  It is my impression that most people have the necessary spare time
>  around semester breaks, and christmas. Extending the initial period  
> to
>  six months ensures that it always crosses one of these peak times.

The original idea when we designed CDR was that documents are already  
publicly scrutinized elsewhere before they are submitted to CDR, so  
the 6 weeks were targeted at last-minute changes that typically arise  
in the last few weeks.

Note, though, that you can always ask for a longer period when  
submitting a CDR, and we are flexible to change these periods on the  
fly. CDR should be a service to the community, not a stumbling block  
that stifles good work just because we interpret rules by the word,  
and not by their spirit.

However, I have the impression that many people understand the period  
of six weeks indeed as the _main_ discussion period for a proposal. If  
we hear a lot of people wishing for extending the default initial  
period to 6 months (or so), we can certainly change the CDR rules to  
reflect the needs of the community.

Do other people have any opinions on this?

> 2. I also suggest the permission to update finalized CDRs under the
>   following constrain:
>
>     A section talking about current practises CDR may be updated to
>     reflect adoption of the documentation's content that has occured
>     since the publication of the CDR. The change must not alter the
>     meaning or the intent of the document.
>

This is a problematic suggestion because it requires somebody to make  
a judgment call whether or not a change alters the meaning or the  
intent of a document. It is likely that the CDR maintainers are not  
experts in each and every domain CDR documents will be about, so this  
would be quite hard to organize.

Instead, consider simply resubmitting a document under as a new CDR,  
and state in that document that it supersedes a previous CDR. We can  
also add to the previous CDR that a new CDR supersedes it. This is the  
preferred way for such changes, because it leaves the judgment whether  
a changed document accurately reflects community agreement completely  
to the community.

Or do you think this is not sufficient?


Pascal

-- 
Pascal Costanza, mailto:pc at p-cos.net, http://p-cos.net
Vrije Universiteit Brussel, Programming Technology Lab
Pleinlaan 2, B-1050 Brussel, Belgium