ietf
[Top] [All Lists]

Re: [IAOC] xml2rfc and legal services RFPs

2011-02-21 13:23:50
John,

Support for the xml2rfc tool was discussed on the IETF mail list when Marshall 
Rose indicated he could no longer support the tool.  Several people volunteered 
to maintain the tool, and they recommended the it be rewritten.  That 
recommendation plus the realization that this tool had become critical to IETF 
operations resulted in the IAOC deciding to issue a public RFP for a rewrite.

The Statement of Work in the xml2rfc RFP was reviewed and modified by the 
people on the tool-development list.  You can read the discussion at:

  http://www.ietf.org/mail-archive/web/tools-development/current/maillist.html

There was an active discussion that resulted in many changes from what was 
first proposed.  While this wasn't the whole community I think it a good 
representation of the people who are interested in the xml2rfc tool.  I will 
send you the list members off line.  Also, we will use a subset of this group 
to review the bids.

The legal services RFP was developed inside the IAOC/Trust.  To be honest, I 
didn't see very much value in doing a community review for this.  We will 
consider it in the future.  I do note that this work was previously done 
without a public RFP (dating back to when Wilmer-Hale was providing the 
service) so I think this is a better process than what was used earlier.

In the future, we will strive to give more notice to the community on planned 
RFPs.

Bob

p.s. I will forward your specific comments to tools-discuss.


On Feb 19, 2011, at 8:49 AM, John C Klensin wrote:

Hi.

This is not an attempt to derail either of these RFPs, nor is it
a formal appeal (request for review).  However, these two RFPs
raise an issue that may be worth some consideration.

The clear intent of the discussions leading up to RFC 4071/ BCP
101, and some of text in that document, was that the IASA was to
act with maximum transparency to the community and openness to
community comment.   It is especially important that substantive
decisions be open for community review and discussion before
they are made because, especially for those that are eventually
represented by contracts, there is no mechanism for review later.

IASA has recently issued two RFPs -- for legal services and for
a reprogrammed version of xml2rfc -- with no advance indication
to the community (at least that I can find) that they were
coming or opportunity for the community to review draft
provisions.  The clear expectation is that proposals will be
submitted (on a fairly short timeframe) and that the IAD and
IAOC will do whatever they do to evaluate the proposals and
establish contracts.   I don't know whether that is harmful in
these particular cases, but I don't believe it is how the
community had intended that things be done.

FWIW, community discussion might have improved at least the
XML2RFC one -- either the details of the RFP or community
confidence that it addresses/includes the right specifications.
For example, of the very large number of extensions or additions
that have bee requested over the years, the RFP selects two
(explicit line breaks in titles) and alternate anchors for
citations/references) but ignores the others.   I know that the
ability to easily index and cross-reference items in bulleted
lists, to easily generate numbers for ABNF productions (rules)
and build an index of productions, to have indexing reflect
section (and other subdivision) numbers rather than page numbers
(as various RFC style guides have required for years and that
becomes particularly important as people contemplate
non-paginated output formats), the ability to properly reference
books and journal articles without resorting to odd tricks
involving seriesInfo, and better handling of widows and orphans
(especially with regard to section title-text and lists with
undented headings top my personal list, but there are certainly
others.   

In addition, one of the two extensions that was specified
involves the addition of a new format-specific directive that is
exclusive to xml2rfc, not the DTD/Schema, thereby making
equivalent processing by other, XML-standard, processors
problematic and violating the fundamental principle that generic
markup does not specify formatting.   Perhaps community members
might have been able to propose design models that are more
friendly to XML principles and other tools (even I can think of
one or two).

The two extensions that were chosen may be the most important
ones, and there may be a reason why two extensions were chosen
rather than one or five.  But it seems to me that the community,
and perhaps even the text of the RFP and the proposals that will
arrive in response, would have benefited from some opportunity
for review and discussion.

Can we at least agree to more openness about draft RFP contents
in the future?  Or get an explanation from the IAOC as to why
the procedural model now seems to include issuance of RFPs
without any opportunity for community review of their
substantive provisions?

thanks,
  john





_______________________________________________
IAOC mailing list
IAOC(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/iaoc

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf