WG4 Business Plan: Parts 1 and 4 Maintenance

[Rex Jaeschke]

I wrote, "I propose then that for the foreseeable future maintenance of
Parts 1 and 4 be limited to DR processing and that the next COR set for
these Parts not be produced for at least two years from the date of the COR2
set closure (2011-09-28)."

Murata-san replied, "The convener sees no rationales for this proposal,
especially because some important defects about fonts have been on the table
for more than 2 years."

What I am suggesting is simply that we continue the approach we've used for
two editions thus far: Produce a base standard, then produce a COR and/or
AMD, and then produce a consolidated base standard. In theory, this has been
a 2-year cycle, and would have given us editions of 29500 in 2008, 2010, and
2012. However, do to contradictions we found between COR1 and AMD1, we took
a lot longer to produce the 2nd edition, and what was going to be 29500:2010
was delayed until the following year. Nonetheless, the 2nd edition has 2
years-worth of DR resolutions in it as will the 3rd edition.

So, as we closed COR2 on 2011-09-28, I proposed that we close COR3 no sooner
than 2 years from then.

Now Murata-san wrote that some important DRs have been open for more than 2
years. As to whether or not they are "important" begs the question, "By
whose measure are they considered to be important?" and this leads me back
to marketplace relevance. I am not aware of any marketplace-relevant
implementer, consumer, or national body having informed WG4 that without
resolutions "real soon" that any commercial interests will be harmed.

As to a DR being open for 2 years, I say, "So what?" Elapsed time alone is
not a measure of whether something is important. A DR will only be closed
when enough resources are applied to it, and resource priorities are often
driven by  marketplace relevance.

There is also the issue of where in an implementer's cycle a COR or reprint
is published. Specifically, it may take some years before a fix from a COR
actually appears in an implementation. So while we might hurry to fix
something, that does not guarantee such a fix will appear anytime soon.

That said, suppose we find one or a few errors that we all agree need
fixing. I'm not at all opposed to our publishing a COR containing just those
few DR resolutions, and getting it out as soon as the ballot process allows.
Then the bulk of the DRs that get resolved will follow later in a
(presumably much larger) COR. In such a scenario, I would argue strongly
against publishing a consolidated reprint until after the 2nd COR has been
adopted. So, my recommendation can easily be adapted to allow one or more
"emergency" CORs to be published sooner than the 2 years I proposed. But to
work on such a priority project we'd have to define the criteria for an
"emergency" situation. The time it's been open is insufficient on its own.

The reason for maintaining the current 2-year cycle is predictability. The
outside world needs to know what our plans are and to be comfortable with
our edition-generation pattern. In fact, so do we as active participants, so
we can budget out time and expenses.

Rex

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.vse.cz/pipermail/sc34wg4/attachments/20120111/048e9521/attachment-0001.htm>