ISO/IEC JTC1/SC22/WG5 N1940

                  Convener's report from SC22 meeting, 
                      Geneva, 10-11 September 2012

                      John Reid, 29 September 2012
                    

1. Introduction

I found this to be the most enjoyable and productive SC22 that I have 
attended so far. Six countries were represented (US, UK, Netherlands, 
Japan, Denmark, Canada) and conveners of seven WGs were present 
(Cobol, Fortran, Ada, C, Prolog, Vulnerabilites, Ruby).

2. Remote Participation

An experiment in remote participation was made at this meeting. All the remote
participants were in the US, which must have been hard for them, given the
time difference. Only one stayed for the whole first day (9-17 in Geneva,
21-5 in Hawaii). The system (provided by ISO) worked well. It uses a telephone
conference call for audio and a web browser for visuals. The screen of the
host PC is echoed to participants, so it is just as if they were looking at
a screen in a lecture room. And the host can be changed at any time.

I was not convinced that I want to participate in SC22 meetings remotely - the
people interactions at this meeting were very helpful. However, John Benito
uses it for his WGs and spoke very positively about it. He limits remote
participation to 3 hours per day, with those 3 hours chosen to best suit the
time zones and preferences of the participants. We should perhaps consider
doing it. For example, it might be useful for developing our new TS.

3. Eliminating the CD stage of standards processing

We are not the only people to object to the idea of eliminating the CD
stage of standards processing. It looks as if it will become optional.
SC22 felt that it should be able to set its own policy on optionality
and that it should be able to decide on a project-by-project basis,
probably at the time of New Work Item approval. There are cases where 
omitting the CD stage would be desirable, for example, for a revision 
that has no new technical content and was just incorporating corrigenda 
and making editorial improvements.

4. Three-year limit on corrigenda

JTC 1 proposes that corrigenda be allowed for only three years after 
the publication of a standard. SC22 objects to this, given that language 
standards have a longer life and defects may be found after more than 
three years.

5. eCommittee (LiveLink)

ISO has mandated that all subcommittees and working groups host themselves 
on eCommittee (LiveLink). This is a significant problem for working groups 
because open access will be lost. SC22 is very concerned about this and 
passed two resolutions on it - one explaining the problem and objecting 
- the other asking why this is being done.

6. Presentations from IEC and ISO 

At the end of the meeting, we heard presentations from the IEC (Gabriel 
Barta) and ISO (Trevor Vyze, Maho Takahashi and Hannah Ekberg). Gabriel 
showed a good understanding of what it is like to construct a language 
standard, but was not familiar with the detail of how ISO manages them.
It was clear from the ISO presentations that most standards are different 
from ours. This explains the proposal to eliminate the CD stage since it 
is not needed for the vast majority of standards. SC22 made the point 
that it is needed for most of ours.