Date: Mon, 12 Mar 2018 11:37:52 +0000
From: Andrew Josey <ajo...@opengroup.org>
Message-ID: <f973b802-10e5-4ab8-ab69-b7aac30a3...@opengroup.org>
| We usually say if we have one to continue next time.
Yes, I know, and I have observed that - and thought (in my
somewhat clumsy way) to have mentioned that - that case is
not an issue.
| Otherwise the simple rule is to assume is next in sequence.
OK, I could see that things generally proceed approximately
sequentially, but there have to be exceptions ... like the minutes
often say "we will return to this in a future call" (and some issues
look to have been in that state for a very long time) - it would be
nice to have some advance warning when that future call has
suddenly arrived (and sequencing will not help there.)
eg: there's bug 0000249 ($'...' in the shell). and 0000953 (aliases,
for which, IMO, the resolution should just be to delete them from
the stanadard, they're ugly, but never mind) which I assume will
one day be discussed again. But they have been "outstanding
issues" so consistently, for so long now, that it seems like they'fve
just fossilised, and spending any time on them would be wasting it
(for 249 I don't even understand what the real issue is - that is not
the original report, but the "reopened")
There's also the issue of ...
| That being said, of you have a specific bug you would want to discuss then
| we are usually willing to bring that to the top of the stack if you dial in
which Nick also mentioned (incidentally, thanks to both of you for
the replies.)
First, that is certainly not what I was hinting at - there is no particular
bug on the list that I want to see resolved before any others (there are
some I'd like to see resolved, and others I have no particular interest
in one way or the other).
What raised this for me was issue 1075, where, had I known it
was to be discussed (or for that matter, even that the issue existed)
I would have commented before it was decided that the proposed
new text was suitable, as (as you will have seen) I do not believe it
is. What's more, that is not the first time I believe this kind of thing
has happened - and it just seems wasteful to me, when I could have
commented ahead of it being dealt with, instead of after...
However, that offer, which I assume applies to everyone, not just me,
also allows for issues to be addressed out of order, with no notice
to those not on the call, which is exactly what I was hoping could be
avoided - no problem with elevating an issue that is important to
someone so it can be discussed on a call when they are going to
be able to participate, just let the list know, in advance, that it is
happening.
| We could add a part to the end agenda each week to minute what
| we expect to address next time
That I think would help. It would (at least for me) act as something
of a reading list for the week.... Then if I have anything to offer it could
be done in advance.
Also if you will all pardon some outrageous suppositions I am about
to make (well, one anyway) - it seems to me, and again this is a guess
since I have never been on one of your calls, as if you often spend
quite a bit of phone time doing wordsmithing on actual text that is to
be changed in the standard.
If that is what happens, then, frankly, it is a colossal waste of
everyone's time (those who participate) - I have been involved with
both face to face, and tele- meetings that have attempted to do
this many times, and it never proceeds as quickly as it should.
Meetings are good at two things - they allow people to react
more quickly to views that are inconsistent with their own thoughts
(or express agreement when they are) and so allow the contentious
issues to be revealed, and perhaps the germ of a suitable resolution
to appear (and non contentious issues to be quickly handled), and
second, they provide the forum to allow someone who disagrees
with an issue to quickly observe that the overall opinion is not in
favour of their opinion, and that either they need to quickly provide
something very convincing to sway opinion, or accept that their
view has simply failed, and to accept defeat. This latter is much
harder over e-mail (and similar) as there it is very tempting for
everyone to believe that the great "silent majority" is supporting thir
obviously correct viewpoint.
Beyond those, meetings suck as a way of making progress, and
they are particularly poor as a way of generating text (everyone
tends to have opinions on things that don't really matter).
This supposition comes from the rate of progress of issues on
the phone calls - I'd typically expect that you should spend (given
the relatively small number of participants) no more than 5, perhaps
10 when there is really something to discuss, minutes on each
issue.
On "first reading" (unless it is a very good bug report perhaps,
which most are not, or unless it is very simple "spelling mistake on
line .....") all that should be done with any issue is to decide whether
it is worth working on at all, and then assign someone to draft some
proposed text (this could be 10 mins or so of discussion).
Next time (which is a call after the text is available, not "next week"
necessarily) then either "that's good. done", or "we need to change the
part about ..." and it gets deferred to a later next time, which could happen
more that once. Each such issue should take 5 minutes tops during
each call. Initial drafting should take place in the privacy of the tasked
individual's workstation, and once that person is happy with the result,
be added as a note to the issue - where others can suggest improvements.
By the time of the "next time" call, often, everyone should be in agreement.
Acting this way each phone call (90 mins right?) should be touching
on more than 10 bug reports, not the 2 or 3 that seems like is currently
the case (and sometimes less than one). Those 10 or more are not
all going to be resolved, but 4-5 of them should be (every time.)
kre
ps: on the subject of issue order, I did notice that issue 1188 got
dealt with, very quickly, and not at all in sequential order .... :-)
