Open Stack

Zane Bitter <zbitter at redhat.com> writes:

> To be clear, the thing that stopped us from automatically including it
> was that there was no consensus that it met the criteria, which
> exclude words that describe a general class of Geographic feature. I
> regret that you didn't get an opportunity to discuss this; I initially
> raised it in response to you and I both being pinged[1], but we
> probably should have tried to ping you again when discussions resumed
> during office hours the next day. FWIW I never thought that Pike
> should have been automatically included either, but nobody asked me at
> the time ;)

Thanks, I suppose it's better late than never to have this discussion.

Happily, the process does not require that the TC come to a consensus on
whether a name fits the criteria.  In establishing the process, this was
a deliberate decision to avoid the TC having exactly that kind of
discussion because we all have better things to be doing.  That is why
this is the sole purview of the election official.

We should remember that the purpose of this process is to collect as
many names as possible, weeding out only the obvious non-conforming
candidates, so that the whole community may decide on the name.

As I understand it, the sequence of events that led us here was:

A) Doug (as interim unofficial election official) removed the name for
   unspecified reasons. [1]

B) I objected to the removal.  This is in accordance with step 5 of the
   process:

     Once the list is finalized and publicized, a one-week period shall
     elapse before the start of the election so that any names removed
     from consideration because they did not meet the Release Name
     Criteria may be discussed. Names erroneously removed may be
     re-added during this period, and the Technical Committee may vote
     to add exceptional names (which do not meet the standard criteria).

C) Rico (the election official at the time) agreed with my reasoning
   that it was erroneously removed and re-added the name. [2]

D) The list was re-issued and the name was once again missing.  Four
   reasons were cited, three of which have no place being considered
   prior to voting, and the fourth is a claim that it does not meet the
   criteria.

Aside from no explanation being given for (A) (and assuming that the
explanation, if offered, would have been that the name does not meet the
criteria) the events A through C are fairly in accordance with the
documented process.

I believe the following:

* It was incorrect for the name to have been removed in the first place
  (but that's fine, it's an appeal-able decision and I have appealed
  it).

* It was correct for Rico to re-add the name.  There are several reasons
  for this:

  * Points 1 and 2 of the Release Name Criteria are not at issue.

  * The name refers to the human geography of the area around the summit
    (it is a name of a place you can find on the map), and so satisfies
    point 3.

  * I believe that point 4, which it has been recently asserted the name
    does not satisfy, was not intended to exclude names which describe
    features.  It was a point of clarification that should a feature
    have a descriptive term, it should not be included, for the sake of
    brevity.  Point 4 begins with the length limitation, and therefore
    should be considered as a discussion primarily of length.  It
    states:

      The name must be a single word with a maximum of 10 characters.
      Words that describe the feature should not be included, so "Foo
      City" or "Foo Peak" would both be eligible as "Foo".

    Note that the examples in the text are "Foo City" and "Foo Peak" for
    "Foo".  Obviously, that example would be for the "F" release where
    "City" and "Peak" would not be candidates.  Therefore, point 4 is
    effectively silent on whether words like "City" and "Peak" would be
    permitted for the "C" and "P" releases.

  * The name "Pike" was accepted as meeting the criteria.  It is short
    for "Massachusetts Turnpike".  It serves the same function as a
    descriptive name and serves and precedent.

  * I will absolutely agree that point 4 could provide more clarity on
    this and therefore a subjective evaluation must be made.  On this
    point, we should refer to step 4 of the Release Naming Process:

      In general, the official should strive to make objective
      determinations as to whether a name meets the Release Name
      Criteria, but if subjective evaluation is required, should be
      generous in interpreting the rules. It is not necessary to reduce
      the list of proposed names to a small number.

    This indicates again that Rico was correct to accept the name,
    because of the "generous interpretation" clause.  The ambiguity in
    point 4 combined with the precedent set by Pike is certainly
    sufficient reason to be "generous".

* While the election official is free to consult with whomever they
  wish, including the rest of the TC, there is no formal role for the TC
  in reducing the names before voting begins (in fact, the process
  clearly indicates that is an anti-goal).  So after Rico re-added the
  name, it was not necessary to further review or reverse the decision.

I appreciate that the TC proactively considered the name under the
"really cool" exception, even though I had not requested it (deeming it
to be unnecessary).  Thank you for that.

Given the above reasoning, I hope that I have made a compelling case
that the name meets the criteria (or at least, warrants "generous
interpretation") and would appreciate it if the name were added back to
the poll.

Thanks,

Jim

[1] http://eavesdrop.openstack.org/irclogs/%23openstack-tc/%23openstack-tc.2019-08-08.log.html#t2019-08-08T15:02:46
[2] http://lists.openstack.org/pipermail/openstack-discuss/2019-August/008334.html