From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Fri Jun 01 2007 - 12:55:34 CDT
Mark Davis wrote:
> On 5/30/07, Philippe Verdy <verdy_p@wanadoo.fr> wrote:
>> There are disputed entries that were part of CLDR 1.4 by error :
>> it was not even possible to avoid them to be released, because
>> the proposal period for CLDR 1.4 was extremely short (about one month),
>> it was not announced clearly in advanced (a single message posted
>> to the Unicode list, that was not delivered to every one), and then
>> inaccessible for a part of that period (lots of technical problems
>> of performance on the server, even during that period).
>> The result was that the CLDR 1.4 contained many entries fro which it was
later impossible to vote AGAINST their addition.
>
> This is not the case. Previous values can be overridden;
> see the process documentation.
Actually I know that I can *propose* an override and vote for it, but with
the current automated minimum ratios needed to perform a change, it will be
difficult to change something with just two votes in favour of the change,
even if the previous data was an error, and this is explained in a bug
report (or in the online CLDR forum, now viewable also in the RSS feed).
So I fear that a lot of English terms that were standardized in CLDR 1.4
(for which I vote against before the CLDR 1.4 release, because I could not
even make an alternative proposal, as the submission phase was so much
unusable before it was closed), will still persist in CLDR 1.5, because
there won't be enough votes to overturn the current value. Many entries in
CLDR 1.4 did not even reach any minimum consensus: all we could do was to
vote against them, but it was too late and a single vote from the initial
proposer was accepted and put into the release because of absence of an
alternative
This was a major problem last year, and this past release received very
negative opinions, or it has been used now in several projects as if they
were effectively right. We can still see today, people trying to change
correct data by replacing it with the incorrect data because it was part of
the CLDR 1.4 release.
I had lot of works to do (and needed lot of patience, with submission times
in the tool that took sometimes 12 minutes to complete for a single change
since the end of April... any change was overloading the server for several
minutes, so I had to wait and perform most of the work in the middle of the
night when the server was finally accessible) trying to catch all those
errors and completing many new entries for French, and proposing several
alternatives that vetters could also support (this means that some of my
submissions were acceptable for me, but I voted for another one that I still
prefer).
There remains proposals that could not be deleted (they keep one vote on
them, because apparently my own account used a different numeric vetter id
that has changed over time, or because of possible past SQL bugs in the CLDR
tool). The addition of a DELETE option (for entries we were alone to
propose) came very late (but it was really needed to avoid creating
confusion among vetters, as nobody would really want to vote for these
bogous inputs.
(Note that blank votes are still not counted as votes against all existing
proposals, it is considered like "no opinion, every other vote is
acceptable, vetters should still have the possibility to cast their vote
explicitly for a "vote against all these proposals" option, counted like
other explicit proposals: if this option gets the majority, then no proposal
will be accepted in the release before a final consensus is asked directly
to vetters within the forum; for now, all we can do is to discuss the
problematic item in the forum/RSS feed).
This archive was generated by hypermail 2.1.5 : Fri Jun 01 2007 - 12:58:19 CDT