[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Second call for votes for the Lenny release GR

Hello Bdale,

Thanks a lot for this mail. I had been planning to write a mail asking for 
a standpoint on the vote, but I'm glad I waited long enough for this to 
arrive first.

On Monday 22 December 2008, Debian Project Secretary wrote:
> It is clear that there are flaws with the way the current ballot is
> constructed.  In particular, mixing choices that need supermajority with
> choices that don't seems likely to always lead to confusion.

That is not the only problem. The other issue is the mixing of options 
which only affect the release of Lenny with options that clearly have a 
long term impact. And the question whether or not (and if so how) the 
option to formally allow firmware to be part of Debian should lead to a 
change in our foundation documents.

> However, after thinking long and hard about this, I can find no real
> constitutional basis for terminating the current vote.

I agree that this is a real problem. Could we in theory have a GR to rule 
that the current vote is declared to be abandonded and its results void?
Given procedural timings that could of course only happen after the fact.

> I strongly encourage *every* Debian Developer to vote on this ballot. 
> The beauty of our Debian voting system is that ranking the options
> allows us to convey a great deal of information about how we feel,
> beyond just picking a winning choice.  By carefully placing "Further
> Discussion", we can indicate which options we think are acceptable and
> in what rank above it, and which options we think are unacceptable and
> in what rank below it.
> For example, if you think this ballot is a mess and don't want to
> figure out what all the options mean, then the right choice is to vote
> a ballot that ranks "Further Discussion" ahead of all other options. 
> If you choose *not* to vote, then you are in effect saying that *any*
> of the options presented would be ok with you.  A vote for Further
> Discussion tells the world that you think we should start over and try
> again with a better set of choices.  That is a completely ok result for
> the project.  It wouldn't "solve" anything, but it would do no harm.
> I also strongly encourage every voter to take the time to read,
> understand, and rank *all* of the presented options.  Even if you
> choose to rank some or all of them below Further Discussion, ranking
> the various choices helps to convey your thinking about which choices
> would be better or worse, regardless of whether you think any of the
> choices are what you really want.

Although I agree that concordet is normally great when it comes to ranking 
choices, I see a fundamental problem with this in this particular case, 
and that would be my main reason why I still feel this vote should be 
declared void.

Let's take this vote that was CCed to debian-vote today (accidentally or 
not) as example:
>[ 6 ] Choice 1: Reaffirm the Social Contract
>[ 5 ] Choice 2: Allow Lenny to release with proprietary firmware [3:1]
>[ 5 ] Choice 3: Allow Lenny to release with DFSG violations [3:1]
>[ 4 ] Choice 4: Empower the release team to decide about allowing DFSG
>      violations [3:1] 
>[ 2 ] Choice 5: Assume blobs comply with GPL unless proven otherwise
>[ 3 ] Choice 6: Exclude source requirements for firmware (defined) [3:1]
>[ 1 ] Choice 7: Further Discussion

Excellent: the vote is a mess; we all already knew that.

However, the voter's 2nd preferred "real" option requires a 3:1 super 
majority. But because it is ranked below FD in this ballot, this ballot 
will not actually count towards that! If there is a significant number of 
voters who do this, it effectively means that all of the options 
requiring 3:1 are effectively buggered as it means they stand a much 
bigger chance of being dropped _before_ they get ranked.

Filtering out FD votes for an (informal?) analysis of the vote seems no 
solution to me as 1) it would always be unofficial and 2) it would not 
respect those votes where FD was used to mean anything other than "this 
vote is a mess" (as in my own vote).

A possible "solution" for that could be to drop all super-majority 
requirements from this vote, but again, there is probably no 
constitutional basis that could allow that. And if that were to be done 
for some informal analysis it would only increase the invalitity of any 
such analysis.

Your thoughts on the above would be appreciated.


P.S. I personally would accept, and respect, and very much welcome a joint 
decision by the secretary and DPL to abandon the current vote and declare 
it void. Even if not explicitly covered by the constitution it would IMO 
be much more in the spirit of the same constitution and our project than 
to let the current farce continue.

Attachment: signature.asc
Description: This is a digitally signed message part.

Reply to: