Open Stack

----- Original Message -----
> From: "Matt Fischer" <matt at mattfischer.com>
> To: "Steve Gordon" <sgordon at redhat.com>
> 
> I was a first year track chair this year and I think that "your vote
> doesn't count" is not an accurate description from what I did. I can only
> speak for my thoughts and our track, but votes certainly were part of the
> process. Some talks had no votes or primarily negative votes, they were not
> considered much. But then you end up with lots of talks, well more than the
> 11 we could pick, with a good number of positive votes. We get to see
> counts and averages, is a talk with 65 votes and average of 2.5 better than
> a talk with 72 votes and an average of 2.4? You're splitting hairs there,
> so we can only use them as a rough guide for interest in the topic and
> speaker. Also, if we had simply picked the top 11 (by averages), you'd have
> ended up with an unbalanced track too many talks on the same topics for
> example or by the same people.

The assertion that "your vote doesn't count" came from emails from not one but two former track chairs in this thread, so it's certainly the case for at least some tracks. I myself am not actually questioning the process as it exists today, putting those comments aside at least, but rather whether the process is documented.

> Our goals were many, but included considering:
> 
>    - how were the votes? high votes? high score? etc
>    - does the talk fit into this track? is it too advanced/too broad/too
>    narrow?
>    - is it probably a sales pitch?
>    - are we covering the right things here? Does it fit into the goals of
>    this track.
>    - is the topic interesting to attendees? We try to think about what the
>    audience is for the track and go from there.
>    - is this a repeat from a previous year? Some talks are submitted with
>    very similar sounding titles (although sometimes updates on xxx talks are
>    good)
>    - does anyone know the speaker? are they active in the community? an
>    engaging speaker? a new fresh face that would bring a different
>    perspective?
>    - is this a duplicate talk? For example, out of the 11 talks we can
>    pick, we don't have space for 4 talks on Chef, so lets pick one thats good
>    and broad enough and fits this track.
>    - do any of these talks include any locals who would not normally get a
>    chance to talk or travel if this was in NA or Europe?

This list is similar to what I have heard from other former track chairs (not withstanding the comments from those that have responded in this thread) but this appears to be communal knowledge shared among a subset of the community. My question again is whether it is written down somewhere and if not should it be.

> These are how I considered/weighed the talks and the bottom line is that I
> assure there is no secret cabal ignoring everyone's wishes and jamming the
> schedule onto you. (If there is, I don't yet have my robes and secret book,
> please send). In fact I know of at least one talk that included many
> "luminaries" that we did not pick that I'm sure upset people. I had a talk
> in another track that I thought was a shoe-in that wasn't picked and I
> think many of us are in that boat. This process took about 6-8 hours of my
> time and we had a smaller track with about 90 talks to look through, many
> of the other chairs had way more work.
> 
> So I hope that sheds some light on how the process worked at least for my
> track.

The concern I raised was not that there is a secret cabal but that the process as it exists is not well known by those voting, nor in many cases those making submissions, because it isn't formally documented anywhere nor is there necessarily consistency across tracks (from three former chairs in this thread so far we have your well though out list, votes don't count at all, votes don't count at all and I only accepted submissions from Asia - that's quite a diverse range of ways of handling it). 

I *personally* understand the process but only via word of mouth and emails like this that I happen to catch. So not to belabor the point, is the process documented somewhere and is this something we should be highlighting to voters/submitters? Clearly I think the answer is yes, but I'm interested in arguments as to why it shouldn't be documented.

Thanks,

Steve

> On Sun, Aug 30, 2015 at 2:22 PM, Steve Gordon <sgordon at redhat.com> wrote:
> 
> > ----- Original Message -----
> > > From: "Richard Raseley" <richard at raseley.com>
> > > To: community at lists.openstack.org
> > >
> > > On 08/28/2015 10:24 AM, Stefano Maffulli wrote:
> > > > If we had the data we may be able prove this assumption by checking for
> > > > example if the higher amount of votes went to the proposals pushed by
> > > > corporations with an organized marketing machine.
> > >
> > > That would be an interesting use for the data.
> > >
> > > > That's what I've always done too. I ignore votes as a track chair.
> > > >
> > > > I think the voting process is a celebration of our community, a party,
> > a
> > > > ritual to get into the 'summit season; it's not a useful tool to
> > evaluate
> > > > proposals.
> > >
> > > I understand that the track chairs have wide discretion in the selection
> > > of sessions, which seems appropriate. That being said, I am a little
> > > surprised at the casual nature with which current and former track
> > > chairs have talked about how they outright 'ignore votes'.
> > >
> > > As a foundation member (I assume voting is restricted to foundation
> > > members), I was under the impression that my vote would always count at
> > > least a little bit (e.g. as a small part of some weighted score). If
> > > that is not the case I think it would be appropriate to set those
> > > expectations, as I am guessing that may others might be under the same
> > > misapprehension.
> > >
> > > Regards,
> > >
> > > Richard
> >
> > +1, while I've known this for a number of cycles I regularly encounter
> > members of the Foundation who don't. It's not like this is ever broadcast
> > anywhere except casually in email threads such as this one (usually with
> > the implication that the person asking the question should somehow have
> > known/expected their vote wouldn't count), have I missed it and there is in
> > fact a public page documenting the talk selection process?
> >
> > Thanks,
> >
> > Steve