On Tue, Feb 5, 2013 at 1:31 AM, Samer Mansour <samer...@gmail.com> wrote:
> Picking a logo, hmmm.
>
> First part is dates.
> We could find our dates by working backwards:
> 1. When is the estimated release date of AOO4?
> 2. When do we need to have a logo picked (consider time to integrate).
> 3. When do we need to start a final vote.
> 4. If/When do we need a preliminary/elimination vote.
> 5. When is proposal deadline.
> Consider how much time for each of the above steps, advertise the schedule
> so that everyone voting knows when voting will occur and won't miss it.
>
> Second part is voting:
> How do we pick a logo by voting:
> - We could do rounds of elimination voting.  Like American Idol, the
> weakest X  number of logo gets eliminated.  You can then vote again on the
> next round for a logo that survived if your first logo pick did not.
> - Who gets to vote? Help me on this one, I'm newb, is it all the mailing
> lists? Dev list, etc. one person a vote.
>

Previous discussions got bogged down on this point.  Some said that we
should not base it on popularity, but on objective standards and an
expert evaluation of the logos.  Others said we should vote.  IMHO, we
want to aim for two things:  a logo that is objectively good, but also
one that is accepted by the community.  So personally I think some
form of voting is necessary, though maybe with an opportunity for
experts to weigh in and make their arguments.

Also, there are ranked voting schemes we can use that would avoid the
need for multiple rounds.  For example, single transferable voting:
http://en.wikipedia.org/wiki/Single_transferable_vote

> Third part is criteria:
> - What are the minimum criteria for the logo.
> - What are the constraints of the logo, if any, ie. technical.
>

We started to put together a formal call for proposals last year:
https://cwiki.apache.org/confluence/display/OOOUSERS/Brand+Contest+RFP
but it lost steam.

One approach that I suggested as to keep it in two phases:

Phase I: Only a basic logo needs to be provided, bitmap or vector
format.  We then vote to determine the top 3 choices.

Phase II: The designers of the top 3 choices then refine their ideas
and provide a more complete logo package, which might require vector
source, splash screen, web logo, profile picture for Twitter, etc.  We
define a package of required variations that must be included.   We
then vote among the the three submissions.

With this approach we make it easier for anyone to enter.  They don't
need to commit to a lot of design variations up front.  They only do
the extra stuff if they are in the top 3.

Another point was what the winner(s) might get.  We don't have the
ability to offer cash prizes, unless a sponsor steps forward to offer
one.  But we could reward the winner(s) with some recognition, in a
press release, blog interview, podcast, or something similar.

Regards,

-Rob

> From the sounds of it, we were not proposing finished work at this point in
> time.  Still a lot of bouncing ideas,  I do feel like our designs are
> beginning to conform to a higher than all of us design from our feedbacks.
> Maybe voting will be easy, who knows.  I'm optimistic here.
>
> Feedback is welcome, specially in areas like voting.  We can start
> populating this, such as dates.
>
> Samer
>
> On Feb 4, 2013 6:11 AM, "Jürgen Schmidt" <jogischm...@gmail.com> wrote:
>>
>> On 2/1/13 6:03 PM, Kadal Amutham wrote:
>> > I have edited the wiki , giving numbers to all the logos. So all of us
> can
>> > have a look and vote as follows giving the ranking. Kinldy use alphabets
>> > for ranking meaning A as the best, B next one and Z the last.
>>
>> before we start voting on the logos I would suggest that we define some
>> basic rules and a timeline (when to start, how long the vote is open
>> etc.). We have some very interesting and good proposals but in the end
>> only one can make it. We should also consider a complete brand, means a
>> proposal that can be applied everywhere (logo, file icons and app icon,
>> webpage, intro, about, ...)
>>
>> Just my opinion
>>
>> Juergen

Reply via email to