Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Solr Wiki" for change 
notification.

The following page has been changed by HossMan:
http://wiki.apache.org/solr/LogoContest

------------------------------------------------------------------------------
  = Apache Solr: Logo Contest =
+ 
+ The Apache Solr Logo Contest has been concluded.  This page is being 
preserved both as a record of the contest, and as background information for 
other Apache projects that may want to hold similar contests
  
  [[TableOfContents]]
  
@@ -97, +99 @@

  
  http://people.apache.org/~ryan/solr-logo-options.html
  
+ == Results ==
+ 
+ The winning entry was "apache_solr_a_red.jpg" submitted by "Michiel" ...
+ 
+ 
https://issues.apache.org/jira/secure/attachment/12394264/apache_solr_a_red.jpg
+ 
+ The key email threads can be found here...
+ 
+  * 
[http://mail-archives.apache.org/mod_mbox/lucene-solr-user/200811.mbox/%[email protected]%3e
 Community vote] 
+  * 
[http://mail-archives.apache.org/mod_mbox/lucene-solr-dev/200812.mbox/%[email protected]%3e
 Committer vote]
+  * 
[http://mail-archives.apache.org/mod_mbox/lucene-solr-user/200812.mbox/%[email protected]%3e
 Results Announcement]
+ 
+ Various other email threads from this time frame may have additional 
information.
+ 
+ == Lessons Learned ==
+ 
+ This section is for anyone in the community to make notes on any opinions 
(good or bad) they have about the process used for the contest, or the outcome. 
 These notes may be helpful to other projects in the future.
+ 
+  * Something that wasn't anticipated was what to do when a submission turned 
out to be very similar to an existing logo.  The opinion of the PRC was that 
the logo which received the most points in the community vote was too similar 
to the logo for Solaris for Solr to use it, so it was ultimately disqualified, 
but this raised some questions about awarding prizes.
+  * In hindsight, an STV/IRV based voting system for the community preferences 
would have been a much better idea then the ranked point system we used.  
(STV/IRV was what was used ultimately used for the committer vote).  Using STV 
to let the community pick a top 10 list from all submissions, and then 
constraining the committers to those 10 using IRV probably owuld have made the 
most sense.
+  * the "variant" submissions factor wound up posing a somewhat thorny 
problem.  multiple subtle variations on the same theme did "split the vote" for 
those themes (as anticipated) but in the end the winning logo came from a 
family of entries that were very popular.  it's not entirely clear whether 
people voting for all of the variations of a logo they liked wound up 
artificially overshadowing other logos that didn't have variations since people 
were only allowed to pick 5. (ie: would some of the other logo themes have have 
scored better if they had multiple variations).  STV based vote counting (in 
which voters can rank *all* choices) would probably help here ... but future 
contests may want to consider having people vote on "families" of logos that 
are similar.  (in hindsight: the processed used was somewhat akin to letting 
people rank their type 5 choices for who should be president, but letting each 
candidate put their name on the ballot as many times as they want in dif
 ferent fonts)
+ 

Reply via email to