[MGNLPN-241] Preview does not apply traits correctly Created: 30/Nov/15  Updated: 01/Dec/15  Resolved: 01/Dec/15

Status: Closed
Project: Magnolia Personalization
Component/s: Preview App
Affects Version/s: 1.2.2
Fix Version/s: None

Type: Bug Priority: Critical
Reporter: Antti Hietala Assignee: Unassigned
Resolution: Not an issue Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: PNG File sportstation-variants.png    
Template:
Acceptance criteria:
Empty
Task DoD:
[ ]* Doc/release notes changes? Comment present?
[ ]* Downstream builds green?
[ ]* Solution information and context easily available?
[ ]* Tests
[ ]* FixVersion filled and not yet released
[ ]  Architecture Decision Record (ADR)
Bug DoR:
[ ]* Steps to reproduce, expected, and actual results filled
[ ]* Affected version filled

 Description   

The Preview-as-visitor functionality does not apply traits correctly. Magnolia does not serve the variants that it should.

To reproduce on demo:

  1. Create two variants for /sportstation. The order of variants is important.
    • London: Country=United Kingdom AND Cookie WeatherLocality=London
    • UK: Country=United Kingdom
  2. Preview the page as visitor. Set Country=United Kingdom.
    • Expected: London variant is served. Both variants have equal weight (1 matching trait) so we serve whichever variant comes first (London).
    • Actual: UK variant is served although it is second in the variant order.
  3. Move the UK variant to first place, above London.
  4. Preview as visitor. Set Country=United Kingdom and WeatherLocality=London
    • Expected: London variant is served. London now has more weight (2 matching traits) so we serve it first regardless of variant order.
    • Actual: UK variant is served.


 Comments   
Comment by Antti Hietala [ 01/Dec/15 ]

Not an issue. We don't use weighted votes. Per pmundt:

There is no so such thing as weighed voters. I thought we had it because we use the default AndVoting. But the name is misleading: The javadoc says Return the maximum vote but only if all voters voted positive.. And that's what happens during eval. So add-up-votes are currently not supported by the system. The only thing that can be done, is adding a level to a voterset, i.e. weighing a rule set higher than a segment (not in the UI though atm).

Given this, the system is working correctly. The first matching variant wins. Order is significant.

Generated at Mon Feb 12 06:35:42 CET 2024 using Jira 9.4.2#940002-sha1:46d1a51de284217efdcb32434eab47a99af2938b.