Pablo-WMDE added a comment.

  I feel the conflict originates from discussing ways to achieve two (+) 
different things.
  
  Looking at what is written and what we can anticipate (subjectively then) to 
be its intention, this story primarily cares about communicating to the user 
and enabling them to report problems.
  
  Part of this communication is the (sticking to the wording from the story) 
"Error message" - as of now in master this is the `type` of the first of the 
`applicationErrors`.
  
  These types are arbitrarily created by us and they aggregate root causes - 
this inevitably leads to some loss of insights but offers the possibility to 
enumerate them (there are only so many).
  
  During the reporting of the bug we'll also put into the report the "Debug 
information" (full list of the `applicationErrors` and their "previous 
exceptions") - allowing us to gain EventLogging-like insights incl. the 
information which was "lost" when we aggregated the problems.
  
  From a dev perspective it is absolutely desirable to get hold of 
EventLogging-like insights about errors without relying on the user filing a 
ticket - but that can't be the job of this task in my eyes.
  
  The patch Lucas created 
<https://gerrit.wikimedia.org/r/c/mediawiki/extensions/Wikibase/+/585514/> is 
roughly in line with what I expected to come out of this when we talked about 
it in task break-down.
  
  The way we "decide which error condition currently collected in the generic 
error screen should get its own specific error screen", can only be informed by 
the concrete problems found in the "Debug information" of reported problems the 
way this stands, true.
  
  When/if we address that (or maybe continuously until then) I think it makes 
sense to further the distinction of logic and runtime errors as top level error 
types and consciously letting our specialized errors inherit from one or the 
other - this will help (on top of more logic) decide which errors are worth 
sending over the wire to some centralized logged infrastructure, which do not 
(a bit different compared to a "backend" project IMO).

TASK DETAIL
  https://phabricator.wikimedia.org/T249120

EMAIL PREFERENCES
  https://phabricator.wikimedia.org/settings/panel/emailpreferences/

To: Lucas_Werkmeister_WMDE, Pablo-WMDE
Cc: Michael, Lucas_Werkmeister_WMDE, Aklapper, Pablo-WMDE, Oblanco79, 
Alter-paule, Beast1978, Un1tY, Sarai-WMDE, Hook696, Daryl-TTMG, RomaAmorRoma, 
E.S.A-Sheild, darthmon_wmde, Kent7301, Meekrab2012, joker88john, CucyNoiD, 
Nandana, NebulousIris, Gaboe420, Versusxo, Majesticalreaper22, Giuliamocci, 
Adrian1985, Cpaulf30, Lahi, Gq86, Af420, Darkminds3113, Bsandipan, Lordiis, 
GoranSMilovanovic, Adik2382, Th3d3v1ls, Ramalepe, Liugev6, QZanden, 
LawExplorer, WSH1906, Lewizho99, Maathavan, _jensen, rosalieper, Scott_WUaS, 
Wikidata-bugs, aude, Lydia_Pintscher, Mbch331
_______________________________________________
Wikidata-bugs mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs

Reply via email to