Ticket #302 (closed todo: done)

Opened 2 years ago

Last modified 16 months ago

Placement Scopes first draft

Reported by: ichthyo Owned by: ichthyo
Priority: grave Milestone: 0integration
Component: lumieraProc Keywords: design data session
Cc: Blocked By: #309, #430, #455, #662
Blocking: #108, #303, #323, #328, #356, #440

Description

Mostly I'm through with this design work; but it needs to be re-visited and finished for now. Details can be ommited regarding how to link a meta-clip into a Sequence.

Change History

comment:1 Changed 2 years ago by ichthyo

  • Blocking 303 added

comment:2 Changed 2 years ago by ichthyo

  • Blocking 108 added

comment:3 Changed 2 years ago by ichthyo

  • Blocked By 309 added

comment:4 Changed 2 years ago by ichthyo

  • Blocked By 311 added

comment:5 Changed 2 years ago by ichthyo

  • Blocking 323 added

comment:6 Changed 2 years ago by ichthyo

  • Blocking 328 added

comment:7 Changed 2 years ago by ichthyo

  • Status changed from new to accepted

now working on the Placement-Scope topic

comment:8 Changed 2 years ago by ichthyo

  • Blocking 356 added

comment:9 Changed 2 years ago by ichthyo

  • Blocked By 430 added

comment:10 Changed 2 years ago by ichthyo

  • Blocking 440 added

comment:11 Changed 2 years ago by ichthyo

  • Blocked By 455 added

(In #455) oops... consequences can be fatal, thats not just a minor issue. While normally the LUID-based Hash-ID doesn't allow to create zero LUIDs, there is a rather straight forward way to sneak in such a value from within PlacementRef -- which is exactly what we need. Is this a hack? It looks ok-ish to me, as it happens right from within implementation level and right there where it's needed (the default ctor of PlacementRef).

Thus I'm headed towards using a 0-LUID to mark invalid -- but this needs to be complemented by a dedicated check on each index insert

comment:12 Changed 17 months ago by ichthyo

  • Blocked By 662 added

comment:13 Changed 16 months ago by ichthyo

  • Blocked By 311 removed

The planning of timeline-sequence binding (#311) isn't finished yet, but I think this won't cause any significant changes to the design of the placement scopes. I consider the latter stable now, its even mostly implemented and passes the basic tests.

comment:14 Changed 16 months ago by ichthyo

  • Status changed from accepted to closed
  • Resolution set to done
Note: See TracTickets for help on using tickets.