« first day (3054 days earlier)      last day (126 days later) » 

00:05
REFRESH!
[Minesweeper] 96 Games Played. 59 Bombs Used. 11735 Moves Performed. 13 New Users
[rubberduck-vba/Rubberduck] 1 issue comments
 
5 hours later…
04:54
haha perf is ridiculous now
still missing the parsing.xml, investigating...
for real though... this dataflow stuff is bloody awesome, but the most significant gain was made by introducing a staging schema and similar tables as in the v2 schema, except I'd have GUIDs as foreign key IDs, instead of the v2 integers - all the FKs are worked out pre-insertion, and dumped from a DataTable with a SqlBulkInsert.
still, I wasn't expecting to be able to get this processing down to sub-10 seconds, let alone in 5.
..and that's local :D
 
5 hours later…
10:02
I actually thought about orchestrating the parsing and resolving process via data flow several times, but never got around to it.
10:59
Anyway, if we want to support incremental updates in RD 3, we might rather want to use a reactive approach in order to simplify invalidation and keep updates to a minimum.
 
7 hours later…
17:56
> Modifications:
1. Fixes a latent Let/Set code block creation bug: creating a code block from parameterized Get prototypes fail.
2. Adds `DeclarationType.PropertyLet` and `DeclarationType.PropertySet` to the list of valid prototypes for generating code.

Item 1: Discovered while working with `CodeBuilder` to resolve #5886.
Item 2: The resolution to #5886 _should_ look a lot like `IntroduceGetAccessorQuickFix`. However, `IntroduceGetAccessorQuickFix` is not using the `CodeBuilder`. Th
 
5 hours later…
23:50
[rubberduck-vba/Rubberduck] 1658 stars vs. [decalage2/oletools] 2232 stars

« first day (3054 days earlier)      last day (126 days later) »