nth
nth
Thanks @mistadikay! I've been a bit busy but will look into testing and merging this soon. Have you noticed any issues so far since you opened this PR?
Could you try this advice and see if it's faster? ```elisp (advice-add 'org-roam-db-update-file :around (defun +org-roam-db-update-file (fn &rest args) (emacsql-with-transaction (org-roam-db) (apply fn args)))) ``` This should batch all SQLite...
From my profiling, when there are a lot of timestamps then `org-entry-properties` (used by `org-roam-db-node-p`) is slow. I tried `org-entry-get` which should do what we need to check for "ROAM_EXCLUDE"...
I haven't tested that commit yet. To clarify, you're saying it makes saving even slower?
The workaround I'm doing now is enabling `magit-delta` unless there are large diffs in the buffer: ```lisp (defvar nth/magit-delta-point-max 50000) ;; Disable mode if there are too many characters (advice-add...
Thanks @dandavison for looking into this. My diff is ` 11 files changed, 1850 insertions(+), 24757 deletions(-)` and using `time-magit-status` I got: ``` magit-delta off magit-status took 0.5 seconds magit-delta...
That's strange. What did you change in your development copy?
What you proposed makes sense. If `AttributeHook` has a `type` then `SoftSetHook` should as well. Please go ahead with the PR for `virtual-dom`. Once that's merged we'll make changes to...
Nice work :+1:. I also like Option 2 better so hope that'll get merged.