add etherpads to wiki pages

1 files changed, 52 insertions, 0 deletions

diff --git a/2022/talks/rolodex.md b/2022/talks/rolodex.md
index ffd84627..d47ca3d2 100644
--- a/2022/talks/rolodex.md
+++ b/2022/talks/rolodex.md
@@ -182,6 +182,58 @@ but can also trigger Emacs to execute code as well.
   3. The Hyperbole markdown creates links that execute queries
 
   4. This results in a minimal but useful Zettelkasten.
+# Discussion
+
+## Notes
+
+-   My blog: <https://tilde.town/~ramin_hal9001>
+
+## Questions and answers
+
+-   Q: Why is the time-stamp not implemented as an Org mode PROPERTIES
+    entry? (e,g, :CREATED:)
+    -   A: Hyperbole pre-dates Org-Mode, although the maintainers have
+        made efforts to make Hyperbole compatible with Org-Mode as much
+        as possible. You could ask Bob Weiner directly, but it could
+        just be for backward compatibility, trying to keep the
+        formatting for current Hyperbole users. You could raise that as
+        an issue, they may be willing to include a config option
+        allowing you to specify the time-stamp format.
+-   Q: why Hyperbole/HyRolo over Org-Roam? (I don\'t use either, just
+    curious)
+    -   A: HyRolo and Hyperbole require no other software beyond code
+        builtin to Emacs.  For example, when I first built Org-Roam, it
+        did not work properly for me and I had to modify the build
+        process to get it set up.  With Hyperbole, you install one
+        package and you can start working.
+    -   A: \^this, and I find it to be a lighter-weight solution. I was
+        able to get it working without depending on SQLite or doing any
+        indexing. Hyperbole is also a more general solution that can be
+        applied to a wider range of use cases than just Zettelkasten.
+-   Q: How does this scale to very large data bases?
+    -   A: It works very well with fairly large personal databases.  No
+        one has ever complained about performance.  Generally, people
+        are surprised how fast it is given that there is no separate
+        indexing in the background.
+    -   A: I personally do not have a large database so I don\'t know
+        for sure. But it is basically as efficient as Grep is, and I
+        have used Grep on multiple-gigabyte files without noticing it
+        being too slow. Modern computers are fast enough that indexing
+        isn\'t required for reasonable performance on smaller databases.
+-   Q: The demo displayed how to search occurances of certain keywords
+    in a giant single-document text database. But what about other open
+    (or not open) Emacs buffers? Think of IRC chars, emails, etc.
+    -   A: Set the hyrolo-file-list variaable to include any directory
+        of files you want to search.
+        -   I mostly referred to non-file buffers.
+            -   Searching through (for example) an IRC buffer is a
+                different command than searching through a directory of
+                files, and this makes sense for the \"Rolodex\" use
+                case, since typically your database will be a file, and
+                not an in-memory buffer. But you can create a hyperlink
+                button that triggers an ordinary \"isearch\"-like
+                command the same way you would execute other Emacs
+                commands.
 
 [[!inline pages="internal(2022/info/rolodex-after)" raw="yes"]]