aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--etc/README1
-rw-r--r--etc/bug-notes.txt32
2 files changed, 33 insertions, 0 deletions
diff --git a/etc/README b/etc/README
index 5fdc2042..c7f97d9b 100644
--- a/etc/README
+++ b/etc/README
@@ -18,6 +18,7 @@ example test protocol for example proof scripts
notes.txt Misc notes
+bug-notes.txt Test cases for Emacs or PG bugs
cvs-tips.txt Notes on cvs with PG project
debugging-tips.txt Notes on debugging
profiling.txt profiling
diff --git a/etc/bug-notes.txt b/etc/bug-notes.txt
new file mode 100644
index 00000000..e00a23e7
--- /dev/null
+++ b/etc/bug-notes.txt
@@ -0,0 +1,32 @@
+-*- outline -*-
+
+$Id$
+
+Test cases for PG and/or Emacs bugs.
+
+----------------
+
+* XEmacs bug: buffer-syntactic-context-depth returns weird values
+
+Seems to depend on previous history. Test in Coq buffer:
+
+ X
+ (* comment one *)
+ Y
+ (* comment two *)
+ Z
+
+Evaluate (buffer-syntactic-context-depth) at X, Y, then Z.
+Values 0, 1, 2. Evaluate at point Y. Now get 0.
+Perhaps caches previous value, and bases parse on moving point
+forwards from previous value? Anyway, doesn't do well with
+block comments. Also bad with line comments, use same test
+case with buffer in lisp mode, except with lisp comments.
+
+Shame, would be nice to use this to help parse lisp-like
+syntax for PAs, fitting in with present scheme.
+
+
+
+
+