Update todo

src/backup/todo.txt

@@ -1,15 +1,19 @@
 ! What precise workflow do I need for me to be viable to use this?
 ! From a user (novice) point of view, how does it look like?
 
-- Rename: Ctrl-C instead?
-- If I want to make the ID / pointer design right, (mostly I think for coroutines) I think all the pointers need to be accessed through a broad ID interface without exposing pointers. This seems like a giga hassle.
-    - What would we do if the pointer disappered? should Null work as dev null in some circumstances or should all the operations be done one console?
-    - Seems like over reaction but probably the "correct" thing to do?
+
+How to go about search/replace, opening code and other considerations
+- We can use sed + find to search and replace, the automatic reopen should do the job
+- Maybe also we can List all files recursively instead of opening them, how fast is that???
+- For fuzzy find number of files is the problem - most likely just getting them in one place is the biggest problem that can be optimized
+- Opening code is problematic for large / weird files, they stall for a while, getting them buffered up is hard. And they stall the execution when they go overboard. Maybe just skipping on filesisze on something would be enough, or making a very large skip list. Another solution would be lazy loaded buffers + determining what kind of file we are dealing with on open before loading
+- Search everything, like fuzzy panel, on every key stroke we grep immediately and get new results, cancel old, reset
 
 Use session 2
-- I can't change browser currently, maybe syntax like :Set InternetBrowser "firefox"
+- Need configs I can't change browser or vcvarsall currently, maybe syntax like :Set InternetBrowser "firefox"
 - BUILD annoying: Console or use the other window / the window which is already being used
 
+
 Debug session:
 - Should highlight main buffer when clicking on status?
 - Report errorf - use coroutine dialogs