[net.emacs] Bugs in Gosling's

kimbo (04/28/83)

I hope this is not a comercial use of the net but...
If I'm going to have to pay for my emacs, I'd like to get rid of a few
of my pet bugs, namely:
1) If an abbreviation is expanded inside a save-excursion (with dot at
   the end of the abbreviation before expansion) dot is left at the
   beginning of the expansion.  E.g. abrv => abbreviation
					 ^   ^
   where the carats represent the cursor's position.
2) Why doesn't save-excursion also save the current regular expression.
   Without this (or some other mechanism), you can't write mlisp that
   uses looking-at unless you also write your own interface to the
   regular expression operations that remembers what the user's last
   re was.
3) Why no looking-at-reverse?  Is this related to...
4) Reverse regular expressions seem to be full of bugs.
Finally, does it do any good to list bugs here?  Is anyone listening?

		Kimbo B. Peebles-Mundy, NBI Inc.
		{allegra,ucbvax}!nbires!kimbo

chris.umcp-cs@UDel-Relay (04/29/83)

From:  Chris Torek <chris.umcp-cs@UDel-Relay>

	From: hplabs!hao!cires!nbires!kimbo@Ucb-Vax

	If I'm going to have to pay for my emacs, I'd like to get
	rid of a few of my pet bugs, namely:

	1) If an abbreviation is expanded inside a save-excursion
	   dot is left at the beginning of the expansion.

I can't tell you how to fix it, but I CAN tell you exactly what's
going on:  abbreviations are expanded by first deleting the original
word, then inserting the new word.  The deletion moves the mark that
holds the old "dot" value back, while the insertion does not move it
forward.  This applies to a lot of situations, like setting a variable
to (dot) and then munching text.  I spent some time rewriting something
to get around that once.

	2) Why doesn't save-excursion also save the current regular
	expression.

Good question.

	3) Why no looking-at-reverse?  Is this related to...

Probably 'cause no one's needed it badly enough yet to write it.

	4) Reverse regular expressions seem to be full of bugs.

I think the FSM-code doesn't work right for reverse searches, though
I don't know why.  I'd have to do WORK to fix that one....

	Finally, does it do any good to list bugs here? Is anyone
	listening?

There are plenty of people LISTENING, just not the Unipress people.
Still, it might be a good idea for someone to collect a list of known
bugs.  (And fixes?  Would it be legal to distribute changes for the
current (read: old, free) version of Emacs?  How many of you who've
fixed things have kept a ChangeLog?)

					- Chris