Hello all,
I have just upgraded slime to the current version from cvs and I have the following questions:
1. In emacs, let's say I have two open windows in the same frame. The first is the REPL and the second some lisp file. When I macroexpand an expression with C-c RET, say from the REPL, the result of the macroexpansion appears in the window with the lisp file. Up to here, fine, but, when I press `q' to close the macroexpansion, the window where it appeared is deleted (in my example only the REPL remains). In previous versions, the window remained intact and its former contents appeared again.
I find that the old behaviour was much more convenient. I think the current behaviour is expected only if there is only one window originally and the macroexpansion itself causes the second window to open.
So, is it a bug or is it intentional? If the latter is true, is there a simple way for me to revert the old behaviour (I suppose changing something in slime.el or what?)
2. I recently read in this list about some fuzzy completion patches. I thought that those just added new behaviour and I could use the old behaviour. But either I got a false impression or I cannot find the correct settings. Again, how can I revert to the old behaviour?
If this matters, I use GNU Emacs 22.0.50.1 on linux, with SBCL 0.9.12
Thanks in advance,
-- Giorgos
- In emacs, let's say I have two open windows in the same frame. The first is the REPL and the second some lisp file. When I macroexpand an expression with C-c RET, say from the REPL, the result of the macroexpansion appears in the window with the lisp file. Up to...
i've fixed it to handle these situations better. hopefully the version at
darcs get --partial http://www.common-lisp.net/project/cl-wdim/darcs/slime/
will satisfy all requirements. not tested it throughly yet, but it seems to handle well all the cases i could think of: always closes the temp buffer window if it was opened by temp buffer activation and the window config haven't changed since then.
these will eventually reach the official repo when i squeeze the time to send cvs diffs and someone applies them.
- I recently read in this list about some fuzzy completion patches. I thought that those just added new behaviour and I could use the old behaviour. But either I got a false impression or I cannot find the correct settings. Again, how can I revert to the old behaviour?
when slime-fuzzy-completion-in-place is nil (the default in the official cvs) then the fuzzy stuff should behave mostly like before. but we use it with slime-fuzzy-completion-in-place enabled.
we did our best to make the gui more user friendly, but user-friendlyness is a subjective thing. what specifically do you find annoying (other then the unavoidable (hopefully) short-time annoyance coming from any change in the behaviour)?
hth,
Hello,
"Attila Lendvai" attila.lendvai@gmail.com writes:
- In emacs, let's say I have two open windows in the same frame. The first is the REPL and the second some lisp file. When I macroexpand an expression with C-c RET, say from the REPL, the result of the macroexpansion appears in the window with the lisp file. Up to...
i've fixed it to handle these situations better. hopefully the version at
darcs get --partial http://www.common-lisp.net/project/cl-wdim/darcs/slime/
will satisfy all requirements. not tested it throughly yet, but it seems to handle well all the cases i could think of: always closes the temp buffer window if it was opened by temp buffer activation and the window config haven't changed since then.
I did not expect that this had to do with the completion patches. I'll try your latest version from your darcs repository and let you know if I still notice anything that troubles me.
these will eventually reach the official repo when i squeeze the time to send cvs diffs and someone applies them.
- I recently read in this list about some fuzzy completion patches. I thought that those just added new behaviour and I could use the old behaviour. But either I got a false impression or I cannot find the correct settings. Again, how can I revert to the old behaviour?
when slime-fuzzy-completion-in-place is nil (the default in the official cvs) then the fuzzy stuff should behave mostly like before. but we use it with slime-fuzzy-completion-in-place enabled.
we did our best to make the gui more user friendly, but user-friendlyness is a subjective thing. what specifically do you find annoying (other then the unavoidable (hopefully) short-time annoyance coming from any change in the behaviour)?
Well, it is a minor thing but, when the completion list appears, the source window is not updated with the first applicable completion like it used to.
I really liked this, especially combined with one little addition by Antonio Menezes Leitao in a post here from 8 May 2006, which enabled me to press space to confirm the completion, add the space in the source window and keep on writing. Most of the time, I did not even move my eyes from the source window -- I looked at the completions window only if the first completion was not what I intended to type in.
Anyway, I'll try to evaluate slime-fuzzy-completion-in-place a bit more and see if I can provide you with any feedback. Don't expect code for the time being :-) I am just learning Common Lisp and I know next to nothing about Emacs-Lisp. Should I contact you personally? I have the impression that I am just adding noise here and I would not like to further annoy the mailing list.
hth,
--
- attila
Thank you very much for you answer, and thank you everybody for slime.
Best regards, -- Giorgos
Giorgos Pontikakis gnp@freemail.gr writes:
I have the impression that I am just adding noise here and I would not like to further annoy the mailing list.
Please keep this on-list!
Cheers,
-- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs."
-
Attila Lendvai -
Giorgos Pontikakis -
Nikodemus Siivola