home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!think.com!yale.edu!qt.cs.utexas.edu!cs.utexas.edu!gateway
- From: ophof@SERVER.uwindsor.ca (Scott Ophof)
- Newsgroups: comp.mail.elm
- Subject: List of enhancements
- Date: 28 Dec 1992 03:29:35 -0600
- Organization: CS Dept, University of Texas at Austin
- Lines: 88
- Sender: daemon@cs.utexas.edu
- Message-ID: <9212280932.AA14332@SERVER.uwindsor.ca>
- NNTP-Posting-Host: cs.utexas.edu
-
-
- Having read the documentation on "elm", and in character with
- the feeling that emanates from it (to give the user something
- much better than other available MUAs), I'd like to offer some
- suggestions and requests for improvements to "elm".
-
- Friendlier and more "intelligent" display of ones folders:
- Instead of that scrolling of the list, and of course discovering
- that the folder one wanted has already gone off-screen before being
- able to fully read it, why not present the display in screenfuls,
- each of which can be selected SEPARATELY. And when the last folder
- is visible, then *PLEASE* NOT automatically "roll off the end"?
- - This could be done using the "<" and ">" keys to select which page
- one wants to skip to for display, and RETURN to quit the display.
- - And how about allowing one to use the TAB key to skip to each
- folder name in turn, with the SPACE key informing "elm" of the
- folder one wishes to change to?
- - And the cursor keys (and/or the "hjkl" keys) to assist in this
- positioning of the cursor in the right spot?
-
- What about allowing invocation of "elm" from within itself?
- It's strange that one cannot take a "side-look" at some item in
- another folder than the one currently referenced without first
- having to quit that folder, "change" to the other folder, wait for
- its index to be built, find the item, invoke the pager, SPACE down
- to what one is looking for, quit the item, "change" back to the
- original folder, wait again for the (unnecessary) index re-build,
- etc., etc...
- And it's even more frustrating when having to do a side-check while
- building an item to *send*! :-(
-
- Unix boasts being able to be used from almost any terminal
- imaginable. It's strange then that no application I've come across
- up to now realizes that when a user wants to skip straight to (say)
- the 4th page by hitting SPACE 4 times, one still needs to wait for
- each of those pages to be laboriously built and displayed onscreen
- (line by molasses-slow line at 2400 bps...) before going on to the
- next. Isn't there any way for a program to detect that it has been
- requested to already display the next screen before it has ended
- updating the current screen?
- My question here is whether it's at all possible in Unix, and if so,
- why it has never been implemented? It seems so *obvious* to me...
- Unix is a full-duplex system, so it *should* be possible. And I
- know from experience that it *is* possible and common on half-
- duplex systems...
-
- As to aliases, how about enhancing that facility to allow users to
- include such things like which folder to store items in related to a
- specific alias, phone numbers, surface addresses, user-definable
- fieldnames and -values? Especially if the facility were not
- integral to "elm" anymore, but invocable from "elm" as a separate
- utility, which then could also be used by other applications.
- If this were implemented, then the very limited and not-really-so-
- useful relevant parameters in ".elmrc" would at last become
- obsolete. :-)
- And maybe a separate screen to display field-names and -values of an
- alias entry, both for input and display? And being able to cycle
- through the entries forwards and backwards?
-
- Another thing is the possibility of retaining the index of a folder
- (as an on-disk file) as long as the relevant folder doesn't change.
- Skipping the index-rebuild each time for folders containing many
- items would save quite some wall-clock and CPU time (at the expense
- of some disk space, of course).
- In addition, an option like:
- saveindex=[yes|no|number]
- (where "number" would be the minimum number of items at which to
- start saving the index) would still give the user control over this
- option.
-
- A more friendly built-in browser when reading items would also be
- appreciated. One where paging commands are supported, both
- forwards, but especially BACKwards! And at least per screen and
- half-screen, preferably in any increment the user wishes.
- And NOT automatically "run off the end", *PLEASE*!
- The built-in pager does NOT seem to support such commands, nor do
- "more" (and "less")...
- BTW, I do *not* see "vi" or "e-macs" (or rough equivs) as viable
- alternatives.
-
- Please, implement redraw for each and every separate kind of screen
- displayable, not only for the index; some phone lines are really
- *terrible*!
-
-
- Regards.
- $$/
-
-