On 10/19/07, Drew Adams <drew.adams@oracle.com> wrote:<snip>
Making files available in the buffer list means visiting them, even if they
are not displayed. So if by "open" he means visit, then what he wants is
impossible
That's what I figured, but I hoped there might be a layer
of abstraction between the list of buffers and the actual
buffers that would allow the buffer to be lazy opened
when the buffer was actually visited. So the buffer
list would be a combination of real buffers and
buffers that could exist by visiting a file if the user
tried to view that buffer.
The use case is that I have a makefile that
generates dependencies for an application
and those dependencies are scattered around
a directory tree with other files that aren't
dependencies. When I'm working on the source
code, I sometimes want to open a file that I know
starts with xyz, but I don't know its directory or
what it ends with. It would be nice to use buffer
name completion to find the file. What I do today
is a tags search for something I know is in that
file, but sometimes I can't think of exactly what's
in it. If I could just type C-x b xyz<tab> and see
a list of all of the (pseudo)buffers that begin
with xyz or search my ECB alphabetized
buffer list for the file, that would be great.
So I think he's gotten some help for his problem from both of us, depending
on what he really wants. ;-)
Yes, thank you all for your suggestions. I'll try them out.