🚀 go-pugleaf

Joe Wells writes:
 > It would be really incredibly nice if it were possible to give
 > vm-confirm-quit a value that would mean VM would ask before quitting a
 > folder if the folder were larger than a certain size in bytes.  I propose
 > that a numeric value for vm-confirm-quit should have this meaning.

Good idea.  I think size should be in number of messages rather
than bytes, though.

In article <9210021722.AA05654@wendy-fate.UU.NET> kyle@uunet.uu.net (Kyle Jones) writes:

   Joe Wells writes:
    > It would be really incredibly nice if it were possible to give
    > vm-confirm-quit a value that would mean VM would ask before quitting a
    > folder if the folder were larger than a certain size in bytes.  I propose
    > that a numeric value for vm-confirm-quit should have this meaning.

   Good idea.  I think size should be in number of messages rather
   than bytes, though.

What is the relative effect of number of messages vs. number of bytes on
how long it takes VM to load a folder?

-- 
Joe Wells <jbw@cs.bu.edu>
Member of the League for Programming Freedom --- send e-mail for details

Hi-

In article <JBW.92Oct3193217@bigbird.bu.edu> jbw@bigbird.bu.edu (Joe Wells) writes:

   In article <9210021722.AA05654@wendy-fate.UU.NET> kyle@uunet.uu.net (Kyle Jones) writes:
      Good idea.  I think size should be in number of messages rather
      than bytes, though.

   What is the relative effect of number of messages vs. number of bytes on
   how long it takes VM to load a folder?

It should really allow either setting dependent on the symbol's
variable value. In fact, I'd like to see it execute the symbol's
function cell, and act on the result. That way I could set my own test
function.

I would not suggest this be the default behavior, just that it
overrride the default vm-confirm-quit behavoir. Is there interest in
this?

-Rens

--
  __   ___  ,    __  
 /__) /__  /| / (    |  J. Laurens Troost    Lehman Brothers Technical Services
/  \ (___ / |/ __)   |  *Opinions expressed herein are mine. Mine, I tell you!*
-------------------------------------------------------------------------------
INET: rens@shearson.com        VOICE: (212) 464-3705        FAX: (212) 464-2040

Joe Wells writes:
 > In article <9210021722.AA05654@wendy-fate.UU.NET> kyle@uunet.uu.net (Kyle Jones) writes:
 > 
 >    Joe Wells writes:
 >     > It would be really incredibly nice if it were possible to give
 >     > vm-confirm-quit a value that would mean VM would ask before quitting a
 >     > folder if the folder were larger than a certain size in bytes.  I propose
 >     > that a numeric value for vm-confirm-quit should have this meaning.
 > 
 >    Good idea.  I think size should be in number of messages rather
 >    than bytes, though.
 > 
 > What is the relative effect of number of messages vs. number of bytes on
 > how long it takes VM to load a folder?

A large number of bytes primarily means trouble if you have a
slow disk or are loading over NFS.  If you;re tight on physical
memory you'll page some as VM scans the buffer sequentially
looking for message separators.  But Emacs will have just loaded
the whole file into memory so very little of it should have been
paged out, so paging should be low.  Things may get bad later in
the session because of gap motion, but VM mostly avoids this at
startup.

A large number of messages means VM must spend time building
the message lists, parsing headers and generating the summary.
Number of messages is much more significant here, because each
message, regardless of size, adds about the same amount of work.

Rens Troost writes:
 > In article <JBW.92Oct3193217@bigbird.bu.edu> jbw@bigbird.bu.edu (Joe Wells) writes:
 > 
 >    In article <9210021722.AA05654@wendy-fate.UU.NET> kyle@uunet.uu.net (Kyle Jones) writes:
 >       Good idea.  I think size should be in number of messages rather
 >       than bytes, though.
 > 
 >    What is the relative effect of number of messages vs. number of bytes on
 >    how long it takes VM to load a folder?
 > 
 > It should really allow either setting dependent on the symbol's
 > variable value. In fact, I'd like to see it execute the symbol's
 > function cell, and act on the result. That way I could set my own test
 > function.

Too complicated for the typical user, I think.  Size in messages
is the best approximation I think.

	Joe> It would be really incredibly nice if it were possible to
	Joe> give vm-confirm-quit a value that would mean VM would ask
	Joe> before quitting a folder if the folder were larger than a
	Joe> certain size in bytes.  I propose that a numeric value for
	Joe> vm-confirm-quit should have this meaning.

      Kyle> Good idea.  I think size should be in number of messages
      Kyle> rather than bytes, though.

    Joe> What is the relative effect of number of messages vs. number of
    Joe> bytes on how long it takes VM to load a folder?

  Kyle> A large number of bytes primarily means trouble if you have a slow
  Kyle> disk or are loading over NFS.

Or are visiting your mailbox via ange-ftp.

-- 
Joe Wells <jbw@cs.bu.edu>
Member of the League for Programming Freedom --- send e-mail for details