On gripes with refer(1) and it's accumulate setting.

[Sigurd Hermann]

Hey, I'm a semi-proficient user of GNU Groff. I really like this system
of writing.  I do have a (small)-ish complaint. So, the way refer(1)
works is usually great.  I love that it's fast, the format is easy to
understand(after you've read the manpage like three different times
to remember the way the label expressions work and all that).  At any
rate. There's one funny limitation that I don't have any context for why
it is there.  It has to do with the setting `accumulate`. I love that
it does what it says on the tin, but it has the annoying limitation that
I can't refer (easily) to the same book or paper multiple times without
either (1) doing this:

    .[
    $LIST$
    .]

and then doing the reference again, with a different page number, but
this is annoying, since unless I plan ahead it can look weird.  Or (2),
doing this:

    .[
    %A Robert Milner
    %A Robert Harper
    %A David MacQueen
    %A Mads Tofte
    %T The Definition of Standard ML, Revised Edition
    %D May 1997
    %L Standard ML
    %P 50--75
    %O ISBN: 9780262631815
    .]
    .[
    %A Robert Milner
    %A Robert Harper
    %A David MacQueen
    %A Mads Tofte
    %T The Definition of Standard ML, Revised Edition
    %D May 1997
    %L Standard ML
    %P 89--120
    %O ISBN: 9780262631815
    .]

Sidenote: I don't know how long this book really is, it's just an
example :O.  Which, while a lot of typing instead of just doing the
regular keyword citation that you normally have to do, does actually do
what I want.  But. I want to do this:

    .\" pretend of course that I've included whatever bibliography file
    .\" before citation
    .[
    Standard ML
    %P 50--75
    .]
    .[
    Standard ML
    %P 89--120
    .]

The limitation was thankfully documented in the refer(1) manpage:

> The fields components specifies additional fields to replace or
> supplement those specified in the reference. When references are being
> accumulated and the keywords component is non-empty, then additional
> fields should be specified only on the first occasion that a particular
> reference is cited, and will apply to all citations of that reference.

So I'd like, if possible, understand why this limitation exists, and
if there is a simpler way of coping with it other than just copying the
same fields over and over again, since it seems that will be what I will
have to be doing for the foreseeable future.