We have had endless discussions about labels for objects in the LMFDB.
This is different: I want to know how to *name* an object, when
referring to it in a paper. I'm currently interested in this for
elliptic curves (no surprises there) but let's first think about
number fields.
Take a random number field, say Global Number Field 7.1.170783979.1 as
at
http://www.lmfdb.org/NumberField/7.1.170783979.1. Suppose that I
want to refer to this field in a paper. I can say "the number field
with LMFDB label 7.1.170783979.1". Using the hyperref package I can
even make it so that the label is a link to the URL. But I have not
given the field a name so if I need to refer to it again I just have
to call it $F$ or $K$ and use that letter; this does not work if I
have a table of 20 fields each with an LMFDB label and want to refer
to any of them later. I have tried using $F_{7.1.170783979.1}$ or
$F7.1.170783979.1$ but they do not look great.
For elliptic curves over Q I might refer to one in conversation as
"11a1" (though never as "eleven dot a2", sorry -- see the URL
http://www.lmfdb.org/EllipticCurve/Q/11a1 which is quietly 'corrected'
to
http://www.lmfdb.org/EllipticCurve/Q/11/a/2 -- but this is not
supposed to be about labels!). I can try defining $E=11a1" but a
referee or editor might not like that. It's OK to name such a curve
as $E_{11a1}$, I think. (Remember that I need to refer to not just 1
or 2 curves in my paper, but a couple of dozen, and I want to give
each one a name which shows its label if possible.)
That becomes unwieldy over number fields: my current paper has a
whole lot of curves specifically mentioned, including the one with
label 2.0.8.1-9.1-CMa1 for example. Shall I call it
$E_{2.0.8.1-9.1-CMa1}? Or $E2.0.8.1-9.1-CMa1$ or
$E.2.0.8.1-9.1-CMa1$ or $E[2.0.8.1-9.1-CMa1]$?
John