proposed changes to CSL to permit AGU-style reference sorting

CSL Development
21

Hi,

I am starting to understand the logic here. From Bruce’s email:

    The way we do it now is the content of the conditional "type"
    attribute is a list; so one ore more.
    
    <if variable="author editor" match="any" min="3"/>

This formulation of more complex conditions with min and max seems clean
to me. I assume that both can be in a single if to set an upper and
lower bound?

I imagine that match=“any” means match on first existing variable
(author first, then editor). Perhaps this should be match=“first”? I
presume there would also be a match=“all” which would count all the
"author"-type variables listed?

One thing I note though is that it is a bit strange to do essentially a
"substitute" in this using a variable list, but to also have a
which is used in to select which variable to use.
Is it possible to just use a similiar mechanism in ? Perhaps
macros are a problem with this? Perhaps limiting conditionals to just
testing on variables and not macros will cause problems with as
well at some point (e.g., counting name parts picked out by macros???)?
I am not sure these are really important issues, just pointing out
things that could be problems down the road.

Other than that, this formulation of conditionals works for me.
combined with appropriate min and max in declarations to pick
out particular authors, this looks like it should solve the problems
with AGU format. What will happen with the options “et-al-min”, etc.?
It seems to me that the cleanest thing would be to remove them all at
once, but this involves modifying existing CSL’s.

Also, at some point I mentioned some confusions regarding
name-as-sort-order - what it really means, first vs. all, etc. Perhaps
now would be a good time to clear this up in the schema since we will be
making changes already. It seems to me that there are two needs that
people had the idea of addressing with name-as-sort-order. One is that
journals often differ on whether they want the first name to appear
before or after the last name in references. This ordering also may
change on which author is in question (first versus 2-infinity). There
needs to be a way to specify this in CSL. This could be down with
something like name-as-sort-order (though it might need to take
more/different values to represent all cases).

Another competing issue is how to deal with non-western names that might
use the first name for ordering. I think this should be dealt with
differently with something in Zotero that indicates which name is THE
name. But maybe there is something that needs to be in CSL about this.

Cheers,
David–

David M. Kaplan
Charge de Recherche 1
Institut de Recherche pour le Developpement
Centre de Recherche Halieutique Mediterraneenne et Tropicale
av. Jean Monnet
B.P. 171
34203 Sete cedex
France

Phone: +33 (0)4 99 57 32 27
Fax: +33 (0)4 99 57 32 95
http://www.ur097.ird.fr/team/dkaplan/index.html

22

Hi,

Sorry that it has been a while since this was discussed. I am wondering
if these changes for dealing with AGU-style reference sorting are
somewhere on the back burner or fell off the truck. The more I look,
the more I find that almost all science journals implicitly or
explicitly use “AGU ordering” if they use named references (as opposed
to numbered references).

Last I remember, it looked like we were arriving at a consensus on how
thing should work here:

http://sourceforge.net/mailarchive/message.php?msg_name=1213093067.6500.28.camel%40piura

Cheers,
David

23

We need to hear from the Dan/Simon, Johan, Andreas, Liam, etc. about
their thoughts on this, and a commitment to implement it.

I’d really like to clean up these remaining issues so we can call CSL
1.0, but it seems people are distracted with other things.

IIRC, the proposal is to allow a min and max attribute on a conditional?

Bruce

24

Hi,

Responding to Bruce’s email a while back (I was on vacation), the idea
for making AGU formats work involved a few things:

  1. A way to say in only print authors m thru n.
  2. A way to test for the number of authors being greater than or less
    than a certain number.
  3. This would allow removing some of the et-al options in favor of
    macros that would be more flexible long term.
  4. More precision on what name-as-sort-order really does. In
    particular, there needs to be something that clearly controls whether
    author first names appear before or after last names in a list of
    authors.

Some pseudo-code for 1 and 2 above proposed by Johan is below (I
suggested something a bit more general involving general tests for some
quantity greater than some other, but this would do the trick I think):

--> 3,4,5,6,…
--> 0,1,2,3
--> 3
--> 3,4,5,6

we end up with this:

Cheers,
David

25

I don’t recall this as being a real issue here. I’ve not yet seen any
problem with how the current system works on this, and it has the
advantage that it’s at least in theory international-friendly.

Bruce

26

As for 1, 2, and 3, I’m somewhat hesitant to introduce this complexity
if it’s only used in a few styles, although if we decide we want these
features, I don’t think there are any serious issues. For 1, I’d
suggest that, if we intend on adding an element to , we
should just tack this on there. For 2, the conditionals proposed in
the previous emails (e.g., ) seem like necessary evils if we need this kind of
processing. While up to this point, every attribute on has
represented a separate condition, it’s probably true that having
attributes that don’t represent conditions is better than making
implementations parse out the attributes.

I agree with Bruce that 4 is unlikely to present a serious issue. name-
as-sort-order basically controls whether the last name appears before
the first name, although it also takes into account that, with some
foreign names (e.g., Mao Zedong), the last name shouldn’t come before
the first name.

Simon

27

Hi,

Responding to Bruce’s email a while back (I was on vacation), the idea
for making AGU formats work involved a few things:

  1. A way to say in only print authors m thru n.
  2. A way to test for the number of authors being greater than or less
    than a certain number.
  3. This would allow removing some of the et-al options in favor of
    macros that would be more flexible long term.
  4. More precision on what name-as-sort-order really does. In
    particular, there needs to be something that clearly controls whether
    author first names appear before or after last names in a list of
    authors.

As for 1, 2, and 3, I’m somewhat hesitant to introduce this complexity
if it’s only used in a few styles, although if we decide we want these
features, I don’t think there are any serious issues.

Yeah, my feelings as well. I seem to remember David mentioning that
there were likely a number of styles that would need this
functionality, but I don’t know if he was very specific about how
many.

I recall Johan gave this a thumbs up earlier. Any other
thoughts/commitments to implement?

For 1, I’d suggest that, if we intend on adding an element to , we
should just tack this on there.

Yeah, we need this regardless. So we might as well add this.

Bruce