Court spec issues

[Frank Bennett <biercenator@gmail.com>]

I did some more work on the Legal Resource Registry this weekend. Highlights:

* Rendering and exports can now be controlled via a plugin. This
permits individual consumers of LRR data (Free Law Project, OASIS,
MLZ) to refactor the data independently in a repo fork, in the form
that each requires.

* Japanese courts now show English translations in title attribute
revealed on cursor rollover.

I'm also close to implementing a dedicated page for neutral citation
forms in US courts.

While pulling it all together, I spotted a few issues that might
benefit from discussion. They are listed below, in case they are
useful to the group.

***

# Purposing and scope issues

(1) CountListener uses the data for parsing citations out of plain
text. To validate and disambiguate citations, the date range of
reporters and the courts covered by them are useful information.
Should the specification include a framework for handling this
information, or is it out of scope?

(2) MLZ uses the data for the rendering of citations. For consistency,
a full controlled list of court names and IDs is useful. Should the
specification include guidance on the composition of such a controlled
list, and for casting court IDs?

(3) Should the specification cover the method of citing unreported
cases for a given court?

(4) Should the specification cover the pinpoint method for individual
resources (such a page or paragraph)?

(5) Is the aim to assure sufficient information in a cite bundle to
identify the target resource? Or is it to specify the core information
needed to fully describe a resource? Or both?


# Feature specification issues

Features can have several classes of characteristics. Which of these
should be specified for a given field in a given source?

(6) Whether the field is required or optional in OASIS cite data for
that source.

(7) Whether or not the field is *essential* for identifying a specific
resource (the case name comes to mind as something we could expect to
required, but which is non-essential if other details are complete).


# Multilingual issues

(8) Are transliterations and translations of court and reporter names
within the scope of the specification? If so, is the aim to assign
uniform values to individual resources, or only to provide a framework
for individual cites into which arbitrary values can be written?

(9) Where there is an official translation of a resource (for example,
a court-sponsored reporting service), should that status of the
translation be recorded in the cite data?


# Court Identifier issues

(10) Is the court identifier scheme of the LRR (loosely based on
URN:LEX) acceptable?

(10) If so, is it sufficient to adopt a local logic for the segments
of a court identifier, or is a more rigidly consistent set of
requirements necessary?

(11) Specifically for the US federal system, is it acceptable to adopt
the current pattern of setting a named segment for the federal
jurisdiction, and placing all bodies outside of federal authority in
the global national namespace? That is, US Supreme Court =
us;federal;supreme.court; and California Supreme Court =
us;ca;supreme.court.

(12) When the constitutional foundation of court changes (say, Article
1 -> Article 3), should its identifier be changed?

(13) When the name of a court changes (with or without a change in
constitutional foundation), should its identifier be changed?


# Maintenance/workflow issues

(12) Should the specification provide a procedure for recommending and
adopting changes when new services (such as a newly introduced neutral
citation form, or the appearance of a new court) are introduced?

***

Frank