SDD goals

Kevin Thiele kevin.thiele at PI.CSIRO.AU
Wed Nov 1 17:43:44 CET 2000


Original: Eric Zurcher <Eric.Zurcher at pi.csiro.au>
To: <TDWG-SDD at USOBI.ORG>
Sent: Tuesday, October 31, 2000 11:53 AM
Subject: Re: TDWG-SDD XML proposals of Kevin Thiele


| One thing that might be helpful would be for this group to come up with a
| short set of basic design goals that indicate what our objectives really
| are. My understanding is that the developers of XML itself found such a
| set of goals to be very useful.

How's this for a start?

SDD shall allow any description of an object, item or taxon to be marked up
for its data content
SDD shall comprise a single standard, but be flexible enough to allow a wide
variety of transformations
SDD shall allow capture of all standard data elements currently allowed by
Lucid, DELTA [...add to this list...] etc.
SDD shall allow but not require rich attribution of all data elements
SDD shall allow hierarchical structuring of documents and collation of data
from one document into another
SDD shall be straightforwardly usable over the Internet.
SDD documents shall be straightforward to parse
SDD markup shall be performable by hand
The design of SDD shall be formal and concise.
Terseness in SDD markup is of minimal importance.

One that I don't fully understand from the XML list is:
| The number of optional features in XML is to be kept to the absolute
| minimum, ideally zero.

How does this apply to us? The flexibility we need is such that options are
essential - just as in Lucid, DELTA etc where character sets, dependencies,
subsidiary files etc are all optional. In fact, I've argued that
non-optional (required) data elements should be kept to a minimum - and
sdd0.3 is based on this working principle, hence the non-requiredness of
e.g. character lists.

| But perhaps the biggest challenge with any such
| standard is that it needs to be very comprehensive and flexible - ideally,
| it should be able to store everything that all of the "subsidiary" formats
| is capable of representing, and do so in such a way that X<->Y
| translations are as lossless as possible.

See goals 2&3 above.

Cheers - k




More information about the tdwg-content mailing list