EML for Mere Mortals

[Chad Berkley]

Hello David,

I've been meaning to write a review of your document for a month or so 
but I've been on vacation and at various meetings.  I apologize for the 
delay...but better late than never.  I'm going to start out with what I 
see as being wrong with the document then try to give you some 
suggestions on how it should be rewritten.

If I understand correctly, this document is supposed to be a high level 
look at EML that a scientist or data manager could use to fill out EML 
fields and to understand the granularity and purpose behind those 
fields.  I have to say that "EML for Mere Mortals" as it currently 
exists falls short of this goal.

If we are truly trying to give the scientist a top level overview, then 
why is there so much discussion of XML structure and syntax?  I think 
this document could be written without mentioning the acronym "XML" a 
single time.  The sentence "...EML has been designed as a form of 
XML..." is completely wrong and I think this is the wrong thesis 
statement to base this document on.  EML is a content model that happens 
to be implemented in XML.  It could be implemented in any number of 
other markup languages or binary formats.  The content model itself (not 
the implementation) is truly what should be the center of the discussion 
in this document.  Possibly as an appendix, some implementation details 
on how to write an EML document marked up in XML would be useful, but it 
does not belong in the body of this document.

As far as your discussion of the content model goes, I do not think 
referring to it as a "Taxonomy" is at all intuitive or correct.  In 
fact, it's downright confusing because of the actual sub-module of EML 
that deals with biological taxonomy.  It took me several re-reads to 
figure out exactly what you were trying to do with the taxonomy analogy 
(and I'm still not completely sure I understand).  I just don't think 
it's leading in an easy-to-understand direction.

The premise behind the the conversations between Pat, emlMaven and Eco 
Informatics, while well intentioned, frankly made me queasy.  I think 
this device has been over used and not well executed in this document. 
The names of the characters themselves insult me. I think the purpose of 
this device (to give some real world problems and use EML to solve them) 
is a sound one, however I think it would be better to have a set of 
scenarios maybe with multiple implementations and multiple outcomes than 
to have three imaginary characters with less-than-normal names 
conversing.  The "Metalog" device also fails in the same respect.

What I would like to see come out of this document is a generalized map 
of the content model cross-linked to different types of ecological 
research projects. It could have sections like "Spatially referenced 
time series research project" and show what parts of EML would be used 
to correctly document this type of project.  There may be some overlap 
in the sections, but I think that is ok because it will show how many 
parts of EML are generalized for multiple purposes.  I don't think using 
the XML Spy images of the schema is very useful for someone not used to 
it.  It may be painful for you, but I think that some other program 
(Illustrator or the like) should be used to create a nice form view of 
the EML document that is being created for the given scenario.  Ths form 
view does not have to show the entire document (or even document 
segment) in one image, it could be broken up in a visually pleasing way.

I think that the extraneous cheese should be removed from the document. 
  It is distracting to the reader and really just adds more white noise 
though which the reader must sort.  I think a softened professional tone 
should be used throughout.  Not as professional as an IBM AS-400 manual 
but not as soft as Winnie the Pooh either.  Something like the tone in 
an Oreilly book would do nicely.

In a document like this, it is extremely important not to give the 
reader the idea that he is being talked down to.  No one likes to be 
talked down to.  For this reason, I would recomend changing the title to 
something less patronizing.  The tone of the introductory paragraphs 
could also use some work in this area.

A few more tidbits....The XML tags should also be removed.  I don't 
think they are appropriate in this document.  Also, the phrase "computer 
literate" should never be used.  It is a cliche` with basically no 
meaning.  I would also like to see an acknowledgements section as many 
people have worked on EML and deserve to be recognized (NSF should 
probably also be recognized as the grantor...ask Matt or James about 
that).  Also, I don't think that ripping NSF with one-liners in an NSF 
funded grant is a super good idea.

Hope this helps.  I'm not trying to be overly-critical just calling it 
like I see it.  Let me know if you want me to clarify anything that I've 
said.

chad

-- 
-----------------------
Chad Berkley
National Center for
Ecological Analysis
and Synthesis (NCEAS)
berkley at nceas.ucsb.edu
-----------------------