[DAS2] das2 diagrams, questions

[Andrew Dalke]

Allen:
> I have a few questions, mostly targeted at Andrew, regarding the 
> current
> HTML version of the spec on the biodas.org site.  It hasn't been 
> updated in
> about 5 months, and looks pretty out of date.

Strange.  The last changes were in August.

> * Is the HTML document in sync with the "new_spec.txt" document in CVS?

It should not be.  That was a text document I was working on back in
Jan/Feb as part of the updated to the current version of the spec.  I've
removed it from CVS.

(Even though I know it's CVS, my fingers keep typing "svn" :)


> * There is mention of a "fasta" command, and its fragment is linked 
> from the
> ToC of the genome retrievals document, but it does not appear in the
> document.  Does this command exist?  My understanding from conference 
> calls
> is that the sequence/fasta/segment/dna stuff has all merged into the
> "segment" response. Is this correct?

That is correct.  There is a segments request.  Passing "format=fasta"
to a segment request returns the sequence in FASTA format.

I didn't catch that line when I was doing the changes.  I've removed it
from CVS.

> * The "property" command seems to have disappeared.  Is that correct?  
> Are
> property keys no longer URIs?  Also the "prop-*" feature filters could 
> be
> better described, it is not clear to me if they are meant as some sort 
> of
> replacement for the property command.

The property command has disappeared.  Notes are at
   das2-teleconf-2005-11-28.txt

It was replaced by two things.  One is the key/value PROP table, which 
is
meant to store simple string data.  It should be considered to be
user-editable, eg, as a property sheet.  The "prop-*" commands are used
to search that table.

The other the non-DAS namespace'd XML extensions.  For example,

<FEATURE ...>
   ...
   <PROP key="gene_region_length" value="5398" />
   <fly:map xmlns:fly="http://flybase.org/"
      physical_map="4: 26,994..32,391[-]"
      cytogenetic_map="4: 101F1--102A1" />
</FEATURE>

In this case there is no default search mechanism.  Instead the server
may declare that it implements a map-specific search extension to the
DAS query language, or a new search interface, and clients which 
understand
the extension can add support for it.


					Andrew
					dalke at dalkescientific.com