InCommon metadata support

[Tom Scavo <>]

On Thu, Jan 9, 2014 at 4:32 PM, Schwoerer, Bradley J 
<>
 wrote:
> I was reworking our local metadata file generation script and part of it 
> was to do some xml tidy type tasks.  As I was working through the 
> generation of our local file, I had decided to use the incommon federation 
> metadata file as a sanity check.   After comparing the input and output 
> files for the incommon metadata, I did notice an almost 500KB (7% 
> reduction) in file size.   This was mostly from just doing all of the 
> xmlns:* declarations at the EntitiesDescriptor node and removing all of the 
> duplicates declarations.  Just because it piqued my interest, I was looking 
> to see if I could do further reduction.  With removing the two spaces used 
> for indention and just leaving the newline, it reduced it another 500KB.

Those are some very interesting numbers, Bradley.

> I am not sure how many times per day the metadata file is downloaded

I have some scripts that compute this but I haven't run them in
awhile. Should probably dust them off and give them a spin :-)

> but I thought I would bring it to someones attention of the potential 
> savings in bandwidth by making a few changes to the generated metadata file.

Noted. As Scott and others often remind us, however, we're nowhere
close to a size that matters, so it may not be worth the development
effort to realize these modest savings. (In any case, the software
that builds the XML relies on an XML library that we wouldn't want to
touch ourselves.) Also, there is a pilot involving per-entity metadata
in our future, so there may be some benefit in having each entity
descriptor in the aggregate be self-contained. I don't know if that's
true, however. Given the work that you've done, can you say one way or
the other? Does each entity descriptor stand alone?

Thanks,

Tom