VOSpace v1.15 PR comments

[Norman Gray]

On 2009 Jul 13, at 15:45, Dave Morris wrote:

> Norman Gray wrote:
>
>> Sect 3.5.4.2: "Local NFS transfers".
>
> This example originally came from me, so I'd better explain why .
>
>> This talks of transfers within a specific local NFS filesystem.  
>> This is so local that it strikes me as odd that anyone would want  
>> to register an IVORN to describe it -- why would anyone not on this  
>> local system ever need to look this up?

To be clear: I'm not intending to make a RESTful-http-only claim here  
-- we've talked about this before, you know my opinions, and I  
appreciate the complicated constraints here.  I'm not suggesting the  
sort of major change that would imply.

> Second case is an expert user who has more detailed knowledge of the  
> system.

That's fine -- I broadly understand this scenario, more so than I did  
from the document.

However, it does strike me as a bit of a layering violation if the  
(global) VO registry is being involved in any way in an (essentially  
local) NFS optimisation.  If this mechanism is used by deployed  
implementations, then it clearly doesn't cause too much chaos, and  
I'll simply suggest it's a wart to be addressed by v2.0.  If it isn't  
used, but is merely an anticipated use-case, I'll simply question  
whether it really needs to be here and leave it at that.

I suppose, then, my comment changes into a suggestion that some  
rationale for the (apparent?) layering violation be a little more  
explicit in the document.

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
Dept Physics and Astronomy, University of Leicester, UK