Re: Working Group Last Call: Y.1541 QOSM

Hi Jukka,

thanks for your suggestions, see below...

At 02:42 AM 10/27/2008, Jukka Manner wrote:
>...Authors, please update the draft and take into account the 
>following items I found in my own chair's review:
>
>- Some of the references hyperlinks do not work in the html version

This seems to be a quirk in the html-conversion routine
when working from plain text.  The html version I created
from the .xml file has working links, so I'll try to replace
the auto-generated html with my own html file...


>- The text looks a bit weird since you have typically twice the 
>reference, e.g., "As stated above, [Y.1541] [Y.1541] is ..."

yes, fixed.


>- Is parameter M measured in octets? Please specify explicitly. Same 
>for Bp, actually.

OK, done.


>- Restoration priority: you describe some candidate time values that 
>are based on current technology. Wouldn't it be good to add this 
>time-to-restore as an optional value in the object, say, instead of 
>having 42 bits unused, this field could indicate in milliseconds how 
>fast the restoration SHOULD be. The network can of course decide 
>otherwise. A value of all 0's would mean "Unspecified". This would 
>allow the spec to be future proof. The values you give in the spec 
>could be default values to be used.

I discussed this with a key co-author, and we'll make the following
changes:

Add Reference to Rec. Y.2172 for short descriptions of standardized
restoration priorities.

Clarify that the Reserved octets (24 bits) may be used for a restoration
time specification, or an extent (like probability) of restoration,
or both of these, (or these and something else, etc.).
IOW, the unused Reserved Field is intended for the sort of thing you
suggest, but we're not ready to give all the space to restoration
time yet...

hope this helps, the draft is on the way,
Al