The submitted proposal for the efficient video streaming of 3GP files 
project is very interesting and thought provoking. The proposal 
included a brief introduction of the common problem of packet loss while video 
streaming, followed by a proposed solution, which I believe to be 
viable and efficient. A way to improve this proposal would be to mention the 
advantages of using this approach as opposed to other alternatives such 
as error correction and data retransmission.

	This proposal peaked several questions/ideas about the topic, which I 
believe to be necessary components of a good proposal:

 	The introduction section states that the client usually has to carry 
on if packets get lost in transit. An example of exactly what the 
client does do to carry on (e.g. error correction techniques available, 
NACK, ignore, etc.) would prove useful.
 	In the hint track section, the author states that the hint track 
provides a bridge between the media data and the packet data. This 
immediately sparked my interest and led to the following questions: 
 	How are the hint tracks encoded?
 	What algorithms are used to predict distortion for different packet 
levels? 
 	What is the method for determining the best packet level?
 	How much of an improvement is expected with hinted tracks versus 
extra work/bandwidth involved?
 	The author informed that the format of hint tracks for the RTP 
protocol is being standardized and that there is also a proposed format of 
hint tracks for MPEG-2 transport streams. The website for MPEG-4 file 
formats white paper gives detailed information about hint tracks and a 
defined format for RTP streams: 
http://www.chiariglione.org/mpeg/technologies/mp04-ff/index.htm
Will this well defined format easily convert to 3GP file streams as 
advertised?
 	The background section failed to list any groups that are doing 
research in this area or any specifics of standardization groups or 
authorities. Listed below are links to some groups that are doing work in this 
field and may prove helpful in the development of this project:
 	Communications Network Research Institute (CNRI): 
http://www.hamilton.ie/ncnrc/workshop_june2005/VideoOverWLAN.pdf
 	Dept. of Computer Science, University College Dublin, Belfield, 
Dublin, Ireland:
http://www.eeng.dcu.ie/~perryp/pub/cranley2.pdf
 	The author proposes to compare the streaming performance of using 
hint tracks versus not using hint tracks. 
 	Has the author decided on what types of performance measurements to 
use in order to validate the comparison metrics mentioned in the report? 
Some suggestions may be:
 	The Comparing Alternatives method together with a confidence interval 
> 90%.
 	ANOVA testing to ensure that all results are legitimately attained 
(extraneous errors in measurement are accounted for).