C332 Referee report
The strategy explored in this paper is sound but the paper needs additional work 
before publication. My main difficulty is the rather simple algorithm -- ray tracing -- 
which consists of independent and essentially uncorrelated tasks. Thus it does not test key 
issues of inter-processor synchronization, which are important in many distributed 
computing applications.

C332 Letter
Dear XYZ,
We enclose 2 referee reports on:
C332:
This is the style of paper that would be suitable for Concurrency: Practice and 
Experience but both referees agree that current paper is incomplete as it stands. We 
would be happy to consider a revised paper that addresses issues raised by the referees. 
Please include a memo with your resubmission describing any changes.
We thank you for your interest in Concurrency: Practice and Experience.

C348 Referee Report
This is a potentially interesting paper but I cannot recommend publication in its 
current form. In spite of the authors' claim there are many publications in this field as 
long as you include those use "processes/threads" and not just "objects". Thus further 
work must make progress in this context before publication.
I cannot understand the novelty of the new method from the description given.  
Again the rather academic problems and quaint parallel processor make it impossible to 
draw interesting deductions.

C348 Letter (Berman report on its way)
Dear XYZ
We enclose 3 referee reports on:
C348:
The referees have a common set of criticisms -- saying that the new ideas are not 
clearly stated or convincingly investigated. We would be happy to consider a revised 
paper that addresses issues raised by the referees. Please include a memo with your 
resubmission describing any changes.
We thank you for your interest in Concurrency: Practice and Experience.

C355 Letter
Dear XYZ,
We enclose a referee report on your paper
C355:
We would be pleased to publish and ask if you would like to make any 
modifications based either on the referee's comments or recent insights that you might 
have had. We will this revised paper without further review.
We thank you for your interest in Concurrency: Practice and Experience.

C357 Referee Report
This is a rather interesting paper, which should definitely be published with 
perhaps modest revision. I have four comments.
1) The good performance is not surprising and is exhibited by all problems 
whose computational complexity is quadratic in problem size. The best known 
example of this class is the N body problem with N particles interacting with a 
long-range force such as gravity.
2) Comparing the author's box algorithm with the well-known fast multipole 
method for N bodies suggests a modification of the correlation box algorithm. 
Namely one sets up a recursive hierarchy of boxes with presumably increasing 
number of points in boxes of increasing size. However rather than calculating 
all or zero points as now with strict cutoff, one calculates in all boxes but in 
outer boxes, one only calculates a (random) sample of points. The number of 
points is chosen to equalize errors as a function of ln (r) when lnCm (r) is 
plotted against lnr. 
In general I suggest comparing of correlation and N body problems with 
attention to new faster ways of calculating them.
3) I think there is an error in second line of equ 1) where N and N(subscript s) 
are interchanged
4) Although I enjoyed paper, I think it could be usefully shortened.

C357 Letter
Dear XYZ,
We enclose a referee report on
C357:
We hope you can send us a modestly revised paper that we will publish without 
further review.
We thank you for your interest in Concurrency: Practice and Experience.

C358 Report
This is a well-written paper on an interesting subject. I am familiar with the 
general resource management area but not with this particular topic. From my 
point of view, I find the artificial test problems very unsatisfactory. Surely one 
should have some real application in mind with a corresponding set of tasks. I 
would recommend against publishing paper unless the authors explain why and 
where their algorithms would be used. This would then allow you to derive or 
measure the task parameters, which will allow meaningful tests of the strategy.