Well presented material - gives a good feel for the various 
implementation schemes in each area used by the various OS.....

Could have used some more emphasis in the conclusion showing
how each of the OS failed to meet the requirements (especially
Amoeba and WebOS) but the information is implied in the prior
sections .....

i feel that the conclusion made about not achieving global and
system wide OS was too strong - given that only 4 OS were
brought into consideration. but other than that it was well written.

======================================================================

Import - The topic certainly seemed to be extremely exciting, and
I read the paper with enthusiasm and found it to be really 
worthwhile!
 
Novelty - The paper tries to bring out the "requirements" for an 
OS meeting certain objectives, and provides a summary of the 
challenges that underlie the design of such an OS, and thus 
justifies its stated objectives.

Quality - The paper suffers from numerous grammatical errors, 
specially in the portions relating to "Amoeba", which sometimes 
make it difficult to understand what is being stated.

Overall - I found it to be a focussed paper, and it does a good 
job of bringing out the requirements of an OS that meets the 
requirements of the WWW, and offers an insight into some of the 
challenges that lie ahead for those interesred in building it or 
knowing about it. The paper should certainly be accepted.

======================================================================

   The paper presents good reasons why the research in the area of 
operating systems is important, and why it should continue.
There are a few points, however,  where the paper can be improved. 
   The title gives the impression that all the work will relate 
to the operating systems dealing with WWW. But were Mach and Amoeba
really created with that goal in mind ? Chronologically, their
implementation was completed before the WWW became a widely used
phenomena. It seems that the title and the contents in some ways
do not really correspond to each other. However, the authors presented
chosen research well, and gave sound explanations  regarding the positive 
and negative aspects of the implementation. 
   One thing that can be improved is the discussion of performance.
The distributed operating systems are very much involved with 
the performance as a requirement, or at least one of the issues. 
It would be nice to see some comparative results that would give a better
understanding that distributive systems do in fact provide users/clients
with better computing power, or to see which implementation is a better
approach and for what target.
   Since some of the presented project papers date back to late eighties
and early nineties, there is not much novelty in the presented observations.
It would be better to see more papers on more current research.
   One small cosmetic issue (no points should be taken off for that):
it is easier to read the paper if all the margins are aligned.

======================================================================

The paper was generally well written. The title did not seem to really
reflect the information presented in the paper. I liked the general flow
of the paper although some of the sections were quite obviously written
by different people. The introductory sections did not always accurately
indicate the information that followed it. For example, section 3
indicates that it will cover the intent and assumptions made by the
different operating systems. However, section 3's OS specific parts did
not really address the assumptions of Amoeba, or Mach and did not
address the intent of NOW. Overall it was above average.

======================================================================

You seem to be trying to assess these operating systems against a a set
of principles for distributed system design.  You must have first
examined whether these principles have relevance to the context in which
you are functioning.  For example, Tanenbaum's idea of a transparent
file system, and transparent login can not be taken with out
modification and applied to the web.  This comparison appears very
ad-hoc and artificial.  All along the paper, you seem to be so busy
talking about Tanenbaum's hypotheses that you forget to mention the
actual context -- you claim to present the "intent and assumptions" made
by these OSes, but you do not do so at all.  And, it is perfectly okay
for an operating system not to follow Tanenbaum's principles in every
aspect, but do well in those few which matter in that context... and you
can not compare a system against principles it wasn't designed to
meet.... clearly, mach or amoeba's distributed virtual memory was *not*
designed to run over the web... again, your examination of why
Tanenbaum's principles are relevant is lacking.  Finally, it all seems
that the following parts -- "Tanenbaum's principles", "WWW", and "Five
different distributed Oses" -- have been clubbed together artificially;
so you need a stronger thesis.  Also, the paper is full of typos,
section headers are indistinguishable from text.

======================================================================

1, Operating systems for the WWW (BCGL)

Import:
This paper describes operating systems for the WWW, it is an important
topic to the OS community. Distributed operating systems are becoming
widely used in the real world. Many computer researchers are 
working on how to design a perfect one which meets many requirements,
such as dynamically allocate resources,fault tolerance,protection and 
scalability. 

Novelty
The authors give a description of distributed OS requirements and four
case studies, Amoba,Mach,Now and WebOS. Although their analysis are good 
and reasonable,I think all these ideas come from the reference papers.

Quality
This paper has good quality, the observations and criticisms are
sound.These showed the authors have a broad knowledge and can give a
thoroughly analysis of the topic.

Overall
This is a good survey paper. It is well structured and easy to 
be understood. But the writing style looks so sloppy, if this is a real
conference, we could not accept this paper until it conforms to the
format.


======================================================================