Changeset 2337 for Papers/cpp-asm-2012


Ignore:
Timestamp:
Sep 25, 2012, 2:47:10 PM (7 years ago)
Author:
mulligan
Message:

Removed the referee's comments that have already been addressed from the reviews.txt file.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • Papers/cpp-asm-2012/reviews.txt

    r2334 r2337  
    11----------------------- REVIEW 1 ---------------------
    2 1) The main weakness of the paper is the detailed presentation of lemma statements.  I believe it would be better to focus on the high-level insights, from which competent readers could then reconstruct the lemma statements and proofs.  Following this advice would open up substantial space, which I believe could profitably be used to combine this paper with the simultaneous CPP submission by the same group.  It would be nice to have a single conference-length paper that explains all the key insights.
    3 
    4 2) One contribution of this paper is the novel way of reasoning about assembly programs that may do address arithmetic.  From the abstract, it is not clear that the current development supports no such safe arithmetic, but rather the approach is conjectured to support easy extension for that kind of code.  I think this is a reasonable position, but the speculative nature of the motivation ought to be made clearer earlier.
    5 
    6 Page 3, "We only remark", perhaps add "that" after?  (and there is a double "of of" shortly after)
     21) The main weakness of the paper is the detailed presentation of lemma statements.
     3I believe it would be better to focus on the high-level insights, from which competent readers could then reconstruct the lemma statements and proofs.  Following this advice would open up substantial space, which I believe could profitably be used to combine this paper with the simultaneous CPP submission by the same group.
     4It would be nice to have a single conference-length paper that explains all the key insights.
     5
     62) One contribution of this paper is the novel way of reasoning about assembly programs that may do address arithmetic.
     7From the abstract, it is not clear that the current development supports no such safe arithmetic, but rather the approach is conjectured to support easy extension for that kind of code.
     8I think this is a reasonable position, but the speculative nature of the motivation ought to be made clearer earlier.
    79
    810Page 4, "preinstruction" definition: why does "ADD" get two arguments, instead of merging the two "fixed-length vectors"?  Or perhaps I misunderstand what this last term means.
    9 
    10 Page 6, "problems consists", both shouldn't end in "s"
    11 
    12 Page 8, "In plain words, the type of assembly": last word should probably be in code font
    13 
    14 Page 10, "lemmas appears", both shouldn't end in "s"
    15 
    16 Page 12, extraneous comma, "corresponding, source"
    17 
    18 Page 14, "to constraint" should be "to constrain"
    19 Page 14, "modes than the" should be "modes that the"
    20 
    2111
    2212----------------------- REVIEW 2 ---------------------
     
    4333
    4434----------------------- REVIEW 3 ---------------------
    45 page 2, line 4: giveth --> given, taketh --> taken ?
    46 page 9, Huth et al [15] --> Tuch et al [15] ?
    4735
    4836The term "total correctness" used in Sec. 3.4 and 3.5 is confusing,
     
    9684      descriptions.
    9785
    98    c. p.3: "heavily exploit", why not just "exploit"?
    99 
    100    d. p.4: suggest replacing "and so on" with a more accurate description.
    101 
    10286   e. p.6: "The function assembly_1_psuedo_instruction ... is essentially
    10387      the composition of the two passes": is it or isn't it?
     
    10589   f. p.6: "... markedly different ways": well, all of the instructions jump
    10690      to one place or another, so they're not that different!
    107 
    108    g. p.6: "MCS-51's limited code memory"; why not state the limits?
    10991
    11092   h. p.14: "this policy is the only possible policy _in theory_ that can
     
    198180    fact, these sentences seem out of place.
    199181
    200 17. It is not clear why both references [5] and [6] are needed. Nor why both
    201     references [7] and [8] are needed.
    202 
    20318218. Research Questions
    204183
     
    218197x. Minor typos / issues
    219198
    220    a. p.3: "we proved the implementation of the assembler": inconsistency of
    221       verb tense.
    222 
    223    b. p.3: The subsection heading "Overview of the paper." is unnecessary.
    224 
    225199   c. p.4: It would be helpful if the abbreviated definitions of
    226200       preinstruction, instruction, and pseudoinstruction included more
    227201       elements in common with themselves and with the other examples
    228202       (notably DEC).
    229 
    230    d. p.10: "eject out" -> "project"
    231 
    232    e. p.10: "lemmas appears" -> "lemmas appear"
    233 
    234    f. p.11: "Huch" -> "Tuch"
    235 
    236    g. p.14: "constraint" -> "constrain"
Note: See TracChangeset for help on using the changeset viewer.