Transformation Tool Contest 2017 Open Peer Review FamiliesToPersons (Responses)
 Share
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

 
View only
 
 
ABCDEFGHIJKLMNOPQR
1
TimestampWhat is your name?
Which solution are you reviewing?
How do you rate the overall quality of the solution?
How many failures did you encounter? According to the case description, a failure is a test failure even though the solution should pass.
How many limitations did you encounter? According to the case description, a limitation is a test failure because the test requires features that are not supported.
How do you rate the understandability of the solution?
How confident are you about your evaluation?
What did you like about the solution?
What did you dislike about the solution?
Do you have other comments about the solution?
Do you have any feedback for the solution submitters to improve their solution or paper?
2
6/8/2017 21:37:57Georg HinkelYage2Could not testCould not test22
The rule visualization is a good starting point, though in the current state, it is hard to imagine the transformation logic from them.
The solution description is in a very early state. There is no source code repository or a SHARE image to test the solution or at least review the code. Thus, I can only judge based on the paper.
Please add at least a source code repository. There are many things that I am unable to tell based on the solution paper. For instance, how does the solution break down the last name of a person? I cannot see anything in that regard. I would assume this is done by some Java glue code, but I am anything but sure.
There are many things that need to be improved, starting with an introduction into the paper, until a more precise evaluation than simply claiming that it is the fastest solution ever. From the solution itself, I cannot really say anything, except to ask again for a source code repository and instructions how the tool is to be used.
3
6/10/2017 1:01:49Georg HinkelEVL+Strace3Could not testCould not test23
The general approach of EVL to show constraint violations in a separate window of the editor is very interesting.
I don't really understand what is the added value of using Epsilon here, because the solution seems very, very verbose and quite any case is handled explicitly. So essentially, you do not really have an advantage over classic general-purpose languages, but lose a lot of advantages. The implementation for the automated version feels a bit like a hack.
I could not execute the solution automatically. Integrating the solution with the benchmarx framework would be nice.
The name Strace is very confusing, because you would expect that this is the name of a tool where in truth, it is an API generated by Modebite. Perhaps it would be more intuitive if you renamed the solution to expose the Modebite tool more explicitly. Further, the section in the paper that introduces the generated trace models is quite verbose, but I guess that will change anyhow as the post-proceedings employ a page limitation.
4
6/11/2017 21:07:14Tassilo HornSDMLib33
I don't know. 3 tests fail but the paper does not explain why. It actually doesn't even tell that there are failing tests.
44
Very fast, pretty concise, solves almost all test cases, and I like the visualizations of the patterns.
It's not bidirectional at all. The forward and backward transformations are completely separate.

I think the approach to encode "check-if-exists-or-else-create" semantics with optional subpatterns and NACs will quickly become unwieldly for more complex/realistic metamodels.
The solution paper suffices to explain the solution reasonably well, however it
doesn't meet the requirements mentioned in the case description. That is, it
doesn't mention which test cases it passes expectedly or unexpectedly, and
likewise it doesn't mention if the three failing tests fail due to a tool
limitation, a bug, or whatever. (Well, it actually doesn't mention that there
are failing tests at all.)
The solution paper suffices to explain the solution reasonably well, however it
doesn't meet the requirements mentioned in the case description. That is, it
doesn't mention which test cases it passes expectedly or unexpectedly, and
likewise it doesn't mention if the three failing tests fail due to a tool
limitation, a bug, or whatever. (Well, it actually doesn't mention that there
are failing tests at all.)
5
6/11/2017 21:07:43Leila Samimi-DehkordiYage2
Since I could not find any implementation for the method (the SHARE demo and SCM link were missed), the number of failures is not clear.
Since I could not find any implementation for the method (the SHARE demo and SCM link were missed), the number of limitations is not clear.
13
Since there exists no implementation for the Families 2 Persons by Yage and I can only find the description paper on the TTC site, I think the tool is not mature enough to have a true judgement.
The results of the approach are not presented in the description paper and the solution is not clearly described. May be it is needed first to explain the approach requirements, which is not considered in the paper.
In this approach, it is required to program the forward and backward directions of the transformation separately. Therefore, the compatibility of the directions is not checked.
-
6
6/11/2017 21:14:08Leila Samimi-DehkordiSDMLib3
Since I could not find any implementation for the method (the SHARE demo and SCM link were missed), the number of failures is not clear.
Since I could not find any implementation for the method (the SHARE demo and SCM link were missed), the number of limitations is not clear.
44
My solution to this question is based on the description paper. The performance of the approach is considerable in comparison to the well-known bx approaches such as eMoflon and medini. The way of defining the corr model is interesting for me.
The format of the model is not EMF-based and SDMLib class model cannot support aggregation; therefore, it may not be a feasible approach for more complex cases.
This method needs considerable effort for writing the transformation code in comparison to the declarative bx approaches, because both the forward and backward directions must be written. Another point which must be considered is that how the compatibility of forward and backward transformations can be checked.
-
7
6/11/2017 21:21:04Tassilo HornNMF3
I don't know because I haven't been able to run the solution. But according to the author, there are probably 3 failing tests.
According to the author, the remaining failures are caused by inappropriately translating EMF change events to NMF's change events. This is only an EMF-interoperability concern which could be addressed with a bit of work.
33
The ability to add temporary attributes to every element (called temporary stereotypes).
- runs only on Windows
- the rules are all very simple but all complexity is in helpers
The paper doesn't help much in understanding the solution. I've posted a review at https://github.com/eMoflon/benchmarx/issues/62 where Georg already responded with good comments which resolved many of my questions and could be used to improve the final paper.
The paper doesn't help much in understanding the solution. I've posted a review at https://github.com/eMoflon/benchmarx/issues/62 where Georg already responded with good comments which resolved many of my questions and could be used to improve the final paper.
8
6/14/2017 15:18:10Kevin LanoNMF3
Unable to access the executable solution
Unable to access executable solution
34
Based on established theory of lenses.
Low-level coding of the solution in C#
Page 3: some more explanation of the meaning of the synchronisation block diagrams is needed.

The machinery of the code seems excessively complex for such a simple transformation.

Page 4: Technology incompatibilities between NMS and EMF seem to be the major problem and cause of inefficiency.
9
6/14/2017 21:25:26Kevin LanoFunnyQT4
Unable to access the demo
Unable to access the demo
23
Concise and well-structured
Cryptic notation makes solution difficult to understand.
Page 4: "exectued"

Page 5: the use of synthetic ids is an example of the Auxiliary Correspondence Model bx design pattern.
10
6/26/2017 15:19:02Albert ZuendorfUML-RSDS40245
The OCL is quite straigt forward
I believe the solution does not handle the change of gender nor the deletion of elements.
Nicecool
11
6/28/2017 13:50:41Albert ZuendorfEVL+Strace40035
New approach for me. Very Interesting
Looks a bit lengthy.
Looking forward to the presentation.
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Loading...
 
 
 
Form Responses 1