• Andrea Omicini
    Andrea Omicini, 01/04/2018 20:11

    Dear Andrea Omicini,

    We are pleased to inform you that your paper 

    Twenty Years of Coordination Technologies: State-of-the-Art and Perspectives

    has been ACCEPTED for publication at COORDINATION 2018.

    Please, carefully consider the reviews below to improve your paper. In order to do this, you are allowed to use up to two more pages w.r.t. the original submission.

    The camera ready final version is due on 

    April 15, 2018. 

    This is a FIRM deadline for the production of the proceedings. Please note that we have very limited time to get the proceedings ready. Further instructions for the preparation of the camera-ready final version will be sent soon.

    At least one of the authors must register for COORDINATION to present the paper. Otherwise, this may prevent inclusion in the proceedings. We hope, of course, that many of your co-authors will be able to participate in the conference. Please forward this email to them, if applicable.

    Again, congratulations on having your paper accepted! We are looking forward to an exciting program and to seeing you soon in Madrid: June 18-21!

    With best regards Giovanna di Marzo Serugendo and Michele Loreti


     REVIEW 1 

    PAPER: 10 TITLE: Twenty Years of Coordination Technologies: State-of-the-Art and Perspectives AUTHORS: Giovanni Ciatto, Stefano Mariani, Andrea Omicini, Franco Zambonelli and Maxime Louvel

    Overall evaluation: -1 (weak reject)


     Overall evaluation 

    In this paper the history of the COORDINATION conference is analyzed mainly with respect to the coordination technologies proposed at the Conference throughout the years, where by "technology" it is intended a "software artefact offering an API exploitable by other software to coordinate its components".

    The paper starts with a brief overview of the data related to the Conference (nr. of papers, nr. of citations etc.) and then focuses on the technologies proposed in the various editions of the Conference. 

    The paper does not take into consideration models, architectures, language definitions, verification techniques and case studies-on the use of models, languages and technologies-presented at the Conference.

    The criteria for the analysis of the technologies are mainly (1) availability-on the web-of the proposed technology, (2) latest updates, (3) known deployments, (4) installation procedures and (5) basic usage (i.e. whether the SW compiles and runs some tests).

    On the basis of the analysis carried out in Authors draw some conclusions, basically that the current state-of-the-art of coordination technologies is neither discomforting nor encouraging.

    I have serious problems with this paper.

    First of all, what is the main aim of the paper and the target audience? If the aim was to draw the history of the COORDINATION Conference, why considering mainly-actually only-coordination technologies? Why not addressing also the other dimensions, like modelling and verification technologies as well as applications (case studies)? And why writing a paper for something that IMHO should be mainly of interest for the Conference Steering Board (so that a report to the Board would have been enough)?

    If, instead, the aim was to shed some light on coordination technologies, say for practitioners, then why considering only those which have been presented at COORDINATION? Why not addressing coordination technologies in general, mentioning also those which were not coming from the Conference (e.g. BIP, or SCEL, just to mention some)?

    On a more technical side, I have difficulties in understanding and accepting Sect. 2, and in particular Table 1. The table shows some indicators for each edition of the Conference, from 1996 to 2017, and the Section elaborates on the indicators with some statistical considerations. 

    Now, first of all, the average and standard deviation are weighted; the actual weights that have been used are not reported in the paper but it is written that older figures weight more than recent ones. The reported reason for this choice for the weights is that "the older papers had more time to generate citations and downloads". Frankly I fail to understand the rationale and correctness of this choice: why should the citations to older papers have more value than those to newer ones? If I  really had to discriminate citations on the basis of the age of the papers-which, by the way, I would not give for granted I should!-I would give more weight to those to newer papers, in order to compensate for the shorter time available for collecting citations. And similarly for downloads. 

    But, what I really did not understand, is why also the average No. of papers has been computed weighting the papers in the same way! So, a weighted average of 24,36 is provided for the No. of papers whereas the average No. of papers is 20.53. Why 34 papers of 1996 should weight more than 34 papers of 2002, and 14 papers of 2002 should carry more value than 14 papers of 2017? So, Sect. 2 is rather controversial, to say the least, in my opinion.

    Then, the rest of the paper is concerned with discussing the various technologies, the selection criteria and the tests which have been performed on the selected technologies. Once again, the scientific quality of the tests, and consequently of the related results, is rather fair, not to say low. The main tests have to do with compiling the SW and executing it, as well as reporting the status of the distribution (i.e. things like the last update, or information on deployments or success stories as reported in the distribution site). No details are given on the kinds of execution which have been performed: have just "Hallo World" codes been executed or serious benchmarks and stress tests have been executed? Unfortunately no information is provided explicitly ... Nevertheless, this part of the paper provides some information which might be useful for practictioners, although the presentation is more similar to posts to a user forum than to what one would expect from a scientific paper.

    The last section (Sect. 3, Discussion) is also rather problematic: it is kind of driven by Figures 5 to 7, but often the relationship between the text in the section and the figures it refers to is not completely clear. In addition, the figures themselves are not always clear; for instance, in Fig. 6, what does the directed arc labelled "-localities" from TuCSoN to Klava mean? Does it mean that one difference between TuCSoN and Klava is that the latter has no localities whereas the former has?[notice the orientation of the arrow and the "-" (minus) before "localities"] Well, as far as I know, Klava is (a Java package that implements the run-time system for X-Klaim, the high level language based on Klaim which is) based on localities! ...

    Finally, the style and use of english language should be improved.

    In conclusion, I think the paper has not an clear aim, neither a clear target audience and has several technical weaknesses. Consequently it should be rejected. On the other hand, it might provide some information for practictioners, although very limited and rather incomplete. So, I propose a weak reject.

    Some detailed comments:

    pag. 2: For the latter --> For the last

    pag. 3: was gathered --> were gathered

    pag. 4: takes in Madrid --> takes place in Madrid

    pag. 7: acceptation --> sense ?

    pag. 7: You speak about 47 papers, but only 35 technologies (and references!) are reported in Table 2. Maybe I missed something ...

    pag. 11: later described --> described later

    pag. 11: How does the nomber 16 (total nr. of technologies) relate to the number given on page 7 (47)? At this point I am a bit confused ...

    pag. 13: versione --> version

    pag. 17: is dates back --> dates back

    pag. 20: Implemantation of CRIME --> The implemantation of CRIME

    pag. 20: the web page has been updated last in 2010 --> the last update of the web page dates back to 2010

    pag 24: The sentence "From the description of the selected technologies we gathered, three are the main features motivating their evolution:" is unclear.


     REVIEW 2 

    PAPER: 10 TITLE: Twenty Years of Coordination Technologies: State-of-the-Art and Perspectives AUTHORS: Giovanni Ciatto, Stefano Mariani, Andrea Omicini, Franco Zambonelli and Maxime Louvel

    Overall evaluation: 2 (accept)


     Overall evaluation 

    This paper surveys the coordination technologies that have been presented and discussed at all the previous editions of the COORDINATION conference. With the aim of shedding light on the current status of these technologies, the authors analyze them in an attempt to understand success stories, limitations, and possibly reveal the gap between actual technologies, theoretical models, and novel application needs.

    Of course, the topic of the paper is certainly relevant for the COORDINATION conference and for the research community attending the conference. The presentation is quite good and the paper is in general well-written. I think that the authors have done a good and complete job and that the paper should be accepted, although some improvements listed below are needed. Furthermore, it would also be interesting to know whether other "alive" coordination technologies have been published in other venues. This would be valuable for our research community. Maybe the authors could add some paragraphs mentioning the closest technologies to the research lines analyzed in the paper. 

    I finish by pointing out some clarifications about the X-Klaim/Klava technology and the Klaim model. This should impact, at least, sections 2.2 (pag. 12) and 2.3 (pag. 18); in the latter one, as well as in Fig. 6, you can also mention Klaim's type system for controlling access (see below). X-Klaim is a prototypical programming language derived from the coordination model Klaim. Klaim has been expressly designed for the development of network-aware distributed applications with mobile code. Originally, X-Klaim was the technology supporting Klaim. Afterward, it also embedded object-oriented features following the approach of O'Klaim and MoMi. Klava is a Java library implementing the X-Klaim run-time system. It provides a suitable API for programming according to Klaim/X-Klaim. Preliminary presentations of the Klaim language can be found in the proceedings of TAPSOFT 1997 and COORDINATION 1997, where it was named LLinda (Locality based Linda). The language was introduced in TAPSOFT 1997 as a programming paradigm stemming from Linda, that supports distributed programming with explicit localities and where computational agents can migrate from one computing environment to another. In COORDINATION 1997, a type system for LLinda was introduced that permits statically checking access rights violations of mobile agents. Types are used to describe processes intentions (read, write, execute, ...) relatively to the different localities they are willing to interact with or they want to migrate to. The type system is used to determine the operations that processes want to perform at each locality, to check whether they comply with the declared intentions and whether they have the necessary rights to perform the intended operations at the specific localities. The official Klaim website is http://music.dsi.unifi.it/klaim.html. The standard bibliographic references are:

    • Klaim A Rocco De Nicola, Gian Luigi Ferrari, Rosario Pugliese:

    KLAIM: A Kernel Language for Agents Interaction and Mobility. IEEE Trans. Software Eng. 24(5): 315-330 (1998) B Lorenzo Bettini, Viviana Bono, Rocco De Nicola, Gian Luigi Ferrari, Daniele Gorla, Michele Loreti, Eugenio Moggi, Rosario Pugliese, Emilio Tuosto, Betti Venneri: The Klaim Project: Theory and Practice. Global Computing 2003: 88-150

    • X-Klaim & Klava C Lorenzo Bettini, Rocco De Nicola, Rosario Pugliese:

    Klava: a Java package for distributed and mobile applications. Softw., Pract. Exper. 32(14): 1365-1394 (2002)

    • MoMi D Lorenzo Bettini, Betti Venneri, Viviana Bono:

    MOMI: a calculus for mobile mixins. Acta Inf. 42(2-3): 143-190 (2005) Thus, A and B should be used in place of 16, C in place of 17, and D in place of 11.

    Minor points

    • pag. 1: "take a a step back" --> "take a step back".
    • Most of the items occurring in listings (see e.g. pag. 2) lack the final punctuation mark.
    • It seems to me that Sec. 1.1 and Sec. 1.2 (at least, the first paragraph) contain some repetitions. 
    • As far as I can see, in many cases (e.g. Table 1, Fig. 1 and 2, and the comments thereof), "the most cited paper" and "the most download paper" (btw, "downloaded" instead of "download") should rather be "the number of citations of the most cited paper" and "the number of downloads of the most downloaded paper", respectively.
    • Table 2: Name: X-Klaim; Year: 2002; Model: Klaim + O'Klaim; Webpage: http://music.dsi.unifi.it/klaim.html + http://music.dsi.unifi.it/xklaim. This would be also more coherent with Table 3.
    • pag. 12: "For instance, the O’Klaim language presented in 12 evolved into the X-Klaim project 13..." --> "For instance, the Klaim A,B and O’Klaim 12 languages evolved into the X-Klaim project C,13"
    • pag. 17, first paragraph: you write "... we were able to successfully compile and execute the code. No source code is provided and ...": apparently, there is a contradiction between the ability to compile the code and the unavailability of source code. Please, clarify. 
    • pag. 17, about Reo: "... data may flow from the input ports of components to the ioutput ones of others.": I think it should be the other way round.
    • pag. 17, last line: remove the parenthesis.
    • pag. 18, line 4: "Reo an distributed systems" --> "Reo and distributed systems".
    • pag. 21, line 6: "with a SBT" --> "with an SBT".
    • pag. 24, line 4: "in.a" --> "in a".
    • references 56 and 57 are the same.

     REVIEW 3 

    PAPER: 10 TITLE: Twenty Years of Coordination Technologies: State-of-the-Art and Perspectives AUTHORS: Giovanni Ciatto, Stefano Mariani, Andrea Omicini, Franco Zambonelli and Maxime Louvel

    Overall evaluation: 1 (weak accept)


     Overall evaluation 

    The paper offers a survey of coordination technologies from the point of view of the COORDINATION conference itself. It classifies and analyses methods and paradigms reported along the 20-years long conference series, looking at their roots and evolution, with a particular focus on the ones which eventually lead to real technologies. The limits imposed by the perspective adopted are clear. However, even if one could easily argue that there is more on coordination than presented at COORDINATION (for example a number of contributions inspired by process and mobile calculi were typically published elsewhere), the conference is a landmark in the area and the perspective adopted in the survey provides much more than “historical” curiosity.

    The survey is well structured and planned: from the classification of bibliometrics data collected from the publisher, to the identification of the approaches which materialised into concrete technologies (or at least aimed at) and a “hands-on” verification of their present status. The methodology is clearly described; the resulting survey informative and, in general, a pleasant reading.

    I am not sure whether COORDINATION is willing to accept this kind of papers, lacking an actual technical contribution to the field. Maybe the topic fits better in a journal publication or as a keynote/invited lecture. Such is not my decision, however. In any case I agree the survey may raise an interesting discussion in the conference, maybe even in the direction which is somehow overlooked in this work: the one that could try to address the question implicitly raised in the introduction but, unfortunately, left unanswered in the paper. That is “the role of coordination models should expectedly grow along with the relevance of coordination technologies within ICT system: instead this is apparently not happening — yet” .

    Actually, my main criticism to the paper is the lack of a discussion section in which past and current trends on coordination and its role within software engineering could be identified and critically addressed. The analytic dimension in the paper is its weakest side.

    DETAILS

    Pag 24, line 4: “in.a particular” —> “in a particular 

    References: most of them are incomplete, namely all that refer to the proceedings of COORDINATION. LNCS is not mentioned and the LNCS volume omitted. I guess this is due to a wrong use made of DBLP … MUST be corrected if the paper is accepted.


     REVIEW 4 

    PAPER: 10 TITLE: Twenty Years of Coordination Technologies: State-of-the-Art and Perspectives AUTHORS: Giovanni Ciatto, Stefano Mariani, Andrea Omicini, Franco Zambonelli and Maxime Louvel

    Overall evaluation: 2 (accept)


     Overall evaluation 

    The submission comprises an extensive survey of the technologies presented in the 20 years of the COORDINATION conference series. First, it summarizes the number of papers published per edition, their citations and downloads (figures taken from SpringerLink’s BookMetrix service), and tries to delimit phases in the conference, and tendencies.

    Then, from the almost four hundred papers published, the survey focus on those actually presenting a coordination technology (i.e. those papers mentioning a software artifact or tool), resulting in some 35 technologies, which are individually analyzed in the survey. Many of them, esp. for the initial days (e.g. Sonia, Laura…), have completely disappeared, and they cannot be found online, while others required deprecated software (old Eclipse or Java versions) for running. Finally, eleven “surviving” coordination technologies are selected, and they are discussed more at length, trying to identify and discuss their relations (evolution, shared features, influence, differences, etc.)

    The submission represents a valuable piece of work from its authors, giving a panorama of the COORDINATION conference over the years, and therefore of the coordination paradigm itself. It may be more descriptive, collecting bibliographic data and facts (online reachability) than comprehensive (relating technologies, emphasizing novelties, interesting features and deficiencies) but I assume this is derived for an attempt of being “neutral” with respect to the technologies, and therefore to the community behind them. Complete neutrality is impossible, and I am certain that nobody will be completely happy of their position in the general picture, but I think the submission would be a good way of celebrating the conference’s 20th edition (however, I don’t know if other surveys have been submitted).

    Just one final comment, the submission is slightly longer than indicated in the call: 30 pages instead of 25 + 2 of references, but the excess comes from the 5 pages of the references section, which are unavoidable given the number of papers mentioned in the survey.

  • Andrea Omicini
    Andrea Omicini, 19/11/2018 18:51

    Dear Giovanni, Stefano, Andrea, Franco and Maxime,

    we are glad to invite you to submit an extended version of your survey appeared in the Proceedings of COORDINATION 2018:

    Twenty Years of Coordination Technologies: State-of-the-Art and Perspectives

    to the Special Issue in Journal on Algebraic Methods in Programming for the 20 Years of Coordination Models and Languages. This special issue is dedicated to the presentation of surveys describing important results and successful stories that originated in the context of COORDINATION. Info about the Special Issue are available at the following link:

    https://www.journals.elsevier.com/journal-of-logical-and-algebraic-methods-in-programming/call-for-papers/20-years-of-coordination-models-and-languages

    Important dates

    • Deadline for submissions: 15 January 2019
    • First round reviews: March 2019
    • Submission of revised papers: April 2019
    • Notification of acceptance: June 2019
    • Camera-ready version: August 2019

    For the preparation of your contribution please take carefully into account all the comments that you received by the reviewers of your submission to COORDINATIO 2018 and the discussions during your presentation. Your extended version should contain at least 30% of new material to be considered for publication.

    Manuscripts should be formatted according to the rules of the Software Quality Journal, available by following the "Instructions for Authors" link on the journal webpage (https://www.elsevier.com/journals/journal-of-logical-and-algebraic-methods-in-programming/2352-2208/guide-for-authors).

    With best wishes, Michele and Giovanna

Partita IVA: 01131710376 - Copyright © 2008-2021 APICe@DISI Research Group - PRIVACY