Dear all,
In preparation of our TC6 meeting in Coimbra, here is the report on the IFIP TC6 journal prepared by the TC6 committee on "publication/QualityOfConference/LinkWithSpringer" set up in Wroclaw.
We are looking forward to an interesting discussion in Coimbra.
Best regards, Guy
_______________________________________________________________________________
IFIP TC6 committee on publication/QualityOfConference/LinkWithSpringer
Committee members: R. Boutaba G. Leduc, chair G. Pujolle H. Rudin
Topics to be addressed by the committee:
- IFIP TC6 Journal - Quality of Conferences - Link with Springer
Only the first issue (the IFIP TC6 Journal) is addressed here, although some obvious link with Springer is also implicitly there. We did not discuss the "Quality of Conferences" issue, but noticed anyway that this topic seems to have moved to another TC6 committee chaired by Arun Iyengar.
Table of contents
- Open-access - Electronic-only versus electronic+paper - Copyrights - Quality - Reliable archives - Web site - Business model - The journal
1. Open access
The journal should be open access (http://en.wikipedia.org/wiki/Open_access)
2. Electronic-only versus electronic+paper
Electronic-only is enough. Anyway, if needed an open-access electronic journal can be published simultaneously as a hardcopy, either by an official publisher or not.
3. Copyrights
Copyrights are retained by the author, i.e. not even transfered to IFIP. This is the main asset of open access: authors keep control on their work. Authors are likely to appreciate this feature very much. 'No copyrights to IFIP' also means 'no problem with SSBM (Springer Science & Business Media)'.
We propose to replace the "copyrights transfer form" by a "Code of ethics". This will solve the problem of verbatim republication in absence of copyrights transfer. We would ask the author to sign it instead of a copyright transfer form. We would also require the author to use the URI of the paper (i.e. the link to the journal DL) instead of putting the paper on his/her personal web page (after all the journal DL would be freely accessible to anyone).
If an IFIP conference paper (for which IFIP may still hold the copyrights in some cases) is extended and published in the TC6 journal, the traditional problem of reuse of copyrighted material applies. This problem will disappear when most IFIP TC6 conferences will be published in the IFIP DL (and not any more by SSBM).
[Question] How can we use the IFIP name and logo for the journal without having to apply to the usual IFIP bylaws (principally the copyrights transfer).
4. Quality
Authors of best papers from our best conferences (Middleware, WWW, Networking, ...) will be invited to submit extended paper versions to the journal. We need a strong editorial board.
We should first provide and agree on a list of (broad) topics and at least one expert per topic to be part of the editorial board. An expert can be associated with several topics. We need to fix a ratio between the number of editorial board members and the number of "submitted+invited" papers: suppose we get 40 papers to review per year (at the beginning at least), 20 editorial board members are enough. This would lead to 2 papers per year per editorial board member. Note that the editorial board is not a list of reviewers, but a list of people responsible for handling the review process, including the selection of reviewers.
The rule should be: only renowned and reliable people should be in the editorial board. This list will set the tone of the journal. No focus on one sub-community in particular (e.g. the performance community). Fair representation of American, Asian and European members. The purpose of this journal is not to set up an "IFIP mafia" against other mafias such as SIGCOMM or INFOCOM. If we are considered as another mafia, we will surely fail. We should be open-minded.
For the open call, the papers should be submitted to the Editor in Chief who will dispatch the submitted papers fairly to the editorial board members.
[Variant] The IEEE Surveys and Tutorials uses an on-demand reviewing procedure by the editorial board. Members are invited to check the list of submitted papers and give reviewing preferences. We could also implement that.
Authors will receive immediate acknowledgment of submission and will be contacted within at most two weeks by the responsible editor who will handle their paper.
Short reviewing delays (3 months).
We can define a maximum number of very best papers per conference (depending on the quality level of the conference) that are invited to submit an extended version to the IFIP TC6 journal. Order of magnitude: 2% of the submitted papers. For a big conference like Networking, this would mean 9 papers.
[Question] Is there a problem with IEEE/IFIP conferences?
5. Reliable archives
Example of the open access 'Logical Methods in Computer Science' Journal (http://www.lmcs-online.org/index.php): Disk archive and a hardcopy of the contents is maintained by the Department of Theoretical Computer Science at the Technical University of Braunschweig, Germany, and also by a large number of mirror sites around the world. Something similar should be in place for the TC6 journal.
6. Web site
We have three ways:
A. Experienced centre
We delegate this responsibility to a experienced centre, which is already managing one or several (preferably open access) journals. The cost, if reasonable, can be covered by IFIP TC6 for example. A possible example is the GET in France which already publishes "Annals of Telecommunications" and has some resources to do so. There are other possibilities.
B. Google-like solution
Same as A, with Google handling the process. If the author retains the copyrights (i.e. does not transfer the copyrights to IFIP (and SSBM)), Google should not raise any copyrights issue here. Probably no cost for TC6 either. I don't know if Google would accept to handle the typesetting which is resource consuming.
Though less appealing than Google, there are other possibilities, such as ISTE (International Society for Technology in Education, http://www.iste.org/).
C. One of us is responsible for it with some support from his institution or IFIP (?)
Free, open source software is available for those wishing to start up new journals
For example the Open Journal Systems (OJS) http://pkp.sfu.ca/ojs/ developed by the Public Knowledge Project http://www.pkp.ubc.ca/index.html. While OJS is designed for academic publishing, it can be used by anyone.
System Requirements for OJS: To run OJS 2.x, your web server will need: - PHP 4.2.x or later (including PHP 5.x) with MySQL or PostgreSQL support - A database server: MySQL 3.23 or later OR PostgreSQL 7.1 or later - UNIX-like OS recommended (such as Linux, FreeBSD, Solaris, Mac OS X, etc.). OJS 2.0.2 and above supports Windows servers.
Another open source software to run a conference or journal: - OpenConf: http://www.zakongroup.com/technology/openconf.shtml
7. Business model
Depending on the solutions, the costs may differ. The following costs are foreseen: - typesetting costs (perhaps the highest of all costs involved) - managing the web site (storage of papers, interface and reviewing process) - ensuring the reliability of the archives - possibly handling printed copies (perhaps only for reliable archiving)
This is not a big deal for just a journal, but it may become substantial if this grows to a full DL. Perhaps an institution could handle it (e.g. The MIT manages the 'Logical Methods in Computer Science' Journal).
If Google (or ISTE) manages it, they get revenues from advertisements (or use our journal as an advertisement for other things).
If a university manages it, we don't know.
We need some continuity guarantee also. The journal should not depend on the presence of an individual (e.g. a professor) at a given institution. People move.
[Question] Could we estimate the typesetting costs? Guy Pujolle will ask 'Annals of Telecom' to have estimates of the needed resources/costs for this journal which publishes more or less the same number of papers as our TC6 journal would. Note that IEEE charges $60 for the typesetting of a page (and the total cost for an extra page is at least $200).
8. The journal
Name of the journal: IFIP Journal on Networking? [Is it broad enough?]
[Variant]: IFIP Journal on Communications Systems [May be too broad, we don't really address the physical layer in TC6]
Acronym: JONET? JoN?
Logo [Otto, any suggestion?]
ISSN Number needed.
Structure of the board (inspired by other Open Access journals):
- Steering committee
It will be responsible for maintaining the philosophy of the journal in the long run and to decide any changes in the direction of the journal. For example, the steering committee should appoint the editor-in-chief (EiC), then the associate EiC in consultation with EiC. The steering committee has also to approve appointed area editors and updates to the scope of the journal. The steering committee can be composed of the founders of the journal. The EiC and other members below should be renewed periodically. Maybe the steering committee should also be renewed, e.g. (re)elected by TC6 on a regular basis.
- One Editor-in-Chief (EiC)
- One Associate Editor-in-Chief
He can handle special issues for example. He will definitely have to handle all those papers for which the editor in chief has a conflict of interest. The associate editor in chief should also take over in case the editor in chief is unavailable.
- Editorial Board
Small group of 20 people max. In case of conflict of interest, the EiC may act as editorial board member.
- Liaison board
If the need arises, we can have liaisons to the following bodies:
Liaison to IFIP TC6, ACM SIGCOMM, IEEE ComSoc, IEEE Computer Society Liaison to conferences such as Networking, etc Liaison to some non open-access journal (e.g. IEEE, ACM, COMNET, ...) Liaison to W3C consortium Liaison to DBLP, http://www.informatik.uni-trier.de/%7Eley/db/ Liaison to Google Liaison to University libraries Liaison to the directory of open access journals (http://www.doaj.org/) Liaison to CoRR: see http://arxiv.org/corr/home Journal of the Computing Research Repository (CoRR) Liaison to the public knowledge project http://pkp.sfu.ca/harvester/ ...
- Web Master
- Periodicity, Call for papers and Invitations to submit
We will start with a quaterly-published Journal, which will carry full versions of the best papers presented in the TC6 sponsored conferences. Assuming about 10 papers per issue, we believe that we will have a good supply of good papers. In the second year of its operation, we will start accepting papers from the open literature that will permit us to build it up to a monthly publication. The final goal is to alternate issues, one with papers from TC6 conferences, and one from the open call. Several good quality IFIP TC6 conferences take place every year. The total number of submitted papers is about 2000. The number of presented papers in a conference that will be invited for publication in the new Journal will be proportional to the number of submitted papers to the conference. In this way, more papers from prestigious conferences (such as the Networking series that typically attracts 450 submissions) will be published in the Journal than from a lesser conference, which typically does not attract many papers. Assuming that we accept 1 paper for every 50 submitted papers, we will have about 40 papers per year. (These ratios can be changed depending on our discussions). The acceptance procedure will be as follows. Once the selection of the papers has been concluded, we will invite a number of authors to submit full papers. The papers will be sent to referees along with the original reviews done for the conference. The refereeing process should take at most 3 months. Papers from the open call will follow the normal reviewing process.
We need criteria to evaluate the quality of a conference (our second mandate, not really addressed at this stage). Perhaps some impact factor criterion. Anyway, if we apply the principle above, namely inviting the top 2% of every TC6 conference (where 2% is applied to the number of submitted papers), this may be fair with respect to all TC6 WGs and events. By invitation we mean 'invitation to submit an extended version which will be carefully reviewed'. Workshops are not considered here.