The Steering Committee of PODC and the Managing Editor (ME) of Distributed Computing have agreed to a long-term collaboration for the purpose of producing an issue of Distributed Computing (DC) every year based on a selection of papers from the previous year's PODC.
This document describes the guidelines for this collaboration.
Each year the Program Committee (PC) of PODC will select a small number
(four to six) of papers among those accepted, for invitation to a special
issue of DC. The existence of this special issue will be mentioned
in the PODC Call for Papers. (The identity of the invited papers
will not be announced publicly to avoid embarrassment in case an invited
paper does not ultimately appear in the special issue.)
(2) Criteria for inviting papers:
The primary criterion for choosing the papers to invite is quality: only the very best papers accepted for the conference should be invited. However, the PC may also consider other criteria, such as breadth of coverage. Thus, it may be that one of the top-ranked papers is not invited, while a lower-ranked --- but still highly-ranked --- paper is. (The numerical ranking of papers that the PC often uses as an initial basis for its deliberations will generally have little to do with the eventual choice of papers to be invited, as this can change dramatically during the PC meeting.)
Each PC will decide on its own how to deal with the question of inviting
papers co-authored by PC members. One suggestion is that PC members
whose papers are candidates for invitation to the special issue should
not participate in the final discussion that determines the papers to be
(3) Guest Editor (GE):
The privilege and responsibility of editing the special issue belongs *ex officio* to the PC Chair. Carrying out this task is understood as a normal duty at the time of accepting the position of PODC PC Chair. The specific responsibilities of the PC Chair as GE of the special issue are:
The special issue should be ready for publication for the first issue of DC after twelve months from when the conference was held. This is a tight but achievable schedule. To meet it, the following timeline should be followed:
The process should start when the PC meets, not after the conference is held. (There is typically a delay of four months between the two.)
Immediately after the PC meeting, the GE will invite the author(s) of each selected paper to submit, within four months (i.e., around the time of the conference), a full version of the conference paper for the special issue. (If some of the invited authors decline, and the GE feels that there are good alternative papers that were dropped due to space limitations, he/she may, with the consent of the PC, issue additional invitations.)
The papers must be fully refereed within twelve months of the conference. This period includes a second round of refereeing, for those papers that may need it.
Papers that are unduly delayed because the authors are slow in submitting the full version, or slow in revising their paper, may end up not appearing in the special issue: It is not fair to hold hostages the other papers to one unresponsive author. The paper can appear later in a regular issue of DC, with a note indicating that it had been selected for the special issue but appears separately due to delays.
The GE must ensure timely response by the referees, by negotiating clear deadlines with them in advance, checking up regularly by email and phone if the deadlines are missed, and building some slack for possible delays into these deadlines. In addition, to ensure fast and high-quality refereeing, the GE may find it useful to rely heavily on members of the PC as referees, since these people are familiar with the paper, at least in a preliminary form.
A timeline for the actions involved in editing a special issue appears
in (7) below.
(5) Standards of refereeing:
These will be the usual standards of the journal. Each paper must
be refereed by at least two (and preferably three) qualified reviewers.
The ultimate decision for whether the paper is accepted, revised-and-reconsidered,
or rejected, rests with the GE. He/she must interpret, weigh and
evaluate the referee reports. The decision is not necessarily based
on a simple majority vote of the referees' views. Of course, the
GE should not override a clearly expressed view of the referees without
(6) Interaction between GE and ME:
The ME has no say on the choice of papers to be invited (except insofar as he/she also happens to be a member of the PC).
As soon as the PC decides which papers to invite and the GE secures the agreement of their authors, he/she informs the ME which papers have been invited.
The ME has no say on the choice of referees or on the final decision
for each paper invited. However, the GE will provide to the ME full
documentation for these decisions: copies of the submitted papers
(and revisions, if any), the referee reports, the responses of the authors,
the decisions communicated to the authors etc. These should be communicated
to the ME as they occur, so that the ME can help monitor progress towards
the production of the special issue.
(7) Guest Editor's timeline:
Time 0 is the PC meeting. All others times (in months) are relative to that. The schedule completes within 16 months of the PC meeting, which is approximately 12 months after the conference.
Note that this schedule does not allow for delays. To meet it,
the GE must build slack *within* its constraints. For
example, the schedule assumes a 3-month delay for referee reports. To achieve this, experience suggests that it may be a good idea to request reports within 1.5 months.
|0:||PC selects papers to invite.|
|1:||GE completes negotiations with authors of invited papers and extracts commitments for full version.|
|4:||Full versions at hand.|
|5:||Referee assignment completed.|
|8:||1st round referee reports at hand; first round decisions.|
|12:||Revisions at hand.|
|15:||2nd round referee reports at hand; second round (final) decisions.|
|16:||DC-formated versions ready.|