TUGboat follows the American Mathematical Society's statement of ethics for publications: part IV on the AMS Ethical Guidelines page. Here is the text, to avoid any possible confusion:
Editors are responsible for the timely refereeing of articles and must judge articles by the state of knowledge at the time of submission. Editors should accept a paper for publication only if they are reasonably certain the paper is correct.
The contents of [a] submitted manuscript should be regarded by a journal as privileged information. If the contents of a paper become known in advance of publication solely as a result of its submission to or handling by a journal, and if a later paper based on knowledge of the privileged information is received anywhere (by the same or another journal), then any editor aware of the facts must refuse or delay publication of the later paper until after publication of the first—unless the first author agrees to earlier publication of the later paper.
At the time a manuscript is submitted, editors should notify authors whenever a large backlog of accepted papers may produce inordinate delay in publication. A journal may not delay publication of a paper for reasons of an editor's self interest or of any interest other than the author's. The published article should bear the date on which the manuscript was originally submitted to the journal for publication, together with the dates of any revisions. Editors must be given and accept full scientific responsibility for their journals; when a demand is made by an outside agency for prior review or censorship of articles, that demand must be resisted and, in any event, knowledge of the demand must be made public.
Both editors and referees must respect the confidentiality of materials submitted to them unless these materials have previously been made public, and above all may not appropriate to themselves ideas in work submitted to them or do anything that would impair the rights of authors to the fruits of their labors. Editors must preserve the anonymity of referees unless there is a credible allegation of misuse.
All mathematical publishers, particularly those who draw without charge on the resources of the mathematical community through the use of unpaid editors and referees, must recognize that they have made a compact with the community to disseminate information, and that compact must be weighed in their business decisions.
The Society will not take part in the publishing, printing or promoting of any research journal where there is some acceptance criterion, stated or unstated, that conflicts with the principles of these guidelines. It will promote the quick refereeing and timely publication of articles accepted to its journals.
For TUGboat: 1) read "TUG organization" for "Society" in the last item; 2) submission and revision dates are not included in published articles (per third paragraph), since TUGboat, unlike AMS journals, has no significant backlog.
If any questions or concerns, please email TUGboat@tug.org.