Footnotes regarding publication list

These are just footnotes. Footnotes are there to be skipped, otherwise they would not be footnotes. However, some people are tempted to read footnotes. (Perhaps because we keep wondering: what would we miss if we never read footnotes?) You have been warned:

The manuscript numbers on the publication list are there just to provide reference and stable URL, not for the purpose of the counting number of publications. In particular, the list below also contains technical reports.
Why technical reports? I sometimes make technical reports public for early dissemination, while the publications are still being reviewed. I subsequently do not remove the reports from this page, to preserve the numbering on this page, the chronology of the work, and the free availability of the ideas.
Conferences and journals. Most of my refereed publications are in conferences; I do not regularly submit my work to journals. The reason is that top conferences in Computer Science tend to have 10-25 pages of length and they represent complete articles. They are typically peer-reviewed by carefully selected program committee. Moreover, the work tends to get more visibility in selective conferences than in journals. For more information, please see:
- An article on tenure review and conferences
- Statistics on acceptance rates of journals in engineering
- Acceptance rates of various conferences
- Statistics from Academic Search. In November 2010, the 4 top Computer Science conferences were SOSP, SIGCOMM, MOBICOM, POPL, SIGGRAPH (of which POPL is in programming languages) and the top 4 "Computer Science" journals listed were Science, Nature, CCR - Computer Communication Review, Machine Learning, and Artificial Intelligence.
Author's versions: I am posting here "author's versions" of manuscripts. Most of the refereed work is published by publishers such as ACM and Springer, so their web sites contain the "publisher's versions" of this work. The publishers like to call the version on their web sites "authoritative", which means that they do not include any occasional corrections that I may have found after the publication date. I am grateful to the publishers (in particular those making large profit margins) for generously allowing me to post my own work (which is about the research that they did not fund), on the web pages of the institution that pays them tens of thousands of CHF for library subscriptions annually.