Hi Scott & Jason, Jason McIntyre wrote on Sat, Nov 02, 2019 at 06:16:49PM +0000: > On Sat, Nov 02, 2019 at 12:32:30PM -0500, Scott Cheloha wrote:
>> Cite the paper describing the timing wheel. PDF here: >> >> http://www.cs.columbia.edu/~nahum/w6998/papers/ton97-timing-wheels.pdf >> >> The authors have an older paper describing the same idea, from 1987: >> >> http://www.cs.columbia.edu/~nahum/w6998/papers/sosp87-timing-wheels.pdf >> >> but the 1997 version contains additional information gathered over ten >> years of praxis. The typesetting is also way nicer in 1997. >> >> Once question about the .Rs block: why is the issue number rendered >> before the volume number? Isn't it usually the other way around? >> IIRC the issue is a subset of a given volume. > the ordering of these blocks is kind of set in stone. so mandoc will do > what groff does. i guess a reordering will involve persuading other > people that it makes sense. That is true. Then again, it does not seem unlikely that we could persuade people in this case. Whether volume and issue are used together at all may depend on the discipline (for example, in high energay physics, it appears that usually only volume and page numbers are used for citing journals). But *if* both are used together, i suspect you are right that the other order makes more sense. If you think it is worth it, tell me off-list and i'll post a trivial patch to change the order to [email protected]. Changing the order in mandoc is also trivial and would be done at the same time if the groff folks like the idea. Whatever that leads to, given both .%V and .%N only render a bare number, which certainly shouldn't be changed because i suspect using only one at a time may be the dominant use case, either order is not very clear when using both together. So i suggest quoting the numbers in exactly the same way as given in the original paper: .%V vol. 5 .%N no. 6 >> Oh, another question: is it standard to mention the city of publication >> for an academic journal in a citation? ToN is published from Piscataway, >> NJ, fwiw. > i don;t know, but i think the amount of info used in these blocks is > determined by the person adding it - how much info is useful? for rfcs, > for example, we settled on a standard blurb that both gave credit and > allowed readers to find the docs. whether you need all the refs you've > given i guess it decided by you (i think often a title suffices). I agree with Jason. For journals, in particular for well-known ones, the city of publication is often omitted because it is not really needed for getting a copy. In this case, i would probably omit it. For books, the city is often provided in citations, almost as a matter of convention. For more obscure publications, it may even be important. >> ok? > no objection from me. Same here, but i recommend the change explained above. Also, you are deleting information from the source code comment that would probably be misplaced in the manual: "Scheme 7". You might wish to leave that piece of information in the code unless it happens to be incorrect or irrelevant. Yours, Ingo >> Index: share/man/man9/timeout.9 >> =================================================================== >> RCS file: /cvs/src/share/man/man9/timeout.9,v >> retrieving revision 1.46 >> diff -u -p -r1.46 timeout.9 >> --- share/man/man9/timeout.9 14 Apr 2019 08:51:31 -0000 1.46 >> +++ share/man/man9/timeout.9 2 Nov 2019 17:14:48 -0000 >> @@ -282,3 +282,15 @@ These functions are implemented in the f >> .Xr splclock 9 , >> .Xr tsleep 9 , >> .Xr tvtohz 9 >> +.Rs >> +.%A George Varghese >> +.%A Anthony Lauck >> +.%B Hashed and hierarchical timing wheels: efficient data structures for \ >> +implementing a timer facility >> +.%I IEEE/ACM >> +.%J Transactions on Networking >> +.%V 5 >> +.%N 6 >> +.%P pp. 824\(en834 >> +.%D December 1997 >> +.Re >> Index: sys/kern/kern_timeout.c >> =================================================================== >> RCS file: /cvs/src/sys/kern/kern_timeout.c,v >> retrieving revision 1.60 >> diff -u -p -r1.60 kern_timeout.c >> --- sys/kern/kern_timeout.c 2 Nov 2019 16:56:17 -0000 1.60 >> +++ sys/kern/kern_timeout.c 2 Nov 2019 17:14:48 -0000 >> @@ -45,9 +45,7 @@ >> /* >> * Timeouts are kept in a hierarchical timing wheel. The to_time is the >> value >> * of the global variable "ticks" when the timeout should be called. There >> are >> - * four levels with 256 buckets each. See 'Scheme 7' in >> - * "Hashed and Hierarchical Timing Wheels: Efficient Data Structures for >> - * Implementing a Timer Facility" by George Varghese and Tony Lauck. >> + * four levels with 256 buckets each. >> */ >> #define BUCKETS 1024 >> #define WHEELSIZE 256
